University of Southern California Dissertations and Theses

Fate of antibiotic resistance in anaerobic membrane bioreactors

(USC Thesis Other)

Fate of antibiotic resistance in anaerobic membrane bioreactors

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Fate of Antibiotic Resistance in Anaerobic
Membrane Bioreactors
by
Ali Zarei Baygi
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Engineering (Environmental Engineering))
August 2020
Copyright 2020 Ali Zarei Baygi
ii

Acknowledgments
I would like to thank my supervisor Dr. Adam Smith, without whom I would not be able to complete my
PhD journey. His significant insight and knowledge, greatest supports, and excellent advices made this
success possible. I would like to also express my sincere gratitude to my research committee members:
Drs. Amy Childress, Daniel McCurry, and James Boedicker for the thoughtful comments and
recommendations on this dissertation.
Furthermore, I would like to express my deepest appreciation to the people who worked with me on
this project, Dr. Moustapha Harb, Dr. Lauren Stadler, and Phillip Wang. Their valuable assistance, and
thoughtful ideas made my PhD journey more pleasant. A special Thanks to my family for all their
unconditional support. I would also like to acknowledge the following sources of financial supports for
this research: Viterbi Graduate School Fellowship, and Water for Agriculture Grant from the USDA
National Institute of Food and Agriculture.

iii

Table of Contents
Acknowledgments ........................................................................................................................................... ii
List of Tables ................................................................................................................................................... vi
List of Figures ................................................................................................................................................. vii
Abstract ........................................................................................................................................................... x
1. Anaerobic treatment for mitigation of antibiotic resistance spread to the environment ............................. 1
1.1 Introduction .................................................................................................................................................. 1
1.2 Antibiotic resistance in conventional and advanced anaerobic wastewater treatment processes ............. 3
1.2.1 Conventional anaerobic wastewater treatment processes ..................................................................... 5
1.2.2 AnMBR as an emerging technology to manage spread of antibiotic resistance ..................................... 7
1.3 Comparison of anaerobic treatment with aerobic treatment in regard to antibiotic resistance .............. 10
1.4 Overview of dissertation ............................................................................................................................. 12
References ................................................................................................................................................................ 13
2. Intracellular versus extracellular antibiotic resistance genes in the environment: prevalence, horizontal
transfer, and mitigation strategies .................................................................................................................. 19
Abstract .................................................................................................................................................................... 19
2.1 Introduction ................................................................................................................................................ 20
2.2 Prevalence of iARGs versus eARGs in different environments.................................................................... 23
2.2.1 Impact of environmental characteristics ............................................................................................... 23
2.2.2 Impact of ARG characteristics ................................................................................................................ 26
2.3 Horizontal transfer of iARGs versus eARGs ................................................................................................ 28
2.3.1 Conjugation of iARGs ............................................................................................................................. 29
2.3.2 Transduction of iARGs ............................................................................................................................ 34
2.3.3 Natural transformation of eARGs .......................................................................................................... 36
2.3.4 Impact of selective pressure on horizontal transfer of iARGs and eARGs ............................................. 39
2.4 Impact of advanced treatment technologies on abundance of iARGs and eARGs .................................... 41
2.4.1 Disinfection ............................................................................................................................................ 41
2.4.2 Membrane filtration............................................................................................................................... 45
2.5 Conclusion and Future Research Directions ............................................................................................... 47
References ................................................................................................................................................................ 50
3. Evaluating Antibiotic Resistance Gene Correlations with Antibiotic Exposure Conditions in Anaerobic
Membrane Bioreactors .................................................................................................................................. 58
Abstract .................................................................................................................................................................... 58
3.1 Introduction ................................................................................................................................................ 59
3.2 Materials and methods .............................................................................................................................. 60
3.2.1 Bench-scale anaerobic membrane bioreactor operation ...................................................................... 60
3.2.2 Antibiotic Quantification ........................................................................................................................ 61
3.2.3 ARG quantification by qPCR ................................................................................................................... 62
iv

3.2.4 ARB quantification ................................................................................................................................. 63
3.2.5 Statistical analysis methods ................................................................................................................... 63
3.3 Results and discussion ................................................................................................................................ 64
3.3.1 AnMBR performance and antibiotic removal was robust throughout operation ................................. 64
3.3.2 AnMBR effluent exhibited a unique ARG profile compared to the biomass and membrane biofilm ... 65
3.3.3 Most ARGs followed increasing trends in biomass during incremental antibiotic addition ................. 68
3.3.4 10 µg/L antibiotic concentrations induced spikes in total effluent ARG abundance ............................ 72
3.3.5 AnMBR effluent ARB largely unaffected by different antibiotics and concentrations .......................... 76
3.3.6 Correlation between antibiotics, ARGs, and ARB indicated the presence of multi-drug resistance..... 77
References ................................................................................................................................................................ 79
4. Microbial community and antibiotic resistance profiles of biomass and effluent are distinctly affected by
antibiotics addition to an anaerobic membrane bioreactor .............................................................................. 85
Abstract .................................................................................................................................................................... 85
4.1 Introduction ................................................................................................................................................ 86
4.2 Materials and methods .............................................................................................................................. 88
4.2.1 Configuration of bench-scale AnMBR .................................................................................................... 88
4.2.2 Quantification of ARGs by qPCR............................................................................................................. 89
4.2.3 Quantification of antibiotics by LC-MS .................................................................................................. 89
4.2.4 Microbial community analysis ............................................................................................................... 90
4.2.5 Data analysis........................................................................................................................................... 90
4.3 Results and Discussion ................................................................................................................................ 91
4.3.1 AnMBR system performance was robust during antibiotics addition ................................................... 91
4.3.2 AnMBR biomass was dominated by different ARGs than the AnMBR effluent .................................... 92
4.3.3 Microbial community analysis ............................................................................................................... 96
4.3.4 Correlations between ARGs and microbial community structure indicate a potential for HGT ......... 102
4.4 Conclusions ............................................................................................................................................... 107
References .............................................................................................................................................................. 107
5. Membrane fouling inversely impacts intracellular and extracellular antibiotic resistance gene abundances
in the effluent of an anaerobic membrane bioreactor.................................................................................... 113
Abstract .................................................................................................................................................................. 113
5.1 Introduction .............................................................................................................................................. 114
5.2 Material and Methods .............................................................................................................................. 116
5.2.1 Configuration of reactor and preparation of membrane modules with different levels of fouling .... 116
5.2.2 Quantification of EPS and SMPs ........................................................................................................... 117
5.2.3 DNA extraction and ARGs quantification ............................................................................................. 118
5.2.4 Data analysis......................................................................................................................................... 119
5.3 Results and Discussion .............................................................................................................................. 120
5.3.1 Fouling improved treatment performance of the AnMBR .................................................................. 120
5.3.2 AnMBR membrane biofilms may provide more conducive conditions for ARG transfer than suspended
biomass.............................................................................................................................................................. 123
5.3.3 Membrane fouling increased the abundance of iARGs and decreased the abundance of eARGs in the
AnMBR effluent ................................................................................................................................................. 126
5.3.4 Log removal of total ARGs was not impacted by the extent of fouling ............................................... 130
5.3.5 Risk assessment is needed to evaluate/compare the threat of iARGs and eARGs ............................. 131
v

References .............................................................................................................................................................. 132
6. Conclusions ......................................................................................................................................... 137
6.1 Overview ................................................................................................................................................... 137
6.2 eARGs may be as important as iARGs in dissemination of antibiotic resistance in the environment ...... 138
6.3 Selective pressure posed by influent antibiotic induced antibiotic resistance in both biomass and effluent
of AnMBRs ............................................................................................................................................................. 139
6.4 In the presence of antibiotics, HGT is the main reason for variation of ARG profiles .............................. 140
6.5 Membrane fouling distinctly impact ARG and community profiles of biomass and effluent of an AnMBR
141
6.6 Future Research ........................................................................................................................................ 142
References .............................................................................................................................................................. 144
Appendices .................................................................................................................................................. 147
Appendix A: Supporting information for chapter 2 ............................................................................................... 147
References ......................................................................................................................................................... 152
Appendix B: Supporting information for chapter 3 ............................................................................................... 154
B.1 Bench-scale anaerobic membrane bioreactor operation ......................................................................... 154
B.2 Antibiotic quantification ............................................................................................................................ 156
B.3 Quantification of ARGs using real time qPCR ............................................................................................ 158
References ......................................................................................................................................................... 166
Appendix C: Supporting information for chapter 4................................................................................................ 167
C.1 Quantification of antibiotics by LC-MS ...................................................................................................... 167
C.2 COD Mass Balance in the Reactor ............................................................................................................. 180
Reference .......................................................................................................................................................... 181
Appendix D: Supporting information for chapter 5 ............................................................................................... 182
D.1 Anaerobic membrane bioreactor operation ............................................................................................. 182
D.2 Membrane cleaning .................................................................................................................................. 183
D.3 Analysis methods ...................................................................................................................................... 184
D.4 Antibiotic quantification by LC-MS ........................................................................................................... 184
D.3 DNA extraction efficiency ......................................................................................................................... 185
D.4 Fouling reduced antibiotic concentrations in the effluent of the AnMBR ............................................... 188
References ......................................................................................................................................................... 190

vi

List of Tables
Table 2.1 Conjugation frequency of iARGs. ................................................................................................................. 31
Table 2.2 Transduction frequency of iARGs. ............................................................................................................... 35
Table 2.3 Natural transformation frequency of eARGs. .............................................................................................. 37

Table 5.1 Log removal of different ARGs achieved by membranes at different levels of fouling. LF, MF and HF are low
fouled, medium fouled and highly fouled membranes. ............................................................................................ 131

Table S1.1 Absolute abundance of iARGs and eARGs in different environments. .................................................... 147
Table S1.2 Log removal of iARGs and eARGs in secondary clarifier effluent achieved by different disinfection
methods. .................................................................................................................................................................... 150

Table S2.1 Synthetic wastewater composition.......................................................................................................... 155
Table S2.2 Targeted antibiotic properties and MS data acquisition parameters...................................................... 157
Table S2.3 Forward and reverse primers and qPCR thermocycling conditions of all ARGs, intI1 and rpoB gene. ... 159
Table S2.4 The abundance of targeted ARGs (Copy/rpoB) in biomass and biofilm during the addition of increment
concentrations of sulfamethoxazole, erythromycin and ampicillin. Abundance and errors respectively represent the
mean values and standard deviations calculated according to the triplicate qPCR results. ..................................... 160
Table S2.5 The abundance of targeted ARGs (Copy/mL) in effluent during the addition of increment concentrations
of sulfamethoxazole, erythromycin and ampicillin. Abundance and errors respectively represent the mean values
and standard deviations calculated according to the triplicate qPCR results ........................................................... 161

Table S3.1 Synthetic wastewater composition.......................................................................................................... 169
Table S3.2 Forward and reverse primers and qPCR thermocycling conditions of all ARGs, intI1, and rpoB gene. .. 170
Table S3.3 Biomass ARG abundances (Copy/rpoB) in the pre-antibiotics, antibiotics loading, and post-antibiotics
periods. Abundance and errors respectively represent the mean values and standard deviations calculated from
triplicate qPCR results. ............................................................................................................................................... 170
Table S3.4 Effluent ARG abundances (Copy/mL) in the pre-antibiotics, antibiotics loading, and post-antibiotics
periods. Abundance and errors respectively represent the mean values and standard deviations calculated from
triplicate qPCR results. ............................................................................................................................................... 171
Table S3.5 Correlation analysis results between the first 100 most abundant OTUs and ARGs in biomass and effluent
of the AnMBR. Correlated ARGs had strong significant correlation (p < 0.05; and ρ > 0.7 or ρ < −0.7) with the assigned
OTUs. Green rows show correlation in biomass samples and the yellow rows show correlation in effluent samples.
.................................................................................................................................................................................... 172

Table S4.1 Synthetic wastewater composition.......................................................................................................... 182
Table S4.2 Targeted antibiotic properties and MS data acquisition parameters...................................................... 185
Table S4.3 Forward and reverse primers and qPCR thermocycling conditions of all ARGs, intI1, and rpoB gene. .. 187

vii

List of Figures
Figure 2.1 Contribution of intracellular and extracellular ARGs (iARGs and eARGs) to ARG profile in different
environments. .............................................................................................................................................................. 24
Figure 2.2 Fate of intracellular and extracellular ARGs (iARGs and eARGs) in receiving aquatic environment. ........ 26
Figure 2.3 Horizontal transfer of ARGs through conjugation, transduction, and natural transformation. Inserted
tables show frequency of each horizontal gene transfer mechanism in different environments. ............................. 29
Figure 2.4 Log removal of intracellular and extracellular ARGs (iARGs and eARGs) class at (a) and (b) different chlorine
exposure, (c) and (d) different UV fluences (AS is abbreviation for amplicon size). ................................................... 44

Figure 3.1 (A) Performance of AnMBR in COD removal and biogas production during the addition of
sulfamethoxazole, erythromycin and ampicillin at increasing concentrations; (B) Fate of sulfamethoxazole,
erythromycin and ampicillin in AnMBR. Error bars represent the standard deviation of the results obtained from
replicate samples. ........................................................................................................................................................ 65
Figure 3.2 Abundance of targeted ARGs in the biomass, biofilm (Copy/rpoB) and effluent (Copy/mL) during the
addition of 250 (µg/L) sulfamethoxazole. ARGs are sorted by the biomass abundance (highest to the lowest). Bar
charts and error bars respectively represent the mean values and standard deviations calculated according to the
results from three samples collected on at different days during the addition of 250 (µg/L) sulfamethoxazole and
each of their triplicate qPCR results. For biofilm results, mean values and standard deviations were calculated
according to the triplicate qPCR results. ...................................................................................................................... 66
Figure 3.3 (A) Abundance of targeted ARGs (Copy/rpoB) in the biomass (inserted diagram shows abundance of
targeted ARGs (Copy/rpoB) in the biomass excluding ermF gene); (B) Abundance of targeted ARGs (Copy/mL) in the
effluent, during the addition of increment concentrations of sulfamethoxazole, erythromycin and ampicillin. Bars
represent the mean values of three temporal sampling points (except for 10 µg/L erythromycin and 50 µg/L
ampicillin which represent two samples, and 0 µg/L ampicillin which represents one sample) collected during the
addition of each increment concentration of antibiotics and each of their triplicate qPCR results. .......................... 70
Figure 3.4 Abundance of antibiotic corresponding genes in biomass (Copy/rpoB) and effluent (Copy/mL) when (A)
and (D) sulfamethoxazole; (B) and (E) erythromycin; (C) and (F) ampicillin was added to AnMBR. Inserted diagrams
were used to magnify the genes with low ab abundance. Bars and error bars represent the mean values and standard
deviations, respectively, of three temporal sampling points (except for 10 µg/L erythromycin and 50 µg/L ampicillin
which represent two samples, and 0 µg/L ampicillin which represents one sample) collected during the addition of
each increment concentration of antibiotics and each of their triplicate qPCR results.............................................. 73
Figure 3.5 Network analysis representing the correlations between antibiotics concentration (pink circles), ARGs
(teal circles), class 1 integrons (blue circles) and ARB (yellow circles) (A) in biomass; (B) in effluent. SMX, ERY, AMP
and TET stand for sulfamethoxazole, erythromycin, ampicillin and tetracycline; respectively. RB also stands for
resistant bacteria. A solid connection shows strong, significant correlations (ρ > 0.7 or ρ < -0.7; and p<0.05), and a
dashed line represents weak correlations (0.3 < ρ < 0.7 or -0.7 < ρ < -0.3; and p<0.05). ........................................... 78

Figure 4.1 Abundance of targeted genes in the (a) biomass (copies/rpoB) and (b) effluent (copies/mL) of the AnMBR
throughout the experimental period. The x-axis represents the days after steady performance of the AnMBR was
reached. Day 1 represents pre-antibiotics period, days 6, 14, 20, 27 and 35 represent antibiotics loading period (area
bordered by red dashed line), and day 46 represents post-antibiotics period. The markers in (a) represent abundance
of rpoB (copies/mL; secondary y-axis). ........................................................................................................................ 93
Figure 4.2 (a) Non-metric multidimensional scaling (NMDS) and (b) Inverse Simpson index for the biomass and
effluent of the AnMBR throughout the experimental period. The red arrow in the NMDS plot indicates the significant
shift of the effluent samples after antibiotics addition. In the Inverse Simpson plot, the bars for antibiotic loading
represent the average of the diversity index in n = 5 samples during the loading period (n = 1 for the pre- and post-
antibiotics period). Error bars for the pre- and post-antibiotics period represent the standard deviation calculated by
Mothur for each sample, and for the antibiotics loading period represents the standard deviations of the averages
in n = 5. ......................................................................................................................................................................... 98
viii

Figure 4.3 Relative abundance of the (a) biomass and (b) effluent microbial community at the family level throughout
the experimental period. Day 1 represents pre-antibiotics period, days 6, 14, 20, 27 and 35 represent antibiotics
loading period (area bordered by red dashed line), and day 46 represents post-antibiotics period. ...................... 100
Figure 4 4 Network analysis representing the positive correlations between ARGs (purple circles) and microbial
structure (OTUs) with ≥ 0.5% relative abundance in at least one sample in the (a) biomass and (b) effluent of the
AnMBR. A connection shows strong significant positive correlation (p < 0.05; and ρ > 0.7). The bubble size is indicative
of relative abundance. ............................................................................................................................................... 106
Figure 5.1 Performance of the AnMBR for COD removal and biogas/methane production. Phase 1 and 2
of operation were used to achieve different levels of fouling on the membrane modules (Phase 1 also
represents a control experiment where each membrane had the same level of fouling); Phase 3 (main
experimental period) is the period in which each membrane was operated under different fouling levels.
LF, MF, and HF represent low, medium, and high fouled membranes in Phase 3, respectively. ............. 121
Figure 5.2 (a) Variation of the transmembrane pressure (TMP), (b) concentration of EPS, and (c) concentration of
SMP content in suspended biomass (during Phase 3; n=5), MF, and HF membrane biofilms (at the end of Phase 3).
Phase 1 and 2 of operation were used to achieve different levels of fouling on the membrane modules (Phase 1 also
represents a control experiment where each membrane had the same level of fouling); Phase 3 (main experimental
period) is the period in which each membrane was operated under different fouling levels. LF, MF, and HF represent
low, medium, and high fouled membranes in Phase 3, respectively. TMP data is the daily average calculated from
recorded data (every minute) via LabVIEW. .............................................................................................................. 122
Figure 5.3 Abundance of targeted ARGs and intI1 gene in (a) and (c) suspended biomass during Phase 3, and (b) and
(d) MF and HF membrane biofilm (at the end of Phase 3). (a) and (b) are absolute abundance, (c) and (d) are relative
abundance (normalized to rpoB). MF and HF represents medium and highly fouled membranes, respectively. HF
outer and inner layers respectively represent the HF membrane biofilm in the vicinity of biomass and membrane
surface. The averaged absolute ARG abundances in the outer and inner layers of the HF membrane biofilm were
equal to the absolute ARG abundances of the whole cake layer. ............................................................................. 124
Figure 5.4 Absolute abundance of intracellular and extracellular (a) sul1, (b) sul2, (c) ermF, (d) ermB, (e) ampC, and
(f) oxa-1, in the effluent of LF, MF and HF membranes during Phase 3. LF, MF and HF represent low, medium and
highly fouled membranes in Phase 3, respectively. .................................................................................................. 128
Figure 5.5 Possible mechanism of passage of less rigid cells through the mature biofilm due to high transmembrane
pressure (TMP), and adsorption of eARGs by EPS and SMP. ..................................................................................... 129

Figure S2.1 Abundance of targeted ARGs in the biomass, biofilm (Copy/rpoB) and effluent (Copy/mL) during the
addition of 250 (µg/L) (A) erythromycin; (B) ampicillin. ARGs are sorted by the biomass abundance (highest to the
lowest). Bar charts and error bars respectively represent the mean values and standard deviations calculated
according to the results from three samples collected on at different days during the addition of 250 (µg/L) of
antibiotics and each of their triplicate qPCR results. For biofilm results, mean values and standard deviations were
calculated according to the triplicate qPCR results. .................................................................................................. 164
Figure S2.2 (A) Absolute abundance of total and antibiotic resistant bacteria; (B) antibiotic resistant bacteria
normalized to total bacteria; in the effluent of AnMBR during the addition of sulfamethoxazole, erythromycin and
ampicillin at increment concentrations. Error bars represent the standard deviation of the results obtained from
replicate samples. SMX, ERY, AMP and TET stand for sulfamethoxazole, erythromycin, ampicillin and tetracycline;
respectively. RB also stands for resistant bacteria. ................................................................................................... 165

Figure S3.1 Rarefaction curves for (a) biomass and (b) effluent microbial community samples of the AnMBR. ..... 175
Figure S3.2 Performance of the AnMBR in (a) COD removal and biogas production, and (b) antibiotic removal
efficiency. ................................................................................................................................................................... 176
Figure S3.3 PCA of ARG profiles in the biomass and effluent of the AnMBR ............................................................ 177
Figure S3.4 Relative abundance of (a) methanogens and (b) syntrophic bacteria in the biomass of the AnMBR
throughout the experimental period. Day 1 represents pre-antibiotics period, days 14, 20, 27 and 35 represent
antibiotics loading period (area bordered by red dashed line) and day 46 represents post-antibiotics period. ..... 178
ix

Figure S3.5 Relative abundance of the (a) Biomass and (b) effluent Microbial Community in genus Level throughout
the experimental period. Day 1 represents pre-antibiotics period, days 14, 20, 27 and 35 represent antibiotics loading
period and day 46 represents post-antibiotics. ......................................................................................................... 179
Figure S3.6 COD mass balance in the reactor. ........................................................................................................... 180
Figure S4.1 Schematic diagram of the bench-scale anaerobic membrane bioreactor. ............................ 183
Figure S4.2 Effluent concentration of (a) sulfamethoxazole (SMX), (b) erythromycin (ERY), and (c) ampicillin (AMP)
during Phase 3. LF, MF and HF represent low, medium and highly fouled membranes, respectively. .................... 189
Figure S4.3 Absolute abundance of intracellular and extracellular (a) tetO, (b) tetW, (c) intI1, and (d) rpoB, in the
effluent of LF, MF and HF membranes during Phase 3. LF, MF and HF represent low, medium and highly fouled
membranes in Phase 3, respectively. ........................................................................................................................ 189

x

Abstract

Antibiotic resistance infection is one of the biggest threats to human health, which is currently
responsible for over 700,000 death annually worldwide. The world health organization (WHO) has called
for urgent action to avert an antimicrobial crisis, and the U.S. Centers for Disease Control and Prevention
(CDC) has identified addressing antibiotic resistance as a national priority. The CDC has proposed five core
actions to better prepare the United States for an antibiotic resistance pandemic, one of which is keeping
antibiotic resistance from entering the environment. Wastewater treatment plants (WWTPs) as the main
interfaces between the build and natural environment, has been identified as a primary source of
antibiotic resistance spread into the environment. Conventional wastewater treatment technologies have
not been designed to mitigate release of antibiotic resistance, and more advanced technologies are
required. Anaerobic membrane bioreactors (AnMBRs) are an emerging biotechnology that can provide
similar treatment performance to aerobic processes, while promoting energy and nutrient recovery.
AnMBRs have unique features that can also impact release of antibiotic resistance, however, they
remained unexplored in this regard. Here, first we evaluated the role of influent antibiotics on antibiotic
resistance gene (ARG) profile of the biomass and effluent of a bench-scale AnMBRs. A gradual increase in
biomass ARG profile, and an initial increase followed by gradual decrease in effluent ARG profile were
observed. ARGs can be transferred vertically and horizontally, however, it was unclear which transfer
mechanisms was responsible for the variation of the biomass and effluent ARG profiles. Next, we designed
another experiment, introducing three antibiotics to the influent of the AnMBR, and evaluated microbial
community structures, ARG profiles, and their potential association. The gradual increase in biomass ARG
profile in the presence of antibiotics, while biomass microbial community structure was not impacted by
antibiotics addition, indicated the greater influence of horizontal gene transfer (HGT) compared to vertical
gene transfer (VGT) in variation of ARG profile in the AnMBR. Effluent microbial community structure,
however, shifted significantly upon initial exposure to antibiotics, probably due to its lower diversity
xi

(richness and evenness) compared to the biomass community. Last, we investigated role of membrane
fouling layer on release of intracellular and extracellular ARGs (iARGs and eARGs) from an AnMBR. Results
reveled that, compared to the biomass, fouling layer provided more conducive condition for HGT. Fouling
layer also increased abundance of iARGs in the effluent, while it reduced eARG abundances. As an
emerging biotechnology, AnMBRs are expected to be the future of waste management technologies, and
this dissertation provides worthwhile information on operational strategies and design for full-scale
AnMBR systems.

1

Chapter 1

1. Anaerobic treatment for mitigation of antibiotic resistance spread
to the environment

1.1 Introduction
Wastewater, when managed with appropriate technologies, is a resource of water, energy, and
nutrients.
1
However, conventional aerobic treatment processes remain the core bioprocess in the vast
majority of treatment plants despite their inability to directly recover energy and nutrients from
wastewater.
2
Transitioning away from inefficient aerobic processes affords the opportunity to implement
full anaerobic bioprocesses to achieve greater resource recovery (i.e., production of methane-rich biogas
and nutrient-rich biosolids) during wastewater management. Anaerobic digestion and other treatment
processes have long been applied for management of wastewater sludges (i.e., primary sludge and waste
activated sludge)
3-5
and for high-strength wastes (e.g., food processing wastewater)
6-8
and has typically
been operated at elevated temperatures above 25  C.
8, 9
New technological advances, including
incorporating membrane separation into anaerobic processes, has revealed that anaerobic processes can
be successfully applied to a wide range of wastewater strengths at a wider range of temperatures
(psychrophilic, mesophilic, and thermophilic conditions).
10, 11
McCarty et al.
12
investigated anaerobic
treatment of domestic wastewater as a potential net generator of energy and reported that the produced
2

energy in mainstream anaerobic wastewater treatment and sludge digestion can exceed the required
energy for plant operation. Despite the considerable advantages that anaerobic treatment offers relative
to conventional aerobic treatment, research is urgently needed addressing the fate of emerging
contaminants, such as antibiotic resistance, during mainstream anaerobic treatment.
Antibiotic resistance is one the most formidable threats to human health and is responsible for over
700,000 death worldwide each year.
13
Recent studies revealed a wide and rapid spread of antibiotic
resistance,
14
that if it continues, would result in antibiotic resistant infections becoming the leading cause
of death by 2050, responsible for over 10 million deaths each year and $100 trillion in costs worldwide.
13

It is important to note that resistance to antibiotics is a naturally occurring phenomenon occurring via
natural selection.
15
Thus, antibiotic resistance is found in essentially all natural environments, including
so called “pristine” environments.
16
However, anthropogenic activities are driving a marked increase in
antibiotic resistance in engineered, clinical, and environmental settings.
13
Antibiotic resistance can be
transferred among and across bacterial species through vertical gene transfer (VGT) and horizontal gene
transfer (HGT).
17
HGT can result in the acquisition of antibiotic resistance by a pathogenic recipient cell
from a benign donor, and is particularly a concern at wastewater treatment plants (WWTPs) given the
high density of microorganisms in biological unit processes.
18, 19
HGT occurs via mobile genetic elements
(MGEs; e.g., plasmids and transposons) that contain one or more antibiotic resistance gene (ARG).
18, 20

One of the major contributing factors to the development of antibiotic resistance is antibiotic selective
pressure caused by antibiotics use, however rational and appropriate.
13
Antibiotics are commonly
employed to manage infections in humans and animals and have been used widely as an animal growth
promoter regardless of a bacterial infection being positively identified.
17, 21, 22
A large portion of consumed
antibiotics in both humans and animals remain unmetabolized and are subsequently excreted and
contaminate domestic wastewater and livestock manure.
23
The high density of microorganisms in
WWTPs, along with the sublethal concentration of antibiotics, provide a favorable environment for
3

promotion of resistance among bacteria.
21, 24
For this reason, WWTPs have been identified as hotspots for
antibiotic resistance spread to the environment.
25
Therefore, it is crucial to investigate the fate of
antibiotic resistant bacteria (ARB) and ARGs in WWTPs and devise operational strategies and/or design
elements that mitigate their dissemination.
1.2 Antibiotic resistance in conventional and advanced anaerobic
wastewater treatment processes
Employing anaerobic processes for mainstream wastewater treatment can help with mitigation of
antibiotic resistance spread to the environment.
26, 27
Due to the lower microbial yield, anaerobic
treatment processes produce less excess sludge (up to 90% less) as compared to conventional aerobic
systems.
28
A broader range and significantly higher concentration of ARB and ARGs are present in the
biomass of WWTPs relative to the effluent.
29, 30
Therefore, the lower sludge production of anaerobic
treatment systems can significantly reduce the release of antibiotic resistance to the environment.
Further, lower growth rate and yield of anaerobic bacteria implies lower biological activity, potentially
resulting in less spread of ARGs within the treatment process.
31
It is important to note that the
environmental and public health risks associated with antibiotic resistance dissemination are also
impacted by exposure pathway (e.g., biosolids land application versus effluent reuse practices) and thus
both effluent and biosolids ARGs are important to evaluate.
One possible drawback in anaerobic treatment processes in regard to antibiotic resistance proliferation
is the high sludge retention time (SRT) needed to achieve desirable carbon conversion.
32
Theoretically
high SRT can promote occurrence of HGT and VGT, resulting in accumulation of ARB and ARGs in biomass
and subsequently production of excess sludge enriched with antibiotic resistance.
32, 33
Neyestani et al.
34

reported that increasing SRT of a sequencing batch reactor (SBR) treating primary effluent of a municipal
4

WWTP raised the relative percentage of sulfamethoxazole (SMX) and trimethoprim resistant bacteria in
the biomass. Zhang et al.
33
observed that increasing SRT from 25 to 50 days, significantly increased both
relative abundance (normalized to 16S rRNA genes) and absolute abundance of ARGs (sul1, sul2, ermF,
ermB, tetG, tetX, mcr-1, czcA, and merA) and intI1 gene (an indicator of HGT)
35
in biomass of anaerobic,
anoxic, and aerobic reactors in a lab-scale A2O-MBR system treating synthetic domestic wastewater.
Further, microbial community structure of biomass samples at the same SRT clustered closely together,
and far from the samples at different SRT in a principal component analysis (PCA). Surprisingly, higher SRT
caused the absolute abundance of ARGs to decrease and the relative abundance of ARGs to increase in
the effluent of the A2O-MBR system. This implies that at higher SRT, fewer 16S rRNA genes (an indicator
of microbial cell counts) were released from the membrane filtration process which might be due to the
production of a more compact foulant layer on the membrane surface at high SRT. The same results were
observed by Sui et al.
36
in an aerobic sequencing batch membrane bioreactor treating swine wastewater.
It was also reported elsewhere that higher SRT can increase the removal efficiency of absolute abundances
of ARGs (sul1, sul2, tetC, tetG, and tetX).
37
However, zhang et al.
38
observed that shorter SRT (20 days
compared to 15 days) resulted in reduction of total ARG abundances during one-phase and two-phases
anaerobic digestion of wastewater sludge. It is noteworthy to mention that ARGs from different classes
did not similarly respond to increase or decrease in SRT.
Despite the potential drawback of High SRT, anaerobic processes; due to less excess sludge production
and lower biological activity; can be considered as one of the leading technologies for mitigation of
antibiotic resistance spread to the environment.
26, 30
However, there are not many studies to date
evaluating antibiotic resistance in anaerobic wastewater treatment processes. In this section, first we
comprehensively summarized existing studies in conventional anaerobic wastewater treatment with
respect to antibiotic resistance. Then, fate of antibiotic resistance in anaerobic membrane bioreactors
(AnMBRs) as an emerging biotechnology was discussed in detail.
5

1.2.1 Conventional anaerobic wastewater treatment processes
Aydin et al.
39-42
published several studies investigating the effect of influent antibiotics on ARG and
microbial community profiles of anaerobic SBRs treating synthetic pharmaceutical wastewater. Different
mixtures of three antibiotics (SMX, tetracycline (TET) and erythromycin (ERY)) at different non-lethal
concentrations were added to the influent of anaerobic SBRs to evaluate the response of resistance
profile. Results of these studies revealed that when concentration of antibiotics increased in the influent
of the anaerobic SBRs, antibiotic selective pressure caused the abundance of some ARGs (tetA, tetB, tetC,
tetE, tetM, tetS, tetQ, tetX, msrA, ermA, ermF, ereA, sul1, sul2, and sul3) to significantly increase in the
effluent.
40
One important outcome of these studies was that the effect of antibiotics mixture on the
resistance profile and microbial community was more pronounced than the cumulative sum of the effect
of individual antibiotics addition.
39, 40
Further, it was observed that addition of SMX, TET, and ERY at
concentrations higher than 1.5 mg/L significantly changed the structure of the biomass microbial
communities. These changes in microbial community structure were mostly due to the accumulation of
volatile fatty acids and soluble microbial products in the reactors resulting in the reduction of COD removal
efficiency by 37% and biogas production by 30%.
42
Variation of biomass microbial community of anaerobic
SBR, specifically Gram-negative bacteria like Acinetobacter and Bacteroidetes, has been also been found
to potentially increase the occurrence of resistance genes in anaerobic SBRs.
39, 41

Another common anaerobic treatment system is up-flow anaerobic sludge blanket reactors (UASBs).
The major advantage of UASB design is spatial separation of acidogenesis from methanogenesis phase
resulting in improvement of the reactor performance.
43
Chelliapan et al.
44
employed a UASB to treat
pharmaceutical wastewater containing tylosin and observed that tylosin was removed effectively (by 95%)
implying that this antibiotic is readily degradable in anaerobic conditions. Interestingly, high concentration
6

of tylosin in the influent of the UASB (20-200 g/L) did not significantly affect COD removal efficiency and
methane production.
Yi et al.
45
constructed a pilot-scale UASB equipped with an enhanced hydrolysis pretreatment system
in a pharmaceutical factory to compare its performance with a full-scale anaerobic reactor (expanded
granular sludge bed) both treating oxytetracycline production wastewater (influent oxytetracycline
concertation of 850 mg/L). Based on the results, the pilot-scale UASB had greater COD and antibiotic
removals compared to the full-scale anaerobic reactor. Interestingly, after a large portion of
oxytetracycline was removed by hydrolysis pretreatment (99.9%), proliferation of ARGs in the UASB
reactor decreased significantly compared to the full-scale anaerobic reactor. Thereby, absolute and
relative abundance (normalized to 16S rRNA gene) of class 1 integrons in the UASB reactor decreased
throughout the experiment. Further, among 13 quantified TET resistance genes, seven of them (tetC, tetG,
tetL, tetM, tetO, tetQ, and tetX) showed significantly positive correlation with intI1 gene abundance,
reducing throughout the experiment. Since class 1 integrons genes are indicative of HGT,
35
it is likely that
hydrolysis pretreatment reduces the transfer ability of these TET genes in UASBs.
In another study by Hou et al.,
46
using UASBs to treat real pharmaceutical wastewater (without any
pretreatment) containing antibiotics (at average concertation of 1000 ug/L for 18 antibiotics), the
absolute abundance of all quantified ARGs (sul1, sul2, tetO, tetQ, tetM, tetW, qnrD, ermB, oxa-1, and oxa-
10) increased significantly during anaerobic treatment. However, the relative abundances of
aforementioned ARGs remained approximately constant throughout the experiment. First, these results
highlight the impact of antibiotic selective pressure on ARG proliferation, and secondly indicates that
promotion of the resistance profile during anaerobic treatment of pharmaceutical wastewater containing
antibiotics might be due to the dissemination of ARB. Studies on conventional anaerobic processes for
mainstream wastewater treatment have indicated the possibility of decreasing dissemination of antibiotic
7

resistance. However, this reduction can be furthered by using advanced technologies like membrane
filtration.
1.2.2 AnMBR as an emerging technology to manage spread of antibiotic
resistance
AnMBR is an emerging biotechnology offering the same advantages as conventional anaerobic
treatment processes, such as energy and nutrients recovery, while achieving the same treatment
performance as aerobic treatment systems (e.g., activated sludge processes).
10, 11, 47
Given that AnMBRs
have the potential to be operated at higher SRT (through controlling biomass concertation via membrane
separation) than conventional anaerobic treatment processes,
47, 48
it is likely that AnMBRs produce less
excess sludge compared to conventional systems, which can significantly mitigate spread of antibiotic
resistance to the environment. Physical size exclusion via membrane separation is another major obstacle
for ARB and ARGs in AnMBRs.
49, 50
Further, ARB and ARGs can be adsorbed to/trapped by the membrane
fouling layer (biofilm) and retained in the reactor, resulting in less antibiotic resistance release to the
environment.
49, 51
For example, it has been observed that concentrations of ARGs (tetW and tetO) in the
effluent of WWTPs equipped by membrane bioreactors were 1-3 log lower than that detected in the
effluent of conventional WWTPs.
29
Slipko et al.
52
investigated the role of different membranes ranging
from microfiltration (MF) to reverse osmosis (RO) in extracellular linear DNA and plasmid ARG removal.
Surprisingly, membranes with neutral charge showed higher log removal of ARGs compared to negatively
charged membranes. Given that ARB and ARGs are negatively charged,
49
it is likely that membranes with
negative charge alleviate release of antibiotic resistance through repulsion of sludge particles, which is in
contrast with the observation of this study. More research is required to elucidate the mechanism of ARG
removal by charged membranes. Testing different membranes revealed that, although the ultrafiltration
(UF) membrane with molecular weight cut-off (MWCO) 300,000 Da was not able to remove more than
8

10% of ARGs and extracellular linear DNA molecules, the UF membrane with MWCO of 20,000 Da
significantly decreased these constituents in the effluent by 95%. Free linear DNA molecules were able to
permeate through membrane pores significantly smaller than their size (approximately one order of
magnitude) through elongation or deformation. Results also revealed that linear free DNA molecules had
greater elongation ability than circular plasmid containing ARGs.
52, 53
Since these tests were carried out as
short-term experiments, the role of membrane fouling was not taken into account in this study.
AnMBR technology are also considered as a promising method to treat wastewater containing
antibiotics.
54
Although biodegradation is the major antibiotic removal mechanism in AnMBRs, the role of
membrane filtration and the foulant layer should not to be neglected.
50, 55
Interestingly, a recent study
indicated that addition of antibiotics (SMX and TET) at concentrations of 100 and 1000 ug/L to the influent
of an aerobic MBR, made the foulant layer denser and more compact, resulting in reduction of the
membrane fouling cycle from 25 (no antibiotics) to 8 and 4 days, respectively.
51
However, this severe
fouling layer resulted in significant reduction of ARGs release from the aerobic MBR. Addition of antibiotic
to the influent of AnMBR was also reported to decrease sludge floc size, increase the production of
extracellular polymeric substances (EPSs) and soluble microbial products (SMPs), which are among the
major contributing factors of membrane fouling, and shift microbial community structure towards
bacterial groups such as phyla of Firmicutes, Proteobacteria, Chloroflexi and Bacteroidetes which make
most contribution to membrane fouling.
51, 54
Therefore, the trade-offs between increasing antibiotic
resistance removal efficiency and membrane fouling need further research.
Cheng et al.
49
externally connected three membrane housings (MF PVDF membranes with nominal pore
size of 0.3 um) to an anaerobic reactor (operated at 35  C) and harvested each of them at different
transmembrane pressure (TMP) of 20, 40, and 60 kPa. Then the membrane housings were disconnected
from the system and were employed in a filtration test to evaluate the removal efficiency of ARB and ARGs
at various degrees of fouling. Three pathogenic ARB (Escherichia coli PI-7, Klebsiella pneumoniae L7, and
9

E. coli UPEC-RIY-4) and their associated plasmid-borne ARGs ( b la N DM ‐ 1, b la CT X ‐ M ‐ 1 5, and O X A ‐ 4) were
prepared in separate mediums (bacteria medium and plasmid medium) and then filtered through the
membranes. Results revealed that the average particle size of ARB in this study were in the range of 1664
to 2209 nm; however, all of the plasmids were smaller than 565 nm. Both ARB and their associated ARGs
were negatively charged, with zeta potential of ARB at 9-15 mV and ARGs at 22-28 mV. The log removal
values of ARGs using membranes with foulant layers were approximately 1-fold higher than virgin
membranes. However, no significant changes in ARG removal were observed among membranes with
different degrees of fouling. Further, the abundance of ARGs (normalized to the surface area) in the
foulant layers was significantly higher than that attached to the virgin membranes, suggesting that
adsorption into the foulant layers plays a crucial role in detention of ARGs in the reactor. Given that higher
degrees of fouling on the surface of membranes results in reduction of effective pore size of the
membrane
56-58
, no significant differences in ARG removal among the membranes with different degrees
of fouling imply that size exclusion is not as important as adsorption in ARG removal. One possible
explanation for adsorption of ARGs to the foulant layer is high concentration of EPS and SMPs. It has been
reported elsewhere that EPS and SMP in foulant layers have the ability to bind with ARGs.
51
Due to the
high molecular weight and cross-linked structural properties of EPS and SMPs, they have a high binding
ability through ion bridging interaction, hydrophobic interaction, and polymer enlargement.
59
Strong
correlations were also found between the absolute abundance of ARGs and both EPS and SMP in the
membrane foulant layer of an MBR treating antibiotic containing wastewater, indicating the decisive role
of EPS and SMPs in ARB and ARG removal.
51
Concentration polarization might be another factor
preventing ARB and ARGs to escape AnMBRs.
60
Regarding ARB removal, surprisingly the highest log
removal value achieved by both the virgin membrane and the most fouled membrane (60 kPa). The
abundance of ARB attached to the fouled membranes was significantly higher compared to the virgin
membrane, while it was approximately the same in foulant layers of membranes with different degrees
10

of fouling. The cause of this phenomenon might relate to the fact that the size of particles in the foulant
layer decreased from the top to the bottom, making a funnel like structure.
57, 61
Therefore, given that
antibiotics can decrease the rigidity of cell walls,
62
the funnel like structure of the foulant layer may
provide conditions which facilitate passage of less rigid cells through the membrane pores. Higher removal
of ARB by the most fouled membrane might be due to the irremovable blockage of membrane pores.
Kappell et al.
63
used a fluidized bed anaerobic reactor operated at 20 C connected to an external
ceramic membrane (nominal pore size of 0.05 um) to treat real primary clarifier effluent and evaluate ARG
removal. Results showed that the absolute abundances of all quantified ARGs (ermB, tetO and sul1) and
intI1 gene decreased slightly ( 0.5 log removal) in the effluent. However, these abundances further
decreased by 3.5 log after filtration through the external membrane. Interestingly, when ARG results
were normalized against 16S rRNA genes, no significant differences were observed between the ARG
profile of the influent (primary clarifier effluent) and final effluent. It might imply that the sampling
method of this study failed to take extracellular ARGs into account. Analysis of microbial diversity revealed
a significant decrease in richness and diversity of the community after membrane filtration. Further, an
NMDS plot showed that influent samples clustered separately from the final effluent. Most importantly,
effluent microbial community structure selected toward some pathogenic bacteria, which highlights the
potential health risk of releasing AnMBR effluent to the environment or practicing reuse.
1.3 Comparison of anaerobic treatment with aerobic treatment in
regard to antibiotic resistance
Yi et al.
45
compared a full-scale anaerobic reactor to a full-scale aerobic activated sludge reactor, both
treating the same oxytetracycline production wastewater, and observed that the relative abundance of
intI1 and TET resistance genes were considerably lower in the anaerobic system (intI1: 1.61  10
-2
in
11

anaerobic and 1.71  10
0
in aerobic reactor; TET resistance genes: 6.95  10
-2
in anaerobic and 1.68 
10
0
in aerobic reactor) which indicates the advantage of anaerobic processes in reducing the potential of
ARG dissemination. It was also reported elsewhere that the amount of increase in ARG abundances due
to the antibiotics addition in the anaerobic reactor was lower than that in aerobic reactor.
64
Du et al.
27

employed an A2O-MBR system treating municipal wastewater and observed that absolute abundance of
ARGs (tetG, tetX, and sul1) and class 1 integrons genes first decreased gradually along the anaerobic and
anoxic reactors and then increased significantly in the aerobic reactor. Given that biological activity of
microorganism under anaerobic conditions is comparatively lower,
28
it is likely that proliferation of ARGs
in anaerobic and anoxic reactors was slower than that in the aerobic reactor.
31
The final effluent of the
membrane module revealed that the A2O-MBR system was able to remove ARGs from the influent
municipal wastewater by 0.67-4.73 log. Reductions of sul1, tetW, and intI1 were positively correlated to
reduction of 16S rRNA genes. Since abundance of 16S rRNA genes can approximate microbial cell counts
and biomass concentration, it seems that the major reason for total ARG reduction in the effluent of the
A2O-MBR system was due to the retention of biomass in the reactors by membrane separation.
27

Christgen et al.
26
compared anaerobic reactors (UASB and anaerobic hybrid reactor) with an aerobic
reactor and also with a combination of anaerobic (as a pretreatment step) and aerobic reactors to treat
real domestic wastewater and evaluate the fate of ARGs. Both systems employing anaerobic reactors had
approximately the same performance in terms of COD and ARG removal. Metagenomic analysis revealed
that compared to the influent, the relative abundance of total ARGs decreased significantly by 62%,
82%, and 85% in anaerobic, aerobic, and anaerobic-aerobic reactor effluents, respectively. Lower ARG
removal efficiency in the anaerobic reactors might be due to the higher level of total suspended solids
(indicated by 16S rRNA gene abundance) in the effluent of anaerobic processes compared to the aerobic
effluents. It has been reported elsewhere that anoxic reactors can increase the absolute abundance of
ARGs, and only after removing biosolids (reduction in 16S rRNA gene abundance) via secondary
12

clarification did ARG abundances decline.
65
In Christgen et al.,
26
the most abundant types of ARGs detected
in the influent domestic wastewater were respectively genes encoding tetracycline, sulfonamide, multi-
drug resistance, macrolide, chloramphenicol, β-lactam, bacitracin, aminoglycoside, and acriflavine. The
abundances of tetracycline, aminoglycoside, and β-lactam resistance genes were higher in the effluent of
anaerobic reactors compared to the aerobic and anaerobic-aerobic reactors implying the appropriate
conditions of anaerobic treatment processes for these genes to spread. None of the treatment systems
were able to reduce sulfonamide nor chloramphenicol ARGs. Further, the relative abundance of some
ARG types, most importantly multi-drug resistant genes, significantly increased after treatment in aerobic
and anaerobic-aerobic reactors. This significant increase indicates the higher potential of multidrug
resistance to be proliferated during aerobic treatment compared to anaerobic treatment. In regard to the
resistance mechanisms, influent domestic wastewater was dominated by efflux pump and target
modification ARGs. In the effluents of all treatment systems, relative percentage of efflux pump ARGs
increased, however, target modification ARGs decreased. Since resistance genes which code for efflux
pumps are often found on MGEs,
22, 66, 67
selecting toward efflux pump highlights the importance of HGT
during treatment processes. Moreover, results showed that anaerobic treatment selected for ARGs
coding for inactivation mechanism. Based on the literature, enzyme inactivation as a method to resist
against antibiotics is common among anaerobes such as B. fragilis, B. melaninogenicus, C. ramosum, and
Clostridium clostridiiforme.
68
These observations imply that treatment processes play an imperative role
in fate of antibiotic resistance.
1.4 Overview of dissertation
This dissertation specifically focuses on the fate of antibiotic resistance in AnMBRs treating low-strength
wastewater. We began with a literature review of conventional and advanced anaerobic treatment
processes for mitigation of antibiotic resistance dissemination to the environment (Chapter 1) and then
13

investigated prevalence, horizontal transfer, and mitigation strategies of intracellular and extracellular
ARGs (iARGs and eARGs; Chapter 2). Next, the impact of influent antibiotic concentration on the ARG
profiles of both the biomass and effluent of a bench-scale AnMBR was evaluated (Chapter 3).
69

Observations during this study motivated an in-depth investigation of the association between ARGs and
microbial community structure of an AnMBR operated at a high influent concentration of antibiotics
(Chapter 4). Molecular analyses including high-throughput sequencing of 16S rRNA genes and qPCR
targeting genes conferring resistance to sulfonamides (sul1 and sul2), macrolides (ermF and ermB), β-
lactams (oxa-1 and ampC), and tetracycline (tetW and tetO), as well as intl1 which encodes for class 1
integrons, were used to evaluate the microbial community structure and ARG profile. Results of the first
two experimental studies indicated that the membrane foulant layer may significantly influence the
release of antibiotic resistance from AnMBRs. Therefore, we operated an AnMBR equipped by three
independent membranes at different levels of fouling and evaluated both iARG and eARG profiles in the
permeates generated by each membrane (Chapter 5).
References
1. Zeeman, G.; Kujawa, K.; De Mes, T.; Hernandez, L.; De Graaff, M.; Abu-Ghunmi, L.; Mels, A.; Meulman,
B.; Temmink, H.; Buisman, C., Anaerobic treatment as a core technology for energy, nutrients and water
recovery from source-separated domestic waste (water). Water Sci. Technol. 2008, 57, (8), 1207-1212.
2. Chan, Y. J.; Chong, M. F.; Law, C. L.; Hassell, D., A review on anaerobic–aerobic treatment of industrial
and municipal wastewater. Chem. Eng. J. 2009, 155, (1-2), 1-18.
3. Kim, J.; Park, C.; Kim, T.-H.; Lee, M.; Kim, S.; Kim, S.-W.; Lee, J., Effects of various pretreatments for
enhanced anaerobic digestion with waste activated sludge. Journal of bioscience and bioengineering 2003,
95, (3), 271-275.
4. Zhang, T.; Yang, Y.; Pruden, A., Effect of temperature on removal of antibiotic resistance genes by
anaerobic digestion of activated sludge revealed by metagenomic approach. Applied microbiology and
biotechnology 2015, 99, (18), 7771-7779.
5. Ju, F.; Li, B.; Ma, L.; Wang, Y.; Huang, D.; Zhang, T., Antibiotic resistance genes and human bacterial
pathogens: co-occurrence, removal, and enrichment in municipal sewage sludge digesters. Water Res.
2016, 91, 1-10.
14

6. Zhang, C.; Su, H.; Baeyens, J.; Tan, T., Reviewing the anaerobic digestion of food waste for biogas
production. Renewable and Sustainable Energy Reviews 2014, 38, 383-392.
7. Amha, Y. M.; Sinha, P.; Lagman, J.; Gregori, M.; Smith, A. L., Elucidating microbial community adaptation
to anaerobic co-digestion of fats, oils, and grease and food waste. Water Res. 2017, 123, 277-289.
8. Yi, J.; Dong, B.; Jin, J.; Dai, X., Effect of increasing total solids contents on anaerobic digestion of food
waste under mesophilic conditions: performance and microbial characteristics analysis. PloS one 2014, 9,
(7), e102548.
9. da Silva Martins, A.; Ornelas Ferreira, B.; Ribeiro, N. C.; Martins, R.; Rabelo Leite, L.; Oliveira, G.;
Colturato, L. F.; Chernicharo, C. A.; de Araujo, J. C., Metagenomic analysis and performance of a mesophilic
anaerobic reactor treating food waste at various load rates. Environ. Technol. 2017, 38, (17), 2153-2163.
10. Smith, A.; Skerlos, S.; Raskin, L., Anaerobic membrane bioreactor treatment of domestic wastewater
at psychrophilic temperatures ranging from 15 C to 3 C. Environmental Science: Water Research &
Technology 2015, 1, (1), 56-64.
11. Smith, A. L.; Skerlos, S. J.; Raskin, L., Psychrophilic anaerobic membrane bioreactor treatment of
domestic wastewater. Water Res. 2013, 47, (4), 1655-1665.
12. McCarty, P. L.; Bae, J.; Kim, J., Domestic wastewater treatment as a net energy producer–can this be
achieved? In ACS Publications: 2011.
13. Molnar, A., Antimicrobial resistance awareness and games. Trends in microbiology 2019, 27, (1), 1-3.
14. Theuretzbacher, U., Global antibacterial resistance: The never-ending story. Journal of global
antimicrobial resistance 2013, 1, (2), 63-69.
15. Martínez, J. L., Antibiotics and antibiotic resistance genes in natural environments. Science 2008, 321,
(5887), 365-367.
16. Martinez, J. L., Environmental pollution by antibiotics and by antibiotic resistance determinants.
Environ. Pollut. 2009, 157, (11), 2893-2902.
17. Normark, B. H.; Normark, S., Evolution and spread of antibiotic resistance. Journal of internal medicine
2002, 252, (2), 91-106.
18. Miller, J. H.; Novak, J. T.; Knocke, W. R.; Pruden, A., Survival of antibiotic resistant bacteria and
horizontal gene transfer control antibiotic resistance gene content in anaerobic digesters. Frontiers in
microbiology 2016, 7, 263.
19. von Wintersdorff, C. J.; Penders, J.; van Niekerk, J. M.; Mills, N. D.; Majumder, S.; van Alphen, L. B.;
Savelkoul, P. H.; Wolffs, P. F., Dissemination of antimicrobial resistance in microbial ecosystems through
horizontal gene transfer. Frontiers in microbiology 2016, 7, 173.
20. Huang, H.; Chen, Y.; Zheng, X.; Su, Y.; Wan, R.; Yang, S., Distribution of tetracycline resistance genes in
anaerobic treatment of waste sludge: The role of pH in regulating tetracycline resistant bacteria and
horizontal gene transfer. Bioresource technology 2016, 218, 1284-1289.
15

21. Karthikeyan, K.; Meyer, M. T., Occurrence of antibiotics in wastewater treatment facilities in
Wisconsin, USA. Sci. Total Environ. 2006, 361, (1-3), 196-207.
22. Baquero, F.; Martínez, J.-L.; Cantón, R., Antibiotics and antibiotic resistance in water environments.
Curr. Opin. Biotechnol. 2008, 19, (3), 260-265.
23. Pruden, A.; Pei, R.; Storteboom, H.; Carlson, K. H., Antibiotic resistance genes as emerging
contaminants: studies in northern Colorado. Environ. Sci. Technol. 2006, 40, (23), 7445-7450.
24. Rizzo, L.; Manaia, C.; Merlin, C.; Schwartz, T.; Dagot, C.; Ploy, M.; Michael, I.; Fatta-Kassinos, D., Urban
wastewater treatment plants as hotspots for antibiotic resistant bacteria and genes spread into the
environment: a review. Sci. Total Environ. 2013, 447, 345-360.
25. Guo, J.; Li, J.; Chen, H.; Bond, P. L.; Yuan, Z., Metagenomic analysis reveals wastewater treatment
plants as hotspots of antibiotic resistance genes and mobile genetic elements. Water Res. 2017, 123, 468-
478.
26. Christgen, B.; Yang, Y.; Ahammad, S.; Li, B.; Rodriquez, D. C.; Zhang, T.; Graham, D. W., Metagenomics
shows that low-energy anaerobic− aerobic treatment reactors reduce antibiotic resistance gene levels
from domestic wastewater. Environ. Sci. Technol. 2015, 49, (4), 2577-2584.
27. Du, J.; Geng, J.; Ren, H.; Ding, L.; Xu, K.; Zhang, Y., Variation of antibiotic resistance genes in municipal
wastewater treatment plant with A 2 O-MBR system. Environmental Science and Pollution Research 2015,
22, (5), 3715-3726.
28. Low, E. W.; Chase, H. A., Reducing production of excess biomass during wastewater treatment. Water
Res. 1999, 33, (5), 1119-1132.
29. Munir, M.; Wong, K.; Xagoraraki, I., Release of antibiotic resistant bacteria and genes in the effluent
and biosolids of five wastewater utilities in Michigan. Water Res. 2011, 45, (2), 681-693.
30. Zarei-Baygi, A.; Harb, M.; Wang, P.; Stadler, L. B.; Smith, A. L., Evaluating Antibiotic Resistance Gene
Correlations with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors. Environ. Sci.
Technol. 2019, 53, (7), 3599-3609.
31. Pei, R.; Cha, J.; Carlson, K. H.; Pruden, A., Response of antibiotic resistance genes (ARG) to biological
treatment in dairy lagoon water. Environ. Sci. Technol. 2007, 41, (14), 5108-5113.
32. Xia, S.; Jia, R.; Feng, F.; Xie, K.; Li, H.; Jing, D.; Xu, X., Effect of solids retention time on antibiotics
removal performance and microbial communities in an A/O-MBR process. Bioresource technology 2012,
106, 36-43.
33. Zhang, J.; Yang, M.; Zhong, H.; Liu, M.; Sui, Q.; Zheng, L.; Tong, J.; Wei, Y., Deciphering the factors
influencing the discrepant fate of antibiotic resistance genes in sludge and water phases during municipal
wastewater treatment. Bioresource technology 2018, 265, 310-319.
34. Neyestani, M.; Dickenson, E.; McLain, J.; Robleto, E.; Rock, C.; Gerrity, D., Impacts of solids retention
time on trace organic compound attenuation and bacterial resistance to trimethoprim and
sulfamethoxazole. Chemosphere 2017, 182, 149-158.
16

35. Fluit, A.; Schmitz, F., Class 1 integrons, gene cassettes, mobility, and epidemiology. European Journal
of Clinical Microbiology and Infectious Diseases 1999, 18, (11), 761-770.
36. Sui, Q.; Meng, X.; Wang, R.; Zhang, J.; Yu, D.; Chen, M.; Wang, Y.; Wei, Y., Effects of endogenous
inhibitors on the evolution of antibiotic resistance genes during high solid anaerobic digestion of swine
manure. Bioresource technology 2018, 270, 328-336.
37. Ma, Y.; Wilson, C. A.; Novak, J. T.; Riffat, R.; Aynur, S.; Murthy, S.; Pruden, A., Effect of various sludge
digestion conditions on sulfonamide, macrolide, and tetracycline resistance genes and class I integrons.
Environ. Sci. Technol. 2011, 45, (18), 7855-7861.
38. Zhang, J.; Liu, J.; Lu, T.; Shen, P.; Zhong, H.; Tong, J.; Wei, Y., Fate of antibiotic resistance genes during
anaerobic digestion of sewage sludge: Role of solids retention times in different configurations.
Bioresource technology 2019, 274, 488-495.
39. Aydin, S.; Ince, B.; Ince, O., Assessment of anaerobic bacterial diversity and its effects on anaerobic
system stability and the occurrence of antibiotic resistance genes. Bioresource technology 2016, 207, 332-
338.
40. Aydin, S.; Ince, B.; Ince, O., Development of antibiotic resistance genes in microbial communities
during long-term operation of anaerobic reactors in the treatment of pharmaceutical wastewater. Water
Res. 2015, 83, 337-344.
41. Aydin, S.; Ince, B.; Ince, O., Application of real-time PCR to determination of combined effect of
antibiotics on Bacteria, Methanogenic Archaea, Archaea in anaerobic sequencing batch reactors. Water
Res. 2015, 76, 88-98.
42. Aydin, S.; Cetecioglu, Z.; Arikan, O.; Ince, B.; Ozbayram, E. G.; Ince, O., Inhibitory effects of antibiotic
combinations on syntrophic bacteria, homoacetogens and methanogens. Chemosphere 2015, 120, 515-
520.
43. Seghezzo, L.; Zeeman, G.; van Lier, J. B.; Hamelers, H.; Lettinga, G., A review: the anaerobic treatment
of sewage in UASB and EGSB reactors. Bioresource technology 1998, 65, (3), 175-190.
44. Chelliapan, S.; Wilby, T.; Sallis, P. J., Performance of an up-flow anaerobic stage reactor (UASR) in the
treatment of pharmaceutical wastewater containing macrolide antibiotics. Water Res. 2006, 40, (3), 507-
516.
45. Yi, Q.; Zhang, Y.; Gao, Y.; Tian, Z.; Yang, M., Anaerobic treatment of antibiotic production wastewater
pretreated with enhanced hydrolysis: Simultaneous reduction of COD and ARGs. Water Res. 2017, 110,
211-217.
46. Hou, J.; Chen, Z.; Gao, J.; Xie, Y.; Li, L.; Qin, S.; Wang, Q.; Mao, D.; Luo, Y., Simultaneous removal of
antibiotics and antibiotic resistance genes from pharmaceutical wastewater using the combinations of up-
flow anaerobic sludge bed, anoxic-oxic tank, and advanced oxidation technologies. Water Res. 2019, 159,
511-520.
17

47. Smith, A. L.; Stadler, L. B.; Love, N. G.; Skerlos, S. J.; Raskin, L., Perspectives on anaerobic membrane
bioreactor treatment of domestic wastewater: a critical review. Bioresource technology 2012, 122, 149-
159.
48. Lew, B.; Tarre, S.; Beliavski, M.; Dosoretz, C.; Green, M., Anaerobic membrane bioreactor (AnMBR) for
domestic wastewater treatment. Desalination 2009, 243, (1-3), 251-257.
49. Cheng, H.; Hong, P.-Y., Removal of antibiotic-resistant bacteria and antibiotic resistance genes affected
by varying degrees of fouling on anaerobic microfiltration membranes. Environ. Sci. Technol. 2017, 51,
(21), 12200-12209.
50. Harb, M.; Hong, P.-Y., Anaerobic membrane bioreactor effluent reuse: A review of microbial safety
concerns. Fermentation 2017, 3, (3), 39.
51. Zhu, Y.; Wang, Y.; Zhou, S.; Jiang, X.; Ma, X.; Liu, C., Robust performance of a membrane bioreactor for
removing antibiotic resistance genes exposed to antibiotics: role of membrane foulants. Water Res. 2018,
130, 139-150.
52. Slipko, K.; Reif, D.; Wögerbauer, M.; Hufnagl, P.; Krampe, J.; Kreuzinger, N., Removal of extracellular
free DNA and antibiotic resistance genes from water and wastewater by membranes ranging from
microfiltration to reverse osmosis. Water Res. 2019, 164, 114916.
53. Latulippe, D. R.; Zydney, A. L., Separation of plasmid DNA isoforms by highly converging flow through
small membrane pores. J. Colloid Interface Sci. 2011, 357, (2), 548-553.
54. Cheng, D.; Ngo, H. H.; Guo, W.; Liu, Y.; Chang, S. W.; Nguyen, D. D.; Nghiem, L. D.; Zhou, J.; Ni, B.,
Anaerobic membrane bioreactors for antibiotic wastewater treatment: performance and membrane
fouling issues. Bioresource technology 2018, 267, 714-724.
55. Harb, M.; Wei, C.-H.; Wang, N.; Amy, G.; Hong, P.-Y., Organic micropollutants in aerobic and anaerobic
membrane bioreactors: Changes in microbial communities and gene expression. Bioresource technology
2016, 218, 882-891.
56. Zarei-Baygi, A.; Moslemi, M.; Mirzaei, S. H., The combination of KMnO4 oxidation and polymeric
flocculation for the mitigation of membrane fouling in a membrane bioreactor. Sep. Purif. Technol. 2016,
159, 124-134.
57. Drews, A., Membrane fouling in membrane bioreactors—characterisation, contradictions, cause and
cures. Journal of membrane science 2010, 363, (1-2), 1-28.
58. Lin, H.; Peng, W.; Zhang, M.; Chen, J.; Hong, H.; Zhang, Y., A review on anaerobic membrane
bioreactors: applications, membrane fouling and future perspectives. Desalination 2013, 314, 169-188.
59. Drews, A.; Lee, C.-H.; Kraume, M., Membrane fouling-a review on the role of EPS. Desalination 2006,
200, (1-3), 186-188.
60. Baker, R.; Wijmans, J.; Athayde, A.; Daniels, R.; Ly, J.; Le, M., The effect of concentration polarization
on the separation of volatile organic compounds from water by pervaporation. Journal of membrane
science 1997, 137, (1-2), 159-172.
18

61. Xiong, J.; Fu, D.; Singh, R. P.; Ducoste, J. J., Structural characteristics and development of the cake layer
in a dynamic membrane bioreactor. Sep. Purif. Technol. 2016, 167, 88-96.
62. Formosa, C.; Grare, M.; Duval, R. E.; Dague, E., Nanoscale effects of antibiotics on P. aeruginosa.
Nanomedicine: Nanotechnology, Biology and Medicine 2012, 8, (1), 12-16.
63. Kappell, A. D.; Kimbell, L. K.; Seib, M. D.; Carey, D. E.; Choi, M. J.; Kalayil, T.; Fujimoto, M.; Zitomer, D.
H.; McNamara, P. J., Removal of antibiotic resistance genes in an anaerobic membrane bioreactor treating
primary clarifier effluent at 20° C. Environmental Science: Water Research & Technology 2018, 4, (11),
1783-1793.
64. Liu, M.; Zhang, Y.; Yang, M.; Tian, Z.; Ren, L.; Zhang, S., Abundance and distribution of tetracycline
resistance genes and mobile elements in an oxytetracycline production wastewater treatment system.
Environ. Sci. Technol. 2012, 46, (14), 7551-7557.
65. Luo, Y.; Yang, F.; Mathieu, J.; Mao, D.; Wang, Q.; Alvarez, P., Proliferation of multidrug-resistant New
Delhi metallo-β-lactamase genes in municipal wastewater treatment plants in northern China.
Environmental Science & Technology Letters 2013, 1, (1), 26-30.
66. Webber, M.; Piddock, L., The importance of efflux pumps in bacterial antibiotic resistance. J.
Antimicrob. Chemother. 2003, 51, (1), 9-11.
67. Piddock, L. J., Multidrug-resistance efflux pumps? not just for resistance. Nature Reviews Microbiology
2006, 4, (8), 629.
68. Bawdon, R. E.; Crane, L. R.; Palchaudhuri, S., Antibiotic resistance in anaerobic bacteria: molecular
biology and clinical aspects. Reviews of infectious diseases 1982, 4, (6), 1075-1095.
69. Zarei-Baygi, A., Harb, M., Wang, P., Stadler, L. B., & Smith, A. L., Evaluating Antibiotic Resistance
Gene Correlations with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors.
Environmental science & technology 2019, 53(7), 3599-3609.

19

Chapter 2

2. Intracellular versus extracellular antibiotic resistance genes in the
environment: prevalence, horizontal transfer, and mitigation
strategies

Abstract
Antibiotic resistance genes (ARGs) are present as both intracellular and extracellular fractions of DNA in
the environment. Due to the poor yield of extracellular DNA in conventional extraction methods, previous
studies have mainly focused on intracellular ARGs (iARGs). In this review, we evaluate the
prevalence/persistence and horizontal transfer of iARGs and extracellular ARGs (eARGs) in different
environments, and then explore advanced mitigation strategies in wastewater treatment plants (WWTPs)
for preventing the spread of antibiotic resistance in the environment. Although iARGs are the main
fraction of ARGs in nutrient-rich environments, eARGs are predominant in receiving aquatic
environments. In such environments, natural transformation of eARGs occurs with a comparable
frequency to conjugation of iARGs. Further, eARGs can be adsorbed by soil and sediments particles,
protected from DNase degradation, and consequently persist longer than iARGs. Collectively, these
characteristics emphasize the crucial role of eARGs in the spread of antibiotic resistance in the
environment. Fate of iARGs and eARGs during advanced treatment technologies (disinfection and
20

membrane filtration) indicates that different mitigation strategies may be required for each ARG fraction
to be significantly removed. Finally, comprehensive risk assessment is needed to evaluate/compare the
effect of iARGs versus eARGs in the environment.
2.1 Introduction
Antibiotic resistance is one the most formidable threats to human health, responsible for over
700,000 death worldwide each year.
1
Recent studies have revealed a wide and rapid spread of antibiotic
resistance.
1, 2
Unabated, this spread could result in antibiotic resistant infections becoming the leading
cause of death by 2050, resulting in over 10 million deaths each year and $100 trillion in associated costs
worldwide.
3
It is important to note that resistance to antibiotics is a natural phenomenon that occurs via
natural selection.
4
Thus, antibiotic resistance is found in essentially all natural environments, including so
called “pristine” environments.
5
However, anthropogenic activities are driving a marked increase in
antibiotic resistance in engineered, clinical, and environmental settings.
3
One of these anthropogenic
activities which significantly contributes to the development of antibiotic resistance is antibiotic selective
pressure posed by antibiotics use, however rational and appropriate.
6-8
Antibiotics are commonly
employed to manage infections in humans and animals and have been used widely as an animal growth
promoter regardless of a bacterial infection being positively identified.
6, 9, 10
It has been reported that the
global antibiotic consumption has increased 35% in the first decade of the 21
st
century.
11
A large portion
of consumed antibiotics (up to 75%) in both humans and animals remain unmetabolized and are
subsequently excreted, resulting in potential contamination of the receiving environment.
12
Antibiotics
provide a selective pressure in the receiving environment which contribute to the promotion/transfer of
resistance among bacteria.
6, 13, 14

Bacterial resistance to antibiotics occurs via acquisition of antibiotic resistance genes (ARGs). ARGs
are present as both intracellular and extracellular DNA (iDNA and eDNA).
15, 16
It is important to note that
21

eDNA originates from iDNA during the lysis of dead bacterial cells and active secretion from live bacterial
cells. Some studies have categorized ARGs into three different groups: intracellular ARGs (iARGs), free
extracellular ARGs (free eARGs), and particle-associated or adsorbed eARGs (eDNA can be adsorbed to
cells or particles).
17, 18
Free eARGs (also known as cell-free ARGs) are defined as those which pass through
a 0.22 µm filter, while iARGs and adsorbed eARGs (also known as cell-associated ARGs) do not pass
through a 0.22 µm filter.
19
In the present review, we refer to cell-associated ARGs as iARGs and cell-free
ARGs as eARGs. The abundance and diversity of iARGs and eARGs in different environments may vary
significantly.
15, 17, 20-22
iARGs are the predominant form in nutrient-rich environments that support
bacterial growth, such as waste streams, while eARGs are present in higher abundance in soil and
sediments where nutrient deprivation is common.
16, 23, 24

Antibiotic resistance can be transferred among and across bacterial species through vertical gene
transfer (VGT) and horizontal gene transfer (HGT).
9
Horizontal transfer of ARGs, which is known as the
main antibiotic resistance proliferation mechanism in the environment, can result in the acquisition of
antibiotic resistance by a pathogenic recipient cell from a benign donor.
25, 26
Horizontal transfer of iARGs
may occurs through conjugation and transduction, while eARGs predominantly spread through natural
transformation.
25-27
Conjugation is the transfer of iDNA via pili which requires cell to cell contact.
27, 28

Transduction also transfers iDNA, although via bacteriophage infection.
29, 30
Natural transformation is the
direct uptake of eDNA by competent cells.
18, 27, 31
The impact of natural transformation on the spread of
antibiotic resistance was previously underestimated, but recent studies have revealed that this transfer
mechanism plays an important role in dissemination of ARGs in the environment.
4, 15, 19, 32
Therefore, it is
important to track frequency of HGT mechanisms separately to assess the potential risks associated with
each ARG fraction (iARGs and eARGs) in receiving environments.
Another important distinction between iARGs and eARGs is their stability in different environments.
Although iARGs can benefit from the protection provided by cell membranes and cell walls to endure
22

harsh environmental conditions, such conditions may result in cell death and release of iARGs into the
extracellular environment. It has been reported that eARGs can bind to other particles which in turn
protects from degradation by nucleases, consequently persisting for a long duration of time in the
environment.
16, 24
Although free eARGs can be uptaken more easily by competent cells,
33, 34
adsorbed
eARGs remain available for natural transformation.
18, 32
High prevalence and transferability of eARGs
implies their critical role in proliferation of antibiotic resistance. However, the majority of previous studies
have focused on the intracellular fraction of ARGs and have neglected eARGs. The is primarily a result of
conventional DNA extraction methods having poor yields of eDNA.
17, 35
More recent advances in eDNA
extraction methods have been developed to be more representative of iDNA and eDNA fractions in
environmental samples. Further, due to their distinct characteristics, iARGs and eARGs may have differing
responses to mitigation strategies in wastewater treatment plants (WWTPs). Therefore, better
understanding of these two fractions of ARGs is crucial to devising appropriate strategies to mitigate the
spread of antibiotic resistance in the environment.
In the present review, first we compare the prevalence/persistence of iARGs versus eARGs in
different environments, and then decipher how and with what frequency each ARG fraction can be
transferred horizontally to recipient bacteria. This information indicated the crucial role of eARGs in
dissemination of antibiotic resistance in the environment. We then compare the impact of advanced
treatment technologies (disinfection and membrane filtration) on the release of iARGs and eARGs in the
effluent of WWTPs. Finally, we discuss research gaps and suggest priority research areas to increase our
understanding of prevalence, HGT, and mitigation strategies of iARGs and eARGs.
23

2.2 Prevalence of iARGs versus eARGs in different environments
2.2.1 Impact of environmental characteristics
Biotic environmental characteristics (e.g., enzymatic degradation, microbial community structure,
nutrient availability), and abiotic environmental characteristics (e.g., temperature, pH, moisture content,
presence of adsorbents) are the primary factors influencing the prevalence and persistence of iARGs and
eARGs in different environments.
15, 20-23
iDNA is persistent in environments where bacterial growth is
maintained. Therefore, nutrient availability is a crucial factor in iDNA prevalence and persistence. Another
factor which can increase persistence of iARGs in the environment is HGT,
36
such as via conjugation of
iARGs.
37
For example, increased host diversity for conjugation has been shown to contribute to survival of
plasmids.
38
However, for conjugation to occur fast enough to compensate for iARG loss (through release
of iARGs from living or dead cells), a complex community is required.
36, 37
As a result, abundance of iARGs
are usually higher than eARGs in nutrient-rich environment with high microbial complexity. Zhang et al.
19

reported that the absolute abundance of targeted iARGs was 3 orders of magnitude higher than that of
eARGs in municipal wastewater. In another study on municipal wastewater, Yuan et al.
17
observed that
the absolute abundance of targeted iARGs, and eARGs to be 3.4  10
11
copies/L, and 6.5  10
9
copies/L,
respectively. In cattle manure, Zhang et al.
15
reported that the absolute abundance of targeted iARGs were
generally 2-3 orders of magnitude higher than eARGs. Comparable observations have been reported for
swine manure.
39, 40
Collectively, the absolute abundances of iARGs in both wastewater and manure are 2-
3 order of magnitude higher than that of eARGs (Figure 2.1).

24

Figure 2.1 Contribution of intracellular and extracellular ARGs (iARGs and eARGs) to ARG profile in
different environments.
Prevalence of iARGs versus eARGs in aquatic environments such as rivers and marine sediments
reportedly follow a different trend than that of in the above-mentioned environments. In contrast to
waste streams, low nutrient availability in aquatic environments likely results in cell lysis and subsequent
release of iDNA/iARGs, resulting in enrichment of eARGs. Mao et al.
16
reported that in the Haihe River the
dominant fraction of DNA in sediment was eDNA (iDNA: 77 ± 13 ug/g, eDNA: 97 ± 20 ug/g), while iDNA
was predominant in water (iDNA: 9.7 ± 1,5 ug/mL, eDNA: 2.2 ± 0.8 ug/mL). In sediment of aquaculture
farms, however, Yuan et al.
21
found the concertation of iDNA (2.8-17.1 ug/g) to be significantly higher
than eDNA (0.9-6.2 ug/g), which was in accordance with observations in sediment of a livestock waste
management structure (and not natural marine sediments).
15
This might be due to the high nutrient level
of aquaculture farm sediments (compared to the low nutrient availability in river and natural marine
sediments) which provide suitable conditions for growth and survival of bacteria. The absolute
abundances of targeted iARGs and eARGs in Haihe River sediment were reported to be 1.2  10
10
copies/g,
and 3  10
10
copies/g, respectively.
16
In Haihe River estuarine sediment, lower abundance of both iARGs
0
20
40
60
80
100
Tap Water Raw Municipal
Wastewater
Activated
Sludge
Manure Aquatic
Sediment
Contribution of each ARG fraction to total
targeted ARGs %
iARGs eARGs
25

and eARGs were detected, but eARGs remained at higher concentration than iARGs (Figure 1).
41
Gue et
al.
42
reported that in the estuary of the Yangtze River, the absolute abundance of targeted eARGs in
sediment was higher than water samples. In the sediment samples, eARGs accounted for 78% of the total
targeted ARGs, while in water samples the contribution of eARGs was only 5%. These results indicate the
imperative role of environmental conditions (e.g. nutrient availability) in prevalence of different fractions
of ARGs in the environment.
Stability of eDNA in soils and sediments is another factor resulting in high prevalence. Enzymatic
degradation by DNases is the main mechanism of eDNA degradation.
23, 24, 36
In the environment, DNases
associated with active bacteria (and not extracellular DNases) are primarily responsible for eDNA
degradation.
43
It has been reported that nuclease activity is temperature dependent with lower activity
reported at low temperatures.
24
Rapid desiccation, high salt concentration, and low pH are other factors
which may slow enzymatic degradation of eDNA.
44, 45
eDNA has a high binding ability and can be adsorbed
by soil colloids and particles with capacity as high as 10
3
ug/g (Figure 2).
46
eDNA can bind more strongly
to clay particles than to sand, indicating clay particles may provide higher protection of eDNA against
DNase degradation.
24, 47
Therefore, the type of soil can also impact rate of DNA degradation in the
environment. Bacterial eARGs released from living or dead cells can persist for up to 70 days in soils.
24
In
aquatic environments, different persistence times have been reported for eARGs, from a few hours, to a
few months.
16, 23, 24
Mao et al.
16
investigated the persistence of different fractions of DNA in water and
sediments of the Haihe River and reported that both iDNA and eDNA in the water column were below the
detection limit after the first week. In sediment, the degradation rate of eDNA was constant over 20 weeks
of observation. The degradation rate of iDNA was initially similar to that of eDNA during the first 6 weeks,
however, it increased significantly by 12-folds in the following weeks. As a result, iARGs were not
detectable after 8 weeks, while eARGs were still present in sediments after 20 weeks.
26

Figure 2.2 Fate of intracellular and extracellular ARGs (iARGs and eARGs) in receiving aquatic environment.
2.2.2 Impact of ARG characteristics
Characteristics of ARGs (e.g., plasmid-borne or chromosomal) are other factors impacting the
prevalence and persistence of iARGs and eARGs in different environments.
23
Since plasmid-borne ARGs
have a higher mobility compared to chromosomal ARGs, higher concentrations of the former are expected
in the extracellular environment. For instance, the absolute abundances of tetM and tetW were reported
to account for 73% of the total targeted eARGs in swine manure, while their contribution to the iARG
profile was only 20%, implying the prevalence of these ARGs on mobile genetic elements (MGEs).
39
In tap
water, tetC was among the most abundant ARGs in eDNA, while its contribution to the iARG profile was
negligible.
20
On the other hand, in cattle manure, the absolute abundance of intracellular tetO was found
to be 4-5 orders of magnitude higher than its extracellular form.
15
In swine manure, Sui et al.
39
indicated
that both ermB and ermF were prevalent in iDNA, while they were rarely detected in eDNA. The drastically
higher absolute abundance of intracellular forms of an ARG compared to extracellular forms suggests
location on chromosomal DNA, rather than MGEs. In municipal wastewater, the contribution of each ARG
subtype to iARG and eARG profiles was approximately similar.
17, 19
Yuan et al.
17
reported that 98% of
variation of both iARGs and eARGs in a WWTP treating municipal wastewater was due to changes in the
relative abundance of bacterial genera, implying that despite the secretion of plasmids, abundance of
27

iARGs versus eARGs depends on the fate of their bacterial hosts. In sediments, ARG fractions (iARGs and
eARGs) followed the same trend as their DNA fractions (iDNA and eDNA) both in the Haihe River and an
aquaculture farm,
16, 21
implying that the presence of intra- and extracellular forms of ARG are highly
influenced by the compositional pattern of microbial DNA in the sediments. sul1 and sul2 were among the
most abundant ARGs both in iDNA and eDNA of tap water, municipal wastewater, manure, and sediments
indicating the ubiquity of sulfonamide resistance genes in both intra- and extracellular forms.
15, 17, 19, 20, 39,
40

Characteristics of DNA molecules also impact persistence of ARGs in the extracellular environment.
23

Some eARGs are more recalcitrant to DNases than others, likely due to their sequence and structural
features.
24, 44
Smaller DNA fragments has been reported that adsorb preferentially to soil particles, and
are therefore mainly protected from DNase degradation.
48
Further, due to the higher surface charges and
more molecular flexibility, chromosomal DNA is more adsorptive than plasmid DNA.
15, 46
Therefore, it is
highly likely that extracellular chromosomal DNA persists longer than extracellular plasmid DNA in the
environment. As a result, lower detection frequency of eARGs located on plasmid DNA (compared to those
mainly located on chromosomal DNA) is expected in soil and sediments. Yuan et al.
21
reported that
detection frequency of plasmid-borne ARGs was much lower than chromosomal ARGs in aquaculture
sediment eDNA. However, Mao et al.
16
observed that extracellular 16s rRNA degraded faster than eARGs
in sediments, which might imply that chromosomal DNA may degrade faster than plasmid DNA in the
extracellular environment. Zhang et al.
19
evaluated the persistence of iARGs and eARGs in secondary
clarifier effluent, and O 3 disinfected effluent of a WWTP treating municipal wastewater and reported that
after 25 days, the absolute abundance of total targeted iARGs decreased in both effluents, with greater
reduction in disinfected effluent. A significant increase in the absolute abundance of all targeted eARGs,
however, was observed for both secondary clarifier and disinfected effluents. In both effluents,
extracellular blaPSE-1 showed the highest increase (about 3 logs), while extracellular ermB was
28

approximately constant during 25 days of storage. Wang et al.
49
also reported that during storage of
WWTP effluent, the absolute abundance of extracellular ermB and tetO remained constant, however,
extracellular blaTEM and tetW increased significantly by 2 and 1 logs, respectively.
2.3 Horizontal transfer of iARGs versus eARGs
HGT is a prevalent antibiotic resistance proliferation pathway in the environment and of concern for
the potential spread of antibiotic resistance to pathogenic microorganisms. Tomova et al.
50, 51
observed
identical sequences of qnrA1, qnrB1, and qnrS1 in Escherichia coli extracted from both clinical patients
and marine bacteria in the same area, implying the potential of HGT between marine bacteria and human
pathogens. Both intra- and extracellular MGEs (iMGEs and eMGEs) play important roles in the movement
and propagation of ARGs through HGT. Zhou et al.
52
employed metagenomics and observed that in
activated sludge, 2.4-17.4% of eARGs were flanked by more than one eMGE. In particular, co-locations of
several eMGEs with extracellular sulfonamide and tetracycline resistance genes were found, highlighting
the potential widespread proliferation of these ARGs in the environment. In swine manure treatment,
more significant correlations were observed between eMGEs and eARGs than iMGEs and iARGs.
39
In
sediment, Zhao et al.
41
reported higher absolute abundance of iMGEs than eMGEs. Interestingly, in the
same sediment samples, the absolute abundance of eARGs were higher than iARGs. This might imply that
eMGEs carry more ARGs compared to iMGEs, indicating the high mobility of eARGs.
29

Figure 2.3 Horizontal transfer of ARGs through conjugation, transduction, and natural transformation.
Inserted tables show frequency of each horizontal gene transfer mechanism in different environments.
HGT is an ancient mechanism, however, the rate at which it occurs has significantly increased over
time.
25
It has been reported that a considerable portion of most bacterial genomes consist of horizontally
acquired genes.
25, 27, 53
HGT of iDNA/iARGs may occur through conjugation and transduction, while
eDNA/eARGs can be transferred through natural transformation, as described in greater detail in the
following sections.
2.3.1 Conjugation of iARGs
Conjugative transfer is mediated by cell to cell contact via a pili or via adhesins and a pore
(interbacterial junction) for iDNA to pass through (Figure 3).
27
Conjugation is the most studied HGT
mechanism and is commonly associated with plasmids
25
primarily because small MGEs like plasmids can
transfer faster than a whole chromosome (it may take more than 1 hour for a whole chromosome to
30

transfer, however it rarely occurs because the interbacterial junction would break down during that time
span).
27, 28
Estimates suggest that more than 50% of known plasmids can be transferred via conjugation.
36

Conjugation can occur between the same bacterial species, but may also occur between unrelated
populations with substantial taxonomic distance, although at a lower frequency.
25, 27
Conjugative
elements such as plasmids and transposons can facilitate the transfer of antibiotic resistance by picking
up ARGs and carrying them to recipient cells.
27, 28
Insertion sequences (ISs) are the simplest transposable
elements that can be integrated into non-transmissible elements, making them transmissible.
27

Conjugation of MGEs carrying ARGs has been frequently reported in various environments (e.g., soil and
water), and food, plant, animal, and clinical bacteria.
25, 36, 54, 55
Co-location of ARGs in MGEs significantly
increases the conjugative transfer of multi-drug resistance among and across bacteria in different
environments.
25, 52, 56, 57

Conjugation of antibiotic resistance plasmids occurs with different frequencies (Table 1). Factors that
may impact conjugation frequency are: (1) genetic characteristics of donor and recipient cells (e.g., pilus
expression), (2) characteristics of plasmid (certain plasmids are self-transmissible and some have an
extremely broad host range, significantly impacting the frequency of conjugation, (3) mating assay
employed (filter mating, broth mating, biparental mating, chip mating, etc.), and (4) incubation conditions
(contact opportunity, nutrient availability, temperature, etc.). Inoue et al.
58
found the conjugation
frequency of an antibiotic resistance plasmid from E. coli strain to certain bacteria isolated from activated
sludge varied between 8.8  10
-7
to 1.3  10
-2
transconjugants (the recipient cell that horizontally acquired
plasmid during conjugation)/recipient cells, with Acinetobacter sp. YAA and Sphingomonas paucimobilis
551 having the highest conjugation frequency, > 3  10
-3
. This implies some bacterial species are more
susceptible to receive plasmids through conjugation. In activated sludge, Enterobacter, Acinetobacter,
Pseudomonas, and Aeromonas are known to be the most susceptible genera to be transconjugant.
59-61

Dang et al.
62
reported that the conjugation frequency of different conjugative resistance plasmids isolated
31

from sediment samples of the Haihe River ranged from 8.5  10
-7
to 7.5  10
-1
transconjugants/recipient
cells, with pGA52 having the highest and pS21 having the lowest conjugation frequency. This indicates
that the type of plasmid plays an important role in conjugation frequency of ARGs. In regard to mating
assay, conjugation transfer of a multi-drug resistance plasmid among marine sediment bacteria for broth
mating versus plate mating was reported to be similar.
54
However, chip-mating provides more interaction
between bacteria, resulting in higher conjugation frequency compared to other mating assays.
60

Incubation conditions also play crucial roles in conjugation frequency. The average conjugation frequency
of an antibiotic resistance plasmid from an E. coli strain to activated sludge bacteria in a membrane
bioreactor (MBR) was reported to be 2.8  10
-5
transconjugants/recipient cells.
63
However, Li et al.
60
used
microfluidic cell culturing (more chance of cells interaction) to quantify conjugation frequency of an
antibiotic resistance plasmid from an E. coli strain as donor to activated sludge bacteria as recipients, and
revealed that the conjugation frequencies varied between 2.5  10
-3
to 3.4  10
-2

transconjugants/recipient cells. These imply incubation in an MBR reduces the cells contact opportunity
compared to laboratory cultivation. Other incubation conditions can also significantly impact conjugation
transfer frequency of plasmids. Grabow et al.
64
reported that conjugation frequency of antibiotic
resistance plasmids among E. coli strains in raw dam water occurred with a rate of 1000 times lower than
when the bacteria were grown in nutrient-rich media in laboratory. It has been also reported that rising
temperature from 20 °C to 35 °C increased conjugation frequency by 2.3 logs.
Table 2.1 Conjugation frequency of iARGs.
Antibiotic resistance plasmid Donor source Recipient source Culture method Frequency Reference
Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge (E. coli
HB101)
Broth mating 3 10
-3
– 6 10
-3

58

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(Acinetobacter calcoaceticus
AH)
Broth mating 7 10
-6
– 3 10
-5

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(Acinetobacter sp. YAA)
Broth mating 7 10
-3
– 2 10
-2

32

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge (Alcaligenes
sp. YAJ)
Broth mating 2 10
-5
– 6 10
-4

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(Pseudomonas putida BH)
Broth mating 5 10
-4
– 2.3 10
-3

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(Sphingomonas paucimobilis
551)
Broth mating 9 10
-3
– 3 10
-2

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(Aeromonas sp. S-18)
Broth mating 5 10
-6
– 8 10
-6

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(Burkholderia cepacia S-11)
Broth mating 4 10
-4
– 5 10
-3

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(paucimobilis S-5)
Broth mating 3 10
-6
– 7 10
-6

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli C600 Activated Sludge
(paucimobilis S-8)
Broth mating 5 10
-7
– 2.6 10
-6

Plasmid pGA45 (blaIMI-3) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (7.81±7.15) 10
-3

62

Plasmid pA15 (blaTEM-1, aacC2,
ermB, mphA)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (5.33±0.61) 10
-1

Plasmid pGA52 (blaTEM-1,
ermB, mphA)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (7.48±1.91) 10
-1

Plasmid pG4 (blaTEM-1, aacC2,
tetA, ermB, mphA, sul2, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (6.71±2.21) 10
-1

Plasmid pG3 (blaTEM-1, aacC2,
ermB, mphA)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (4.36±1.39) 10
-1

Plasmid pJT1 (blaTEM-1, tetA) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (1.30±0.43) 10
-1

Plasmid pFTC1 (tetA) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (2.13±0.82) 10
-2

Plasmid pFE1 E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (6.60±3.83) 10
-3

Plasmid pNA6 (blaTEM-1, sul1,
ereA, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (3.84±3.05) 10
-1

Plasmid pNA11 (blaTEM-1) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (4.26±0.92) 10
-4

Plasmid pNTC6 (tetA) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (2.08±0.68) 10
-1

Plasmid pZTC3 (tetA) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (8.17±0.96) 10
-5

Plasmid pZTC1 (sul1, tetG, intI1) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (3.08±0.91) 10
-2

Plasmid pS21 (strA, strB,
blaTEM-1, tetA, sul2, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (8.48±4.67) 10
-7

Plasmid pS17 (strA, strB,
blaTEM-1, tetA, sul2, aacC2,
mphA, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (2.13±0.50) 10
-6

Plasmid pS26 (strA, strB,
blaTEM-1, tetA, sul2, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (1.27±1.25) 10
-6

Plasmid pS46 (strA, strB,
blaTEM-1, sul2, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (2.50±0.91) 10
-6

Plasmid pYG1 (blaTEM-1, aacC2,
tetA)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (4.53±2.82) 10
-2

33

Plasmid pYE1 E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (3.06±1.63) 10
-3

Plasmid pYJS2 (sul1, intI1) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (8.24±0.83) 10
-2

Plasmid pYJS6 (sul1, tetG, intI1) E. coli CV601 Marine Sediment (E. coli
J53)
Liquid Mating (7.14±2.00) 10
-2

Plasmid pDTC31 (tetA, intI1) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (1.08±0.16) 10
-2

Plasmid pDTC28 (sul1, tetA) E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (1.18±0.27) 10
-2

Plasmid pDTC33 (tetA, strA,
strB, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (1.78±0.59) 10
-2

Plasmid pDTC44 (strA, strB,
tetA, intI1)
E. coli CV601 Marine Sediment (E. coli J53) Liquid Mating (1.82±0.67) 10
-2

Plasmid pSCL (multi-drug) Pseudomonas
fluorescens
Marine Bacteria Broth mating 2 10
-4
– 5.2 10
-3

54

Plasmid pSCL (multi-drug) Pseudomonas
fluorescens
Marine Bacteria Plate mating 3.4 10
-4
– 5.6
10
-3

Plasmid pSCL (multi-drug) Pseudomonas
fluorescens
Marine Bacteria Natural seawater 8.5 10
-7
–
9.1 10
-4

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli K12 Activated Sludge Bacteria MBR 2.8 10
-5

63

Plasmid pKJK5 (trimethoprim) E. coli MG1
655
Activated Sludge Bacteria Microfluidic chip 2.5 10
−3
–
3.4 10
−2

60

Plasmid pSK41 (gentamicin) Staphylococcus
aureus
Sewage Bacteria Plate mating 10
-7
– 1.2 10
-5

65

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli K12 Pseudomonas putida Plate mating 6 10
-7
– 6.4 10
-6

66

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli K12 Pseudomonas putida Plate mating 6 10
-6
– 1.3 10
-5

67

Amoxicillin resistance plasmids Manure
Bacteria
(Exogenous
isolation of
plasmids)
Piggy manure (E. coli strain
CV601)
biparental
matings plate
mating
7 10
-8
– 6 10
-5

68

Sulfadiazine resistance plasmids Manure
Bacteria
(Exogenous
isolation of
plasmids)
Piggy manure (E. coli strain
CV601)
biparental
matings plate
mating
2 10
-8
– 5 10
-5

Tetracycline resistance plasmids Manure
Bacteria
(Exogenous
isolation of
plasmids)
Piggy manure (E. coli strain
CV601)
biparental
matings plate
mating
5 10
-8
– 7 10
-5

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
E. coli J5 Chicken manure (E. coli
CV601)
Filter mating 3.6 10
-4
–
2.4 10
-3

69

Plasmid pIE723 (gentamicin,
kanamycin, and streptomycin)
E. coli C600 Chicken manure (E. coli
CV601)
Filter mating 8.1 10
-5
–
2.1 10
-3

34

2.3.2 Transduction of iARGs
Transduction is the bacteriophage-mediated transfer of iDNA from an infected cell to a recipient cell
(Figure 3).
25, 29
Bacteriophages (or simply phages) are bacterial viruses which pick up and transfer genes
that are advantageous to their microbial host.
25, 30
They can transfer both chromosomal and plasmid
DNA.
29, 70
To achieve a stable bacterial transduction, the translocated DNA must be incorporated into the
recipient chromosome by homologous recombination.
25, 29
It has been observed that some phages have a
wide range of bacterial hosts and can move across different species.
25, 29, 71
The transfer of ARGs by phages
for different bacterial species has been widely reported previously.
72-74
Measuring transduction frequency
presents significantly more complications relative to measuring conjugation frequency because
experimental design of controlled transduction requires identifying and selecting suitable and sensitive
bacterial hosts which support infection with specific phages.
30, 70
Interestingly, phages can be employed
as an alternative antibiotic treatment for human infection.
75, 76
Phages first attach to their bacterial host
and inject their genome. Then, the injected genome seizes the host’s molecular machinery to synthesize
new phages. Finally, new phages lyse the host cell and spread to the environment.
75, 76
Although phage
therapy can effectively lyse the infectious bacteria, it results in accumulation of ARGs in phages.
Recent studies have indicated that phages are a significant reservoir of specific ARGs in
environment.
70, 71, 77-80
Wang et al.
80
targeted 32 ARGs in pig fecal samples and found 11 of them present
in phages, while all 32 ARGs were detected in bacterial DNA. The absolute abundance of total iARGs in
fecal bacteria and phages were around 10
10
and 10
6
copies/g, respectively. The same results were
observed by Calero-Cá ceres
81
in municipal wastewater and anaerobically digested sludge, where the
absolute abundance of total iARGs in bacterial DNA were 10
3
-10
4
times more than that in phage DNA.
Although the absolute abundance of total ARGs in phages are much lower than that in bacterial DNA,
some studies have reported that certain ARGs, such as qnrA and blaTEM, were predominantly found in
35

phage DNA with absolute abundance close to those in bacterial DNA. Some ARGs like ermB, sul1, and fexA,
on the other hand, were rarely detected in phage DNA.
80, 81
Calero-Cáceres and Balcázar
78
reported that
phages from marine habitats are potential reservoir of -lactam, tetracycline, and multi-drug resistance
genes. These observations imply that phages play an important role in dissemination of specific ARGs.
Table 2.2 Transduction frequency of iARGs.
Antibiotic resistance Phage Recipient source Culture method Frequency Reference
Penicillinase resistance
plasmid
Ø80a clinical isolates of
Staphylococcus aureus USA300
Broth Mating 7.9 10
-6
– 1.5 10
-5

73

Tetracycline resistance
plasmid
Ø80a clinical isolates of
Staphylococcus aureus USA300
Broth Mating 4.6 10
-6

Penicillinase resistance
plasmid
ØJB clinical isolates of
Staphylococcus aureus USA300
Broth Mating 9 10
-7
– 5 10
-6

Tetracycline resistance
plasmid
ØJB clinical isolates of
Staphylococcus aureus USA300
Broth Mating 2.8 10
-6

Plasmid pQSR50 (kanamycin,
streptomycin)
T- ØHSIC marine bacteria (HSIC) Triparental Broth
mating
5.1 10
-9

30

Plasmid pQSR50 (kanamycin,
streptomycin)
T- ØD1B marine bacteria (HSIC) Triparental Broth
mating
1.57 10
-8

Antibiotic resistance clone
(streptomycin)
Ø63 Soil bacteira (Bacillus
thuringiensis)
Liquid Mating 2 10
-7
– 4 10
-7

72

Chromosomal resistance
gene (ampicilin)
P1kc freshwater (E. coli) Plate mating (2±0.5) 10
-7

82

Chromosomal resistance
gene (ampicilin)
P1kc freshwater (E. aerogenes) Plate mating (6±0.5) 10
-7

Chromosomal resistance
gene (ampicilin)
T4GT7 freshwater (E. coli) Plate mating <2 10
-9

Chromosomal resistance
gene (ampicilin)
T4GT7 freshwater (E. aerogenes) Plate mating 2 10
-8

Transduction usually occurs with a lower frequency compared to conjugation (Table 2). Volkova et
al.
74
developed a model to estimate the transduction frequency of a -lactam resistance gene (blaCMY-2)
to an E. coli strain, in the gut of cattle and reported that the rate of transduction of blaCMY-2 was 1000
times lower than conjugation of a plasmid carrying blaCMY-2 to the same recipient strain. One factor that
contributes to the rate of transduction frequency is plasmid characteristics (e.g., self-transmissible). Varga
et al.
73
investigated the transduction of -lactam and tetracycline resistance plasmids to clinically isolated
36

bacterial strain and observed the transduction frequency of as high as 1.5  10
-5
and 4.6  10
-6

transductions/plaque-forming for -lactam and tetracycline resistance plasmids, respectively, showing
the higher mobility of -lactam resistance plasmids compared to tetracycline. The type of phage can also
impact the rate of transduction frequency. For instance, Jiang and Paul
30
found the transduction frequency
of 5.1  10
-9
and 1.6  10
-8
transductions/plaque-forming for an antibiotic resistance plasmid by T- HSIC,
and T- D1B phages, respectively, to marine bacteria.
2.3.3 Natural transformation of eARGs
Natural transformation occurs through direct uptake and integration of eDNA (Figure 3).
27, 83
Bacteria
first need to develop a regulated physiological state of competence for natural transformation to occur.
25,
27, 33
Competence development can be induced in response to specific environmental conditions.
25, 31

However, there is no universal consensus showing a bacterial species can develop competence.
27
The
purpose for eDNA uptake by competent cells may be nutrition, repair of chromosomal DNA, and
diversification of genetic material for evolution.
33
Another prerequisite for natural transformation is the
persistence of DNA in the extracellular environment.
24, 27
It has been reported that large chromosomal
and plasmid eDNA were readily degraded in humans by DNase, while smaller plasmid eDNA can remain
largely intact for greater durations.
27
Nevertheless, the degradation of eDNA is dependent on the
environmental conditions (discussed in detail in previous section). In contrast to conjugation, there are
several studies indicating that the efficiency of natural transformation of longer eDNA fragments is higher
than that of smaller eDNA fragments.
34, 84
After entering the recipient cell, the translocated DNA needs to
be integrated in the recipient bacterial genome if it is chromosomal DNA, and if its plasmid DNA, needs to
be integrated or recircularized into a self-replicating plasmid.
25, 27, 33
The latter process is more complex,
and as a result, natural transformation of chromosomal DNA is more efficient than plasmid DNA.
27
Nielsen
et al.
85
used the gene cassette KTG as a marker for transformation of both plasmid and chromosomal
37

eDNA at the same conditions, and reported that transformation frequency of chromosomal eDNA (4.7 
10
-3
transformants/recipient cells) was significantly higher than plasmid eDNA (2.5  10
-5
).
Methods for quantification of natural transformation of eDNA under environmental conditions are
not as well developed as conjugation and transduction. Naquin et al.
86
filtered treated wastewater
samples (containing mecA gene) to obtain free eARGs and then evaluated natural transformation by S.
aureus strain (sensitive to methicillin antibiotic). Results revealed that after exposure to the treated
wastewater free eDNA sample, S. aureus strain was resistant against methicillin, which indicates uptake
of methicillin resistance gene (possibly mecA) during exposure through natural transformation.
Characteristics of recipient cells can play an important role in frequency of natural transformation of
eARGs (Table 3). Qingxiang et al.
87
used a multi-resistance plasmid for natural transformation by Bacillus
subtilis, Bacillus cereus, Bacillus thuringiensis, and Proteus vulgaris. Their results showed that B. cereus
had the highest transformation frequency (10
-3
transformants/recipient cells), followed by B. thuringiensis
(6.7  10
-6
), P. vulgaris (6.6  10
-6
), and B. subtilis (2  10
-7
). Environmental conditions also influence the
rate of natural transformation frequency. Nielsen
85
et al. reported that the transformation frequency of
chromosomal eDNA by a bacterial strain isolated from soli can reach as high as 4.7  10
-3

transformants/recipient cells, which depended on nutrient availability, temperature, eDNA concertation
and soil content.
Table 2.3 Natural transformation frequency of eARGs.
Antibiotic resistance Free/Adsorbed Recipient source Culture method Frequency Reference
Chromosomal resistance gene
(kanamycin, gentamicin)
free eDNA Soil (Acinetobacter
calcoaceticus)
Plate Mating 4.7 10
-3

85

Plasmid pSKTG (kanamycin,
gentamicin)
free eDNA Soil (Acinetobacter
calcoaceticus)
Plate Mating 2.5 10
-5

Chromosomal resistance gene
(kanamycin, gentamicin)
free & adsorbed
eDNA
Soil (Acinetobacter
calcoaceticus)
Soil 5.6 10
-5
– 1.4 10
-7

Plasmid pYN1 (kanamycin) free eDNA Chicken manure
(Bacillus subtilis)
Plate Mating 2 10
-7

87

38

Plasmid pYN1 (kanamycin) free eDNA Chicken manure
(Bacillus cereus)
Plate Mating 10
-3

Plasmid pYN1 (kanamycin) free eDNA Chicken manure
(Bacillus thuringiensis)
Plate Mating 6.7 10
-6

Plasmid pYN1 (kanamycin) free eDNA Chicken manure
(Proteus vulgaris)
Plate Mating 6.6 10
-6

Chromosomal DNA free eDNA Soil (Pseudomonas
stutzeri)
Liquid mating 7 10
-6

18

Chromosomal DNA sand adsorbed Soil (Pseudomonas
stutzeri)
Sand 4 10
-6

Chromosomal DNA free eDNA Soil (Azotobacter
vinelandii DJ77)
Liquid mating 6 10
-5
– 9 10
-5

32

Chromosomal DNA Adsorbed to
silica
Soil (Azotobacter
vinelandii DJ77)
Liquid mating 2 10
-5
– 10
-4

Chromosomal DNA Adsorbed to
NOM
Soil (Azotobacter
vinelandii DJ77)
Liquid mating 2 10
-5
– 9 10
-4

Plasmid pUC19 (ampicillin) free eDNA E. coli DH 5a Plate Mating (7.5±0.6) 10
-4

88

Plasmid RP4 (ampicillin,
kanamycin, and tetracycline)
free eDNA E. coli Mach1-T1 Plate Mating 5 10
-5
– 6 10
-3

89

In soil and sediment environments, eDNA can be free or adsorbed to particles and colloids. The
uptake rate of adsorbed eDNA is expected to be lower than that of free eDNA.
33, 34
Previous studies
reported that after release of eDNA in soil, it quickly becomes unavailable for transformation.
85

Nevertheless, Dong et al.
40
revealed that an antibiotic resistance plasmid can be uptaken by E. coli
competent cell with the efficiency of 1.21  10
3
and 6.7  10
2
transformants/mg plasmid, for adsorbed
(into sediment) and free eDNA, respectively (Table 3). Lorenz and Wackernagel
18
also reported the
comparable natural transformation frequency of extracted DNA conferring resistance to rifampin and
streptomycin, for both sand adsorbed (4  10
-6
transformants/recipient cells), and free eDNA (7  10
-6
).
Lu et al.
32
also observed the average transformation frequency of 6  10
-5
, 5  10
-5
, and 2.5  10
-4

transformants/recipient cells by for free eDNA, eDNA adsorbed to silica, and eDNA adsorbed to natural
organic matter (NOM), respectively (differences were not statistically significant). These results suggest
that adsorbing to particles does not reduce the uptake rate of eDNA.
39

2.3.4 Impact of selective pressure on horizontal transfer of iARGs and
eARGs
Selective pressure posed by antibiotics, heavy metals, etc. can significantly impact the rate of HGT
through three main mechanisms: (1) up-regulation of stress oxygen species (SOS) response due to DNA
damage (antibiotics and heavy metals can damage DNA directly and/or induce the production of reactive
oxygen species (ROS) related genes which results in DNA damage); (2) increasing the permeability of cell
membranes due to membrane damage; and (3) inducing the formation of DNA-searching pilus by
enhancing secretion systems-associated genes.
Rate of conjugation frequency can be impacted by both antibiotics and heavy metals. Shun-Mei et
al.
90
reported that sub-minimal inhibitory concentrations (sub-MIC) of ciprofloxacin (0.015 ug/L) and
levofloxacin (0.03 ug/L) increased the conjugation efficiency by 5- and 6-folds, respectively. Their results
showed that antibiotic addition induced the expression of SOS related genes, however, no association was
found between SOS related genes expression and conjugation efficiency. Ohlsen et al.
65
reported that the
concentration of antibiotics in hospital wastewater (0.03 mg/L) was not sufficient to induce the
conjugation of an antibiotic resistance plasmids in both sewage agar plates (filter mating), and in liquid
sewage in a bioreactor. Lopatkin et al.
13
also indicated that sublethal concentrations of antibiotics from
the most widely used classes do not significantly promote the conjugative transfer of plasmids.
Nevertheless, heavy metals including Cd(II), Zn(II), Hg(II), Cu(II), Ag, and nanoparticles of Ag, Cu, Al, and
Ti, which are commonly used to treat infectious diseases, reportedly promote conjugative transfer of
ARGs in the environment.
67, 91, 92
Lu et al.
67
and Zhang et al.
66
reported that environmentally relevant
concentrations of ions and nanoparticles of silver and copper significantly increased conjugative transfer
of an antibiotic resistance plasmid by up to 3.5-folds through damaging the cytoplasmic membrane and
upregulation of SOS response.
40

Antibiotic selective pressure can also influence the rate of transduction frequency. Modi et al.
71

reported that antibiotic selective pressure results in not only enrichment of phage-encoded antibiotic
resistance, but also expanding the interaction between phages and bacterial species. Fothergill et al.,
93

however, found that antibiotic treatment had a selective effect, inducing production of phages by some
strains of Pseudomonas aeruginosa, while reducing it in other P. aeruginosa strains. Ciprofloxacin was the
only antibiotic capable of promoting production of phages in all isolated strains of P. aeruginosa.
Bielaszewska et al.
94
also reported that among ciprofloxacin, meropenem, azithromycin, rifaximin,
tigecycline, and chloramphenicol, the only antibiotic that induced production of phages by E. coli strains
was ciprofloxacin. Induction of phage production in the presence of antibiotics is likely due to DNA damage
which subsequently triggers the SOS response and induces the phage lytic cycle.
Natural transformation of eARGs is also affected by selective pressure. Mao et al.
16
observed that
natural transformation of a kanamycin resistance gene (Kr) by indigenous bacteria (lacking a Kr gene) in
sediment can be doubled in the presence of kanamycin (20 mg/L). Lu et al.
88
reported that the natural
transformation frequency of an antibiotic resistance plasmid by an E. coli strain was (7.5 ± 0.6)  10
-4

transformants/recipient cells, which increased by 1.4-fold in presence of 2 ug/L of Triclosan. Wu et al.
89

explored the effect of antibiotics (ampicillin, kanamycin, tetracycline, cephalexin, chloramphenicol,
vancomycin, levofloxacin, rifampicin, and clindamycin) at different concentrations on transformation
frequency of PBR 322 and RP4 plasmids by competent E. coli strain. Their results revealed positive
correlations between the concentration of antibiotics (except chloramphenicol) and transformation
frequency of PBR 322 and RP4 plasmids. Selective pressure posed by cephalexin at its MIC had the highest
impact (among all introduced antibiotics) on transformation of both PBR 322 (100-fold) and RP4 (40-fold)
plasmids by E. coli strain. Nano-Al 2O 3 at concentration of 10 mmol/L has been also reported that can
increase natural transformation of antibiotic resistance plasmids up to two logs.
95
All three above-
mentioned mechanisms (up-regulation of SOS response, increase in impermeability, and induction of
41

DNA-searching pilus) were observed to contribute to increasing the natural transformation of ARGs due
to antibiotic selective pressure.
2.4 Impact of advanced treatment technologies on abundance of iARGs
and eARGs
WWTPs serve as one of the primary interfaces between the built and natural environment and can
play a crucial role in halting the spread of antibiotic resistance by preventing dissemination of iARGs and
eARGs. However, conventional WWTPs are not designed to prevent release of ARGs, and more advanced
technologies are needed for efficient ARGs removal. Given the different characteristics of iARGs versus
eARGs, it is likely that they are impacted distinctly by different wastewater treatment configurations.
Zhang et al.
19
reported that A2O processes followed by a settling tank can reduce iARGs and eARGs from
municipal sewage by about 2.3 and 1 log, respectively. Sui et al.,
39
however observed that iARGs and
eARGs were decreased by 2.5 and 3.3 logs, respectively, in a sequencing-batch MBR (with membrane pore
size of 1 um) treating swine manure. These results indicate that each treatment processes may distinctly
affect abundance of iARGs and eARGs. Here we review the impact of disinfection and membrane filtration,
two common advanced wastewater treatment technologies, on abundance of iARGs and eARGs.
2.4.1 Disinfection
Disinfection is one of the most important treatment processes in both water and wastewater
treatment plants to inactivate pathogenic microorganisms (e.g., bacteria and viruses). Disinfection can
also inactive ARGs, however some studies have indicated that chlorination may co-select antibiotic
resistance, resulting in enrichment of ARGs in the final effluent.
96-99
Because iARGs are protected by the
cell membrane and wall, while eARGs are exposed to extracellular environment, disinfection may
distinctly affect their abundance.
42

Impact of chlorination on different fractions of ARGs is mainly related to the exposure condition and
molecular characteristics of ARGs (Figure 4). Liu et al.
96
reported that chlorination of secondary clarifier
effluent at exposure (concentration  contact time) of 240-270 mg ClO 2/L.min did not significantly change
the absolute abundance of total targeted eARGs. Some eARG abundances increased significantly after
disinfection, in particular, extracellular vanA increased by 600-fold. Some eARGs, however, decreased
significantly, such as tetM and gyrA (the most abundant eARGs). Therefore, this research suggests that
chlorination inactivates eARGs selectively, indicating that some eARGs have a higher antioxidant capacity
than others. As a result, the abundance of eARGs with a high antioxidant capacity increased by
chlorination due to the release of iARGs during inactivation of ARB, while eARGs with lower antioxidant
capacity were broken down by the strong oxidizing effect of ClO 2. Surprisingly, chlorination significantly
increased relative abundance of most of iARGs, implying that ClO 2 may co-select ARB. Another possible
explanation of the significant increase in iARG abundances is that chlorination may promote conjugative
transfer of antibiotic resistance plasmids. Guo et al.
100
reported that low chlorine exposure of up to 40 mg
Cl/L.min can increase frequency of conjugation by 2-5 times. The chloramine generated during
chlorination can increase the cell permeability and induce more pilus, which act as a pathway for iARGs
to transfer from donor to recipient cells. Higher chlorine exposure however, has a different effect on ARG
profiles. Zhang et al.
19
found that chlorination of a municipal WWTP effluent with NaClO at exposure of
1260 Cl 2 mg/L.min decreased iARG abundances by (0.4-1.1 log), while it had no significant effect on eARGs
removal. Yoon et al.,
101
however determined that a chlorine exposure of 1200 mg/L.min is enough for 4
log reduction in abundance of both iARGs and eARGs from the effluent of WWTPs. They reported that
chlorination of secondary clarifier effluent at an exposure of up to 750 mg/L.min resulted in a slight
increase in eARG concentration, however, iARGs decreased by 0.5-1.2 log. Increasing the exposure of
chlorine over 750 mg/L.min resulted in reduction of both iARGs and eARGs, with eARGs decreasing more
quickly. Slower reduction of iARGs might be due to cell membrane protection and different chlorination
43

condition for iARGs (under cytoplasmic pH) versus eARGs (in bulk solution under extracellular pH).
101
It is
noteworthy that using high chlorine concertation might results in enrichment of ARGs in receiving aquatic
environments.
UV is another disinfection method which can effectively damage and remove both iARGs and eARGs
(Figure 4). ARG sequence composition and applied fluence are the main factors affecting UV damage.
102

McKinney and Pruden
98
used different fluences of UV on wastewater effluent and found that among 4
targeted ARGs, mecA was the most sensitive, followed by vanA, tetA, and ampC. UV’s effect on iARGs and
eARGs was approximately the same. UV fluences of 200 mJ/cm
2
reduced mecA and vanA for about 4 logs,
however, for ampC to remove by 4 logs, UV fluences of 700 mJ/cm
2
were required. Yoon et al.
101
observed
that during UV disinfection of secondary effluent, damage of eARGs occurred 1.7-fold faster than iARGs
due to the protection of iARGs by cell membranes. They suggested that in secondary effluent UV fluences
of 60-90 mJ/cm
2
, and 100-140 mJ/cm
2
is required for significant eARG and iARG reduction (by 4 log),
respectively. The difference between suggested UV fluences for significant removal of ARGs of this study
compared to the previous study, is likely due to the different size of targeted amplicons for ARG
quantification in qPCR. also observed that the inactivation rate of antibiotic resistance plasmid by UV was
3 times faster for amplicon size of 601 bp compared to 267 bp segment.
44

Figure 2.4 Log removal of intracellular and extracellular ARGs (iARGs and eARGs) class at (a) and (b)
different chlorine exposure, (c) and (d) different UV fluences (AS is abbreviation for amplicon size).
Compared to chlorine and UV, ozonation seems to be less effective on both iARG and eARG removal.
Zhang et al.
19
reported that ozonation of WWTP effluent with exposure of 88 mg/L.min had no significant
change in both iARG and eARG abundances. Another study also reported that ozonation of WWTPs
effluent with an exposure of up to 145 mg/L.min does not damage iARGs.
103
He et al.,
104
however observed
that in PBS buffer, 0.5, and 0.06 mg/L.min O 3 is enough for 4 log removal of chromosomal iARG and eARG,
respectively. These results imply that the wastewater matrix significantly impacts the ability of ozonation
to damage ARGs.
Disinfection of water and wastewater can also impact the rate of HGT via disinfection byproducts
(DBPs). Mantilla-Calderon et al.
97
reported that environmental concentration of DBPs can significantly
stimulate natural transformation of eDNA. Via oxidative stress, DBPs damaged DNA, which promoted the
45

transcription of recA gene. Since recA proteins are involved in both DNA repair and integration of foreign
DNA, higher level of recA increased integration of foreign DNA in the bacterial chromosome which
resulted in higher frequency of natural transformation. It is noteworthy that exposure to oxidants, solar
radiation, etc. can also cause mutagenic damage of DNA, and consequently induce ROS and/or SOS
responses, resulting in promotion of natural transformation. Augsburger et al.
105
reported that solar
disinfection at fluences of 153 mJ/cm
2
can increase the rate of natural transformation of eDNA by 2-fold
through upregulation of both DNA repair representative genes (recA and ddrR) and competence related
genes (comA and pilX).
2.4.2 Membrane filtration
Membrane filtration can significantly reduce release of ARGs from WWTPs to the environment.
106-109

Munir et al.
109
observed that concentrations of ARGs (tetW and tetO) in the effluent of WWTPs equipped
with membrane bioreactors were 1-3 log lower than that detected in the effluent of conventional WWTPs.
Size exclusion is the main removal mechanism during membrane filtration. Therefore, it is expected that
membrane filtration results in a higher removal efficiency of intracellular substances compared to
extracellular. Adsorption to the surface of the membrane is another imperative removal mechanism
during membrane filtration. Surface properties of membranes such as electric charge and hydrophobicity-
hydrophilicity are factors that may impact membrane removal efficiency through adsorption. However,
as soon as a membrane fouling layer forms on the surface of the membrane, surface characteristics of the
membrane itself may have little impact on separation performance. Membrane barriers can also trap
eARGs, however, previous studies reported that plasmids can pass through membrane pores one order
of magnitude smaller than their radius size.
110
Slipko et al.
111
reported that microfiltration (2,300,000 Da)
and ultrafiltration (300,000 Da) membranes cannot remove liner (325,000 Da), and plasmid (2,172,000
Da) DNA, while ultrafiltration membranes with molecular weight cut-off of 20,000 Da can remove both by
46

approximately 3.5 logs. Retention efficiency of free liner and plasmid DNA by ultrafiltration membrane
was 89.4%, and 99.9%, respectively, implying that liner DNA can pass easier through pores than plasmid
DNA. Passage of free liner and plasmid DNA through pores smaller than their size is due to the elongation
and deformation.
Because the membrane barrier removes ARGs selectively (based on the size of the iARG host bacteria,
and also the size of MGEs carrying eARGs), it can drastically change the ARG profile structure of the
permeate side as compared to the feed side. For instance, it has been reported that membrane filtration
can decrease extracellular sul2 with a higher efficiency compared to extracellular sul1,
39, 108
likely because
MGEs carrying sul2 are larger than those that carry sul1.
112
As a result, the effluent ARG profile of aerobic
or anaerobic MBRs is remarkably distinct from that of its biomass.
106, 108
Wang et al.
49
reported that the
average absolute abundance of targeted eARGs (blaTEM, ermB, tetO, and tetW), and extracellular intI1 in
the influent of five full-scale MBRs (with membrane pore size of < 0.4 um)was around 5.6  10
13
copies/mL,
which decreased to 4.5  10
8
in MBR effluents. Extracellular blaTEM was removed by the highest rate
among all targeted genes, while intI1 had the lowest removal rate. Interestingly MBRs have been reported
that are more effective on removal of eARGs than iARGs.
39, 49
Higher removal of eARGs compared to iARGs
in MBRs might be due to fouling layer attached to the membrane surface, which can considerably adsorb
eARGs. Cheng et al.
113
reported that membrane fouling increased the removal rate of plasmid-borne
ARGs, while it decreased ARB removal. The membrane fouling layer, due to the high concentrations of
extracellular polymeric substances (EPSs), and soluble microbial products (SMPs),
113-115
which both have
a high binding ability, can effectively adsorb eDNA and consequently increase eARG removal. It has been
previously reported that the absolute abundance of eARGs in the fouling layer attached to the membrane
was much higher than that of in the biomass in a sequencing-batch MBR treating swine manure.
39

47

Given that ARG removal achieved by the membrane barrier is mainly due to size exclusion, increasing
the size of particles and collides may improve ARG removal. Li et al.
116
used an integrated process including
coagulation (by FeCl 3) followed by microfiltration (pore size of 0.22 um) to remove ARGs from a WWTP
effluent, and reported removal of total ARGs via coagulation, microfiltration, and the integrated process
(coagulation followed by microfiltration) of 1.5, 2.7, and 4 logs, respectively. For eARGs, microfiltration
removal efficiency was only 1.3 logs, while coagulation decreased eARGs by 3.1 logs. Combining both
processes resulted in eARG reduction of 6 logs. Coagulation neutralizes the surface charge of particles and
colloids, resulting in sludge aggregation.
117
Negatively charged eDNA (phosphate group) binds with
positively charged Fe hydroxide and is entrapped in sludge flocs during the flocculation process. Large
flocs are subsequently rejected by the membrane during microfiltration due to size exclusion, resulting in
high removal efficiency of eARGs. However, this mitigation strategy only transfers ARGs from wastewater
to sludge, which increases the risk of antibiotic resistance spread via biosolids land application.
2.5 Conclusion and Future Research Directions
This review revealed that eARGs may be as important as iARGs in dissemination of antibiotic
resistance in the environment. Although iARGs are the main fraction of ARGs in waste streams, eARGs are
predominant in receiving aquatic environments, like river and marine sediments. In regard to prevalence
of iARGs versus eARGs in different environments, most previous studies have relied on qPCR to quantify
these fractions of ARGs. Therefore, more metagenomics-based studies are needed to better understand
the structure and diversity of both iARG and eARG profiles in different environments. Abundances of both
iARGs and eARGs in river and marine environments (which do not provide conditions as suitable as waste
streams for bacterial growth) are not stable and change significantly over time. iARG abundance decreases
due to death of their bacterial hosts. eARG abundance, however, increases through release of iARGs from
dead or living cells. Therefore, one key question is in regard to the duration for which eARGs can persist
48

in different environments. Experiments for evaluating persistence of eARGs in different environments
have been previously conducted via storage of WWTP effluent, however, experiments must also be
carried out with pure isolates of eARGs to avoid eARG enrichment through release of iARGs. Although
release of iARGs from bacteria in natural environments is a common phenomenon, endurance
experiments with pure isolates of eARGs would indicate the actual perseverance of eARGs in different
environments which aids in assessment of the potential risks associated with each fraction of ARGs. This
can then be furthered by considering the release rate of iARGs and the uptake rate of eARGs in different
environments. Another key research direction is employing different methods such as metagenomics and
genomic cross-linking to identify the bacterial cells from which each specific eARG is released.
118
This could
provide insight into the secretion rate of different ARGs from their bacterial hosts.
Previous studies revealed that, although conjugation of iARGs is the dominant HGT mechanism in
WWTPs, natural transformation of eARGs plays a significant role in the spread of antibiotic resistance in
receiving aquatic environments. Nevertheless, no practical experiments have been developed so far to
indicate the rate of natural transformation of eARGs considering different environmental conditions.
Another important research gap here is identifying the bacterial species capable of competence
development. Phages are also a significant reservoir of specific ARGs, which can spread antibiotic
resistance among different species through transduction. The association between different phages and
bacterial species, however, is understudied. Phages can also lyse infectious bacteria, resulting in release
of iARGs, and consequent enrichment of eARGs. Therefore, the relationship between the abundance of
phages and eARGs is an important research direction for future studies. It also remains unclear why some
specific ARGs (e.g., blaTEM) are more prevalent than others among phages. Elucidating the association
between phages and ARGs can provide insights into this matter. Further, to fairly compare frequency of
the three HGT mechanisms (conjugation, transduction, and natural transformation), experiments need to
be carried out at the same conditions (e.g., same mating assay and incubation condition). Performing
49

these experiments in real environments by employing emerging molecular biology methods for tracking
genes could indicate the actual frequency of each HGT mechanism. Another interesting research idea is
the investigation of the simultaneous transfer of ARGs through conjugation, transduction, and natural
transformation, to evaluate how these HGT mechanisms may affect each other.
eARGs and iARGs, according to previous studies, face different fates in WWTPs, suggesting a need
for different approaches to mitigate their release both during wastewater treatment. Membrane fouling
is an effective strategy for eARG removal, while chlorination with low exposure usually results in
enrichment of eARGs. Further, UV disinfection is more efficient for eARG inactivation than iARG.
Therefore, in order to effectively remove both iARGs and eARGs from waste streams, a combination of
different technologies is required. For example, it has been reported that combination of microfiltration,
reverse osmosis (RO), and UV/H2O2 can reduce all targeted iARGs (up to 10 logs) to bellow detection limit
in an advanced water treatment facility treating WWTP effluent.
107
RO has been also observed to remove
eARGs by 7 logs,
102
implying that it is a good approach to combine with other advanced treatment
technologies like advanced oxidation for complete iARG and eARG removal. The combination of
adsorption and photocatalyst degradation have been also reported to significantly reduce eARGs.
119
As it
was mentioned previously, combination of coagulation and microfiltration can also remove both iARGs
(up to 4 logs) and eARGs (up to 6 logs). Beside these approaches, comprehensive research assessing the
risk of iARGs versus eARGs in downstream environments, considering prevalence/persistence and
horizontal transfer of different fraction of ARGs, is urgently needed. With a better understanding of the
relative risk of iARGs and eARGs, we can begin to design and operate treatment systems to reduce the
overall risk of ARG proliferation.
50

References
1. WHO, No time to wait: securing the future from drug-resistant infections. Report to the Secretary-
General of the United Nations. World Health Organization, Geneva 2019.
2. Theuretzbacher, U., Global antibacterial resistance: The never-ending story. Journal of global
antimicrobial resistance 2013, 1, (2), 63-69.
3. Molnar, A., Antimicrobial resistance awareness and games. Trends in microbiology 2019, 27, (1), 1-3.
4. Martínez, J. L., Antibiotics and antibiotic resistance genes in natural environments. Science 2008, 321,
(5887), 365-367.
5. Martinez, J. L., Environmental pollution by antibiotics and by antibiotic resistance determinants.
Environ. Pollut. 2009, 157, (11), 2893-2902.
6. Baquero, F.; Martínez, J.-L.; Cantón, R., Antibiotics and antibiotic resistance in water environments.
Curr. Opin. Biotechnol. 2008, 19, (3), 260-265.
7. CDC., Antibiotic Resistance Threats in the United States. U.S. Department of Health and Human Services
2019.
8. Fair, R. J.; Tor, Y., Antibiotics and bacterial resistance in the 21st century. Perspectives in medicinal
chemistry 2014, 6, PMC. S14459.
9. Normark, B. H.; Normark, S., Evolution and spread of antibiotic resistance. Journal of internal medicine
2002, 252, (2), 91-106.
10. Karthikeyan, K.; Meyer, M. T., Occurrence of antibiotics in wastewater treatment facilities in
Wisconsin, USA. Sci. Total Environ. 2006, 361, (1-3), 196-207.
11. Van Boeckel, T. P.; Gandra, S.; Ashok, A.; Caudron, Q.; Grenfell, B. T.; Levin, S. A.; Laxminarayan, R.,
Global antibiotic consumption 2000 to 2010: an analysis of national pharmaceutical sales data. The Lancet
infectious diseases 2014, 14, (8), 742-750.
12. Pruden, A.; Pei, R.; Storteboom, H.; Carlson, K. H., Antibiotic resistance genes as emerging
contaminants: studies in northern Colorado. Environ. Sci. Technol. 2006, 40, (23), 7445-7450.
13. Lopatkin, A. J.; Huang, S.; Smith, R. P.; Srimani, J. K.; Sysoeva, T. A.; Bewick, S.; Karig, D. K.; You, L.,
Antibiotics as a selective driver for conjugation dynamics. Nature microbiology 2016, 1, (6), 1-8.
14. Berendonk, T. U.; Manaia, C. M.; Merlin, C.; Fatta-Kassinos, D.; Cytryn, E.; Walsh, F.; Bürgmann, H.;
Sørum, H.; Norström, M.; Pons, M.-N., Tackling antibiotic resistance: the environmental framework.
Nature Reviews Microbiology 2015, 13, (5), 310.
15. Zhang, Y.; Snow, D. D.; Parker, D.; Zhou, Z.; Li, X., Intracellular and extracellular antimicrobial resistance
genes in the sludge of livestock waste management structures. Environ. Sci. Technol. 2013, 47, (18),
10206-10213.
16. Mao, D.; Luo, Y.; Mathieu, J.; Wang, Q.; Feng, L.; Mu, Q.; Feng, C.; Alvarez, P., Persistence of
extracellular DNA in river sediment facilitates antibiotic resistance gene propagation. Environ. Sci. Technol.
2014, 48, (1), 71-78.
17. Yuan, Q.-B.; Huang, Y.-M.; Wu, W.-B.; Zuo, P.; Hu, N.; Zhou, Y.-Z.; Alvarez, P. J., Redistribution of
intracellular and extracellular free & adsorbed antibiotic resistance genes through a wastewater
51

treatment plant by an enhanced extracellular DNA extraction method with magnetic beads. Environment
international 2019, 131, 104986.
18. Lorenz, M. G.; Wackernagel, W., Natural genetic transformation of Pseudomonas stutzeri by sand-
adsorbed DNA. Arch. Microbiol. 1990, 154, (4), 380-385.
19. Zhang, Y.; Li, A.; Dai, T.; Li, F.; Xie, H.; Chen, L.; Wen, D., Cell-free DNA: a neglected source for antibiotic
resistance genes spreading from WWTPs. Environ. Sci. Technol. 2018, 52, (1), 248-257.
20. Hao, H.; Shi, D.-y.; Yang, D.; Yang, Z.-w.; Qiu, Z.-g.; Liu, W.-l.; Shen, Z.-q.; Yin, J.; Wang, H.-r.; Li, J.-w.,
Profiling of intracellular and extracellular antibiotic resistance genes in tap water. J. Hazard. Mater. 2019,
365, 340-345.
21. Yuan, K.; Wang, X.; Chen, X.; Zhao, Z.; Fang, L.; Chen, B.; Jiang, J.; Luan, T.; Chen, B., Occurrence of
antibiotic resistance genes in extracellular and intracellular DNA from sediments collected from two types
of aquaculture farms. Chemosphere 2019, 234, 520-527.
22. Zhang, Y.; Niu, Z.; Zhang, Y.; Zhang, K., Occurrence of intracellular and extracellular antibiotic
resistance genes in coastal areas of Bohai Bay (China) and the factors affecting them. Environ. Pollut. 2018,
236, 126-136.
23. Barnes, M. A.; Turner, C. R.; Jerde, C. L.; Renshaw, M. A.; Chadderton, W. L.; Lodge, D. M.,
Environmental conditions influence eDNA persistence in aquatic systems. Environ. Sci. Technol. 2014, 48,
(3), 1819-1827.
24. Nielsen, K. M.; Johnsen, P. J.; Bensasson, D.; Daffonchio, D., Release and persistence of extracellular
DNA in the environment. Environmental biosafety research 2007, 6, (1-2), 37-53.
25. von Wintersdorff, C. J.; Penders, J.; van Niekerk, J. M.; Mills, N. D.; Majumder, S.; van Alphen, L. B.;
Savelkoul, P. H.; Wolffs, P. F., Dissemination of antimicrobial resistance in microbial ecosystems through
horizontal gene transfer. Frontiers in microbiology 2016, 7, 173.
26. Miller, J. H.; Novak, J. T.; Knocke, W. R.; Pruden, A., Survival of antibiotic resistant bacteria and
horizontal gene transfer control antibiotic resistance gene content in anaerobic digesters. Frontiers in
microbiology 2016, 7, 263.
27. Thomas, C. M.; Nielsen, K. M., Mechanisms of, and barriers to, horizontal gene transfer between
bacteria. Nature reviews microbiology 2005, 3, (9), 711.
28. Burrus, V.; Pavlovic, G.; Decaris, B.; Guédon, G., Conjugative transposons: the tip of the iceberg. Mol.
Microbiol. 2002, 46, (3), 601-610.
29. Bergman, J.; Fineran, P.; Petty, N.; Salmond, G., Transduction: The Transfer of Host DNA by
Bacteriophages. 2019.
30. Jiang, S. C.; Paul, J. H., Gene transfer by transduction in the marine environment. Appl. Environ.
Microbiol. 1998, 64, (8), 2780-2787.
31. Johnsborg, O.; Eldholm, V.; Håvarstein, L. S., Natural genetic transformation: prevalence, mechanisms
and function. Res. Microbiol. 2007, 158, (10), 767-778.
32. Lu, N.; Zilles, J. L.; Nguyen, T. H., Adsorption of extracellular chromosomal DNA and its effects on
natural transformation of Azotobacter vinelandii. Appl. Environ. Microbiol. 2010, 76, (13), 4179-4184.
33. Dubnau, D., DNA uptake in bacteria. Annual Reviews in Microbiology 1999, 53, (1), 217-244.
52

34. Palmen, R.; Hellingwerf, K. J., Uptake and processing of DNA by Acinetobacter calcoaceticus–a review.
Gene 1997, 192, (1), 179-190.
35. Eichmiller, J. J.; Miller, L. M.; Sorensen, P. W., Optimizing techniques to capture and extract
environmental DNA for detection and quantification of fish. Molecular ecology resources 2016, 16, (1),
56-68.
36. Lopatkin, A. J.; Meredith, H. R.; Srimani, J. K.; Pfeiffer, C.; Durrett, R.; You, L., Persistence and reversal
of plasmid-mediated antibiotic resistance. Nature communications 2017, 8, (1), 1-10.
37. Lili, L. N.; Britton, N. F.; Feil, E. J., The persistence of parasitic plasmids. Genetics 2007, 177, (1), 399-
405.
38. Hall, J. P.; Wood, A. J.; Harrison, E.; Brockhurst, M. A., Source–sink plasmid transfer dynamics maintain
gene mobility in soil bacterial communities. Proceedings of the National Academy of Sciences 2016, 113,
(29), 8260-8265.
39. Sui, Q.; Chen, Y.; Yu, D.; Wang, T.; Hai, Y.; Zhang, J.; Chen, M.; Wei, Y., Fates of intracellular and
extracellular antibiotic resistance genes and microbial community structures in typical swine wastewater
treatment processes. Environment international 2019, 133, 105183.
40. Dong, P.; Wang, H.; Fang, T.; Wang, Y.; Ye, Q., Assessment of extracellular antibiotic resistance genes
(eARGs) in typical environmental samples and the transforming ability of eARG. Environment international
2019, 125, 90-96.
41. Zhao, Z.; Zhang, K.; Wu, N.; Li, W.; Xu, W.; Zhang, Y.; Niu, Z., Estuarine sediments are key hotspots of
intracellular and extracellular antibiotic resistance genes: A high-throughput analysis in Haihe Estuary in
China. Environment International 2020, 135, 105385.
42. Guo, X.-p.; Yang, Y.; Lu, D.-p.; Niu, Z.-s.; Feng, J.-n.; Chen, Y.-r.; Tou, F.-y.; Garner, E.; Xu, J.; Liu, M.,
Biofilms as a sink for antibiotic resistance genes (ARGs) in the Yangtze Estuary. Water Res. 2018, 129, 277-
286.
43. Blum, S. A.; Lorenz, M. G.; Wackernagel, W., Mechanism of retarded DNA degradation and prokaryotic
origin of DNases in nonsterile soils. Systematic and applied microbiology 1997, 20, (4), 513-521.
44. Nagler, M.; Insam, H.; Pietramellara, G.; Ascher-Jenull, J., Extracellular DNA in natural environments:
features, relevance and applications. Applied microbiology and biotechnology 2018, 102, (15), 6343-6356.
45. Nagler, M.; Podmirseg, S. M.; Griffith, G. W.; Insam, H.; Ascher-Jenull, J., The use of extracellular DNA
as a proxy for specific microbial activity. Applied microbiology and biotechnology 2018, 102, (6), 2885-
2898.
46. Poly, F.; Chenu, C.; Simonet, P.; Rouiller, J.; Jocteur Monrozier, L., Differences between linear
chromosomal and supercoiled plasmid DNA in their mechanisms and extent of adsorption on clay
minerals. Langmuir 2000, 16, (3), 1233-1238.
47. Ogram, A. V.; Mathot, M. L.; Harsh, J. B.; Boyle, J.; Pettigrew, C. A., Effects of DNA polymer length on
its adsorption to soils. Appl. Environ. Microbiol. 1994, 60, (2), 393-396.
48. Ogram, A.; Sayler, G. S.; Barkay, T., The extraction and purification of microbial DNA from sediments.
J. Microbiol. Methods 1987, 7, (2-3), 57-66.
49. Wang, S.; Ma, X.; Liu, Y.; Yi, X.; Du, G.; Li, J., Fate of antibiotics, antibiotic-resistant bacteria, and cell-
free antibiotic-resistant genes in full-scale membrane bioreactor wastewater treatment plants.
Bioresource Technology 2020, 302, 122825.
53

50. Tomova, A.; Ivanova, L.; Buschmann, A. H.; Godfrey, H. P.; Cabello, F. C., Plasmid-mediated quinolone
resistance (PMQR) genes and class 1 integrons in quinolone-resistant marine bacteria and clinical isolates
of Escherichia coli from an aquacultural area. Microbial ecology 2018, 75, (1), 104-112.
51. Tomova, A.; Ivanova, L.; Buschmann, A. H.; Rioseco, M. L.; Kalsi, R. K.; Godfrey, H. P.; Cabello, F. C.,
Antimicrobial resistance genes in marine bacteria and human uropathogenic E scherichia coli from a
region of intensive aquaculture. Environmental microbiology reports 2015, 7, (5), 803-809.
52. Zhou, S.; Zhu, Y.; Yan, Y.; Wang, W.; Wang, Y., Deciphering extracellular antibiotic resistance genes
(eARGs) in activated sludge by metagenome. Water Res. 2019, 161, 610-620.
53. Nakamura, Y.; Itoh, T.; Matsuda, H.; Gojobori, T., Biased biological functions of horizontally transferred
genes in prokaryotic genomes. Nature genetics 2004, 36, (7), 760-766.
54. Chandrasekaran, S.; Venkatesh, B.; Lalithakumari, D., Transfer and expression of a multiple antibiotic
resistance plasmid in marine bacteria. Current Microbiology 1998, 37, (5), 347-351.
55. Licht, T. R.; Wilcks, A., Conjugative gene transfer in the gastrointestinal environment. Adv. Appl.
Microbiol. 2005, 58, 77-95.
56. Chen, B.; Yuan, K.; Chen, X.; Yang, Y.; Zhang, T.; Wang, Y.; Luan, T.; Zou, S.; Li, X., Metagenomic analysis
revealing antibiotic resistance genes (ARGs) and their genetic compartments in the Tibetan environment.
Environ. Sci. Technol. 2016, 50, (13), 6670-6679.
57. Guo, J.; Li, J.; Chen, H.; Bond, P. L.; Yuan, Z., Metagenomic analysis reveals wastewater treatment
plants as hotspots of antibiotic resistance genes and mobile genetic elements. Water Res. 2017, 123, 468-
478.
58. Inoue, D.; Sei, K.; Soda, S.; Ike, M.; Fujita, M., Potential of predominant activated sludge bacteria as
recipients in conjugative plasmid transfer. Journal of bioscience and bioengineering 2005, 100, (6), 600-
605.
59. Jacquiod, S.; Brejnrod, A.; Morberg, S. M.; Abu Al‐Soud, W.; Sørensen, S. J.; Riber, L., Deciphering
conjugative plasmid permissiveness in wastewater microbiomes. Molecular ecology 2017, 26, (13), 3556-
3571.
60. Li, B.; Qiu, Y.; Zhang, J.; Liang, P.; Huang, X., Conjugative potential of antibiotic resistance plasmids to
activated sludge bacteria from wastewater treatment plants. International biodeterioration &
biodegradation 2019, 138, 33-40.
61. Li, L.; Dechesne, A.; He, Z.; Madsen, J. S.; Nesme, J.; Sørensen, S. J.; Smets, B. F., Estimating the transfer
range of plasmids encoding antimicrobial resistance in a wastewater treatment plant microbial
community. Environmental Science & Technology Letters 2018, 5, (5), 260-265.
62. Dang, B.; Mao, D.; Xu, Y.; Luo, Y., Conjugative multi-resistant plasmids in Haihe River and their impacts
on the abundance and spatial distribution of antibiotic resistance genes. Water Res. 2017, 111, 81-91.
63. Yang, D.; Wang, J.; Qiu, Z.; Jin, M.; Shen, Z.; Chen, Z.; Wang, X.; Zhang, B.; Li, J.-W., Horizontal transfer
of antibiotic resistance genes in a membrane bioreactor. J. Biotechnol. 2013, 167, (4), 441-447.
64. Grabow, W.; Prozesky, O.; Burger, J., Behaviour in a river and dam of coliform bacteria with
transferable or non-transferable drug resistance. Water Res. 1975, 9, (9), 777-782.
65. Ohlsen, K.; Ternes, T.; Werner, G.; Wallner, U.; Löffler, D.; Ziebuhr, W.; Witte, W.; Hacker, J., Impact of
antibiotics on conjugational resistance gene transfer in Staphylococcus aureus in sewage. Environ.
Microbiol. 2003, 5, (8), 711-716.
54

66. Zhang, S.; Wang, Y.; Song, H.; Lu, J.; Yuan, Z.; Guo, J., Copper nanoparticles and copper ions promote
horizontal transfer of plasmid-mediated multi-antibiotic resistance genes across bacterial genera.
Environment international 2019, 129, 478-487.
67. Lu, J.; Wang, Y.; Jin, M.; Yuan, Z.; Bond, P.; Guo, J., Both silver ions and silver nanoparticles facilitate
the horizontal transfer of plasmid-mediated antibiotic resistance genes. Water Res. 2020, 169, 115229.
68. Binh, C. T. T.; Heuer, H.; Kaupenjohann, M.; Smalla, K., Piggery manure used for soil fertilization is a
reservoir for transferable antibiotic resistance plasmids. FEMS Microbiol. Ecol. 2008, 66, (1), 25-37.
69. Guan, J.; Wasty, A.; Grenier, C.; Chan, M., Influence of temperature on survival and conjugative
transfer of multiple antibiotic-resistant plasmids in chicken manure and compost microcosms. Poultry
science 2007, 86, (4), 610-613.
70. Brown-Jaque, M.; Calero-Cáceres, W.; Muniesa, M., Transfer of antibiotic-resistance genes via phage-
related mobile elements. Plasmid 2015, 79, 1-7.
71. Modi, S. R.; Lee, H. H.; Spina, C. S.; Collins, J. J., Antibiotic treatment expands the resistance reservoir
and ecological network of the phage metagenome. Nature 2013, 499, (7457), 219-222.
72. LANDÉN, R.; HEIERSON, A.; BOMAN, H. G., A phage for generalized transduction in Bacillus
thuringiensis and mapping of four genes for antibiotic resistance. Microbiology 1981, 123, (1), 49-59.
73. Varga, M.; Kuntová, L.; Pantůček, R.; Mašlaňová, I.; Růžičková, V.; Doškař, J., Efficient transfer of
antibiotic resistance plasmids by transduction within methicillin-resistant Staphylococcus aureus USA300
clone. FEMS Microbiol. Lett. 2012, 332, (2), 146-152.
74. Volkova, V. V.; Lu, Z.; Besser, T.; Gröhn, Y. T., Modeling the infection dynamics of bacteriophages in
enteric Escherichia coli: estimating the contribution of transduction to antimicrobial gene spread. Appl.
Environ. Microbiol. 2014, 80, (14), 4350-4362.
75. Golkar, Z.; Bagasra, O.; Pace, D. G., Bacteriophage therapy: a potential solution for the antibiotic
resistance crisis. The Journal of Infection in Developing Countries 2014, 8, (02), 129-136.
76. Kakasis, A.; Panitsa, G., Bacteriophage therapy as an alternative treatment for human infections. A
comprehensive review. International journal of antimicrobial agents 2019, 53, (1), 16-21.
77. Lekunberri, I.; Villagrasa, M.; Balcázar, J. L.; Borrego, C. M., Contribution of bacteriophage and plasmid
DNA to the mobilization of antibiotic resistance genes in a river receiving treated wastewater discharges.
Sci. Total Environ. 2017, 601, 206-209.
78. Calero-Cáceres, W.; Balcázar, J. L., Antibiotic resistance genes in bacteriophages from diverse marine
habitats. Sci. Total Environ. 2019, 654, 452-455.
79. Colomer-Lluch, M.; Jofre, J.; Muniesa, M., Antibiotic resistance genes in the bacteriophage DNA
fraction of environmental samples. PloS one 2011, 6, (3).
80. Wang, M.; Liu, P.; Zhou, Q.; Tao, W.; Sun, Y.; Zeng, Z., Estimating the contribution of bacteriophage to
the dissemination of antibiotic resistance genes in pig feces. Environ. Pollut. 2018, 238, 291-298.
81. Calero-Cáceres, W.; Melgarejo, A.; Colomer-Lluch, M.; Stoll, C.; Lucena, F.; Jofre, J.; Muniesa, M.,
Sludge as a potential important source of antibiotic resistance genes in both the bacterial and
bacteriophage fractions. Environ. Sci. Technol. 2014, 48, (13), 7602-7611.
82. Kenzaka, T.; Tani, K.; Nasu, M., High-frequency phage-mediated gene transfer in freshwater
environments determined at single-cell level. The ISME journal 2010, 4, (5), 648-659.
55

83. Stewart, G. J.; Sinigalliano, C. D., Detection of horizontal gene transfer by natural transformation in
native and introduced species of bacteria in marine and synthetic sediments. Appl. Environ. Microbiol.
1990, 56, (6), 1818-1824.
84. Zawadzki, P.; Cohan, F. M., The size and continuity of DNA segments integrated in Bacillus
transformation. Genetics 1995, 141, (4), 1231-1243.
85. Nielsen, K. M.; Van Weerelt, M.; Berg, T. N.; Bones, A. M.; Hagler, A. N.; Van Elsas, J., Natural
transformation and availability of transforming DNA to Acinetobacter calcoaceticus in soil microcosms.
Appl. Environ. Microbiol. 1997, 63, (5), 1945-1952.
86. Naquin, A.; Shrestha, A.; Sherpa, M.; Nathaniel, R.; Boopathy, R., Presence of antibiotic resistance
genes in a sewage treatment plant in Thibodaux, Louisiana, USA. Bioresource technology 2015, 188, 79-
83.
87. Qingxiang, Y.; Qiang, W.; Kaiyue, Z., EFFECTS OF ANTIBIOTICS AND METAL IONS EXPOSURE ON THE
NATURAL TRANSFORMATION FREQUENCY OF AN ANTIBIOTIC RESISTANT PLASMID. FEB-FRESENIUS
ENVIRONMENTAL BULLETIN 2017, 5732.
88. Lu, J.; Wang, Y.; Zhang, S.; Bond, P.; Yuan, Z.; Guo, J., Triclosan at environmental concentrations can
enhance the spread of extracellular antibiotic resistance genes through transformation. Sci. Total Environ.
2020, 713, 136621.
89. Wu, H.-y.; Shi, D.-y.; Yang, D.; Yin, J.; Yang, Z.-w.; Li, J.-w.; Yang, W.; Jin, M., Putative environmental
levels of levofloxacin facilitate the dissemination of antibiotic-resistant Escherichia coli via plasmid-
mediated transformability. Ecotoxicology and Environmental Safety 2020, 195, 110461.
90. Shun-Mei, E.; Zeng, J.-M.; Yuan, H.; Lu, Y.; Cai, R.-X.; Chen, C., Sub-inhibitory concentrations of
fluoroquinolones increase conjugation frequency. Microbial pathogenesis 2018, 114, 57-62.
91. Di Cesare, A.; Eckert, E.; Corno, G., Co-selection of antibiotic and heavy metal resistance in freshwater
bacteria. J. Limnol 2016, 75, (Suppl 2), 59-66.
92. Ji, X.; Shen, Q.; Liu, F.; Ma, J.; Xu, G.; Wang, Y.; Wu, M., Antibiotic resistance gene abundances
associated with antibiotics and heavy metals in animal manures and agricultural soils adjacent to feedlots
in Shanghai; China. J. Hazard. Mater. 2012, 235, 178-185.
93. Fothergill, J. L.; Mowat, E.; Walshaw, M. J.; Ledson, M. J.; James, C. E.; Winstanley, C., Effect of
antibiotic treatment on bacteriophage production by a cystic fibrosis epidemic strain of Pseudomonas
aeruginosa. Antimicrob. Agents Chemother. 2011, 55, (1), 426-428.
94. Bielaszewska, M.; Idelevich, E. A.; Zhang, W.; Bauwens, A.; Schaumburg, F.; Mellmann, A.; Peters, G.;
Karch, H., Effects of antibiotics on Shiga toxin 2 production and bacteriophage induction by epidemic
Escherichia coli O104: H4 strain. Antimicrob. Agents Chemother. 2012, 56, (6), 3277-3282.
95. Ding, C.; Pan, J.; Jin, M.; Yang, D.; Shen, Z.; Wang, J.; Zhang, B.; Liu, W.; Fu, J.; Guo, X., Enhanced uptake
of antibiotic resistance genes in the presence of nanoalumina. Nanotoxicology 2016, 10, (8), 1051-1060.
96. Liu, S.-S.; Qu, H.-M.; Yang, D.; Hu, H.; Liu, W.-L.; Qiu, Z.-G.; Hou, A.-M.; Guo, J.; Li, J.-W.; Shen, Z.-Q.,
Chlorine disinfection increases both intracellular and extracellular antibiotic resistance genes in a full-
scale wastewater treatment plant. Water Res. 2018, 136, 131-136.
97. Mantilla-Calderon, D.; Plewa, M. J.; Michoud, G. g.; Fodelianakis, S.; Daffonchio, D.; Hong, P.-Y., Water
disinfection byproducts increase natural transformation rates of environmental DNA in Acinetobacter
baylyi ADP1. Environ. Sci. Technol. 2019, 53, (11), 6520-6528.
56

98. McKinney, C. W.; Pruden, A., Ultraviolet disinfection of antibiotic resistant bacteria and their antibiotic
resistance genes in water and wastewater. Environ. Sci. Technol. 2012, 46, (24), 13393-13400.
99. Lin, W.; Zhang, M.; Zhang, S.; Yu, X., Can chlorination co-select antibiotic-resistance genes?
Chemosphere 2016, 156, 412-419.
100. Guo, M.-T.; Yuan, Q.-B.; Yang, J., Distinguishing effects of ultraviolet exposure and chlorination on
the horizontal transfer of antibiotic resistance genes in municipal wastewater. Environ. Sci. Technol. 2015,
49, (9), 5771-5778.
101. Yoon, Y.; Chung, H. J.; Di, D. Y. W.; Dodd, M. C.; Hur, H.-G.; Lee, Y., Inactivation efficiency of plasmid-
encoded antibiotic resistance genes during water treatment with chlorine, UV, and UV/H2O2. Water Res.
2017, 123, 783-793.
102. Krzeminski, P.; Feys, E.; d'Auriac, M. A.; Wennberg, A. C.; Umar, M.; Schwermer, C. U.; Uhl, W.,
Combined membrane filtration and 265 nm UV irradiation for effective removal of cell free antibiotic
resistance genes from feed water and concentrate. Journal of Membrane Science 2020, 598, 117676.
103. Czekalski, N.; Imminger, S.; Salhi, E.; Veljkovic, M.; Kleffel, K.; Drissner, D.; Hammes, F.; Burgmann,
H.; Von Gunten, U., Inactivation of antibiotic resistant bacteria and resistance genes by ozone: from
laboratory experiments to full-scale wastewater treatment. Environ. Sci. Technol. 2016, 50, (21), 11862-
11871.
104. He, H.; Zhou, P.; Shimabuku, K. K.; Fang, X.; Li, S.; Lee, Y.; Dodd, M. C., Degradation and deactivation
of bacterial antibiotic resistance genes during exposure to free chlorine, monochloramine, chlorine
dioxide, ozone, ultraviolet light, and hydroxyl radical. Environ. Sci. Technol. 2019, 53, (4), 2013-2026.
105. Augsburger, N.; Mantilla-Calderon, D.; Daffonchio, D.; Hong, P.-Y., Acquisition of Extracellular DNA
by Acinetobacter baylyi ADP1 in response to Solar and UV-C254nm Disinfection. Environ. Sci. Technol.
2019, 53, (17), 10312-10319.
106. Baygi, A. Z.; Harb, M.; Wang, P.; Stadler, L.; Smith, A. L., Microbial community and antibiotic resistance
profiles of biomass and effluent are distinctly affected by antibiotics addition to an anaerobic membrane
bioreactor. Environmental Science: Water Research & Technology 2020.
107. Harb, M.; Wang, P.; Zarei-Baygi, A.; Plumlee, M. H.; Smith, A., Background antibiotic resistance and
microbial communities dominate effects of advanced purified water recharge to an urban aquifer.
Environmental Science & Technology Letters 2019.
108. Zarei-Baygi, A.; Harb, M.; Wang, P.; Stadler, L. B.; Smith, A. L., Evaluating Antibiotic Resistance Gene
Correlations with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors. Environ. Sci.
Technol. 2019, 53, (7), 3599-3609.
109. Munir, M.; Wong, K.; Xagoraraki, I., Release of antibiotic resistant bacteria and genes in the effluent
and biosolids of five wastewater utilities in Michigan. Water Res. 2011, 45, (2), 681-693.
110. Latulippe, D. R.; Ager, K.; Zydney, A. L., Flux-dependent transmission of supercoiled plasmid DNA
through ultrafiltration membranes. Journal of membrane science 2007, 294, (1-2), 169-177.
111. Slipko, K.; Reif, D.; Wögerbauer, M.; Hufnagl, P.; Krampe, J.; Kreuzinger, N., Removal of extracellular
free DNA and antibiotic resistance genes from water and wastewater by membranes ranging from
microfiltration to reverse osmosis. Water Res. 2019, 164, 114916.
112. Sköld, O., Sulfonamide resistance: mechanisms and trends. Drug resistance updates 2000, 3, (3), 155-
160.
57

113. Cheng, H.; Hong, P.-Y., Removal of antibiotic-resistant bacteria and antibiotic resistance genes
affected by varying degrees of fouling on anaerobic microfiltration membranes. Environ. Sci. Technol.
2017, 51, (21), 12200-12209.
114. Chen, R.; Nie, Y.; Hu, Y.; Miao, R.; Utashiro, T.; Li, Q.; Xu, M.; Li, Y.-Y., Fouling behaviour of soluble
microbial products and extracellular polymeric substances in a submerged anaerobic membrane
bioreactor treating low-strength wastewater at room temperature. Journal of Membrane Science 2017,
531, 1-9.
115. Drews, A.; Lee, C.-H.; Kraume, M., Membrane fouling-a review on the role of EPS. Desalination 2006,
200, (1-3), 186-188.
116. Li, Z.-H.; Yuan, L.; Gao, S.-X.; Wang, L.; Sheng, G.-P., Mitigated membrane fouling and enhanced
removal of extracellular antibiotic resistance genes from wastewater effluent via an integrated pre-
coagulation and microfiltration process. Water Res. 2019, 159, 145-152.
117. Zarei-Baygi, A.; Moslemi, M.; Mirzaei, S. H., The combination of KMnO4 oxidation and polymeric
flocculation for the mitigation of membrane fouling in a membrane bioreactor. Sep. Purif. Technol. 2016,
159, 124-134.
118. Rice, E. W.; Wang, P.; Smith, A. L.; Stadler, L. B., Determining Hosts of Antibiotic Resistance Genes: A
Review of Methodological Advances. Environmental Science & Technology Letters 2020, 7, (5), 282-291.
119. Yuan, Q.; Zhang, D.; Yu, P.; Sun, R.; Javed, H.; Wu, G.; Alvarez, P. J., Selective Adsorption and
Photocatalytic Degradation of Extracellular Antibiotic Resistance Genes by Molecularly-Imprinted
Graphitic Carbon Nitride. Environ. Sci. Technol. 2020, 54, (7), 4621-4630.

58

Chapter 3

3. Evaluating Antibiotic Resistance Gene Correlations with Antibiotic
Exposure Conditions in Anaerobic Membrane Bioreactors

Abstract
Anaerobic membrane bioreactors (AnMBRs) are an emerging technology with potential to improve energy
efficiency and effluent reuse of mainstream wastewater treatment. However, their contribution to the
proliferation of contaminants of emerging concern, such as antibiotic resistance genes (ARGs), remains
largely unknown. The purpose of this study was to determine the effect of select influent antibiotics at
varying concentrations on the presence and abundance of ARGs in an AnMBR system and its effluent.
Quantification of targeted ARGs revealed distinct profiles in biomass and effluent, with genes conferring
resistance to different antibiotic classes dominating in biomass (macrolides) and effluent (sulfonamides).
Effluent sul1 gene abundance was strongly correlated with abundance of intl1, signifying the potential
importance of mobile genetic elements in ARG release from AnMBR systems. The addition of specific
antibiotics also affected normalized abundances of their related ARGs, exemplifying the potential impact
of selective pressures at both low (10 µg/L) and high (250 µg/L) influent antibiotic concentrations.

59

3.1 Introduction
Despite the longstanding benefits of antibiotic use for the treatment of infectious diseases, their
unwanted persistence in wastewater has intensified antibiotic resistance in natural microbial
communities.
1-3
Approximately 23,000 deaths in the US and 25,000 in Europe have been attributed to
antibiotic resistant bacteria (ARB), along with approximately $1 billion and $1.5 billion in annual
healthcare costs, respectively.
4, 5
Reports have linked antibiotic loading from hospitals and farms to
increases in resistance within downstream surface water, soil, and groundwater.
6
Water reuse (a more
common practice given persistent water scarcity) elevates such threats by reducing buffers between
wastewater and potential human exposure.
7
These circumstances have created a sense of urgency
worldwide, with antibiotic resistance cited as one of the most critical human health risks.
4
Moreover,
sensitive (nonresistant) bacteria can acquire resistance mechanisms from ARB via the horizontal exchange
of mobile genetic elements (MGEs) containing antibiotic resistance genes (ARGs; e.g., plasmids, integrons,
and transposons).
4

Wastewater treatment systems (both municipal and agricultural) are the primary gateway for antibiotic
release to the environment.
6
The high density of microorganisms in these systems promotes antibiotic
resistance proliferation via vertical and horizontal gene transfer when exposed to sub-lethal levels of
antibiotics.
8, 9
Studies have demonstrated that higher solids retention times (SRTs) increase ARG
abundance in activated sludge of both membrane-based and conventional treatment processes.
10-12

Further, studies have evaluated various antibiotic types relative to abundance of their associated ARG,
with some showing positive correlations and others showing negative correlations or none at all.
13-15

Anaerobic treatment remains significantly understudied with respect to antibiotic resistance.
16

Anaerobic membrane bioreactors (AnMBRs) are emerging as the forefront anaerobic technology for
mainstream treatment, promoting both energy recovery and agricultural effluent reuse while achieving
60

similar treatment performance to aerobic processes at a range of operational temperatures.
17, 18

Theoretically, AnMBRs can lessen the spread of antibiotic resistance by minimizing excess sludge
production due to the significantly lower yields of anaerobic microorganisms relative to their aerobic
counterparts in activated sludge. Although this inherently results in extended SRTs (shown to increase
ARG abundance in aerobic biomass), longer SRTs in anaerobic systems may reduce ARG presence.
19, 20

Few studies have evaluated AnMBRs for their ARB and ARG reduction capacity. Notably, recent work
by Kappell et al. exhibited log removal values of greater than 3.5 for sul1, ermB, and tetO during AnMBR
treatment of primary clarifier effluent.
21
Further, recent work by Cheng et al. showed that both ARB and
ARG removal was significantly improved by subcritical membrane fouling in an AnMBR, implying that
membrane biofilms may play an integral role in their reduction.
22
However, no studies to date have
systematically considered the presence and concentration of specific antibiotics on ARG or ARB
abundances in AnMBR biomass and effluent, despite their observed correlations in other wastewater
systems. Further, antibiotic concentration is of particular relevance for evaluating system applicability to
different wastewater source types (e.g., domestic vs. hospital wastewater). In the present study, ARGs
and ARB were quantified in an AnMBR system across varying influent antibiotic concentrations for three
different antibiotic types.
3.2 Materials and methods
3.2.1 Bench-scale anaerobic membrane bioreactor operation
A bench-scale AnMBR consisting of a 5 L working-volume continuously stirred tank reactor (CSTR)
(Chemglass Life Science, Vineland, NJ) was operated at 25 °C. Three separate membrane housings were
submerged in the reactor, each including a flat-sheet silicon carbide (ceramic) microfiltration membrane
(Cembrane, Denmark) with 0.1 µm pore size. The effective membrane area of each module was
61

approximately 0.015 m
2
. The AnMBR was inoculated with sludge from a mesophilic anaerobic digester at
the Joint Water Pollution Control Plant (Carson, CA). AnMBR influent was a synthetic wastewater (Table
S2.1) representative of domestic wastewater in the US.
23
The use of synthetic wastewater in this study
allowed for the direct control and observation of specific antibiotic addition without the incidental
influence of background antibiotic, ARB, or ARG occurrence.
After steady AnMBR performance was reached, defined as consistently low effluent COD (< 40 mg/L),
stable biogas production, and high methane content (> 60%) over at least 10 days of operation, three
antibiotics including sulfamethoxazole (sulfonamide), erythromycin (macrolide), and ampicillin (β-
lactam), were added to the influent in independent sequential phases at incremental concentrations of
10, 50, and 250 µg/L (10 days at each concentration) to represent typical antibiotic level ranges in
domestic and hospital wastewater discharges.
24, 25
Before each antibiotic phase, membrane modules were
removed for physical and chemical cleaning using 0.5% (v/v) NaOCl solution to prevent membrane fouling
from influencing observations between phases (details provided in Appx 2). After membrane cleaning, the
AnMBR was operated for one week prior to commencement of the next antibiotic phase. Details regarding
AnMBR operation and performance analysis methods are provided in the Appx 2.
3.2.2 Antibiotic Quantification
All antibiotics were obtained from Sigma-Aldrich (> 99% purity). Quantification of antibiotics was
achieved using matrix-matched external calibration to correct for any suppression or enhancement effects
of the influent and effluent sample matrices. Sample antibiotic concentrations were analyzed by direct
injection liquid chromatography mass spectrometry with electrospray ionization (LC-ESI-MS) on a 6560
Ion Mobility Quadrupole Time-of-Flight (IM-QTOF) LC-MS system (Agilent) using 1290 Infinity UHPLC, Dual
Agilent Jet Stream (ASJ) ESI, and EclipsePlus C18 column (2.1 mm; 50 mm; 1.8 µm). Method detection
limits (MDLs) and practical quantitation limits (PQL) for each targeted compound were estimated using
62

minimum signal to noise ratios of 3:1 and 10:1, respectively. PQLs were < 0.1 µg/L for all three antibiotics
based on compound-specific optimization of LC-ESI-MS conditions. Details of sample preparation
procedures, optimized LC program, and MS operational conditions are presented in the Appx 2.
3.2.3 ARG quantification by qPCR
AnMBR biomass and effluent were both sampled biweekly for DNA extraction. For biomass samples, 2
mL of suspended sludge was centrifuged at 5,000g for 10 minutes, supernatant decanted, and stored at -
80 °C. A freeze dry system (FreeZone 2.5 Liter Freeze Dryer, Labconco, Kansas City, MO) was used to
lyophilize 50 mL effluent samples, which were then stored at -80 °C prior to DNA extraction. DNA
extraction was performed using a Maxwell 16 Blood DNA Purification kit (Promega, Madison, WI)
according to manufacturer specifications. Extracted DNA concentration and quality were evaluated
spectrophotometrically using a BioSpectrometer (Eppendorf, Hamburg, Germany).
Quantitative polymerase chain reaction (qPCR) was performed on a LightCycler 96 (Roche, Basel,
Switzerland) targeting a suite of 9 ARGs spanning a range of antibiotic classes based on the most common
ARGs observed in domestic wastewater in previous studies.
19, 26, 27
Targeted ARGs included genes
conferring resistance to sulfonamides (sul1 and sul2), macrolides (ermF and ermB), β-lactams (oxa-1,
ampC, and mecA), and tetracycline (tetW and tetO), as well as intl1 which encodes for class 1 integrons.
The rpoB gene was also quantified for ARG normalization due to its ubiquity as a single copy gene,
preventing potential biases with multiple 16S rRNA operon copy numbers. qPCR reactions were carried
out in 20 µL reactions containing 10 µL qPCR master mix (Forget-Me-Not EvaGreen, Biotium, Fermont,
CA), forward and reverse primers at a final concentration of 0.25 µM each, 1 µL of DNA template and
ddH 2O. Each reaction was performed in triplicate. Thermal cycling was varied for each ARG targeted, with
details provided in the Appx 2.
63

3.2.4 ARB quantification
Total bacteria and ARB in the effluent were enumerated using the heterotrophic plate count (HPC)
method.
28
Nutrient agar was used for all HPC plating. Ampicillin, tetracycline, erythromycin, and
sulfamethoxazole were added to the nutrient agar at 16, 16, 50.4, and 18.1 µg/mL, respectively, based on
previously used ARB quantification methods.
29, 30
Effluent samples from the permeate of all three
membrane units were collected in a sterile microcentrifuge tube and diluted with 1X PBS to the expected
range necessary to achieve plate counts in the range of 30-300 colony forming units (CFUs) per plate. All
diluted effluent samples were then plated in duplicate. Plates were incubated at 35 °C for 48 hours prior
to CFU enumeration.
3.2.5 Statistical analysis methods
To determine significance of changes in ARG abundance at the different influent antibiotic
concentrations, a 2-tailed unpaired student’s t-test was carried out between all phase-adjacent data sets.
To evaluate for significant linear correlation between data, Pearson correlation was employed using
MAXSTAT Lite 3.6 over a 95% confidence interval. Correlation analysis was performed between all ARGs,
ARB, and antibiotic concentrations over the entire operational period. Strong and weak correlations were
identified based on the Pearson coefficient (ρ) as ρ > 0.7 or ρ < -0.7 for strong correlation and 0.3 < ρ < 0.7
or -0.7 < ρ < -0.3 for weak correlation. The analysis was carried out for biomass and effluent separately.
64

3.3 Results and discussion
3.3.1 AnMBR performance and antibiotic removal was robust throughout
operation
Chemical oxygen demand (COD) removal in the AnMBR averaged 93  3.1% throughout operation
resulting in an effluent COD of 35  17 mg/L. Mixed liquor suspended solids (MLSS) and mixed liquor
volatile suspended solids (MLVSS) concentrations were relatively constant at 8.2  0.5 g/L and 6.6  0.5
g/L, respectively. Average biogas production was 726  39 mL/d with an average methane content of 76.4
 5.7%. Membrane performance was consistent throughout the experiment with an effective
transmembrane flux of 7 L/m
2
/h and transmembrane pressure of < 25 kPa. Sulfamethoxazole and
ampicillin addition to the influent of the AnMBR both coincided with steady increases in biogas methane
content (p < 0.036; Figure 3.1A). Although it was unclear whether this trend was a result of antibiotic
addition, previous work has shown that methanogens can degrade sulfonamide antibiotics as a co-
substrate.
31
However, sulfamethoxazole COD at the concentrations used in the present study was
negligible relative to influent COD. The impact of antibiotics on methanogen activity deserves further
research. Antibiotic removal rates were high for all three of the antibiotics tested (> 67%; Figure 3.1B).
Removal rates for sulfamethoxazole and erythromycin were in the range of 71-85% and 67-88%,
respectively, with no significant trends. Ampicillin removal rates were significantly higher, starting above
94% upon initial addition and increasing to over 98%. These relatively high parent compound removal
rates are in general accordance with previously reported ranges for anaerobic treatment,
32-35
which in the
cases of sulfamethoxazole and ampicillin, signify an advantage over removal efficiencies of conventional
activated sludge-based systems.
36, 37

65

Figure 3.1 (A) Performance of AnMBR in COD removal and biogas production during the addition of
sulfamethoxazole, erythromycin and ampicillin at increasing concentrations; (B) Fate of sulfamethoxazole,
erythromycin and ampicillin in AnMBR. Error bars represent the standard deviation of the results obtained
from replicate samples.
3.3.2 AnMBR effluent exhibited a unique ARG profile compared to the
biomass and membrane biofilm
The ARG profile of the biomass, biofilm, and effluent revealed significant differences independent of
the specific antibiotic addition phases. Exemplifying this, a comparison of biomass and effluent samples
at 250 µg/L sulfamethoxazole in the influent, along with the corresponding membrane biofilm (taken
66

directly after 250 µg/L phase), revealed that biomass and biofilm ARG profiles were similar to each other
and both remarkably different from the effluent (Figure 3.2). Analysis of subsequent antibiotic phases
(erythromycin and ampicillin) revealed a similar relative distribution of ARG profiles across sample type
(Appx 2, Figure S2.1).

Figure 3.2 Abundance of targeted ARGs in the biomass, biofilm (Copy/rpoB) and effluent (Copy/mL) during
the addition of 250 (µg/L) sulfamethoxazole. ARGs are sorted by the biomass abundance (highest to the
lowest). Bar charts and error bars respectively represent the mean values and standard deviations
calculated according to the results from three samples collected on at different days during the addition
of 250 (µg/L) sulfamethoxazole and each of their triplicate qPCR results. For biofilm results, mean values
and standard deviations were calculated according to the triplicate qPCR results.
ermF was the most abundant targeted ARG in the biomass and biofilm. Previous work has also identified
ermF as being among the most abundant ARGs in anaerobic digesters.
38, 39
The predominant ARGs in the
biomass sample (250 µg/L sulfamethoxazole) in order of normalized abundance were ermF, sul1, tetO,
ermB, oxa-1, sul2, tetW, intI1 and ampC. Results indicated that oxa-1, ampC, tetO, and tetW were all
significantly higher in biofilm samples (p < 0.031) compared to their normalized biomass ARG abundance.
These higher abundances in the biofilm could be due to affinity of specific ARGs to the biofilm matrix as
well as potential differences in microbial community structure. For example, tetracycline resistance genes
Biomass Biofilm Effluent
Effluent ARGs (Copy/mL)
0 ×10
0
5 ×10
1
10
2
1.5 ×10
2
2 ×10
2
2.5 ×10
2
10
4
2 ×10
4
3 ×10
4
4 ×10
4
Biomass/Biofilm ARGs (Copy/rpoB)
0
5
10
15
20
25
225
250
275
300
325
ermF sul1 tetO ermB oxa-1 sul2 tetW intI1 ampC
67

have previously been shown to migrate from the water column to drinking water biofilms,
40
while ampC
has been commonly detected in wastewater biofilms.
2

Effluent ARG profiles were vastly different from those of the biomass and biofilm. Perhaps most
surprisingly, ermF, which accounted for approximately 70% of targeted ARGs in biomass and biofilm,
contributed only 0.2% in the effluent. The lack of ermF presence in the effluent is significant considering
that previous work has shown mesophilic anaerobic digestion to enhance the presence of erm-type
genes.
41
Specifically, it highlights the importance of membrane separation for removal of ermF-harboring
bacteria in AnMBRs, as well as implies a limited ermF presence on extracellular MGEs. The most abundant
genes detected in the effluent (when the influent contained 250 µg/L sulfamethoxazole) were sul1 and
intI1, followed by sul2, tetO, tetW, oxa-1, ermF, ermB, and ampC. Class 1 integrons gene abundance was
significantly greater in the effluent relative to the biomass, suggesting the presence of intI1 on
extracellular MGEs. The low relative abundance of intI1 compared to biomass ARGs is consistent with
previous studies on anaerobic digestion.
38, 42

Recent studies investigating the fate of ARGs in MBR systems have observed similar differences
between effluent ARG distributions and those of the reactor biomass, with effluent ARGs also being
dominated by sul1 in most instances.
12, 21, 26, 43, 44
The significance of sul1 in membrane-separated effluents
is noteworthy, especially given its strong association with class 1 integrons.
45
Previous studies found that
effluent ARG abundances spanned a broad range of values (from 10
3
to 10
7
copies/mL). Our study found
ARG abundances on the lower range of that spectrum (all detected ARGs < 10
4.5
copies/mL). This
observation, in combination with the significantly lower detection of specific genes (such as ermF and
ermB, < 10
2
copies/mL) as compared to previous studies of MBR and conventional systems,
12
implies that
AnMBRs could serve to reduce the overall rates of release of ARGs to the environment.
68

3.3.3 Most ARGs followed increasing trends in biomass during
incremental antibiotic addition
The overall impact of the sequential addition of sulfamethoxazole, erythromycin, and ampicillin at
incremental concentrations (10 µg/L, 50 µg/L, and 250 µg/L) on the biomass ARG profile (normalized to
rpoB) is shown in Figure 3.3A. Although no overall trends in total ARG abundance were observed across
the operational timeframe (i.e., all three antibiotic phases), there were several antibiotic-specific trends
for both total and individual ARG abundance.
Sulfamethoxazole
Total biomass ARG abundance significantly increased with the addition of sulfamethoxazole, as well as
with each concentration increment from 10 to 250 µg/L (p < 0.05). Assessment of ARG abundance without
the dominant influence of ermF (subset Figure 3.3A) revealed a similar trend, increasing to above 58
copies/rpoB at 250 µg/L sulfamethoxazole. Specific genes that increased significantly with
sulfamethoxazole concentration were ermF and sul1. The parallel increase of these two genes in the
biomass despite ermF’s notably lower presence in the effluent compared to sul1 may be due to previously
observed differences in the genotypes of class 1 integrons and their incorporation of sul1 and ermF genes
separately.
46
At the highest concentration of sulfamethoxazole (250 µg/L), both sul2 and oxa-1 genes were
also observed to increase significantly (p < 0.028) compared to 50 µg/L, which suggests that a higher
threshold of activation may cause the emergence of such genes. oxa-1 has previously been observed in
gene cassettes with both sul1 and sul2, implying that there could be a basis for its enrichment by
sulfamethoxazole at high concentrations.
47

Regarding sulfonamide resistance genes, sul1 was consistently found at higher concentrations than sul2
in the biomass ARG profile (Figure 3.4A). The biomass abundance of sul1 before antibiotic addition
averaged 8.3  0.8 copies/rpoB which significantly increased at each incremental concentration of
69

sulfamethoxazole to 16.1 1.0 at 250 µg/L (p < 0.043), corroborating previously observed correlations for
sul1 in both anaerobic and conventional wastewater treatment systems.
13, 48, 49
Conversely, sul2’s relative
abundance was 4.5  0.7 without any notable changes at 10 µg/L and 50 µg/L (p > 0.27). At 250 µg/L,
however, sul2 significantly increased by 37  4.1% (p < 0.038) compared to 50 µg/L. The increased
response of sul1 at lower sulfamethoxazole concentrations could be associated with its common
occurrence on small conjugative plasmids, while sul2 emergence at higher concentrations may highlight
its previously observed presence on larger less-mobile plasmids.
50

Erythromycin
Ten days after the conclusion of the sulfamethoxazole addition phase, erythromycin was added to the
influent in a replicate concentration sequence. Between the last sulfamethoxazole increment (250 µg/L)
and the pre-erythromycin sample, total targeted ARG abundance in the biomass decreased significantly
from 315  38 to 79.4  11.1 copies/rpoB (p < 0.001; Figure 3.3A). This drastic decrease was mainly due
to reductions in ermF and sul1 gene abundance. Given that both ermF and sul1 genes in the biomass were
positively correlated with sulfamethoxazole concentration, it is likely that ceasing its addition to the
influent resulted in a sudden decrease in its selective pressure on the microbial community. The addition
of 10 µg/L erythromycin to the influent increased both sul1 and ermF biomass abundance significantly,
although no significant change was observed for the remaining ARGs.
The targeted erythromycin-associated ARGs were ermF and ermB, both of which confer resistance to
macrolides by antibiotic target alteration (ribosomal target methylase).
51
Although ermF abundance was
found to be approximately an order of magnitude higher than ermB in the biomass (Figure 3.4B), both
genes followed a similar trend during addition of erythromycin. Abundances of ermF and ermB
significantly increased by (p < 0.046) at 10 µg/L erythromycin. However, this increase was then followed
by significant decreases (p < 0.038) at 50 µg/L. No significant changes in erm gene abundances were
70

observed between 50 and 250 µg/L. This was somewhat surprising, given previous reports of erythromycin
at 250 µg/L initiating resistance in a range of bacterial strains.
52
Although both the current and previous
studies have found a lack of correlation between erythromycin concentration and biomass erm gene
abundances,
53
their significant increase at 10 µg/L implies that the antibiotic can affect total microbial
community resistance even at trace levels.

Figure 3.3 (A) Abundance of targeted ARGs (Copy/rpoB) in the biomass (inserted diagram shows
abundance of targeted ARGs (Copy/rpoB) in the biomass excluding ermF gene); (B) Abundance of targeted
ARGs (Copy/mL) in the effluent, during the addition of increment concentrations of sulfamethoxazole,
erythromycin and ampicillin. Bars represent the mean values of three temporal sampling points (except
for 10 µg/L erythromycin and 50 µg/L ampicillin which represent two samples, and 0 µg/L ampicillin which
represents one sample) collected during the addition of each increment concentration of antibiotics and
each of their triplicate qPCR results.
Ampicillin
0
100
200
300
400
500
0 10 50 250 0 10 50 250 0 10 50 250
Biomass ARGs (Copy/rpoB)
ermF ermB sul1 sul2 intI1 oxa-1 ampC tetO tetW
0
1 0
2 0
3 0
4 0
5 0
6 0
7 0
0 1 0 5 0 2 5 0 0 1 0 5 0 2 5 0 0 1 0 5 0 2 5 0
Sulfamethoxazole Erythromycin Ampicillin
0.0E+00
1.0E+05
2.0E+05
3.0E+05
4.0E+05
5.0E+05
0 10 50 250 0 10 50 250 0 10 50 250
Effluent ARGs (Copy/mL)
An tibiotic Concentration ( µg/L)
A
B
71

After the erythromycin phase, total biomass ARGs targeted increased significantly from 91.5  16.3 to
197.6  27.5 copies/rpoB (p < 0.001), primarily due to a rise in ermF gene abundance. These observations
suggest that the final erythromycin concentration of 250 µg/L had an inhibitory effect on even ermF-
associated bacteria. During the ampicillin phase, several of the targeted ARGs, including sul1, sul2, ampC,
tetO, and tetW, increased progressively at 50 and 250 µg/L ampicillin (p < 0.031) (Figure 3.3A).
The -lactam resistance genes targeted in this study, oxa-1 and ampC, both encode for enzymatic
inactivation of antibiotics as their resistance mechanism. Still, oxa-1 and ampC genes followed remarkably
different trends in the biomass ARG profile during ampicillin addition. Although oxa-1 decreased
significantly from 2.46  0.10 copies/rpoB before the addition of ampicillin to 1.1  0.1 at 10 µg/L (p =
0.042), it remained relatively constant thereafter at 50 and 250 µg/L. ampC, however, significantly
increased at each increment of ampicillin (Figure 3.4C). This trend might be due to the fact that ampC and
oxa-1 genes belong to different molecular classes of -lactam resistance genes: ampC directly confers
resistance to ampicillin whereas oxa-1 confers resistance to cloxacillin and oxacillin. Overall, the biomass
ARG profiles of this study revealed that the abundances of ARGs, including those conferring resistance to
the same antibiotic, can follow vastly different trends in relation to increasing antibiotic concentrations
(e.g., oxa-1 vs. ampC). Although the cause of this phenomenon is unclear, it may relate to specific gene
presence on plasmids, plasmid mobility and conjugation, and microbial community dynamics. However,
given the consistent reactor biomass concentration (MLVSS), the short duration of antibiotic phases in
relation to the system SRT (> 300 d), and the stability of microbial communities typically observed in
AnMBRs,
23, 54, 55
it is unlikely that microbial abundance dynamics alone could have resulted in such high
variability in ARG copy number (which ranged from several- to over 10-fold). Therefore, this is an
indication of the importance of ARG presence on plasmids and their subsequent horizontal transfer across
and/or loss from the microbial community.
72

3.3.4 10 µg/L antibiotic concentrations induced spikes in total effluent
ARG abundance
As discussed in section 3.3, the ARG profile of the AnMBR effluent was distinct from that of the biomass.
Due to the potential bias associated with extracellular ARGs in the effluent (post-membrane filtration),
ARG abundances were normalized to volume (mL) for the analysis of all effluent samples.

0
4
8
12
16
20
0 10 50 250
Biom ass ARG s (Copy/rpoB)
Sulfam ethoxazole ( µg/L)
sul1 sul2
0.0 E+00
4.0 E+04
8.0 E+04
1.2 E+05
1.6 E+05
0 10 50 250
Effluent ARG s (Copy/m L)
Sulfam ethoxazole ( µg/L)
sul1 sul2
0.0E+00
1.0E+04
2.0E+04
3.0E+04
0 1 0 5 0 2 5 0
0
25
50
75
100
125
150
0 10 50 250
Biom ass ARGs (Copy/rpoB)
Erythrom ycin ( µg/L)
ermF ermB
0
1
2
3
4
5
0 1 0 50 2 50
0.0E+0 0
1.4E+0 2
2.8E+0 2
4.2E+0 2
5.6E+0 2
0 10 50 250
Effluent ARG s (Copy/m L)
Erythrom ycin ( µg/L)
ermF ermB
0
1
2
3
4
5
0 10 50 250
Biom ass ARGs (Copy/rpoB)
Am picillin ( µg/L)
oxa-1 ampC
0
0.01
0.02
0.03
0.04
0.05
0 1 0 5 0 2 5 0
0 .0E+0 0
1 .0E+0 4
2 .0E+0 4
3 .0E+0 4
4 .0E+0 4
0 10 50 250
Effluent ARG s (Copy/m L)
Am picillin ( µg/L)
oxa-1 ampC
0.0E+00
4.0E+01
8.0E+01
1.2E+02
0 10 5 0 2 50
C
B
A
F
E
D
73

Figure 3.4 Abundance of antibiotic corresponding genes in biomass (Copy/rpoB) and effluent (Copy/mL)
when (A) and (D) sulfamethoxazole; (B) and (E) erythromycin; (C) and (F) ampicillin was added to AnMBR.
Inserted diagrams were used to magnify the genes with low ab abundance. Bars and error bars represent
the mean values and standard deviations, respectively, of three temporal sampling points (except for 10
µg/L erythromycin and 50 µg/L ampicillin which represent two samples, and 0 µg/L ampicillin which
represents one sample) collected during the addition of each increment concentration of antibiotics and
each of their triplicate qPCR results.
Sulfamethoxazole
Effluent ARG profiles during sulfamethoxazole addition were dominated by sul1, sul2, and intI1. The
addition of 10 µg/L sulfamethoxazole to the influent significantly increased the abundance of sul1, sul2,
intI1, and oxa-1 genes (p < 0.026). However, at 50 µg/L sulfamethoxazole, a sharp decrease in the
abundance of all ARGs (p < 0.007) except for ermF and ermB caused a reduction in total effluent ARG
abundance (Figure 3.3B). The abundance of the same ARGs (sul1, sul2, intI1, oxa-1, tetO, and tetW) grew
significantly (p < 0.034) when sulfamethoxazole was increased from 50 to 250 µg/L, although only to
approximately 20% of the 10 µg/L levels.
Both sulfonamide resistance genes (sul1 and sul2) followed a similar trend in the effluent across the
varying sulfamethoxazole influent concentrations (Figure 3.4D). For both, the addition of 10 µg/L resulted
in abundance significantly increasing by 68.5  1.1% and 83.3  2.5% (p < 0.005), respectively. sul1 and
sul2 abundance then dramatically decreased by about 2 orders of magnitude, at 50 µg/L and finally
increased again slightly at 250 µg/L sulfamethoxazole (p < 0.012).
The dominance and common trends of the sul-type genes in the effluent, along with intl1 and oxa-1,
are a strong testament to their likely presence on a single gene cassette encoded for recombination by
intl1.
45
Gene cassettes combining sul1, sul2, and oxa-1, specifically, are well documented and have been
found in several wastewater-associated pathogenic bacteria.
47
Although the abundant presence of such a
cassette-carrying plasmid in AnMBR effluent is cause for further scrutiny, this observation must also be
put in perspective. For example, anaerobic systems have been shown to harbor a significantly lower
74

abundance of intl1-associated ARGs (including sul and tet genes) than activated sludge.
49, 56
Although not
well understood, the reason for this lower ARG abundance may be associated with faster rates of ARG
loss by bacteria under anaerobic conditions.
57

Erythromycin
The abundance of total ARGs in the AnMBR effluent before commencement of the erythromycin run (0
µg/L erythromycin), in contrast to the biomass ARG profile, increased approximately six-fold relative to
the last stage of sulfamethoxazole addition (Figure 3.3B). This spike in ARGs could have been created by
a sudden die-off of a subset of resistant bacteria previously selected for by sulfamethoxazole at 250 µg/L
and the subsequent release of their intracellular DNA. Alternatively, it is possible that the membrane
cleaning performed prior to the erythromycin phase reduced the propensity of the membrane biofilm
matrix to partially or fully retain plasmids/extra-cellular ARGs. This scenario seems likely, especially
considering observations of recent studies documenting the importance of biofilms in MBR systems for
achieving higher removals of ARGs.
22, 44
Effluent ARGs increased further at the first erythromycin
concentration of 10 µg/L, then gradually decreased at subsequent concentrations of 50 and 250 µg/L.
During erythromycin addition, both ermF and ermB genes were more abundant in the effluent than in
the previous run and followed a parallel trend to each other (Figure 3.4E). However, a significant decrease
was observed in total abundance of both ermF and ermB by 60  7.8% and 48  1.3% (p < 0.027),
respectively, at 10 µg/L erythromycin. Subsequent increases in erythromycin concentration to 50 µg/L
and 250 µg/L did not significantly change the abundance of ermF or ermB in the effluent (p > 0.471).
Overall, erm gene abundance dynamics were entirely independent of those of the dominant ARGs in the
effluent, as was the case during the sulfamethoxazole run. The lack of correlation between these
macrolide resistance genes and the remaining ARGs is likely attributable to their minor contribution to
the effluent profile (< 0.2% of total ARGs), which implies their lack of extracellular presence in the biomass
and/or presence on MGEs of a size range large enough to be rejected by the membranes.
75

Ampicillin
Similar to the interim period between sulfamethoxazole and erythromycin addition, ARG abundance
increased significantly prior to ampicillin addition (Figure 3.3B). Total effluent ARGs increased slightly after
the addition of ampicillin at 10 µg/L then decreased to original abundance at 50 µg/L. When the influent
ampicillin concentration was increased to 250 µg/L, however, a drastic drop by approximately an order of
magnitude was observed in total effluent ARG abundance, led by sul1, sul2, intI1, and oxa-1 (p < 0.032).
Considering the low concentrations (< 4 µg/L) of ampicillin detected in the AnMBR effluent, this drop may
have been due to a significant reduction in extracellular plasmid-associated ARGs in the biomass.
As previously observed for the sulfamethoxazole- and erythromycin-associated resistance genes, oxa-
1 and ampC (conferring -lactam resistance) followed a parallel trend in their effluent ARG abundance
during the ampicillin run despite a lack of consistency between the two genes in their corresponding
biomass samples (Figure 3.4F). The highest abundance of oxa-1 and ampC in the effluent occurred at 10
µg/L ampicillin, which was about 3.7 times higher than at 0 µg/L for both genes. oxa-1 then significantly
decreased by 61.9  0.7% and 66.3  1.2% at 50 and 250 µg/L (p < 0.001), respectively, while ampC
decreased significantly by 68  4.3% (p < 0.001) at 50 µg/L.
Overall, a common trend of increase at 10 µg/L and subsequent decrease at 50 and 250 µg/L was
observed for most ARGs in the effluent during all three antibiotic phases (Figure 3.3B). There are two
probable explanations for this phenomenon. First, it is possible that the initial trace level exposure to the
antibiotics increased horizontal gene transfer. This could have led to a temporary increase in extracellular
plasmid DNA, which thereby amplified the harboring of plasmid-based resistance within the biomass
microbial community for the remainder of individual antibiotic phases. If, after initial antibiotic exposure,
the rates of horizontal gene transfer then slowed, this would have again reduced the levels of extracellular
plasmid-based ARGs and lowered the rate of ARGs passing through membranes into the effluent. Second,
76

it is likely that the cleaning of AnMBR membranes prior to commencement of each antibiotic run reduced
the effect of membrane biofilm-based ARG removal, which then gradually increased along with the
development of sub-critical fouling layers.
22

3.3.5 AnMBR effluent ARB largely unaffected by different antibiotics and
concentrations
The results of heterotrophic plate counts (HPC) and ARB plate counts (including sulfamethoxazole,
erythromycin, ampicillin, and tetracycline) revealed a lack of consistent trends between antibiotic
concentration and specific ARB effluent abundance, both in terms of their absolute and HPC-normalized
values (Appx 2, Figure S2.2). One exception was in the case of sulfamethoxazole resistant bacteria, which
increased in relation to total bacterial count with the increasing influent sulfamethoxazole concentrations.
A recent study by Le et al. on an MBR system similarly found that among 19 antibiotics targeted,
sulfamethoxazole was one of three antibiotics that increased with its corresponding ARB.
58
It is important
to note that the heterotrophic enumeration methodology used conventionally (and in this study) does
not accurately account for anaerobic microorganisms.
The overall highest counts among targeted ARB in the effluent were erythromycin resistant bacteria.
Sulfamethoxazole and ampicillin resistant bacteria were lower and generally close to one another in terms
of CFU/mL, while tetracycline resistant bacteria had the lowest absolute plate counts. The dominant
abundance of erythromycin resistant bacteria in the effluent stood in contrast to a relatively minor
presence of erythromycin-associated ARGs, implying that their persistence was achieved through
alternative macrolide and/or multidrug resistance genes than those detected in the AnMBR biomass.
77

3.3.6 Correlation between antibiotics, ARGs, and ARB indicated the
presence of multi-drug resistance
Correlation analysis of the effluent revealed that all ARB types targeted were strongly correlated with
each other (ρ > 0.79, p < 0.011; Figure 3.5), implying that a large fraction of effluent bacteria may have
harbored multidrug resistance. However, sulfamethoxazole resistant bacteria were the only ARB that
showed significant positive correlation to their corresponding antibiotic (ρ = 0.72, p = 0.041). Strong
correlation between the abundance of some ARB and their corresponding ARGs were also observed, such
as ampicillin resistant bacteria and both oxa-1 (ρ = 0.84, p = 0.005) and ampC (ρ = 0.93, p < 0.0001) genes.
This observation might suggest that, in the AnMBR effluent, a considerable portion of oxa-1 and ampC
genes were predominantly located within bacterial cells. The lack of correlation between other ARB in the
effluent and their corresponding ARGs (i.e., sulfamethoxazole and erythromycin resistant bacteria)
highlights one of the main limitations of this study and qPCR-based ARG targeting in general: any
conclusions drawn from this analysis are only representative of the predetermined gene targets used. The
inclusion of more and/or different ARGs could lead to significantly different results in terms of total ARG
abundance and their relative distributions. Metagenomics-based approaches have previously been
developed for ARG screening in wastewater systems and could serve to circumvent such limitations in
future studies.
59, 60

78

Figure 3.5 Network analysis representing the correlations between antibiotics concentration (pink circles),
ARGs (teal circles), class 1 integrons (blue circles) and ARB (yellow circles) (A) in biomass; (B) in effluent.
SMX, ERY, AMP and TET stand for sulfamethoxazole, erythromycin, ampicillin and tetracycline;
respectively. RB also stands for resistant bacteria. A solid connection shows strong, significant correlations
(ρ > 0.7 or ρ < -0.7; and p<0.05), and a dashed line represents weak correlations (0.3 < ρ < 0.7 or -0.7 < ρ
< -0.3; and p<0.05).
The analysis also demonstrated that there were a high number of correlations among ARGs in the
biomass (such as sul2 and ermB) and comparatively fewer among ARGs in the effluent (such as sul1 and
ampC) (Figure 3.5). Abundance of the integrase gene intI1 in the biomass was strongly correlated with
ermF (ρ = 0.72), sul2 (ρ = 0.71), tetO (ρ = 0.79), and tetW (ρ = 0.71) genes (p < 0.0001), while it was not
significantly correlated with sul1 (p > 0.35). In effluent samples, however, intI1 was only correlated with
sul1 (ρ = 0.9, p < 0.0001) and ampC (ρ = 0.7, p = 0.0001) genes. Given the frequent association of sul1 with
class 1 integrons,
45
this correlation between sul1 and intl1 in the effluent is not surprising. Such integrons’
ability to initiate the recombination of a variety of ARGs in associated cassettes,
45
however, is an indication
of the potential presence of additional undetected ARGs. Further, the association of class 1 integrons
(encoded for by intl1) with horizontal gene transfer in anaerobic enviroments
61, 62
suggests a relatively
consistent MGE-based ARG occurrence in the AnMBR biomass. Although this may be expected given the
high density of anaerobic microorganisms in the bioreactor, the lack of consistency between biomass and
79

effluent intl1-ARG correlations implies that only a fraction of these MGEs are being released into the
effluent.
Membrane separation in MBRs has been shown to be a key factor in the reduction of effluent ARGs.
11,
58
Such improved removal of ARGs is likely attributable to their predominant presence within bacterial
cells, as extracellular plasmids have previously been shown to be completely permeable to 0.1 µm
membranes.
63
However, given that ARG-associated plasmids can reach sizes exceeding 90 kb,
64
it is likely
that such large plasmids would be at least partially retained by microfiltration membranes when present
in “supercoiled” form.
65
This, in combination with the reduction in effective pore size caused by
membrane biofilm formation in MBRs,
22, 44, 66
could conceivably lead to significant variability in the
removal of even extracellular ARGs by the AnMBR. Such variability would, indeed, contribute to the
differences observed in this study between the biomass and effluent ARG profiles. The phenomenon
responsible for this variability (i.e., physical retention of extracellular plasmid-located ARGs) could also
serve to accentuate the advantages of AnMBR systems for ARG reduction by providing an additional
barrier to ARG release. Evaluating ARG-associated plasmid retention via membrane separation is an
important topic for future research.
References
1. Michael, I.; Rizzo, L.; McArdell, C.; Manaia, C.; Merlin, C.; Schwartz, T.; Dagot, C.; Fatta-Kassinos, D.,
Urban wastewater treatment plants as hotspots for the release of antibiotics in the environment: a
review. Water Res. 2013, 47, (3), 957-995.
2. Schwartz, T.; Kohnen, W.; Jansen, B.; Obst, U., Detection of antibiotic-resistant bacteria and their
resistance genes in wastewater, surface water, and drinking water biofilms. FEMS Microbiol. Ecol. 2003,
43, (3), 325-335.
3. Zhang, S.; Han, B.; Gu, J.; Wang, C.; Wang, P.; Ma, Y.; Cao, J.; He, Z., Fate of antibiotic resistant cultivable
heterotrophic bacteria and antibiotic resistance genes in wastewater treatment processes. Chemosphere
2015, 135, 138-145.
4. Blair, J. M.; Webber, M. A.; Baylay, A. J.; Ogbolu, D. O.; Piddock, L. J., Molecular mechanisms of antibiotic
resistance. Nat. Rev. Microbiol. 2015, 13, (1), 42.
80

5. Services, U. D. o. H. a. H., Antibiotic resistance threats in the United States. CDC 2013.
6. Baquero, F.; Martínez, J.-L.; Cantón, R., Antibiotics and antibiotic resistance in water environments.
Curr. Opin. Biotechnol. 2008, 19, (3), 260-265.
7. Al-Jassim, N.; Ansari, M. I.; Harb, M.; Hong, P. Y., Removal of bacterial contaminants and antibiotic
resistance genes by conventional wastewater treatment processes in Saudi Arabia: Is the treated
wastewater safe to reuse for agricultural irrigation? Water Res. 2015, 73, 277-290.
8. Rizzo, L.; Manaia, C.; Merlin, C.; Schwartz, T.; Dagot, C.; Ploy, M. C.; Michael, I.; Fatta-Kassinos, D., Urban
wastewater treatment plants as hotspots for antibiotic resistant bacteria and genes spread into the
environment: A review. Sci. Total Environ. 2013, 447, 345-360.
9. Pruden, A.; Larsson, D. J.; Amézquita, A.; Collignon, P.; Brandt, K. K.; Graham, D. W.; Lazorchak, J. M.;
Suzuki, S.; Silley, P.; Snape, J. R., Management options for reducing the release of antibiotics and antibiotic
resistance genes to the environment. Environ. Health Prespect. (Online) 2013, 121, (8), 878.
10. Xia, S.; Jia, R.; Feng, F.; Xie, K.; Li, H.; Jing, D.; Xu, X., Effect of solids retention time on antibiotics
removal performance and microbial communities in an A/O-MBR process. Bioresour. Technol. 2012, 106,
36-43.
11. Sui, Q.; Jiang, C.; Zhang, J.; Yu, D.; Chen, M.; Wang, Y.; Wei, Y., Does the biological treatment or
membrane separation reduce the antibiotic resistance genes from swine wastewater through a
sequencing-batch membrane bioreactor treatment process. Environ. Int. 2018, 118, 274-281.
12. Zhang, J.; Yang, M.; Zhong, H.; Liu, M.; Sui, Q.; Zheng, L.; Tong, J.; Wei, Y., Deciphering the factors
influencing the discrepant fate of antibiotic resistance genes in sludge and water phases during municipal
wastewater treatment. Bioresour. Technol. 2018, 265, 310-319.
13. Gao, P.; Munir, M.; Xagoraraki, I., Correlation of tetracycline and sulfonamide antibiotics with
corresponding resistance genes and resistant bacteria in a conventional municipal wastewater treatment
plant. Sci. Total Environ. 2012, 421, 173-183.
14. Rodriguez-Mozaz, S.; Chamorro, S.; Marti, E.; Huerta, B.; Gros, M.; Sànchez-Melsió, A.; Borrego, C. M.;
Barceló, D.; Balcázar, J. L., Occurrence of antibiotics and antibiotic resistance genes in hospital and urban
wastewaters and their impact on the receiving river. Water Res. 2015, 69, 234-242.
15. Xu, J.; Xu, Y.; Wang, H.; Guo, C.; Qiu, H.; He, Y.; Zhang, Y.; Li, X.; Meng, W., Occurrence of antibiotics
and antibiotic resistance genes in a sewage treatment plant and its effluent-receiving river. Chemosphere
2015, 119, 1379-1385.
16. Harb, M.; Hong, P.-Y., Anaerobic Membrane Bioreactor Effluent Reuse: A Review of Microbial Safety
Concerns. Fermentation 2017, 3, (3), 39.
17. Smith, A. L.; Stadler, L. B.; Love, N. G.; Skerlos, S. J.; Raskin, L., Perspectives on anaerobic membrane
bioreactor treatment of domestic wastewater: a critical review. Bioresour. Technol. 2012, 122, 149-59.
18. Smith, A.; Skerlos, S.; Raskin, L., Anaerobic membrane bioreactor treatment of domestic wastewater
at psychrophilic temperatures ranging from 15° C to 3° C. Environ. Sci.: Water Res. Technol. 2015, 1, (1),
56-64.
81

19. Ma, Y.; Wilson, C. A.; Novak, J. T.; Riffat, R.; Aynur, S.; Murthy, S.; Pruden, A., Effect of various sludge
digestion conditions on sulfonamide, macrolide, and tetracycline resistance genes and class I integrons.
Environ. Sci. Technol. 2011, 45, (18), 7855-7861.
20. Zhang, Q.-Q.; Tian, G.-M.; Jin, R.-C., The occurrence, maintenance, and proliferation of antibiotic
resistance genes (ARGs) in the environment: influencing factors, mechanisms, and elimination strategies.
Appl. Microbiol. Biotechnol. 2018, 102, (19), 8261-8274.
21. Kappell, A. D.; Kimbell, L. K.; Seib, M. D.; Carey, D. E.; Choi, M. J.; Kalayil, T.; Fujimoto, M.; Zitomer, D.
H.; McNamara, P. J., Removal of antibiotic resistance genes in an anaerobic membrane bioreactor treating
primary clarifier effluent at 20° C. Environ. Sci.: Water Res. Technol. 2018, 4, 1783-1793.
22. Cheng, H.; Hong, P.-Y., Removal of antibiotic-resistant bacteria and antibiotic resistance genes affected
by varying degrees of fouling on anaerobic microfiltration membranes. Environ. Sci. Technol. 2017, 51,
(21), 12200-12209.
23. Smith, A. L.; Skerlos, S. J.; Raskin, L., Psychrophilic anaerobic membrane bioreactor treatment of
domestic wastewater. Water Res. 2013, 47, (4), 1655-1665.
24. Karthikeyan, K.; Meyer, M. T., Occurrence of antibiotics in wastewater treatment facilities in
Wisconsin, USA. Sci. Total Environ. 2006, 361, (1-3), 196-207.
25. Kulkarni, P.; Olson, N. D.; Raspanti, G. A.; Rosenberg Goldstein, R. E.; Gibbs, S. G.; Sapkota, A.; Sapkota,
A. R., Antibiotic concentrations decrease during wastewater treatment but persist at low levels in
reclaimed water. Int. J. Environ. Res. Public Health 2017, 14, (6), 668.
26. Munir, M.; Wong, K.; Xagoraraki, I., Release of antibiotic resistant bacteria and genes in the effluent
and biosolids of five wastewater utilities in Michigan. Water Res. 2011, 45, (2), 681-693.
27. Pruden, A.; Pei, R.; Storteboom, H.; Carlson, K. H., Antibiotic resistance genes as emerging
contaminants: studies in northern Colorado. Environ. Sci. Technol. 2006, 40, (23), 7445-7450.
28. Federation, W.E. and APHA, Standard methods for the examination of water and wastewater. APHA:
Washington, D.C., 2005.
29. Zheng, J.; Su, C.; Zhou, J.; Xu, L.; Qian, Y.; Chen, H., Effects and mechanisms of ultraviolet, chlorination,
and ozone disinfection on antibiotic resistance genes in secondary effluents of municipal wastewater
treatment plants. Chem. Eng. J. 2017, 317, 309-316.
30. Pei, R.; Kim, S.-C.; Carlson, K. H.; Pruden, A., Effect of river landscape on the sediment concentrations
of antibiotics and corresponding antibiotic resistance genes (ARG). Water Res. 2006, 40, (12), 2427-2435.
31. Sponza, D. T.; Demirden, P., Treatability of sulfamerazine in sequential upflow anaerobic sludge
blanket reactor (UASB)/completely stirred tank reactor (CSTR) processes. Sep. Purif. Technol. 2007, 56,
(1), 108-117.
32. Wei, C.-H.; Hoppe-Jones, C.; Amy, G.; Leiknes, T., Organic micro-pollutants’ removal via anaerobic
membrane bioreactor with ultrafiltration and nanofiltration. J. Water Reuse Desal. 2015, 6, (3), 362-370.
82

33. Mitchell, S. M.; Ullman, J. L.; Teel, A. L.; Watts, R. J.; Frear, C., The effects of the antibiotics ampicillin,
florfenicol, sulfamethazine, and tylosin on biogas production and their degradation efficiency during
anaerobic digestion. Bioresour. Technol. 2013, 149, 244-252.
34. Dutta, K.; Lee, M.-Y.; Lai, W. W.-P.; Lee, C. H.; Lin, A. Y.-C.; Lin, C.-F.; Lin, J.-G., Removal of
pharmaceuticals and organic matter from municipal wastewater using two-stage anaerobic fluidized
membrane bioreactor. Bioresour. Technol. 2014, 165, 42-49.
35. Xiao, Y.; Yaohari, H.; De Araujo, C.; Sze, C. C.; Stuckey, D. C., Removal of selected pharmaceuticals in
an anaerobic membrane bioreactor (AnMBR) with/without powdered activated carbon (PAC). Chem. Eng.
J. 2017, 321, 335-345.
36. Papageorgiou, M.; Kosma, C.; Lambropoulou, D., Seasonal occurrence, removal, mass loading and
environmental risk assessment of 55 pharmaceuticals and personal care products in a municipal
wastewater treatment plant in Central Greece. Sci. Total Environ. 2016, 543, 547-569.
37. Wei, C.-H.; Wang, N.; HoppeJones, C.; Leiknes, T.; Amy, G.; Fang, Q.; Hu, X.; Rong, H., Organic
micropollutants removal in sequential batch reactor followed by nanofiltration from municipal
wastewater treatment. Bioresour. Technol. 2018, 268, 648-657.
38. Zhang, J.; Chen, M.; Sui, Q.; Wang, R.; Tong, J.; Wei, Y., Fate of antibiotic resistance genes and its drivers
during anaerobic co-digestion of food waste and sewage sludge based on microwave pretreatment.
Bioresour. Technol. 2016, 217, 28-36.
39. Zhang, J.; Liu, J.; Wang, Y.; Yu, D.; Sui, Q.; Wang, R.; Chen, M.; Tong, J.; Wei, Y., Profiles and drivers of
antibiotic resistance genes distribution in one-stage and two-stage sludge anaerobic digestion based on
microwave-H2O2 pretreatment. Bioresour. Technol. 2017, 241, 573-581.
40. Engemann, C. A.; Keen, P. L.; Knapp, C. W.; Hall, K. J.; Graham, D. W., Fate of tetracycline resistance
genes in aquatic systems: migration from the water column to peripheral biofilms. Environ. Sci. Technol.
2008, 42, (14), 5131-5136.
41. Zhang, T.; Yang, Y.; Pruden, A., Effect of temperature on removal of antibiotic resistance genes by
anaerobic digestion of activated sludge revealed by metagenomic approach. Appl. Microbiol. Biotechnol.
2015, 99, (18), 7771-7779.
42. Zhang, J.; Mao, F.; Loh, K.-C.; Gin, K. Y.-H.; Dai, Y.; Tong, Y. W., Evaluating the effects of activated carbon
on methane generation and the fate of antibiotic resistant genes and class I integrons during anaerobic
digestion of solid organic wastes. Bioresour. Technol. 2018, 249, 729-736.
43. Wen, Q.; Yang, L.; Zhao, Y.; Huang, L.; Chen, Z., Insight into effects of antibiotics on reactor
performance and evolutions of antibiotic resistance genes and microbial community in a membrane
reactor. Chemosphere 2018, 197, 420-429.
44. Zhu, Y.; Wang, Y.; Zhou, S.; Jiang, X.; Ma, X.; Liu, C., Robust performance of a membrane bioreactor for
removing antibiotic resistance genes exposed to antibiotics: Role of membrane foulants. Water Res. 2018,
130, 139-150.
83

45. Fluit, A.; Schmitz, F., Class 1 integrons, gene cassettes, mobility, and epidemiology. Eur. J. Clin.
Microbiol. Infect. Dis. 1999, 18, (11), 761-770.
46. Zerva, I.; Alexandropoulou, I.; Panopoulou, M.; Melidis, P.; Ntougias, S. Antibiotic Resistance Genes
Dynamics at the Different Stages of the Biological Process in a Full-Scale Wastewater Treatment Plant,
MDPI 2018, 2, (11), 650.
47. Shin, H. W.; Lim, J.; Kim, S.; Kim, J.; Kwon, G. C.; Koo, S. H., Characterization of trimethoprim-
sulfamethoxazole resistance genes and their relatedness to class 1 integron and insertion sequence
common region in gram-negative bacilli. Microbiol. Biotechnol. 2015, 25, (1), 137-42.
48. Aydin, S.; Ince, B.; Ince, O., Development of antibiotic resistance genes in microbial communities
during long-term operation of anaerobic reactors in the treatment of pharmaceutical wastewater. Water
Res. 2015, 83, 337-344.
49. Harb, M.; Wei, C. H.; Wang, N.; Amy, G.; Hong, P. Y., Organic micropollutants in aerobic and anaerobic
membrane bioreactors: Changes in microbial communities and gene expression. Bioresour. Technol. 2016,
218, 882-891.
50. Sköld, O., Sulfonamide resistance: mechanisms and trends. Drug Resist. Updates 2000, 3, (3), 155-160.
51. Leclercq, R.; Courvalin, P., Bacterial resistance to macrolide, lincosamide, and streptogramin
antibiotics by target modification. Antimicrob. Agents Chemother. 1991, 35, (7), 1267.
52. Ge, B.; Domesle, K. J.; Yang, Q.; Young, S. R.; Rice-Trujillo, C. L.; Jones, S. M. B.; Gaines, S. A.; Keller, M.
W.; Li, X.; Piñeiro, S. A., Effects of low concentrations of erythromycin, penicillin, and virginiamycin on
bacterial resistance development in vitro. Sci. Rep. 2017, 7, (1), 11017.
53. Gao, P.; He, S.; Huang, S.; Li, K.; Liu, Z.; Xue, G.; Sun, W., Impacts of coexisting antibiotics, antibacterial
residues, and heavy metals on the occurrence of erythromycin resistance genes in urban wastewater.
Appl. Microbiol. Biotechnol. 2015 , 99, (9), 3971-3980.
54. Harb, M.; Xiong, Y.; Guest, J.; Amy, G.; Hong, P.-Y., Differences in microbial communities and
performance between suspended and attached growth anaerobic membrane bioreactors treating
synthetic municipal wastewater. Environ. Sci.: Water Res. Technol. 2015, 1, (6), 800-813.
55. Xiong, Y.; Harb, M.; Hong, P.-Y., Performance and microbial community variations of anaerobic
digesters under increasing tetracycline concentrations. Appl. Microbiol. Biotechnol. 2017, 101, (13), 5505-
5517.
56. Diehl, D. L.; LaPara, T. M., Effect of temperature on the fate of genes encoding tetracycline resistance
and the integrase of class 1 integrons within anaerobic and aerobic digesters treating municipal
wastewater solids. Environ. Sci. Technol. 2010, 44, (23), 9128-9133.
57. Rysz, M.; Mansfield, W. R.; Fortner, J. D.; Alvarez, P. J., Tetracycline resistance gene maintenance under
varying bacterial growth rate, substrate and oxygen availability, and tetracycline concentration. Environ.
Sci. Technol. 2013, 47, (13), 6995-7001.
84

58. Le, T.-H.; Ng, C.; Tran, N. H.; Chen, H.; Gin, K. Y.-H., Removal of antibiotic residues, antibiotic resistant
bacteria and antibiotic resistance genes in municipal wastewater by membrane bioreactor systems. Water
Res. 2018, 145, 498-508.
59. Christgen, B.; Yang, Y.; Ahammad, S.; Li, B.; Rodriquez, D. C.; Zhang, T.; Graham, D. W., Metagenomics
Shows That Low-Energy Anaerobic− Aerobic Treatment Reactors Reduce Antibiotic Resistance Gene
Levels from Domestic Wastewater. Environ. Sci. Technol. 2015, 49, (4), 2577-2584.
60. Yang, Y.; Li, B.; Zou, S.; Fang, H. H.; Zhang, T., Fate of antibiotic resistance genes in sewage treatment
plant revealed by metagenomic approach. Water Res. 2014, 62, 97-106.
61. Barraud, O.; Baclet, M.-C.; Denis, F.; Ploy, M.-C., Quantitative multiplex real-time PCR for detecting
class 1, 2 and 3 integrons. J. Antimicrob. Chemother. 2010, 65, (8), 1642-1645.
62. Burch, T. R.; Sadowsky, M. J.; LaPara, T. M., Modeling the fate of antibiotic resistance genes and class
1 integrons during thermophilic anaerobic digestion of municipal wastewater solids. Appl. Microbiol.
Biotechnol. 2015, 1-8.
63. Breazeal, M. V. R.; Novak, J. T.; Vikesland, P. J.; Pruden, A., Effect of wastewater colloids on membrane
removal of antibiotic resistance genes. Water Res. 2013, 47, (1), 130-140.
64. San Millan, A.; Heilbron, K.; MacLean, R. C., Positive epistasis between co-infecting plasmids promotes
plasmid survival in bacterial populations. ISME J. 2014, 8, (3), 601.
65. Li, Y.; Butler, N.; Zydney, A. L., Size-based separation of supercoiled plasmid DNA using ultrafiltration.
J. Colloid Interface Sci. 2016, 472, 195-201.
66. Zarei-Baygi, A.; Moslemi, M.; Mirzaei, S. H., The combination of KMnO4 oxidation and polymeric
flocculation for the mitigation of membrane fouling in a membrane bioreactor. Sep. Purif. Technol. 2016,
159, 124-134.
85
Chapter 4

4. Microbial community and antibiotic resistance profiles of biomass
and effluent are distinctly affected by antibiotics addition to an
anaerobic membrane bioreactor

Abstract
The transfer of antibiotic resistance to pathogenic bacteria is one of the most eminent human health
threats and a concern in water reuse schemes. Anaerobic membrane bioreactors (AnMBRs) are an
emerging wastewater treatment biotechnology that have significant potential for mainstream
wastewater treatment. However, AnMBR effluents remain largely unexplored with respect to their
microbial community composition and their antibiotic resistance profiles. In this study, we operated a
bench-scale AnMBR for the treatment of domestic wastewater containing antibiotics (250 μg/L each of
sulfamethoxazole, ampicillin, and erythromycin) and evaluated microbial community structure and
antibiotic resistance gene (ARG) dynamics in both the biomass and effluent. Results showed that ARG
abundances in the biomass of the AnMBR consistently increased throughout the experiment, while the
effluent ARG abundances saw a sharp increase upon initial antibiotics exposure to the system and then
dropped immediately thereafter. Further, a vastly more variable microbial community was observed in
the AnMBR effluent as compared to the biomass. Several potentially pathogenic genera in the effluent
86
were strongly correlated with the abundance of specific resistance genes (e.g., sul1), as well as a class 1
integrase gene (intl1). Overall, results of this study provide useful insights into the association of ARGs
with microbial community dynamics in AnMBR, which is needed to devise operational and design
strategies to lessen dissemination of antibiotic resistance to the environment.
4.1 Introduction
Antibiotic resistance is an issue of crucial concern as one of the most imminent human health risks. In
the US, antibiotic resistance is currently responsible for over $20 billion in excess health costs and 8 million
additional hospital days.
1
The proliferation of antibiotic resistance in the environment and clinical settings
is dictated by the dissemination of antibiotic resistance genes (ARGs).
2
Resistance spreads due to the
selective pressures raised by antimicrobial compounds and through vertical and horizontal gene transfer
(VGT and HGT) mechanisms among and across different bacterial populations, respectively. Recent
studies have revealed a considerable increase in the number of bacterial species capable of resisting
different antibiotic classes.
1, 3, 4

Wastewater treatment plants (WWTPs) are known reservoirs of ARGs. Studies have revealed that
variable concentrations of various ARGs are released daily from WWTPs to the environment, sometimes
at levels that can even exceed those observed in influent wastewater.
5
Although wastewater can be
considered a resource of energy, water, and nutrients, recovering these resources must be balanced with
protecting the downstream environment, including mitigating emerging threats such as antibiotic
resistance.
6
Anaerobic membrane bioreactors (AnMBRs) are an emerging biotechnology that can recover
energy from wastewater via biogas production and reduce residuals production, while also potentially
playing a role in lessening antibiotic resistance dissemination.
7-9
Given that the biomass of WWTPs
contains both a broader range and higher concentrations of ARGs than do WWTP effluents,
4
the vastly
lower biomass production of AnMBRs as compared to conventional aerobic WWTPs has the potential to
87
significantly reduce the overall release of ARGs to the environment. Further, microorganisms and larger
mobile genetic elements (MGEs) can be effectively retained in the reactor via the membrane barrier and
its associated biofilm layer, reducing dissemination of antibiotic resistance in the effluent.
8

Despite these potential advantages, the fate of antibiotics and ARGs in AnMBRs remains understudied.
In a previous study, we demonstrated that individual influent antibiotics at a range of concentrations (10
to 250 g/L) can significantly alter the abundance of both related and unrelated ARGs in the biomass and
effluent of AnMBRs.
8
It has been reported elsewhere that the impact of a mixture of antibiotics on ARG
profiles can be considerably greater than an individual antibiotic.
10, 11
Therefore, in the present study we
investigated the influence of a mixture of antibiotics on the ARG profile of both the biomass and effluent
of an AnMBR. Although no study to date has been able to verify the reasons for variation of ARG profiles
in the presence of antibiotics,
12, 13
possible explanations could include HGT,
12, 14
changes in microbial
community abundance,
11, 15
or some combination of the two. HGT is known to be a critical factor for
dissemination of antibiotic resistance in the environment
14
and is of particular concern in WWTPs: namely,
ARGs can spread via HGT from non-harmful bacteria to more virulent pathogenic species, thus posing a
serious human health risk. In addition to HGT, the fate of ARGs during wastewater treatment is also
intricately connected to microbial community dynamics. Although previous studies have demonstrated
relationships between biomass microbial communities and their accompanying ARG profiles during the
anaerobic digestion of sewage sludge
16
, no studies to date have investigated such associations in
mainstream anaerobic wastewater treatment systems like AnMBRs. Further, the effluent of AnMBRs may
contain resistant pathogenic bacteria, posing a serious health risk in water reuse schemes. Therefore, a
comprehensive investigation of associations between microbial community dynamics and ARGs in AnMBR
effluents is needed. Consequently, in the present study, both the microbial communities and the ARG
profiles of the biomass and effluent of an AnMBR were investigated before, during, and after the addition
of three antibiotics from different classes.
88
4.2 Materials and methods
4.2.1 Configuration of bench-scale AnMBR
Detailed information of the bench-scale AnMBR has been reported in our previous study.
8
Briefly, the
AnMBR consisted of a continuously stirred-tank reactor (Chemglass Life Science, Vineland, NJ) with a
working volume of 5 L and three separate microfiltration silicon carbide membrane modules (Cembrane,
Denmark) submerged in the reactor. The effective membrane area of each module was approximately
0.015 m
2
and the membrane pore size was 0.1 μm. The AnMBR was seeded with sludge from a mesophilic
anaerobic digester at the Joint Water Pollution Control Plant (Carson, CA). The AnMBR was operated at
25 C and fed with a synthetic wastewater representative of domestic wastewater in the US (Appx 3, Table
S3.2).
17
The experiments of the present study commenced two months after the end of the previous study.
Membrane modules were chemically cleaned both after the end of the prior experiment and before the
commencement of the current study’s experiment. Between cleanings, the AnMBR was continuously
operated with no antibiotics being added to the influent. Steady-state performance of the AnMBR was
reached by the end of this period (defined as consistent COD removal of > 85%, stable biogas production
and methane content of > 60% over at least two weeks of operation). Five days after confirming steady-
state operation, three antibiotics that included sulfamethoxazole (SMX, a sulfonamide), erythromycin
(ERY, a macrolide), and ampicillin (AMP, a β-lactam), were simultaneously added to the influent of the
AnMBR at a concentration of 250 g/L each for a period of one month. Although the estimated total
antibiotic concentration in domestic wastewater is around 50 g/L,
18
a higher non-lethal concentration
11

was used in the present study to emphasize the antibiotic selective pressure impact on microbial
community. The present study was divided into three periods: pre-antibiotics (defined as the time after
steady performance was reached and before antibiotic addition), antibiotics loading, and post-antibiotics.
To monitor the performance of the AnMBR, mixed liquor suspended solid (MLSS), mixed liquor volatile
89
suspended solid (MLVSS), chemical oxygen demand (COD), biogas production, and methane content of
biogas were measured continuously during the experimental period, as described previously.
8
Additional
details on the system and operational parameters were also described previously.
8

4.2.2 Quantification of ARGs by qPCR
Quantitative polymerase chain reaction (qPCR) was performed to quantify targeted ARGs using
procedures described previously.
8
For biomass ARG profiles, 2 mL of mixed liquor was collected biweekly,
centrifuged, and decanted. For effluent ARG profiles, 50 mL of permeate was freeze-dried using a
lyophilizer (FreeZone 2.5 Liter Freeze-Dryer, Labconco, Kansas City, MO). Both biomass and effluent
samples, were then stored at -80 C prior to DNA extraction. DNA extraction was conducted using the
Maxwell 16 Blood DNA Purification kit (Promega, Madison, WI), recommended by the manufacturer for
wastewater sludges, according to manufacturer instructions. qPCR was performed using a LightCycler 96
(Roche, Basel, Switzerland) targeting 8 ARGs commonly found in domestic wastewater
19, 20
including genes
conferring resistance to sulfonamides (sul1 and sul2), macrolides (ermB and ermF), β-lactams (ampC and
oxa-1), and tetracycline (tetO and tetW), as well as a class 1 integrons-associated gene (intl1). Due to
variable operon copy numbers of 16S rRNA in bacteria, a single copy molecular marker gene (rpoB) was
selected for ARG normalization to avoid this bias
21
. Details of thermal cycling and primers for each
targeted gene were provided in Appx 3, Table S3.3.
4.2.3 Quantification of antibiotics by LC-MS
Ten mL of influent and effluent samples were filtered through 0.2 m PTFE syringe filters (Whatman,
GE Healthcare, UK) and then stored at 4 C for no more than 3 days prior to analysis. Antibiotics
quantification was performed using direct injection liquid chromatography electrospray ionization
tandem mass spectrometry (LC-ESI-MS) on a 6560 Ion Mobility Quadrupole Time-of-Flight (IM-QTOF) LC-
90
MS system (Agilent Technologies, Santa Clara, CA) using 1290 Infinity UHPLC, Dual Agilent Jet Stream (ASJ),
and EclipsePlus C18 column (2.1 mm; 50 mm; 1.8 m). Details on method development and calibration
protocols are provided in ESI.
4.2.4 Microbial community analysis
Biomass and effluent extracted DNA samples were sent for sequencing at the Microbial Systems
Molecular Biology Laboratory (University of Michigan, Ann Arbor, MI) where library preparation and
sequencing was performed on the Illumina MiSeq platform using the MiSeq Reagent Kit V2 (2x250 bp
reads). To amplify 16S rRNA gene targets, a universal 16S rRNA gene primer set targeting the V4 region
was used as described previously.
22
High-throughput sequencing results were then analyzed using Mothur
v.1.42.1, with Silva 132 reference database for alignment and Ribosomal Database Project (RDP) reference
taxonomy for classification (rarefaction curves are provided in Appx 3, Figure S3.1). A non-metric
multidimensional scaling (NMDS) plot was conducted to present the distance between samples. To
evaluate the richness and evenness of samples, the Inverse Simpson index was calculated for each
operational phase to serve as a diversity index.
4.2.5 Data analysis
Analysis of molecular variance (AMOVA)
23
was employed to determine the statistical significance of
temporal changes in the microbial community data set. Unweighted principal component analysis (PCA)
was conducted using XLSTAT to compare the distance of different ARG profiles. To determine the
significance of linear relationships between ARG abundances and microbial community profiles,
Spearman’s correlation was conducted using MAXSTAT Pro 3.6 over a 95% confidence interval. Strong
positive correlation was determined based on P-value and Spearman’s coefficients (ρ) where p < 0.05 and
ρ > 0.7.
91
4.3 Results and Discussion
4.3.1 AnMBR system performance was robust during antibiotics addition
The AnMBR was operated at a hydraulic retention time (HRT) of 16 h and solids retention time (SRT) of
300 d. Permeate flux was 7 L/m
2
h (LMH) and transmembrane pressure (TMP) remained lower than 25 kPa
throughout the experimental period. MLSS and MLVSS were relatively constant at 10.6 ± 1.3 and 9.6 ± 0.3
g/L, respectively. Total COD in the influent of the AnMBR averaged 453 ± 32 mg/L. COD removal was 90.0
± 1.8% throughout operation, with antibiotics addition having no significant effect on removal rate. Total
biogas production and methane content were also stable during the experimental period, averaging 736
± 20 and 536 ± 14 mL/d, respectively (Appx 3, Figure S3.2a). Details of COD mass balance in the reactor
were provided in Appx 3, Figure S3.6.
Performance of the AnMBR with regards to antibiotics removal in the effluent (Appx 3, Figure S3.2b)
was similar to a previous study
8
which explored separate addition of the same antibiotics (AMP and SMX
removal range of 89-98% and 69-78%, respectively). One exception was for the case of ERY, which had a
slightly lower removal rate in the present study than previously observed (removal of 40-58% versus 67-
88% in the previous study). In the previous study antibiotics were added individually to the AnMBR and
at incremental concentrations. Thus, it is possible that the simultaneous addition of SMX, ERY, and AMP
at higher concentrations (250 g/L) from the first day of the experiment in the present study resulted in
less effective ERY biotransformation in the system. This may have occurred due to the microbial
community requiring longer acclimation time under the elevated mixed antibiotics conditions.
Erythromycin has shown both higher variability and lower removal rates among mainstream anaerobic
treatment systems, in general, as compared to the other antibiotics used in this study.
24

92
4.3.2 AnMBR biomass was dominated by different ARGs than the AnMBR
effluent
Biomass ARGs were normalized against the rpoB gene (as gene copies/rpoB), while, due to the potential
significance of extracellular ARGs, effluents were normalized against volume (as gene copies/mL). Despite
differing normalization strategies, biomass and effluent ARG profiles and temporal trends were markedly
distinct (Figure 4.1). The only exception was day 46 (post-antibiotics period), where the abundance of all
effluent ARGs increased significantly. The increase in effluent ARGs after the antibiotics loading period
could be due to regrowth of microorganisms harboring ARGs that were non-functional due to regulation
and codon usage bias, which can result in simultaneous ARG presence and antibiotic sensitivity.
25, 26
SMX
and ERY are bacteriostatic antibiotics, halting growth but not necessarily lysing cells, and the only two
antibiotics present in significant concentrations in the effluent.
27
Growth of sensitive ARG-harboring
microorganisms may have been inhibited during the antibiotics phase, resulting in low detection in the
ARG profile. When the antibiotic selective pressure was removed, the inhibited microorganisms may have
increased growth rate, resulting in the observed increase in ARG abundances.
93

Figure 4.1 Abundance of targeted genes in the (a) biomass (copies/rpoB) and (b) effluent (copies/mL) of
the AnMBR throughout the experimental period. The x-axis represents the days after steady performance
of the AnMBR was reached. Day 1 represents pre-antibiotics period, days 6, 14, 20, 27 and 35 represent
antibiotics loading period (area bordered by red dashed line), and day 46 represents post-antibiotics
period. The markers in (a) represent abundance of rpoB (copies/mL; secondary y-axis).
sul1, tetO, and ermF were the most dominant ARGs in the biomass profile, each respectively accounting
for 41.3 ± 3.9%, 26.6 ± 2.4% and 16.7 ± 3.1% of targeted ARGs. Except for sul1, which was also the most
abundant targeted ARG in the AnMBR effluent profiles (accounting for 53.6 ± 12.5%), the dominant ARGs
of the biomass (tetO and ermF) accounted for less than 1% of total targeted genes in the effluent (except
for day 46, as described above). Alternatively, intI1 was the second most abundant gene in the effluent
profile, accounting for 31.0 ± 14.6% of targeted ARGs. Abundances of targeted ARGs throughout the
experimental period in the biomass and effluent of the AnMBR are provided in Appx 3, Table S3.4 and
Table S3.5, respectively.
94
ARG abundance in AnMBR biomass increased throughout operation
After antibiotics addition on day 5, the abundance of class 1 integrons increased significantly in the
biomass profile. However, no significant changes were observed in ARG abundances (Figure 4.1a). Class 1
integrons are associated with HGT due to their presence on mobile genetic elements (MGEs) such as
plasmids and transposons.
28
Thus, the significant increase in the biomass abundance of intl1, one day after
antibiotics addition, could indicate a considerable rise in the presence of MGEs (and HGT). This rise may
correspond to an increase in plasmid-based resistance within the biomass microbial community. This,
along with an increase in the abundance of resistant microorganisms due to the antibiotic selective
pressure, could have subsequently resulted in the significantly higher total biomass ARG abundances
observed after day 6. Specifically, a marked increase in total ARG abundances from 13.3 ± 1.1 on day 6 to
27.3 ± 1.7 copies/rpoB on day 14 was observed. Biomass ARGs that increased on day 14 were ermF, ermB,
sul1, sul2, oxa-1, and tetO. Increases in the abundance of the aforementioned ARGs under selective
pressure of high antibiotics has been previously reported elsewhere.
8, 11
After the initial increase,
abundance of quantified biomass ARGs remained approximately constant at 28.7 ± 2.6 copies/rpoB during
the antibiotics loading period. However, biomass ARGs further increased significantly to 48.3 ± 3.4
copies/rpoB ten days after ceasing addition of antibiotics (day 46). It should be noted that with one sample
for the post-antibiotics period, it is not possible to conclusively evaluate the response of the ARG profiles
to removing the antibiotics selective pressure. Ultimately, both the initiation of antibiotics addition and
its subsequent cessation at the final stage of the experiment likely corresponded with marked increases
in total biomass ARG levels.
AnMBR effluent ARG abundance spiked upon initial antibiotics exposure
Quantified effluent ARG abundances increased approximately 34-fold one day after antibiotics
addition, primarily due to increases in sul1 gene abundance, along with the class 1 integrons-associated
intl1 (Figure 4.1B). Since, these two genes (sul1 and intI1) are commonly co-located on conjugative
95
plasmids, the cause of this phenomenon may relate to the fact that plasmid/extracellular ARGs contribute
significantly to the effluent ARG profiles in membrane-based reactors. To explain this drastic increase in
sul1 and class 1 integrons, it can be inferred that antibiotics addition induced a spike in the rate of HGT in
the biomass that led to a temporary increase in biomass extracellular plasmid DNA. This could have
manifested as a sharp increase in effluent extracellular plasmid DNA and, consequently, a significant rise
in abundance of intI1 and sul1 genes in the effluent ARG profiles. We previously observed an increase in
sul1 and intI1 abundance in AnMBR effluent during individual addition of antibiotics (SMX, ERY and AMP).
8

Further, it has been reported that antibiotic mixtures, as opposed to individual antibiotics, result in more
pronounced changes in ARG abundance.
11
Therefore, the drastic increase in sul1 and intI1 abundance
after simultaneous addition of SMX, ERY, and AMP at a relatively high concentration of 250 μg/L is not
unreasonable. Another possible explanation for this drastic increase might be the considerable release of
MGEs due to the lysis of bacterial cells caused by AMP, which is bactericidal. Eight days after the sharp
spike, total abundance of effluent ARGs decreased by 5-fold, and then decreased another 42-fold on day
20. Effluent ARG profiles remained approximately constant thereafter for the rest of the antibiotics
loading period. These reductions could be due to microbial loss of MGEs such as plasmid DNA along with
lower HGT rates after the initial antibiotics exposure. Another possible explanation for the spike and
subsequent reduction in total effluent ARG abundances could be the development of the membrane
fouling layer (biofilm) and subsequent biofilm-based ARG removal.
8, 29
However, given that the TMP was
consistent during the experimental period, it is possible that the influence of the biofilm on the effluent
ARG profile was negligible.
Comparing the ARG profiles in biomass with those of the AnMBR effluent, it is noteworthy that both
tetO and ermF genes (among the most abundant targeted ARGs in the biomass) were hardly detected in
the effluent. This likely implies that tetO and ermF gene presence on extracellular MGEs was limited while
bacteria harboring these ARGs were concurrently retained in the AnMBR by membrane separation.
96
However, this was not the case for all genes. For instance, the significant increase in biomass abundance
of the oxa-1 gene on day 14 corresponded with a considerable increase in its abundance in the effluent
ARG profile of the same day.
Following the same trend as the biomass ARG profile, effluent ARG abundances also increased after
cessation of the addition of antibiotics. Overall, ARG abundances in both biomass and effluent were
affected significantly by the addition of antibiotics to the influent. However, the mechanisms that dictated
the increases in effluent ARG abundances at the stages of antibiotics addition and cessation, respectively,
may have been vastly different from each other. Specifically, the ARG increases seen at the beginning of
antibiotics addition (days 6 and 14) were driven by different gene combinations (sul1 and intl1) than those
that dominated the gene profile after antibiotics addition ceased on day 46 (predominantly tetO, sul1,
and ermF). Interestingly, the effluent ARG profile of day 46 was also highly similar to that of the biomass
throughout the experiment (Figure 4.1), this observation being confirmed by an unweighted PCA
performed on all of the ARG profiles (Appx 3 Figure S3.3). One possible explanation for this phenomenon
is that the effluent ARG increases on day 6 were caused by elevated occurrence of HGT, while the
increases on day 46 were the result of changes in the microbial composition of the effluent after
antibiotics cessation. Supporting this hypothesis, several anaerobic biomass-associated microbial
populations increased in their relative abundance in the effluent line on day 46. These groups included
Syntrophomonas and Dechloromonas genera that showed strong correlation with biomass-dominating
ARGs (including ermF, ermB and tetO). Microbial dynamics and specific correlations are discussed in more
detail in the following sections.
4.3.3 Microbial community analysis
Biomass relative abundances showed remarkable stability during antibiotics addition
97
Similar to the ARG profiles, the microbial community structure in the biomass was distinct from that of
the effluent. The biomass community structure across the experimental period was highly stable. This
stability of the microbial community even after the addition of multiple antibiotics at concentrations of
250 μg/L is further indication of the ability of AnMBRs to sustain treatment of high antibiotic-containing
wastewaters (e.g., hospital wastewaters). NMDS and AMOVA were employed (Figure 4.2a) to confirm
similarity of the biomass microbial community at the genus level throughout the experimental period.
Results revealed that all biomass community samples clustered closely together with no statistically
significant changes (p = 0.081). Based on the results of our previously published study,
8
it was speculated
that variations in biomass ARG profiles during the addition of antibiotics to the influent could have been
primarily due to changes in the microbial community. However, given the high-level of stability among
community relative abundances before, during, and after antibiotics addition observed in the present
work, it is likely that these antibiotic-influenced changes to the biomass ARG profile are primarily due to
HGT and not microbial community alteration. Occurrence of HGT in anaerobic digesters has been reported
previously.
12, 30

98

Figure 4.2 (a) Non-metric multidimensional scaling (NMDS) and (b) Inverse Simpson index for the biomass
and effluent of the AnMBR throughout the experimental period. The red arrow in the NMDS plot indicates
the significant shift of the effluent samples after antibiotics addition. In the Inverse Simpson plot, the bars
for antibiotic loading represent the average of the diversity index in n = 5 samples during the loading
period (n = 1 for the pre- and post-antibiotics period). Error bars for the pre- and post-antibiotics period
represent the standard deviation calculated by Mothur for each sample, and for the antibiotics loading
period represents the standard deviations of the averages in n = 5.
Bacteroidetes (28.6 ± 1.96%), Chloroflexi (19.7 ± 1.2%), and Proteobacteria (13.2 ± 1.0%) were the most
abundant phyla in the biomass. At the family level (Figure 4.3a), the microbial community was comprised
of 21.3 ± 2.1% unclassified Bacteroidetes and 11.0 ± 0.8% unclassified Chloroflexi. Only 25% of
Bacteroidetes and 44% of Chloroflexi sequences were classified at the family level. Anaerolineaceae (8.59
± 0.50%), Syntrophaceae (5.21 ± 0.33%), Ignavibacteriaceae (4.50 ± 0.27%), Methanoregulaceae (3.70 ±
0.20%), and Methanotrichaceae (2.77 ± 0.18%) were the most abundant classified families in the biomass
community. Most genera associated with Anaerolineaceae are strict anaerobes that are commonly found
in anaerobic treatment system for domestic wastewater
31
and anaerobic digesters. The family
Ignavibacteriaceae contains a single facultative anaerobic genus that is capable of utilizing aromatic
compounds
32
and can also contribute to sulfide oxidation.
33

Relative abundances of both methanogens and syntrophic fatty-acid oxidizing bacteria in the biomass
remained stable throughout operation, averaging 8.68 ± 0.45% and 8.67 ± 0.55% relative abundance
99
across the seven temporal samples (Appx 3, Figure S3.4). Methanolinea (3.73 ± 0.19%) and Methanosaeta
(2.79 ± 0.18%) were the most abundant methanogens, indicating relatively comparable contribution of
hydrogenotrophic (Methanolinea) and acetoclastic (Methanosaeta) methanogenesis. The most abundant
syntrophs were unclassified Syntrophaceae and unclassified Syntrophorhabdaceae, with relative
abundances of 3.96 ± 0.35% and 2.30 ± 0.20%, respectively. The family Syntrophaceae can oxidize long
chain fatty acids to produce acetate and hydrogen,
34
however, the family Syntrophorhabdaceae mainly
oxidizes aromatic compounds, such as benzoate.
35
Based on AMOVA analysis, antibiotics addition did not
significantly affect the relative abundance of methanogens (p = 0.108) or their syntrophic counterparts (p
= 0.095), indicating the robustness of these keystone microbial populations regardless of influent
antibiotics concentration.
100

Figure 4.3 Relative abundance of the (a) biomass and (b) effluent microbial community at the family level
throughout the experimental period. Day 1 represents pre-antibiotics period, days 6, 14, 20, 27 and 35
represent antibiotics loading period (area bordered by red dashed line), and day 46 represents post-
antibiotics period.
101
Dominant effluent microbial groups were affected significantly by antibiotics exposure
Effluent microbial communities were distinct from the biomass, as was clearly elucidated in the NMDS
plot and AMOVA analysis (p = 0.001) (Figure 4.2a). This is not surprising, owing to the fact that the 0.1 μm
membranes used in this AnMBR system likely exclude passage of nearly all microorganisms present in the
biomass. However, the addition of antibiotics to the influent did appear to significantly impact the effluent
microbial community structure (which was not the case for AnMBR biomass). Since transmembrane
pressure was consistent during the operational period at 19.3  2.6 kPa, these changes in effluent
community structure were likely not due to membrane fouling. Firmicutes (63.1%), Proteobacteria
(22.5%), and Bacteroidetes (11.2%) were the most abundant phyla in the pre-antibiotics periods, whereas
the dominant phyla in the effluent during antibiotics addition and in the post-antibiotics period were
Proteobacteria (69.4 ± 7.7%), Firmicutes (15.7 ± 6.0%), and Bacteroidetes (6.47 ± 2.02%). Since these
effluent changes during antibiotics addition did not concurrently change the biomass microbial
community structure, further analysis of microbial diversity was performed to compare the biomass and
effluent communities. Results revealed that the Inverse Simpson index of biomass sample in the pre-
antibiotics period was 8 times higher than the effluent (Figure 4.2b). This higher observed diversity
indicates more evenness and richness among the biomass microbial community structure as compared to
the effluent. The lower diversity of the effluent microbial community may have made this community
more susceptible to inhibition, resulting in the high variability observed upon antibiotics addition.
At the family level (Figure 4.3b), Veillonellaceae (61.07%) was the most abundant population in the pre-
antibiotics period. A significant decrease to 13.3 ± 5.5% relative abundance during the antibiotics loading
and post-antibiotics periods suggests that members of this family are highly susceptible to inhibition from
one or more of the introduced antibiotics. In contrast, the relative abundance of Helicobacteraceae, a
sulfur-oxidizing family, significantly increased after addition of antibiotics from 2.24% in the pre-
antibiotics period to 29.2 ± 5.7% during the antibiotics loading and post-antibiotics periods. Selection of
102
the Helicobacteraceae family after antibiotics addition suggests a high likelihood of multi-drug resistance.
Rhodocyclaceae, another prominent member of the effluent community, increased to 18.3 ± 3.8% after
antibiotics addition. Within Rhodocyclaceae, Zoogloea is an aerobic genus that has been commonly found
in activated sludge systems.
36
The enrichment of this genus in the effluent (Appx 3, Figure S3.5) might be
an indication that aerobic bacteria likely increased their presence over time in the effluent tubes (which
discharged to the open air). However, enrichment of Zoogloea in the effluent of anaerobic reactors has
been reported frequently.
13, 37
Further, several effluent genera have been routinely isolated from biofilm
samples: Novispirillum (Rhodospirillaceae), Arcobacter (Campylobacteraceae), Comamonas
(Comamonadaceae), and Aquabacterium (Comamonadaceae).
38-41
These genera are mostly facultative
anaerobes or capable of growth under anaerobic conditions. Therefore, their presence in the effluent may
be due to seeding and regrowth as a result of membrane permeation by even a relatively small number
of bacteria.
4.3.4 Correlations between ARGs and microbial community structure
indicate a potential for HGT
To further investigate associations of ARGs with microbial communities in the AnMBR, a Spearman’s
correlation analysis was performed between ARG abundances and relative abundance of operational
taxonomic units (OTUs) for the biomass and effluent samples (Appx 3, Table S3.6). We elected to use an
out-based approach for correlation analysis due to the large proportion of unclassified sequences at the
genus level. Network analysis was used to illustrate statistically significant strongly positive correlations
(Figure 4.4) due to the implications that positive ARG-microbial correlations have on ARG association with
specific bacterial groups.
42
The analysis revealed numerous strong positive correlations between certain
ARGs and microbial groups (OTUs) in the biomass. Although statistically significant positive correlation
between a microbial group and a particular ARG cannot be considered as evidence of that group carrying
103
antibiotic resistance, it can indicate the OTUs that are potential host bacteria for ARGs and MGEs.
43

Several of the OTUs in the AnMBR biomass showed strong correlations with multiple ARGs. Therefore,
these groups may have a greater likelihood of serving as potential multi-resistant host bacteria. Based on
this and Table S6, OTU6 (Bacteroidetes), OTU32 (Clostridiales), OTU35 (Verrucomicrobia), OTU84
(Firmicutes), OTU105 (Anaerolineaceae), and OTU134 (Bacteroidetes) in the biomass microbial
community were identified as potential multi-resistant host bacteria for ermF, ermB, sul1, sul2, ampC, and
tetO genes. Occurrence of multi-drug resistant bacteria in anaerobic environments, such as anaerobic
digesters, has been reported previously via strong correlations between microbial community structure
and ARG abundance.
15, 44

Effluent correlations bear strong implications to the bacterial types that are actually entering the
environment through effluent discharge and/or reuse, and therefore are of particular interest. The
correlation analysis resulted in the effluent microbial communities being divided into five distinct
groupings: OTUs with strong correlations to (1) sul1 and intI1, (2) sul2 and oxa-1, (3) ermF, ermB and tetO,
(4) tetW, and (5) genera with no correlations to ARGs. Based on Table S5, Group 1 included OTU25
(Bacteria), OTU27 (Arcobacter), OTU66 (Novispirillum), OTU78 (Comamonas), OTU89 (Flavobacterium),
OTU159 (Acetobacteroides), OTU247 (Bacteria), OTU264 (Rhodococcus), OTU352 (Caulobacter), OTU441
(Sulfuricurvum), OTU501 (Caulobacteraceae), OTU550 (Chryseobacterium), and OTU648 (Bdellovibrio). It
is possible that these groups can serve as potential hosts for sul1-combining class 1 integrons, thus also
being implicated in HGT. sul1 and class 1 integrons have been previously found on the same gene cassettes
in different Arcobacter
46
and Comamonas
47
species in wastewater and soil samples. Strong correlation of
Flavobacterium with these two genes has also been reported in drinking water and soil samples.
48-50

Perhaps most importantly, 7 OTUs of Group 1 are known to harbor pathogenic species. The presence of
class 1 integrons in other pathogenic bacterial strains, including Escherichia coli,
51, 52
Pseudomonas
aeruginosa,
53
and Salmonella spp.,
54
is well established. Thus, the observed dominance of Group 1 by
104
potentially pathogenic groups might be an indication of the transferability of such MGEs to a broader
range of clinically significant bacterial strains. However, one of the limitations of the present study was
the inability to classify microbial groups at the species-level for identification of pathogens.
Group 2 (showing strong correlation to sul2 and oxa-1) included OTU60 (Comamonadaceae), OTU70
(Acinetobacter), OTU111 (Bacteria), OTU163 (Methylophilus), OTU209 (Desulfobulbus), OTU287
(Desulfovibrio), OTU360 (Cryomorphaceae), OTU384 (Spirochaetaceae), and OTU499 (Bosea). Among
group 2, some species of Acinetobacter, Desulfovibrio, Cryomorphaceae, Spirochaetaceae and Bosea
might act as opportunistic pathogens. Acinetobacter spp. are among the six most important multi-drug
resistant organisms in hospitals.
55
They are also frequently used as an antibiotic resistance indicator in
water and wastewater.
56
The presence of sul2
57
and oxa-1
58
genes in different isolates of Acinetobacter
spp. in wastewater samples has been reported previously. Although no putative association between the
other members of group 2 and oxa-1 have been reported in literature, in silico analysis of several
Comamonadaceae species has revealed the presence of open reading frames (ORFs) that correspond to
OXA genes.
59
Further, strong correlation of sul2 genes with Comamonadaceae in treated wastewater,
60

Methylophilus in a soil microbial fuel cell,
61
and Desulfobulbus during anaerobic digestion of cattle
manure
62
have also been reported. Given that the co-location of sul2 and oxa-1 genes on plasmids has
also been previously reported for multiple pathogenic strains,
63, 64
the observed correlations of group 2
with these genes might suggest their potential transferability to such species in the effluent.
Group 3 (strongly correlated with ermF, ermB and tetO) genera included OTU11 (Anaerosinus), OTU21
(Propionispira), OTU48 (Dechloromonas), OTU178 (Rhodocyclaceae), OTU184 (Comamonas), OTU299
(Rhodopseudomonas), and OTU389 (Bacteria). Strong correlations of ermF and ermB genes with a
different tetracycline resistance gene (tetX) have previously been reported in multiple aquatic
environments.
44, 65, 66
Further, the co-location of ermB and tetO genes in multiple Enterococcus spp. has
105
been observed elsewhere.
67
Co-occurrence of ermF and ermB genes with Comamonas,
13
which is typically
regarded as a pathogenic resistant bacteria,
68
has also been reported previously.
The forefront of the threat of antimicrobial resistance is specifically related to its ultimate transfer to
pathogenic bacteria that can reach the environment. Thus, the dynamics of ARGs and MGEs in wastewater
effluent microbial communities that are released to downstream water bodies and in reuse applications
are of critical concern. This concern is further exacerbated when effluents contain significant levels of
multiple pathogenic populations that can acquire these resistance elements through HGT. Correlation
analysis of ARGs with microbial communities in the AnMBR effluent revealed some possible associations
of potential pathogenic groups with at least one ARG. This is relevant, considering that several of these
genera were correlated with Class 1 integrons, which implies a strong potential for HGT in the effluent
environment. However, when putting these observations in the context of AnMBR systems, specifically,
it is important to note that previous studies have reported significantly lower effluent abundances of
pathogenic species in AnMBR effluents than in the effluents of their aerobic MBR counterparts.
69
In the
present study, a sharp decrease in abundances of both ARGs and potentially pathogenic genera in the
effluent were observed after Day 6 of antibiotics addition and remained relatively low until the end of the
experimental period. This suggests that pathogenic and/or resistant bacteria in the effluent were not
sustained beyond the initial antibiotic’s exposure adjustment period. However, It is noteworthy to state
that since the community analysis of the present study was not classified at the species level, no certain
claims can be made about actual pathogenicity of identified OTUs. Therefore, future research is needed
to investigate AnMBR effluents with respect to the association of antibiotic resistance with pathogenic
species.
106

Figure 4 4 Network analysis representing the positive correlations between ARGs (purple circles) and
microbial structure (OTUs) with ≥ 0.5% relative abundance in at least one sample in the (a) biomass and
(b) effluent of the AnMBR. A connection shows strong significant positive correlation (p < 0.05; and ρ >
0.7). The bubble size is indicative of relative abundance.
107
4.4 Conclusions
AnMBRs are an emerging biotechnology for mainstream wastewater treatment with the potential to
enhance energy efficiency and effluent reuse, while also theoretically lessening the spread of antibiotic
resistance to the environment. In this study, a bench-scale AnMBR was employed to treat domestic
wastewater containing antibiotics and investigate the association of microbial communities with ARG
profiles in both biomass and effluent of the AnMBR. The main conclusions of the experiment are as
follows:
• Performance of the AnMBR regarding COD removal, biogas production, and methane yield was
robust during the simultaneous addition of three antibiotics at 250 μg/L.
• ARG profiles and temporal trends in the biomass of the AnMBR were markedly distinct from those
of the effluents.
• Effluent ARG abundance spiked upon initial antibiotics exposure, mostly due to the significant
increase in sul1 and class 1 integrons. It then gradually decreased by around 167-folds and
remained constant during the rest of antibiotics loading period.
• Biomass microbial community structure was unaffected by antibiotics addition and was relatively
uniform throughout the experimental period.
• Antibiotics addition significantly influenced effluent microbial community structure.
• Correlation analysis revealed the existence of potential multi-resistant host bacteria in the
biomass, while also showing that the effluent microbial community contained distinct groups of
bacteria with varied potential mechanisms of resistance.
References
1. R. J. Fair and Y. Tor, Antibiotics and bacterial resistance in the 21st century, Perspectives in medicinal
chemistry, 2014, 6, PMC. S14459.
108
2. T. U. Berendonk, C. M. Manaia, C. Merlin, D. Fatta-Kassinos, E. Cytryn, F. Walsh, H. Bürgmann, H. Sørum,
M. Norström and M.-N. Pons, Tackling antibiotic resistance: the environmental framework, Nature
Reviews Microbiology, 2015, 13, 310.
3. C. L. Ventola, The antibiotic resistance crisis: part 1: causes and threats, Pharmacy and therapeutics,
2015, 40, 277.
4. C. A. Engemann, P. L. Keen, C. W. Knapp, K. J. Hall and D. W. Graham, Fate of tetracycline resistance
genes in aquatic systems: migration from the water column to peripheral biofilms, Environ. Sci. Technol.,
2008, 42, 5131-5136.
5. F. Baquero, J.-L. Martínez and R. Cantón, Antibiotics and antibiotic resistance in water environments,
Current opinion in biotechnology, 2008, 19, 260-265.
6. M. Harb, P. Wang, A. Zarei-Baygi, M. H. Plumlee and A. Smith, Background antibiotic resistance and
microbial communities dominate effects of advanced purified water recharge to an urban aquifer,
Environmental Science & Technology Letters, 2019.
7. M. Harb and P.-Y. Hong, Anaerobic Membrane Bioreactor Effluent Reuse: A Review of Microbial Safety
Concerns, Fermentation, 2017, 3, 39.
8. A. Zarei-Baygi, M. Harb, P. Wang, L. B. Stadler and A. L. Smith, Evaluating Antibiotic Resistance Gene
Correlations with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors, Environmental
science & technology, 2019, 53, 3599-3609.
9. S. Chen, M. Harb, P. Sinha and A. L. Smith, Emerging investigators series: revisiting greenhouse gas
mitigation from conventional activated sludge and anaerobic-based wastewater treatment systems,
Environmental Science: Water Research & Technology, 2018, 4, 1739-1758.
10. S. Aydin, B. Ince and O. Ince, Assessment of anaerobic bacterial diversity and its effects on anaerobic
system stability and the occurrence of antibiotic resistance genes, Bioresource technology, 2016, 207, 332-
338.
11. S. Aydin, B. Ince and O. Ince, Development of antibiotic resistance genes in microbial communities
during long-term operation of anaerobic reactors in the treatment of pharmaceutical wastewater, Water
Res., 2015, 83, 337-344.
12. J. H. Miller, J. T. Novak, W. R. Knocke and A. Pruden, Survival of antibiotic resistant bacteria and
horizontal gene transfer control antibiotic resistance gene content in anaerobic digesters, Frontiers in
microbiology, 2016, 7, 263.
13. J. Tong, A. Tang, H. Wang, X. Liu, Z. Huang, Z. Wang, J. Zhang, Y. Wei, Y. Su and Y. Zhang, Microbial
community evolution and fate of antibiotic resistance genes along six different full-scale municipal
wastewater treatment processes, Bioresource technology, 2019, 272, 489-500.
14. C. J. von Wintersdorff, J. Penders, J. M. van Niekerk, N. D. Mills, S. Majumder, L. B. van Alphen, P. H.
Savelkoul and P. F. Wolffs, Dissemination of antimicrobial resistance in microbial ecosystems through
horizontal gene transfer, Frontiers in microbiology, 2016, 7, 173.
15. F. Ju, B. Li, L. Ma, Y. Wang, D. Huang and T. Zhang, Antibiotic resistance genes and human bacterial
pathogens: co-occurrence, removal, and enrichment in municipal sewage sludge digesters, Water Res.,
2016, 91, 1-10.
109
16. S. Zhang, B. Han, J. Gu, C. Wang, P. Wang, Y. Ma, J. Cao and Z. He, Fate of antibiotic resistant cultivable
heterotrophic bacteria and antibiotic resistance genes in wastewater treatment processes, Chemosphere,
2015, 135, 138-145.
17. A. L. Smith, S. J. Skerlos and L. Raskin, Psychrophilic anaerobic membrane bioreactor treatment of
domestic wastewater, Water research, 2013, 47, 1655-1665.
18. K. Kümmerer, Drugs in the environment: emission of drugs, diagnostic aids and disinfectants into
wastewater by hospitals in relation to other sources–a review, Chemosphere, 2001, 45, 957-969.
19. Y. Ma, C. A. Wilson, J. T. Novak, R. Riffat, S. Aynur, S. Murthy and A. Pruden, Effect of various sludge
digestion conditions on sulfonamide, macrolide, and tetracycline resistance genes and class I integrons,
Environ. Sci. Technol., 2011, 45, 7855-7861.
20. M. Munir, K. Wong and I. Xagoraraki, Release of antibiotic resistant bacteria and genes in the effluent
and biosolids of five wastewater utilities in Michigan, Water research, 2011, 45, 681-693.
21. M. Vos, C. Quince, A. S. Pijl, M. de Hollander and G. A. Kowalchuk, A comparison of rpoB and 16S rRNA
as markers in pyrosequencing studies of bacterial diversity, PLoS One, 2012, 7, e30600.
22. J. J. Kozich, S. L. Westcott, N. T. Baxter, S. K. Highlander and P. D. Schloss, Development of a dual-
index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq
Illumina sequencing platform, Appl. Environ. Microbiol., 2013, 79, 5112-5120.
23. L. Excoffier, P. E. Smouse and J. M. Quattro, Analysis of molecular variance inferred from metric
distances among DNA haplotypes: application to human mitochondrial DNA restriction data, Genetics,
1992, 131, 479-491.
24. M. Harb, E. Lou, A. L. Smith and L. B. Stadler, Perspectives on the fate of micropollutants in mainstream
anaerobic wastewater treatment, Curr Opin Biotechnol, 2019, 57, 94-100.
25. L. M. Carroll, A. Gaballa, C. Guldimann, G. Sullivan, L. O. Henderson and M. Wiedmann, Identification
of Novel Mobilized Colistin Resistance Gene mcr-9 in a Multidrug-Resistant, Colistin-Susceptible
Salmonella enterica Serotype Typhimurium Isolate, mBio, 2019, 10, e00853-00819.
26. A. Yannai, S. Katz and R. Hershberg, The codon usage of lowly expressed genes is subject to natural
selection, Genome biology and evolution, 2018, 10, 1237-1246.
27. P. Wang, M. Harb, A. Zarei-Baygi, L. B. Stadler and A. L. Smith, Comparative analysis of intracellular
and extracellular antibiotic resistance gene abundance in anaerobic membrane bioreactor effluent,
BioRxiv, 2019, 702076.
28. A. Fluit and F. Schmitz, Class 1 integrons, gene cassettes, mobility, and epidemiology, European Journal
of Clinical Microbiology and Infectious Diseases, 1999, 18, 761-770.
29. A. Zarei-Baygi, M. Moslemi and S. H. Mirzaei, The combination of KMnO4 oxidation and polymeric
flocculation for the mitigation of membrane fouling in a membrane bioreactor, Sep. Purif. Technol., 2016,
159, 124-134.
30. H. Huang, Y. Chen, X. Zheng, Y. Su, R. Wan and S. Yang, Distribution of tetracycline resistance genes in
anaerobic treatment of waste sludge: The role of pH in regulating tetracycline resistant bacteria and
horizontal gene transfer, Bioresource technology, 2016, 218, 1284-1289.
110
31. F. T. Saia, T. S. Souza, R. T. D. Duarte, E. Pozzi, D. Fonseca and E. Foresti, Microbial community in a
pilot-scale bioreactor promoting anaerobic digestion and sulfur-driven denitrification for domestic
sewage treatment, Bioprocess and biosystems engineering, 2016, 39, 341-352.
32. M. Rysz, W. R. Mansfield, J. D. Fortner and P. J. Alvarez, Tetracycline resistance gene maintenance
under varying bacterial growth rate, substrate and oxygen availability, and tetracycline concentration,
Environmental science & technology, 2013, 47, 6995-7001.
33. Y. Zhang, L. Zhang, L. Li, G.-H. Chen and F. Jiang, A novel elemental sulfur reduction and sulfide
oxidation integrated process for wastewater treatment and sulfur recycling, Chem. Eng. J., 2018, 342, 438-
445.
34. J. Kuever, The Family Syntrophaceae, The Prokaryotes: Deltaproteobacteria and
Epsilonproteobacteria, 2014, 281-288.
35. T. P. Delforno, A. G. L. d. Moura, D. Y. Okada and M. B. A. Varesche, Effect of biomass adaptation to
the degradation of anionic surfactants in laundry wastewater using EGSB reactors, Bioresource
technology, 2014, 154, 114-121.
36. F. Dias and J. Bhat, Microbial ecology of activated sludge: I. Dominant bacteria, Appl. Environ.
Microbiol., 1964, 12, 412-417.
37. R. Mei, M. K. Nobu, T. Narihiro, K. Kuroda, J. M. Sierra, Z. Wu, L. Ye, P. K. Lee, P.-H. Lee and J. B. van
Lier, Operation-driven heterogeneity and overlooked feed-associated populations in global anaerobic
digester microbiome, Water Res., 2017, 124, 77-84.
38. K. Calderón, B. Rodelas, N. Cabirol, J. González-López and A. Noyola, Analysis of microbial communities
developed on the fouling layers of a membrane-coupled anaerobic bioreactor applied to wastewater
treatment, Bioresource technology, 2011, 102, 4618-4627.
39. L. Oliveira, R. Costa, I. Sakamoto, I. Duarte, E. Silva and M. Varesche, Las degradation in a fluidized bed
reactor and phylogenetic characterization of the biofilm, Brazilian Journal of Chemical Engineering, 2013,
30, 521-529.
40. J. K. Braga, F. Motteran, E. L. Silva and M. B. A. Varesche, Evaluation of bacterial community from
anaerobic fluidized bed reactor for the removal of linear alkylbenzene sulfonate from laundry wastewater
by 454-pyrosequence, Ecol. Eng., 2015, 82, 231-240.
41. I. Naz, D. Hodgson, A. Smith, J. Marchesi, S. Ahmed, C. Avignone-Rossa and D. P. Saroj, Effect of the
chemical composition of filter media on the microbial community in wastewater biofilms at different
temperatures, RSC advances, 2016, 6, 104345-104353.
42. W. Tao, X.-X. Zhang, F. Zhao, K. Huang, H. Ma, Z. Wang, L. Ye and H. Ren, High levels of antibiotic
resistance genes and their correlations with bacterial community and mobile genetic elements in
pharmaceutical wastewater treatment bioreactors, PloS one, 2016, 11, e0156854.
43. J. Zhang, M. Yang, H. Zhong, M. Liu, Q. Sui, L. Zheng, J. Tong and Y. Wei, Deciphering the factors
influencing the discrepant fate of antibiotic resistance genes in sludge and water phases during municipal
wastewater treatment, Bioresource technology, 2018, 265, 310-319.
44. J. Zhang, M. Chen, Q. Sui, R. Wang, J. Tong and Y. Wei, Fate of antibiotic resistance genes and its drivers
during anaerobic co-digestion of food waste and sewage sludge based on microwave pretreatment,
Bioresource technology, 2016, 217, 28-36.
111
45. M. Gillings, Y. Boucher, M. Labbate, A. Holmes, S. Krishnan, M. Holley and H. W. Stokes, The evolution
of class 1 integrons and the rise of antibiotic resistance, J. Bacteriol., 2008, 190, 5095-5100.
46. A. Tomova, L. Ivanova, A. H. Buschmann, H. P. Godfrey and F. C. Cabello, Plasmid-mediated quinolone
resistance (PMQR) genes and class 1 integrons in quinolone-resistant marine bacteria and clinical isolates
of Escherichia coli from an aquacultural area, Microbial ecology, 2018, 75, 104-112.
47. M. Lin, J. Liang, X. Zhang, X. Wu, Q. Yan and Z. Luo, Genetic diversity of three classes of integrons in
antibiotic-resistant bacteria isolated from Jiulong River in southern China, Environmental Science and
Pollution Research, 2015, 22, 11930-11939.
48. M. Duan, H. Li, J. Gu, X. Tuo, W. Sun, X. Qian and X. Wang, Effects of biochar on reducing the abundance
of oxytetracycline, antibiotic resistance genes, and human pathogenic bacteria in soil and lettuce, Environ.
Pollut., 2017, 224, 787-795.
49. I. Vaz-Moreira, O. C. Nunes and C. M. Manaia, Ubiquitous and persistent Proteobacteria and other
Gram-negative bacteria in drinking water, Sci. Total Environ., 2017, 586, 1141-1149.
50. G. Qiu, Y. Song, P. Zeng, L. Duan and S. Xiao, Combination of upflow anaerobic sludge blanket (UASB)
and membrane bioreactor (MBR) for berberine reduction from wastewater and the effects of berberine
on bacterial community dynamics, Journal of hazardous materials, 2013, 246, 34-43.
51. K. Kadlec and S. Schwarz, Analysis and distribution of class 1 and class 2 integrons and associated gene
cassettes among Escherichia coli isolates from swine, horses, cats and dogs collected in the BfT-GermVet
monitoring study, J. Antimicrob. Chemother., 2008, 62, 469-473.
52. S. E. Murinda, P. D. Ebner, L. T. Nguyen, A. G. Mathew and S. P. Oliver, Antimicrobial resistance and
class 1 integrons in pathogenic Escherichia coli from dairy farms, Foodbourne Pathogens & Disease, 2005,
2, 348-352.
53. E. Martinez, C. Marquez, A. Ingold, J. Merlino, S. P. Djordjevic, H. Stokes and P. R. Chowdhury, Diverse
mobilized class 1 integrons are common in the chromosomes of pathogenic Pseudomonas aeruginosa
clinical isolates, Antimicrob. Agents Chemother., 2012, 56, 2169-2172.
54. H. Zhang, L. Shi, L. Li, S. Guo, X. Zhang, S. Yamasaki, S.-i. Miyoshi and S. Shinoda, Identification and
Characterization of Class 1 Integron Resistance Gene Cassettes among Salmonella Strains Isolated from
Healthy Humans in China, Microbiology and Immunology, 2004, 48, 639-645.
55. L. Antunes, P. Visca and K. J. Towner, Acinetobacter baumannii: evolution of a global pathogen,
Pathogens and disease, 2014, 71, 292-301.
56. L. Guardabassi, A. Petersen, J. E. Olsen and A. Dalsgaard, Antibiotic resistance in Acinetobacterspp.
isolated from sewers receiving waste effluent from a hospital and a pharmaceutical plant, Appl. Environ.
Microbiol., 1998, 64, 3499-3502.
57. P. T. P. Hoa, L. Nonaka, P. H. Viet and S. Suzuki, Detection of the sul1, sul2, and sul3 genes in
sulfonamide-resistant bacteria from wastewater and shrimp ponds of north Vietnam, Sci. Total Environ.,
2008, 405, 377-384.
58. M. A. Islam, M. Islam, R. Hasan, M. I. Hossain, A. Nabi, M. Rahman, W. H. Goessens, H. P. Endtz, A. B.
Boehm and S. M. Faruque, Environmental spread of New Delhi metallo-β-lactamase-1-producing
multidrug-resistant bacteria in Dhaka, Bangladesh, Appl. Environ. Microbiol., 2017, 83, e00793-00717.
112
59. L. Poirel, T. Naas and P. Nordmann, Diversity, epidemiology, and genetics of class D β-lactamases,
Antimicrob. Agents Chemother., 2010, 54, 24-38.
60. T. Fernandes, I. Vaz-Moreira and C. M. Manaia, Neighbor urban wastewater treatment plants display
distinct profiles of bacterial community and antibiotic resistance genes, Environmental Science and
Pollution Research, 2019, 26, 11269-11278.
61. X. Zhao, X. Li, Y. Li, Y. Sun, X. Zhang, L. Weng, T. Ren and Y. Li, Shifting interactions among bacteria,
fungi and archaea enhance removal of antibiotics and antibiotic resistance genes in the soil
bioelectrochemical remediation, Biotechnology for Biofuels, 2019, 12, 160.
62. W. Sun, J. Gu, X. Wang, X. Qian and H. Peng, Solid-state anaerobic digestion facilitates the removal of
antibiotic resistance genes and mobile genetic elements from cattle manure, Bioresource technology,
2019, 274, 287-295.
63. A. M. Hammerum, F. Hansen, H. L. Nielsen, L. Jakobsen, M. Stegger, P. S. Andersen, P. Jensen, T. K.
Nielsen, L. H. Hansen and H. Hasman, Use of WGS data for investigation of a long-term NDM-1-producing
Citrobacter freundii outbreak and secondary in vivo spread of bla NDM-1 to Escherichia coli, Klebsiella
pneumoniae and Klebsiella oxytoca, J. Antimicrob. Chemother., 2016, 71, 3117-3124.
64. J. R. Mediavilla, A. Patrawalla, L. Chen, K. D. Chavda, B. Mathema, C. Vinnard, L. L. Dever and B. N.
Kreiswirth, Colistin-and carbapenem-resistant Escherichia coli harboring mcr-1 and blaNDM-5, causing a
complicated urinary tract infection in a patient from the United States, MBio, 2016, 7, e01191-01116.
65. Q. Sui, X. Meng, R. Wang, J. Zhang, D. Yu, M. Chen, Y. Wang and Y. Wei, Effects of endogenous
inhibitors on the evolution of antibiotic resistance genes during high solid anaerobic digestion of swine
manure, Bioresource technology, 2018, 270, 328-336.
66. Y. Chen, J.-Q. Su, J. Zhang, P. Li, H. Chen, B. Zhang, K. Y.-H. Gin and Y. He, High-throughput profiling of
antibiotic resistance gene dynamic in a drinking water river-reservoir system, Water Res., 2019, 149, 179-
189.
67. C.-L. Tremblay, A. Letellier, S. Quessy, D. Daignault and M. Archambault, Antibiotic-resistant
Enterococcus faecalis in abattoir pigs and plasmid colocalization and cotransfer of tet (M) and erm (B)
genes, J. Food Prot., 2012, 75, 1595-1602.
68. U. C. Oppermann, I. Belai and E. Maser, Antibiotic resistance and enhanced insecticide catabolism as
consequences of steroid induction in the gram-negative bacterium Comamonas testosteroni, The Journal
of steroid biochemistry and molecular biology, 1996, 58, 217-223.
69. M. Harb and P.-Y. Hong, Molecular-based detection of potentially pathogenic bacteria in membrane
bioreactor (MBR) systems treating municipal wastewater: a case study, Environmental Science and
Pollution Research, 2017, 24, 5370-5380.

113
Chapter 5

5. Membrane fouling inversely impacts intracellular and extracellular
antibiotic resistance gene abundances in the effluent of an
anaerobic membrane bioreactor

Abstract
Anaerobic membrane bioreactors (AnMBRs) can significantly reduce the release of antibiotic resistance
elements to the environment. The purpose of this study was to elucidate the role of membrane fouling layers
(biofilms) in mitigating the release of intracellular and extracellular antibiotic resistance genes (iARGs and
eARGs) from an AnMBR. The AnMBR was equipped with three membrane modules, each exhibiting a different
level of fouling. Results showed that the absolute abundance of ARGs decreased gradually in the suspended
biomass during operation of the AnMBR. Normalized abundances of targeted ARGs and intI1 were found to be
significantly higher in the fouling layers compared to the suspended biomass, implying adsorption or an
increased potential for horizontal gene transfer of ARGs in the biofilm. Effluent ARG data revealed that the
highly fouled (HF) membrane significantly reduced the absolute abundance of eARGs. However, the HF
membrane effluent concomitantly had the highest absolute abundance of iARG. Nevertheless, total ARG
abundance (sum of iARG and eARG) in effluent of the AnMBR was not impacted by the extent of fouling. These
results suggest a need for a combination of different treatment technologies to effectively prevent antibiotic
resistance proliferation associated with these two ARG fractions.
114
5.1 Introduction
Antibiotic resistance is a serious global health concern estimated to be responsible for over 700,000
deaths per year worldwide.
1
According to a 2019 report by the US Centers for Disease Control and
Prevention (CDC), more than 2.8 million people suffer from antibiotic resistant infections annually in the
US.
2
This information underscores the urgent need for action to significantly reduce the spread of
antibiotic resistance. Wastewater treatment plants (WWTPs) are a hotspot for both antibiotic resistant
bacteria (ARB) and antibiotic resistance genes (ARGs) and subsequent release to the environment.
3, 4
The
high density of microorganisms in WWTPs, along with sublethal concentrations of various antibiotics,
provide an environment that accelerates the spread of resistance among bacteria.
4, 5
Therefore, it is crucial
to investigate the fate of antibiotic resistance in WWTPs and devise operational strategies and/or design
elements that mitigate their dissemination.
Advanced wastewater treatment technologies such as membrane bioreactors (MBRs) can serve to
reduce (by 1-3 log compared to conventional treatment processes) the release of ARB and ARGs from
WWTPs.
6
Anaerobic membrane bioreactors (AnMBRs) are an emerging biotechnology which can produce
a similar particulate-free effluent to MBRs, along with the ability to directly recover energy from
wastewater via methane-rich biogas.
7
Compared with aerobic microorganisms, anaerobic microorganisms
have significantly lower growth rates. This results in lower sludge yields from anaerobic bioprocesses,
which can help to reduce the release of antibiotic resistance to the environment by minimizing land
application of biosolids.
8-10
Nevertheless, fate of antibiotic resistance elements during mainstream
wastewater treatment by AnMBRs remains understudied.
ARGs in aquatic environments are present as both intracellular and extracellular forms of DNA (iDNA
and eDNA).
11
Horizontal transfer of intracellular ARGs (iARGs) predominantly occurs through conjugation
and transduction, while extracellular ARGs (eARGs) spread predominantly through transformation to
115
naturally competent bacteria.
11, 12
iARGs can also spread through vertical gene transfer (VGT). iARGs can
be converted to eARGs via cell death and lysis. The form of the ARG impacts the mechanism by which it
can be transferred, in turn suggesting that the rates of horizontal transfer of iARGs and eARGs in
wastewater systems could be significantly different.
13
Recent studies have revealed that the
transformation of eARGs, specifically, plays a crucial role in the spread of antibiotic resistance in
WWTPs.
14-16
In AnMBRs, since a large fraction of iARGs are retained in the reactor by the membrane
barrier (size exclusion), the effect of effluent eARGs on the spread of antibiotic resistance in receiving
environments may be more prevalent.
9, 10, 15
Further, it has been reported that eDNA can persist in soils
and sediments for up to several years.
11, 13
The persistence of eDNA in soils and sediments may indicate
that soil/sediment-associated ARGs have a high potential for dissemination via horizontal transfer of
eARGs.
11, 13, 16

Along with the membrane barrier, membrane fouling can significantly impact the abundance of ARGs
in AnMBR effluents.
17-19
It has been reported that high biological activity in the membrane fouling layer of
AnMBRs provides additional biodegradation of organic compounds
20
and this high biological activity may
also be a factor that influences ARG effluent abundances. Further, the high density and temporally stable
spatial organization of microbial cells in the fouling layer increases the likelihood of interaction between
cells,
21
which along with the accumulation of mobile genetic elements,
22
may provide a conducive
environment for horizontal gene transfer. Formation of biofilms can also significantly alter filtration
processes by affecting size exclusion and adsorption properties.
17
Therefore, it is likely that the membrane
biofilm also directly influences the proliferation of antibiotic resistance. For instance, it has been reported
that principle components of membrane biofilms, namely extracellular polymeric substances (EPS) and
soluble microbial products (SMPs), in an aerobic MBR treating antibiotic-containing wastewater had a
large capacity to retain ARGs in the reactor.
19
In AnMBR-fouled membrane filtration experiments, Cheng
and Hong observed a positive correlation between the extent of membrane fouling and ARG removal.
17

116
Despite the important observations of this study, the membranes were tested ex-situ over short
operational periods. This is a significant limitation of the work, especially considering the dynamic nature
of biofilm matrices under in-situ filtration conditions. No studies to date have systematically investigated
the role of membrane fouling on release of both iARGs and eARGs from a continuously operated AnMBR.
Here, we evaluated the relative abundances of ARGs in the suspended biomass and membrane biofilms
of an AnMBR fed a synthetic wastewater amended with three antibiotics to provide a selective pressure.
We further evaluated the absolute abundance of iARGs and eARGs in three independent effluents from
three submerged membranes with different levels of fouling. EPS and SMP content of suspended biomass
and membrane biofilms were analyzed to evaluate the potential relationship between these fouling-
associated components and the abundance of iARGs and eARGs in the effluent.
5.2 Material and Methods
5.2.1 Configuration of reactor and preparation of membrane modules
with different levels of fouling
The AnMBR system was operated with three membrane modules with effective membrane area of
0.015 m
2
each and membrane pore size of 0.1 μm. The membranes were submerged in a continuously
stirred reactor (Chemglass LifeScience, Vineland, NJ) with a working volume of 5 L. The seed sludge was
collected from a mesophilic anaerobic digester at the Joint Water Pollution Control Plant in Carson, CA.
The reactor was operated at 25 C and fed a synthetic wastewater (Table S1),
9, 23
containing antibiotics
from different classes including sulfamethoxazole (SMX), erythromycin (ERY), and ampicillin (AMP) at
concentrations of 50 μg/L each. Total antibiotic concentrations in domestic wastewater are generally
around 50 μg/L,
24, 25
however, here we used a higher non-lethal concentration for the purpose of applying
117
a selective pressure. Hydraulic and sludge retention times (HRT and SRT) of the AnMBR were maintained
at 16 h and 300 d, respectively.
After reaching steady-state AnMBR performance, which was defined as consistent chemical oxygen
demand (COD) removal of >85% and stable biogas production and methane content of >60%, the reactor
was operated for 3 weeks using three clean membrane modules (Phase 1; as a control experiment).
Afterwards, two of the membrane modules were removed, cleaned, and placed back into the reactor.
Operation of the AnMBR was then continued for another 3 weeks (Phase 2). Next, one of the previously
cleaned membrane modules was removed, cleaned, and placed back into the reactor. Therefore, starting
on the 7
th
week of the experiment, the AnMBR contained three membrane modules: one without cleaning
for six weeks (highly fouled membrane, HF), one without cleaning for three weeks (medium fouled
membrane, MF), and a freshly cleaned membrane (low fouled membrane, LF). Subsequently, operation
of the AnMBR was continued for another one month to evaluate performance and ARG profiles of the
suspended biomass, biofilms, and three effluents to determine the impact of fouling (Phase 3). COD,
mixed liquor suspended solid (MLSS), mixed liquor volatile suspended solid (MLVSS), biogas production,
and methane content of biogas were also measured continuously during the experiment as detailed
previously.
9, 10
The transmembrane pressure (TMP) of each membrane module was continuously
monitored and recorded using LabVIEW 2014 (Student Edition). 10 mL of both influent and effluent
samples (from each membrane module) were used to analyze antibiotics concertation by direct injection
liquid chromatography mass spectrometry with electrospray ionization (LC-ESI-MS). Additional details on
the AnMBR operation are provided in the Supporting Information (SI).
5.2.2 Quantification of EPS and SMPs
Mixed liquor samples were centrifuged at 12,000 g for 45 minutes and then the supernatant was
decanted and filtered through 0.2 um PTFE syringe filters (Whatman, GE Healthcare, UK) for SMP
118
analysis.
26
For EPS extraction using a modified heat extraction method,
27, 28
the sludge pellets were
resuspended in 1.5 mL phosphate buffer saline (PBS) solution, vigorously vortexed for 15 min, treated
ultrasonically for 3 min, and heated to 80 °C in a water bath for 30 min. The mixtures then were
centrifuged at 10,000 g for 20 min, and the supernatant was collected for EPS analysis. Although EPS and
SMPs consist of polysaccharides, proteins, humic substances, and other macromolecules, they are
typically reported as the sum of polysaccharide and protein concentrations. Polysaccharides and proteins
were quantified using the DuBois
29
and Lowry
30
methods using a Phenol-Sulfuric Acid Assay and a Pierce™
BCA Protein Assay (Thermo Scientific, NY), respectively.
5.2.3 DNA extraction and ARGs quantification
Two mL of suspended biomass samples were collected weekly, centrifuged, decanted, and stored at -
80 °C. Biofilm samples were taken by gently scraping off the cake layer using a plastic sheet at the end of
Phase 3 from the surface of MF and HF membranes (biofilm had not formed on the surface of the LF
membrane). Additionally, given that a thick cake layer was formed on the surface of the HF membrane,
two layers of biofilm samples (outer/top cake layer and inner/bottom cake layer) were obtained from the
biofilm biomass of this membrane module. The outer layer of biofilm contained loose biomass that was
not compressed. The inner layer, however, was compressed and tightly attached to the membrane
surface. This biomass sampling method was consistent with previous studies.
31, 32
Permeate from each
membrane module (150 mL each) were collected to quantify both iARGs and eARGs in the effluents.
Collected effluent samples were spiked with 2  10
6
copies of plasmid pUC19 to serve as an eDNA
extraction efficiency standard. According to previous studies,
11, 14
eDNA is generally defined as DNA which
passes through a 0.22 µm filter, while iDNA is defined as DNA that does not. Effluent samples were filtered
through a sterile 0.22 μm filter membrane (Whatman, GE Healthcare, UK), and processed filters with
residual cells and filtrate were used for iDNA and eDNA extractions, respectively. Before DNA extraction
119
(of biomass and processed filters with residual cells), 8  10
9
copies of YaiO (a single copy per cell E. coli
gene) were spiked into each of the biomass/biofilm samples and processed filters (effluent iDNA) to serve
as an internal standard for DNA extraction efficiency. DNA extraction (of biomass and processed filters
with residual cells) was performed using the Maxwell 16 Blood DNA purification kit (Promega, Madison,
WI), recommended by the manufacturer for wastewater sludges, according to manufacturer instructions.
eDNA extraction was performed using a modified method from a previous study.
14
Briefly, eDNA in the
filtrate was precipitated out using 100% isopropanol at a final mixing ratio of 1:1. The filtrate-isopropanol
samples were incubated overnight at room temperature. Samples were then centrifuged at 12,000 g for
30 min to pellet the eDNA. The pellets were washed with 70% ethanol and centrifuged at 12,000 g for 10
min two consecutive times. After decanting the supernatant, the pellet was air dried for 5 min at room
temperature and then resuspended in 600 ul of TE buffer. The resuspended eDNA samples were then
treated with 60 uL of proteinase K and incubated at 56 °C for 20 min. Finally, the proteinase K treated
eDNA samples were purified using the DNA Maxwell instrument.
Quantitative polymerase chain reaction (qPCR) was conducted using a LightCycler 96 (Roche, Basel,
Switzerland) to quantify target ARGs commonly found in wastewater, which included genes conferring
resistance to sulfonamides (sul1 and sul2), erythromycin (ermF and ermB), β-lactams (oxa-1 and ampC),
and tetracycline (tetW and tetO), as well as the intl1 gene, which encodes for class 1 integrons.
5, 6
The
rpoB gene (a single copy molecular marker gene) was also quantified to represent total cell abundance.
33

Additional details on qPCR reactions, thermal cycling conditions, and primers used are provided in the SI.
5.2.4 Data analysis
2-tailed unpaired student’s t-test was employed to determine the statistically significant differences
between the results with 95% confidence interval. To determine the significance of linear correlation
between the extent of fouling and iARGs, eARGs, and total ARGs, Pearson’s correlation was conducted
120
using MAXSTAT Pro 3.6 over a 95% confidence interval. Strong correlation was determined based on p-
value and Pearson’s coefficients (ρ), where p < 0.05 and ρ > 0.7 or ρ < -0.7.
5.3 Results and Discussion
5.3.1 Fouling improved treatment performance of the AnMBR
The AnMBR was operated at steady-state performance for over 6 months, and the results of the last 75
days of operation, starting from Phase 1 (control experiment), are shown in Figure 5.1. MLSS and MLVSS
were stable throughout the experiment at 7.8 ± 0.4 g/L (mean ± SD) and 6.4 ± 0.5 g/L, respectively. Biogas
production and methane content were also consistent during operation, averaging 693 ± 22 mL/d and 478
± 34 mL/d, respectively. Effluent COD of all three membrane modules was similar during the control
experiment (Phase 1) at 47.2 ± 4.9 mg/L. During Phase 3, however, effluent COD of the MF (25.8 ± 5 mg/L)
and HF (12.4 ± 3.8 mg/L) membranes were both lower than that of the LF membrane (45.2 ± 6.6 mg/L),
indicating the ability of the membrane fouling layer to further reduce COD (Figure 5.1). Smith et al.
similarly observed that fouling can enhance COD removal in the effluent of an AnMBR treating synthetic
domestic wastewater, which coincided with enrichment of methanogens and syntrophic bacteria in the
biofilm community.
20

Membrane permeate flux was maintained at 7 L/m
2
/h throughout the experiment for all three
membrane modules. TMP results revealed a similar trend for all three membrane modules in Phase 1,
where the fouling rate was 0.33 ± 0.01 kPa/day ( TMP/ t). At the start of Phase 3, TMP of the LF, MF,
and HF membranes were 3 kPa, 9 kPa, and 20 kPa, respectively. TMP of each membrane module
continually increased during Phase 3 at fouling rates of 0.32 kPa/day, 0.52 kPa/day, and 0.82 kPa/day,
respectively, ultimately reaching 14 kPa (LF), 22 kPa (MF), and 44 kPa (HF) at the end of this phase (Figure
121
5.2a). The higher fouling rate of the HF membrane in Phase 3 was likely a result of compression of the
cake layer on the membrane surface.
26, 34

Figure 5.1 Performance of the AnMBR for COD removal and biogas/methane production. Phase 1 and 2
of operation were used to achieve different levels of fouling on the membrane modules (Phase 1 also
represents a control experiment where each membrane had the same level of fouling); Phase 3 (main
experimental period) is the period in which each membrane was operated under different fouling levels.
LF, MF, and HF represent low, medium, and high fouled membranes in Phase 3, respectively.
Both polysaccharide and protein components of EPS and SMPs in the fouling layers (MF and HF) were
observed to be significantly higher than those of the suspended biomass (Figure 5.2b and 5.2c). Higher
concentration of EPS and SMP content in the fouling layer of a submerged AnMBR compared to its
suspended biomass has been reported previously.
35
The fouling layer of the HF membrane had roughly
2.5 and 1.5 times more EPS and SMP content, respectively, than the MF membrane, indicating that the
relative overall difference in EPS content was greater than that of SMPs.
122

Figure 5.2 (a) Variation of the transmembrane pressure (TMP), (b) concentration of EPS, and (c)
concentration of SMP content in suspended biomass (during Phase 3; n=5), MF, and HF membrane
biofilms (at the end of Phase 3). Phase 1 and 2 of operation were used to achieve different levels of fouling
on the membrane modules (Phase 1 also represents a control experiment where each membrane had the
same level of fouling); Phase 3 (main experimental period) is the period in which each membrane was
operated under different fouling levels. LF, MF, and HF represent low, medium, and high fouled
membranes in Phase 3, respectively. TMP data is the daily average calculated from recorded data (every
minute) via LabVIEW.
123
5.3.2 AnMBR membrane biofilms may provide more conducive conditions
for ARG transfer than suspended biomass
The most abundant ARGs in the suspended biomass were sulfonamide resistance genes (sul1 and sul2),
followed by ermF and tetO (Figure 5.3a and 5.3c), as was previously observed in studies on ARGs in AnMBR
systems.
9, 10, 36
The total absolute abundance of suspended biomass ARGs (copies/g dry weight (dw)) and
the intI1 gene decreased gradually during Phase 3 (which was confirmed by correlation analysis between
the days of experiment and total absolute abundance of biomass ARGs on each day), which has also been
previously observed for the anaerobic treatment of domestic wastewater.
37
These observations reiterate
the potential advantages of anaerobic treatment (over aerobic) for the mitigation of ARG dissemination
from WWTPs, which may be attributable to lower biomass growth rates in anaerobic systems.
38
sul2 was
the only ARG that did not decrease during Phase 3 and remained approximately constant in suspended
biomass samples. The absolute abundance of the rpoB gene in the biomass was approximately constant
during Phase 3 (averaging 8.74  10
12
± 6  10
11
copies/g dw), indicating that decreases in suspended
biomass ARG abundances were not due to bacterial decay alone.
Consistent with the biomass ARG profiles, sul1, sul2, ermF, and tetO were the most abundant ARGs in
the fouling layers (Figure 5.3b and 5.3d). Comparing ARG profiles of the suspended biomass (in the last
day of the experiment, day 73, which was the closest day to biofilm sampling) with those of the biofilms
revealed a higher abundance of all targeted ARGs in the suspended biomass. However, given that the
average absolute abundance of rpoB (copies/g) in the suspended biomass was around 7 times higher than
the biofilms, the greater absolute abundance of ARGs was predominantly attributable to higher overall
bacterial counts. Despite the higher cell counts in the suspended biomass (in 1 g dw), the absolute
abundance of the intI1 gene in the fouling layers was not significantly different than the suspended
biomass.
124

Figure 5.3 Abundance of targeted ARGs and intI1 gene in (a) and (c) suspended biomass during Phase 3,
and (b) and (d) MF and HF membrane biofilm (at the end of Phase 3). (a) and (b) are absolute abundance,
(c) and (d) are relative abundance (normalized to rpoB). MF and HF represents medium and highly fouled
membranes, respectively. HF outer and inner layers respectively represent the HF membrane biofilm in
125
the vicinity of biomass and membrane surface. The averaged absolute ARG abundances in the outer and
inner layers of the HF membrane biofilm were equal to the absolute ARG abundances of the whole cake
layer.
When both ARG profiles (suspended biomass and biofilms) were normalized to rpoB, the relative
abundance of total ARGs in the biofilm samples (approximately 21 and 42 copies/rpoB in MF and HF
membranes, respectively) were significantly higher (P < 0.0003) than that in the suspended biomass
(approximately 13 copies/rpoB; Figure 5.3c and 5.3d). Similar results were observed for the mixed liquor
and foulant layers of a sequencing batch MBR treating swine wastewater.
15
Accumulation of ARGs in the
biofilm might be due to adsorption and/or greater rates of horizontal transfer of ARGs compared to the
suspended biomass. Adsorption of ARGs by EPS and SMP in fouling layers has also been observed
previously in an aerobic MBR. Greater rates of horizontal transfer of ARGs also seems likely, considering
that previous reports have indicated that biofilms can increase the potential for horizontal transfer of
mobile genetic elements.
22, 39
It has also been reported previously that eARGs are more abundant than
iARGs in biofilms.
21
This, along with the higher likelihood of interactions within microbial consortiums
existing at high cell density in a three-dimensional polymer network (biofilms),
22, 23
may significantly
increase the rate of horizontal ARG transfer. For example, conjugation in biofilms has been reported to
be up to 700-folds more efficient than in planktonic bacteria cell environments.
41

Comparing the ARG profiles of the different fouling layers, both absolute (copies/g) and relative
(copies/rpoB) abundance of ARGs in the HF membrane biofilm were approximately two-fold higher than
that of the MF membrane biofilm (Figure 5.3b and 5.3d). Cheng et al. also observed a higher abundance
of three ARB and their associated plasmid-borne ARGs in the cake layers of membranes with higher
degrees of fouling.
17
This suggests that increased membrane fouling results in higher retention of ARGs
and/or provides conditions conducive to horizontal ARG transfer in the mature cake layer. A previous
126
study by Singh et al. also found that, compared to less developed biofilm layers, horizontal gene transfer
occurs with a higher frequency in mature biofilms.
39

Comparing the ARG profiles in the outer and inner layers of the HF membrane biofilm revealed that the
absolute and relative abundances of all targeted ARGs and the intI1 gene were higher in the inner layer
of the HF membrane biofilm (Figure 5.3b and 5.3d). The only exceptions were sul2 and ermB genes, which
were more abundant in the outer layer. The co-occurrence of sul2 and ermB in anaerobic sludge have
been reported previously.
9
The observed lower absolute abundance of sul2 and ermB in the inner layer of
the HF membrane biofilm could imply that their horizontal transfer rates was not promoted under
attached growth conditions and/or the bacteria harboring these ARGs (sul2 and ermB) were not enriched
as much as other ARG host bacteria in the inner layer. The absolute abundance of the rpoB gene
(normalized to biomass weight; copies/g) was also greater in the outer layer, resulting in a significantly
higher relative abundance of ARGs in the inner layer (approximately 70 copies/rpoB) compared to the
outer layer (approximately 31 copies/rpoB) of the HF membrane biofilm. This, along with higher intl1
abundance, implies the likelihood of a significantly higher accumulation of mobile genetic elements
carrying ARGs in the inner layer as compared to the outer.
39

5.3.3 Membrane fouling increased the abundance of iARGs and
decreased the abundance of eARGs in the AnMBR effluent
As expected, the effluent ARG profiles of the three membrane modules were not significantly different
from one another (p > 0.05) during the control experiment (when fouling rates of all three modules were
identical; Phase 1). In the effluent samples, both iDNA and eDNA profiles were dominated by the sul1
gene, followed by intI1 and sul2 (Figure 5.4 and Figure S4.3). High abundance of the aforementioned genes
in the effluent of AnMBRs has been reported previously.
9, 10
Although all targeted genes (ARGs, intI1, and
127
rpoB) were detected in the effluent iDNA samples, ermB, tetO, tetW, and rpoB in the effluent eDNA
samples were below the detection limit (50 copies/mL, Figure 5.4 and Figure S4.3). In other work
comparing wastewater iDNA and eDNA, Sui et al. also found that tetW and ermB genes were below the
detection limit in eDNA effluent fractions of a pilot-scale sequencing batch MBR and a full-scale anaerobic-
anoxic-oxic system treating swine wastewater.
15
Absolute abundance of total iARGs was consistently
higher than eARGs in all effluent samples of the AnMBR. Higher abundance of iARGs compared to eARGs
has also been reported previously in most aquatic environments.
14, 15

During Phase 3, which consisted of different fouling levels across the three membrane modules, iARGs
were detected with the greatest abundance in the HF membrane effluent (Figure 5.4 and Figure S4.3).
This could have been caused by the higher TMP applied to the HF membrane during Phase 3, which may
have resulted in more passage of less rigid cells through the membrane pores (Figure 5.5). Since antibiotics
can decrease the rigidity of cell walls,
40
the exposure of the suspended biomass and biofilms to the three
antibiotics in the influent may have aided in this phenomenon. Further, it has been reported that particle
size in fouling layers decreases from the outer to the inner layer, making a funnel like structure,
41
which
can also facilitate the passage of less rigid cells. Exceptions to this observation were intracellular sul2 and
ermB genes, which were higher in the permeate of the LF, followed by MF and HF membranes (following
an inverse trend to other genes). These ARGs were also the only ones with lower absolute abundance in
the inner layer of the HF membrane biofilm compared to the outer layer, which may indicate that the
effluent iARG profiles were impacted by the release of ARB from the inner layer of biofilm attached to the
membranes surface (and not the outer). It has been reported previously that by progressive thickening of
the cake layer, bacterial cells in the inner layer (closer to membrane surface) die and detach, therefore
increasing their potential for high-TMP induced permeate washout.
39, 42

128

Figure 5.4 Absolute abundance of intracellular and extracellular (a) sul1, (b) sul2, (c) ermF, (d) ermB, (e)
ampC, and (f) oxa-1, in the effluent of LF, MF and HF membranes during Phase 3. LF, MF and HF represent
low, medium and highly fouled membranes in Phase 3, respectively.
In contrast to the effluent iARGs, the highest quantity of all detectable eARGs were released from the
LF membrane, followed by the MF and the HF membranes (Figure 5.4, and Figure S4.3). One possible
explanation for this phenomenon lies in the significantly higher EPS and SMP content of the HF membrane.
It has been reported elsewhere that EPS and SMP content (both polysaccharide and protein fractions) in
foulant layers have the ability to bind ARGs.
17, 18
Due to the high molecular weight and cross-linked
129
structural properties of both polysaccharides and proteins, they can contribute a significant capacity for
DNA binding through ion bridging interaction, hydrophobic interaction, and polymer enlargement.
43

Strong correlations have also previously been observed between the absolute abundance of ARGs and
both the EPS and SMP content in the membrane fouling layer of an MBR treating antibiotic-containing
wastewater.
19
Therefore, it is highly likely that the lower detection of eARGs in the permeate of the HF
membrane was due to its greater capacity for adsorption of extracellular mobile genetic elements (Figure
5.5). Another possible explanation for high eARG removal by HF membrane is increased rejection of eARGs
by fouled membranes due to the enhancement of steric hindrance and/or charge repulsion.
17
On the other
hand, fouling layers have been also reported as possibly decreasing rejection of organic matter through
enhancement of the concentration polarization effect.
44, 45

Figure 5.5 Possible mechanism of passage of less rigid cells through the mature biofilm due to high
transmembrane pressure (TMP), and adsorption of eARGs by EPS and SMP.
130
5.3.4 Log removal of total ARGs was not impacted by the extent of fouling
The use of internal standards for DNA extraction made it possible to report absolute abundance of ARGs
in both biomass and effluent samples. Based on this information, log removal of ARGs could be calculated
on a volumetric basis. Note that there were no ARGs in the influent synthetic wastewater and thus log
removal is calculated based on suspended biomass versus effluent ARGs, which is consistent with previous
studies.
19, 46
Table 1 shows the log removal of each ARG released from the membrane module (with
effluent ARG calculated as the sum of iARG and eARG concentrations). The highest log removal was
achieved for tetracycline resistance genes (tetO and tetW), which was followed by macrolide resistance
genes (ermB and ermF). intI1, sul1, and ampC had the lowest overall log removal. The log removal of total
ARGs by the LF, MF, and HF membranes were 3.6 ± 0.4, 3.4 ± 0.3, and 3.1 ± 0.4, respectively, and were
not statistically significantly different from one another (p > 0.29). Comparing the LRVs of different
membrane modules revealed that the extent of membrane fouling had negative correlations with total
ermF, tetO, and intI1 removal (p < 0.046, and ρ < -0.7), and positive correlations with total sul2 and ermB
removal (p < 0.049, and ρ > 0.7). Therefore, for each specific ARG, membrane fouling impacts the form
(i.e., iARG versus eARG) but does not significantly impact its total abundance (sum of iARG and eARG) in
the effluent of AnMBRs.
Zhang et al. had previously reported log removal of 3.2 and 4.1 for sul2 and ermB, respectively, as
achieved by an ultrafiltration membrane (pore size of 0.02 um) in an aerobic MBR treating domestic
wastewater,
14
while Zhu et al. observed 2.2, 1.3, and 3.1 log removal of sul1, sul2, and intI1, respectively,
by a microfiltration membrane (pore size of 0.1 um) in a similar MBR system.
19
Wen et al. also reported
log removal of 3.76 and 3.42 for tetracycline and sulfonamide resistance genes, respectively, by a
microfiltration membrane (pore size of 0.2 um) in an anoxic/oxic MBR.
46
Overall, results of the present
study showed comparable log removal for total ARGs to those of the above-mentioned studies (where log
131
removal was also calculated as suspended biomass versus effluent ARG), with specific ARGs (including
tetO, tetW, ermB, and ermF) significantly exceeding previously reported log removal values.

Table 5.1 Log removal of different ARGs achieved by membranes at different levels of fouling. LF, MF and
HF are low fouled, medium fouled and highly fouled membranes.

Log Removal

sul1 sul2 ermF ermB ampC oxa-1 tetO tetW intI1 rpoB
LF 3.27  0.42 4.47  0.46 5.99  0.41 6.33  0.37 3.28  0.16 4.39  0.37 7.73  0.59 7.35  0.42 3.18  0.44 4.11  0.53
MF 3.04  0.33 4.99  0.66 5.67  0.57 7.10  0.27 3.22  0.18 4.21  0.50 7.28  0.56 7.17  0.52 2.74  0.28 3.80  0.65
HF 2.74  0.38 5.01  0.64 5.26  0.29 7.16  0.25 3.13  0.20 3.82  0.57 6.90  0.53 7.01  0.56 2.39  0.21 3.44  0.76
5.3.5 Risk assessment is needed to evaluate/compare the threat of iARGs
and eARGs
The results of the present study revealed that high levels of fouling decreased the abundance of eARGs
and increased the abundance of iARGs in the effluent of the AnMBR. Different responses of eARGs and
iARGs to other treatment technologies such as disinfection, have been also reported previously.
14, 48, 49

This difference in abundance of iARGs and eARGs suggests a need for different approaches (such as a
combination of different technologies) to effectively remove both iARGs and eARGs during wastewater
treatment. However, prior to that, research assessing the risk of iARGs and eARGs in downstream
environments is needed. The lower abundance of eARGs compared to iARGs in the effluent of the AnMBR,
which was consistent with previous observation on WWTP effluents,
13-15
indicates that WWTPs release a
higher abundance of iARGs to the environment. However, significantly higher abundance of eARGs
compared to iARGs has been frequently observed in receiving aquatic environments,
11, 13, 16, 22
possibly due
132
to the higher persistence of eDNA (compared to iDNA) and death and lysis of bacteria. One important
factor to assess and compare the risk of iARGs and eARGs is the frequency with which they transfer to
other bacteria in the environment. Previous studies reported that frequencies of conjugation and
transduction of antibiotic resistance plasmids in aquatic sediment were 10
-1
-10
-7
, and 10
-8
-10
-9
per total
cell count, respectively.
49, 50
It has also been reported that the natural transformation frequency of eARG
in soil samples was 10
-3
-10
-7
resistant cell per total cell count.
51
Although comparing the data of these
studies may indicate that conjugation of iARGs has a higher frequency compared to the natural
transformation of eARGs, these studies were not conducted under similar experimental conditions.
Numerous parameters likely influence the frequency of horizontal transfer of ARGs, such as persistence
of the DNA fraction, microbial diversity, cell density, nutrient availability, etc. Therefore, a comprehensive
study is urgently needed to assess the risk of these two fractions of ARGs (iARGs and eARGs) in different
environmental matrices. With a better understanding of the relative risk of iARGs and eARGs, we can
begin to design and operate treatment processes to reduce the overall risk of ARG proliferation from
wastewater systems.
References
1. Molnar, A., Antimicrobial resistance awareness and games. Trends in microbiology 2019, 27, (1), 1-3.
2. CDC., Antibiotic Resistance Threats in the United States. U.S. Department of Health and Human Services
2019.
3. Guo, J.; Li, J.; Chen, H.; Bond, P. L.; Yuan, Z., Metagenomic analysis reveals wastewater treatment plants
as hotspots of antibiotic resistance genes and mobile genetic elements. Water Res. 2017, 123, 468-478.
4. Rizzo, L.; Manaia, C.; Merlin, C.; Schwartz, T.; Dagot, C.; Ploy, M.; Michael, I.; Fatta-Kassinos, D., Urban
wastewater treatment plants as hotspots for antibiotic resistant bacteria and genes spread into the
environment: a review. Sci. Total Environ. 2013, 447, 345-360.
5. Pruden, A.; Pei, R.; Storteboom, H.; Carlson, K. H., Antibiotic resistance genes as emerging
contaminants: studies in northern Colorado. Environ. Sci. Technol. 2006, 40, (23), 7445-7450.
6. Munir, M.; Wong, K.; Xagoraraki, I., Release of antibiotic resistant bacteria and genes in the effluent
and biosolids of five wastewater utilities in Michigan. Water Res. 2011, 45, (2), 681-693.
133
7. Smith, A. L.; Stadler, L. B.; Love, N. G.; Skerlos, S. J.; Raskin, L., Perspectives on anaerobic membrane
bioreactor treatment of domestic wastewater: a critical review. Bioresource technology 2012, 122, 149-
159.
8. Smith, A.; Skerlos, S.; Raskin, L., Anaerobic membrane bioreactor treatment of domestic wastewater at
psychrophilic temperatures ranging from 15 C to 3 C. Environmental Science: Water Research &
Technology 2015, 1, (1), 56-64.
9. Zarei-Baygi, A.; Harb, M.; Wang, P.; Stadler, L. B.; Smith, A. L., Evaluating Antibiotic Resistance Gene
Correlations with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors. Environ. Sci.
Technol. 2019, 53, (7), 3599-3609.
10. Baygi, A. Z.; Harb, M.; Wang, P.; Stadler, L.; Smith, A. L., Microbial community and antibiotic resistance
profiles of biomass and effluent are distinctly affected by antibiotics addition to an anaerobic membrane
bioreactor. Environmental Science: Water Research & Technology 2020.
11. Mao, D.; Luo, Y.; Mathieu, J.; Wang, Q.; Feng, L.; Mu, Q.; Feng, C.; Alvarez, P., Persistence of
extracellular DNA in river sediment facilitates antibiotic resistance gene propagation. Environ. Sci. Technol.
2014, 48, (1), 71-78.
12. Jain, R.; Rivera, M. C.; Lake, J. A., Horizontal gene transfer among genomes: the complexity hypothesis.
Proceedings of the National Academy of Sciences 1999, 96, (7), 3801-3806.
13. Zhang, Y.; Snow, D. D.; Parker, D.; Zhou, Z.; Li, X., Intracellular and extracellular antimicrobial resistance
genes in the sludge of livestock waste management structures. Environ. Sci. Technol. 2013, 47, (18),
10206-10213.
14. Zhang, Y.; Li, A.; Dai, T.; Li, F.; Xie, H.; Chen, L.; Wen, D., Cell-free DNA: a neglected source for antibiotic
resistance genes spreading from WWTPs. Environ. Sci. Technol. 2018, 52, (1), 248-257.
15. Sui, Q.; Chen, Y.; Yu, D.; Wang, T.; Hai, Y.; Zhang, J.; Chen, M.; Wei, Y., Fates of intracellular and
extracellular antibiotic resistance genes and microbial community structures in typical swine wastewater
treatment processes. Environment international 2019, 133, 105183.
16. Dong, P.; Wang, H.; Fang, T.; Wang, Y.; Ye, Q., Assessment of extracellular antibiotic resistance genes
(eARGs) in typical environmental samples and the transforming ability of eARG. Environment international
2019, 125, 90-96.
17. Cheng, H.; Hong, P.-Y., Removal of antibiotic-resistant bacteria and antibiotic resistance genes affected
by varying degrees of fouling on anaerobic microfiltration membranes. Environ. Sci. Technol. 2017, 51,
(21), 12200-12209.
18. Cheng, D.; Ngo, H. H.; Guo, W.; Liu, Y.; Chang, S. W.; Nguyen, D. D.; Nghiem, L. D.; Zhou, J.; Ni, B.,
Anaerobic membrane bioreactors for antibiotic wastewater treatment: performance and membrane
fouling issues. Bioresource technology 2018, 267, 714-724.
19. Zhu, Y.; Wang, Y.; Zhou, S.; Jiang, X.; Ma, X.; Liu, C., Robust performance of a membrane bioreactor for
removing antibiotic resistance genes exposed to antibiotics: role of membrane foulants. Water Res. 2018,
130, 139-150.
20. Smith, A. L.; Skerlos, S. J.; Raskin, L., Membrane biofilm development improves COD removal in
anaerobic membrane bioreactor wastewater treatment. Microbial biotechnology 2015, 8, (5), 883-894.
134
21. Guo, X.-p.; Yang, Y.; Lu, D.-p.; Niu, Z.-s.; Feng, J.-n.; Chen, Y.-r.; Tou, F.-y.; Garner, E.; Xu, J.; Liu, M.,
Biofilms as a sink for antibiotic resistance genes (ARGs) in the Yangtze Estuary. Water Res. 2018, 129, 277-
286.
22. Molin, S.; Tolker-Nielsen, T., Gene transfer occurs with enhanced efficiency in biofilms and induces
enhanced stabilisation of the biofilm structure. Curr. Opin. Biotechnol. 2003, 14, (3), 255-261.
23. Smith, A. L.; Skerlos, S. J.; Raskin, L., Psychrophilic anaerobic membrane bioreactor treatment of
domestic wastewater. Water Res. 2013, 47, (4), 1655-1665.
24. Verlicchi, P.; Al Aukidy, M.; Zambello, E., What have we learned from worldwide experiences on the
management and treatment of hospital effluent?—An overview and a discussion on perspectives. Science
of The Total Environment 2015, 514, 467-491.
25. Verlicchi, P.; Galletti, A.; Petrovic, M.; Barceló, D., Micro-pollutants in Hospital Effluent: Their Fate,
Risk and Treatment Options. In Emerging Organic Contaminants and Human Health, Springer: 2012; pp
139-171.
26. Zarei-Baygi, A.; Moslemi, M.; Mirzaei, S. H., The combination of KMnO4 oxidation and polymeric
flocculation for the mitigation of membrane fouling in a membrane bioreactor. Sep. Purif. Technol. 2016,
159, 124-134.
27. D’Abzac, P.; Bordas, F.; Van Hullebusch, E.; Lens, P. N.; Guibaud, G., Extraction of extracellular
polymeric substances (EPS) from anaerobic granular sludges: comparison of chemical and physical
extraction protocols. Applied microbiology and biotechnology 2010, 85, (5), 1589-1599.
28. Morgan, J.; Forster, C.; Evison, L., A comparative study of the nature of biopolymers extracted from
anaerobic and activated sludges. Water Res. 1990, 24, (6), 743-750.
29. Dubois, M.; Gilles, K. A.; Hamilton, J. K.; Rebers, P. t.; Smith, F., Colorimetric method for determination
of sugars and related substances. Anal. Chem. 1956, 28, (3), 350-356.
30. Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J., Protein measurement with the Folin phenol
reagent. J. Biol. Chem. 1951, 193, 265-275.
31. Gao, W.; Lin, H.; Leung, K.; Schraft, H.; Liao, B., Structure of cake layer in a submerged anaerobic
membrane bioreactor. Journal of membrane science 2011, 374, (1-2), 110-120.
32. Wang, Z.; Wu, Z.; Yin, X.; Tian, L., Membrane fouling in a submerged membrane bioreactor (MBR)
under sub-critical flux operation: membrane foulant and gel layer characterization. Journal of Membrane
Science 2008, 325, (1), 238-244.
33. Vos, M.; Quince, C.; Pijl, A. S.; de Hollander, M.; Kowalchuk, G. A., A comparison of rpoB and 16S rRNA
as markers in pyrosequencing studies of bacterial diversity. PLoS One 2012, 7, (2), e30600.
34. Lin, H.; Peng, W.; Zhang, M.; Chen, J.; Hong, H.; Zhang, Y., A review on anaerobic membrane
bioreactors: applications, membrane fouling and future perspectives. Desalination 2013, 314, 169-188.
35. Lin, H.; Liao, B.-Q.; Chen, J.; Gao, W.; Wang, L.; Wang, F.; Lu, X., New insights into membrane fouling
in a submerged anaerobic membrane bioreactor based on characterization of cake sludge and bulk sludge.
Bioresource technology 2011, 102, (3), 2373-2379.
36. Kappell, A. D.; Kimbell, L. K.; Seib, M. D.; Carey, D. E.; Choi, M. J.; Kalayil, T.; Fujimoto, M.; Zitomer, D.
H.; McNamara, P. J., Removal of antibiotic resistance genes in an anaerobic membrane bioreactor treating
135
primary clarifier effluent at 20° C. Environmental Science: Water Research & Technology 2018, 4, (11),
1783-1793.
37. Du, J.; Geng, J.; Ren, H.; Ding, L.; Xu, K.; Zhang, Y., Variation of antibiotic resistance genes in municipal
wastewater treatment plant with A 2 O-MBR system. Environmental Science and Pollution Research 2015,
22, (5), 3715-3726.
38. Harb, M.; Wei, C.-H.; Wang, N.; Amy, G.; Hong, P.-Y., Organic micropollutants in aerobic and anaerobic
membrane bioreactors: Changes in microbial communities and gene expression. Bioresource technology
2016, 218, 882-891.
39. Singh, R.; Paul, D.; Jain, R. K., Biofilms: implications in bioremediation. Trends in microbiology 2006,
14, (9), 389-397.
40. Formosa, C.; Grare, M.; Duval, R. E.; Dague, E., Nanoscale effects of antibiotics on P. aeruginosa.
Nanomedicine: Nanotechnology, Biology and Medicine 2012, 8, (1), 12-16.
41. Xiong, J.; Fu, D.; Singh, R. P.; Ducoste, J. J., Structural characteristics and development of the cake layer
in a dynamic membrane bioreactor. Sep. Purif. Technol. 2016, 167, 88-96.
42. Drews, A., Membrane fouling in membrane bioreactors—characterisation, contradictions, cause and
cures. Journal of membrane science 2010, 363, (1-2), 1-28.
43. Ding, Z.; Bourven, I.; Guibaud, G.; van Hullebusch, E. D.; Panico, A.; Pirozzi, F.; Esposito, G., Role of
extracellular polymeric substances (EPS) production in bioaggregation: application to wastewater
treatment. Applied microbiology and biotechnology 2015, 99, (23), 9883-9905.
44. Pan, S.-F.; Zhu, M.-P.; Chen, J. P.; Yuan, Z.-H.; Zhong, L.-B.; Zheng, Y.-M., Separation of tetracycline
from wastewater using forward osmosis process with thin film composite membrane–Implications for
antibiotics recovery. Sep. Purif. Technol. 2015, 153, 76-83.
45. Baker, R.; Wijmans, J.; Athayde, A.; Daniels, R.; Ly, J.; Le, M., The effect of concentration polarization
on the separation of volatile organic compounds from water by pervaporation. Journal of membrane
science 1997, 137, (1-2), 159-172.
46. Wen, Q.; Yang, L.; Zhao, Y.; Huang, L.; Chen, Z., Insight into effects of antibiotics on reactor
performance and evolutions of antibiotic resistance genes and microbial community in a membrane
reactor. Chemosphere 2018, 197, 420-429.
47. Chandrasekaran, S.; Venkatesh, B.; Lalithakumari, D., Transfer and expression of a multiple antibiotic
resistance plasmid in marine bacteria. Current Microbiology 1998, 37, (5), 347-351.
48. Jiang, S. C.; Paul, J. H., Gene transfer by transduction in the marine environment. Appl. Environ.
Microbiol. 1998, 64, (8), 2780-2787.
49. Stewart, G. J.; Sinigalliano, C. D., Detection of horizontal gene transfer by natural transformation in
native and introduced species of bacteria in marine and synthetic sediments. Appl. Environ. Microbiol.
1990, 56, (6), 1818-1824.
50. Lu, N.; Zilles, J. L.; Nguyen, T. H., Adsorption of extracellular chromosomal DNA and its effects on
natural transformation of Azotobacter vinelandii. Appl. Environ. Microbiol. 2010, 76, (13), 4179-4184.
136
51. Nielsen, K. M.; Van Weerelt, M.; Berg, T. N.; Bones, A. M.; Hagler, A. N.; Van Elsas, J., Natural
transformation and availability of transforming DNA to Acinetobacter calcoaceticus in soil microcosms.
Appl. Environ. Microbiol. 1997, 63, (5), 1945-1952.

137
Chapter 6

6. Conclusions

6.1 Overview
The main objective of this dissertation was to evaluate antibiotic resistance in anaerobic wastewater
treatment using anaerobic membrane bioreactors (AnMBRs). Antibiotic resistance is one of the biggest
global concerns of our time, responsible for more than 700,000 death annually worldwide.
1
Wastewater
treatment plants (WWTPs), as a main interface between the built and natural environment, play a crucial
rule in the spread of antibiotic resistance to the environment.
2, 3
AnMBRs are an emerging treatment
biotechnology which can help with mitigation of antibiotic resistance.
4, 5
Therefore, the approach of the
dissertation was to evaluate the impact of operating and design strategies of AnMBRs on antibiotic
resistance. First, we reviewed literature on the prevalence, horizontal transfer, and mitigation strategies
for environmental intracellular and extracellular antibiotic resistance genes (iARGs and eARGs; Chapter
2). Next, a bench-scale AnMBR was built and operated to evaluate the impact of influent antibiotics on
the ARG profiles of the biomass and effluent (Chapter 3).
4
The following chapter investigated the potential
association between antibiotic resistance and microbial community structure of an AnMBR (Chapter 4).
5

Lastly, the role of membrane fouling on release of iARGs and eARGs from an AnMBR was evaluated
(Chapter 5).
138
6.2 eARGs may be as important as iARGs in dissemination of antibiotic
resistance in the environment
ARGs are found as both intracellular and extracellular DNA (iDNA and eDNA).
6, 7
eDNA originates from
iDNA during the lysis of dead bacterial cells and active secretion from live bacterial cells. Therefore, iARGs
are the main fraction in environments which support bacterial growth, such as waste streams. For
example, in wastewater and manure the absolute abundance of iARGs are 2-3 order of magnitudes higher
than eARGs.
6, 8-10
The antibiotic resistance profile in receiving aquatic environments with low nutrient
availability, such as river and marine sediments, however, are usually dominated by eARGs.
7, 11, 12
In such
environments, eARGs can be adsorbed by soil particles and colloids, and consequently protected from
DNase degradation.
7
Horizontal transfer of iARGs and eARGs also occurs through different mechanisms.
iARGs transfer via conjugation and transduction, while natural transformation is the main eARG transfer
mechanism. Conjugation is the most studied horizontal gene transfer (HGT) mechanism which has been
found to occur in nutrient-rich environments like activated sludge with a frequency of up to 10
-2
transconjugants/recipient cells.
13-15
Transduction of iARGs, which is mediated by bacteriophage infection,
occurs with a lower frequency (up to 10
-5
transductions/plaque-forming) compared to conjugation.
16, 17

Natural transformation of eARGs occurs with direct uptake of eDNA and can reach a frequency of up to
10
-3
transformants/recipient cells.
18, 19
It is noteworthy that selective pressure posed by antibiotics, heavy
metals, etc. can increase the rate of all HGT mechanisms.
7, 20, 21
Another distinction between iARGs and
eARGs is that they usually face different fates in WWTPs. Membrane fouling can be more effective in
eARGs removal, while chlorination with low exposure usually results in enrichment of eARGs.
22, 23
Further,
UV disinfection is more efficient for eARG inactivation than iARG inactivation.
24
As a result, in order to
effectively remove both iARGs and eARGs, a combination of multiple processes (e.g., membrane filtration
and advanced oxidation) is required.
139
6.3 Selective pressure posed by influent antibiotic induced antibiotic
resistance in both biomass and effluent of AnMBRs
Influent antibiotics are a selective pressure that can impact the ARG profile in biological wastewater
treatment systems. In this study, the impact of three influent antibiotics including sulfamethoxazole
(SMX), erythromycin (ERY), and ampicillin (AMP) on biomass and effluent ARG profiles of an AnMBR was
investigated. Each antibiotic was introduced to the reactor in independent sequential phases an at
incremental concentration of 10, 50, and 250 µg/L. Eight ARGs including two sulfonamides (sul1 and sul2),
two macrolides (ermF and ermB), two β-lactams (oxa-1 and ampC), and two tetracycline (tetW and tetO),
as well as intl1 which encodes for class 1 integrons were quantified in both biomass and effluent using
quantitative polymerase chain reaction (qPCR). The rpoB gene was also quantified as a gene marker
representing the bacterial count for ARG normalization. According to the results, AnMBR was able to
efficiently remove AMP, SMX and ERY in the effluent by 94-98%, 71-85% and 67-88%, respectively. During
operation, the effluent had a unique ARG profile compared to the biomass, likely due to selective passage
of ARGs through the membrane. Relative abundance of most targeted ARGs in the AnMBR biomass
increased gradually with increasing influent antibiotic concentrations. This indicates that the selective
pressure posed by a specific antibiotic can affect abundance of both related and non-related ARGs. This
increase in relative abundance of ARGs might be due to HGT, changes in microbial community, or some
combination of the two. Effluent ARG profiles first increased significantly upon the initial exposure to
antibiotics at 10 µg/L, and then decreased gradually at subsequent concentration of 50 and 250 µg/L. The
sharp increase in effluent ARG profile at initial trace level exposure might be due to an increase in HGT in
biomass, resulting in enrichment of biomass eARGs which can be easily pass through the clean membrane,
and cause a sharp rise in effluent ARG profile. After the initial exposure, the rate of HGT in biomass
decreased, which resulted in a reduction in the levels of eARGs and lowered the rate of ARGs passing
140
through membranes into the effluent. Correlation analysis revealed strong positive correlation between
sul1 and intI1 in the effluent, signifying the high impact of HGT in effluent ARG profile.
6.4 In the presence of antibiotics, HGT is the main reason for variation
of ARG profiles
The possible association between ARGs and microbial community structure is of particular interest for
two reasons: (1) to indicate the bacterial hosts of ARGs, and (2) to find the reasons for variation of the
ARG profile in the presence of antibiotics. In this study, we simultaneously introduced SMX, ERY, and AMP
at a high concentration of 250 µg/L to the influent of an AnMBR and explored the ARG and microbial
community profiles of biomass and effluent, and their possible association. Results indicated that AnMBR
biomass was dominated by different ARGs than the AnMBR effluent. The biomass ARG profile increased
gradually throughout the experiment, however, microbial community structure was consistent. This
suggests that variation of ARGs in the presence of antibiotics is mainly due to HGT. The sharp increase in
intI1 gene abundance, which is an indicator of HGT, is additional evidence of the significant role of HGT in
fate of ARGs in WWTPs. The effluent ARG profile first increased by 34-folds upon the initial exposure to
antibiotics, and then decreased throughout the experiment. These results indicate that antibiotic mixture,
as opposed to individual antibiotics, results in more pronounced changes in ARG abundance. The effluent
microbial community were distinct from the biomass and were affected significantly by antibiotic
exposure. Significant changes in the effluent microbial community after antibiotics addition, while the
biomass community was consistent, might be due to the lower diversity of effluent community compared
to the biomass (indicated by Simpson diversity index) which make the community more susceptible to
inhibition. Correlation analysis revealed the existence of several potential multi-resistant host bacteria in
the biomass of the AnMBR. In AnMBR effluent, lower correlations were obtained between bacteria and
ARGs, probably due to the higher contribution of eARGs to the effluent ARG profile compared to the
141
biomass. Strong positive correlations were also found between ARGs and several pathogenic groups,
indicating the transferability of antibiotic resistance to a broader range of clinically significant bacterial
strains.
6.5 Membrane fouling distinctly impact ARG and community profiles of
biomass and effluent of an AnMBR
Membrane fouling, due to high biological activity and temporally stable spatial organization of microbial
cells can impact abundance of ARGs in effluent of AnMBRs. In this study, we investigated the release of
iARGs and eARGs from three submerged membrane modules at different levels of fouling (low fouled; LF,
medium fouled; MF, and high fouled; HF) in an AnMBR. We also evaluated the absolute and relative
abundance of ARGs in biomass and fouling layers. Results reveled that membrane fouling significantly
improve performance of AnMBR in COD removal. This might be due to the additional biodegradation of
organics in the fouling layer. Absolute abundance of ARGs in biomass decreased gradually throughout the
experiment which may be attributable to low biomass growth rates in anaerobic systems. Relative
abundance of ARGs in fouling layers were higher than the biomass, indicating that the biofilm provides
more conducive conditions for dissemination of ARGs. The HF membrane contained higher abundance of
ARGs compared to MF membrane, suggesting that a mature biofilm results in higher retention of ARGs
and/or provides more suitable conditions for horizontal ARG transfer. Effluent ARG profiles revealed that
membrane fouling increased the abundance of iARGs, while it decreased eARGs in the AnMBR effluent.
Higher absolute abundance of iARGs in HF permeate might be due to the higher TMP applied to this
membrane, which may have resulted in more passage of less rigid cells through the membrane pores.
Membrane fouling, however, decreased release of eARGs from the AnMBR, probably due to the significant
adsorption of extracellular compounds. HF membrane contained higher extracellular polymeric
substances (EPSs) and soluble microbial products (SMPs) compared to the MF membrane. Given that both
142
EPSs and SMPs have a high binding ability, it is likely that more fouled membranes have a greater capacity
for adsorption of extracellular mobile genetic elements.
6.6 Future Research
This dissertation investigated the fate of antibiotic resistance in AnMBRs. Antibiotic resistance is a
serious global health concern, which impacts more than 2.8 million people in the US and is responsible
for approximately 32,000 deaths annually.
25
One of the core actions proposed by the U.S. Centers for
Disease Control and Prevention (CDC) to better prepare the United States for an antibiotic resistance
pandemic is preventing antibiotic resistance from entering the environment.
25
WWTPs, as a primary
interface between the built and natural environment, can effectively prevent release of antibiotic
resistance to the environment.
2
AnMBR as an emerging treatment biotechnology, which can produce
energy and nutrient-rich biosolids, can also mitigate release of antibiotic resistance.
4, 5
Therefore, the
findings of this work on the fate of antibiotic resistance in AnMBRs can guide future efforts on strategies
for reducing the release of antibiotic resistance to the environment.
We observed that influent antibiotic can significantly increase abundance of ARGs in both biomass and
effluent of AnMBRs.
4
This increase is likely due to the promotion of HGT. However, it remains unclear how
much each HGT mechanism contributes to observed increases in the ARG profile of biomass and effluent.
Previous studies have shown that antibiotic selective pressure can induce all HGT mechanisms.
7, 20, 21

However, no study to date has evaluate the concurrent variation of conjugation, transduction, and natural
transformation frequency in the presence of antibiotics. Further, our results reveled that effluent ARG
profiles sharply increased upon initial exposure to antibiotics (10 µg/L), and then decreased at subsequent
concentrations of 50 and 250 µg/L.
4
Although there are some possible explanations (e.g., temporal
adaptation of the microbial community), future research is needed to elucidate the reasons behind this
phenomenon.
143
ARG-host relationship is another future research direction that needs investigation. We used Spearman
correlation to evaluate the possible associations between ARGs and the microbial community in both
biomass and effluent of an AnMBR.
5
Although statistically significant positive correlation between a
microbial group and a particular ARG cannot be considered as evidence of that group carrying antibiotic
resistance, it can indicate the OTUs that are potential host bacteria for ARGs and MGEs. More accurate
tools are now being explored to evaluate ARG-host relationship, including metagenomics fluorescence-
activated cell sorting, single-cell fusion PCR, and genomic cross-linking methods.
26
Future research efforts
should apply these new molecular biology tools to better characterize ARG-host range in WWTPs. ARG-
host information can also be used to target elimination of antibiotic resistant pathogens in WWTPs. Since
antibiotic resistance is a natural phenomenon that is ubiquitous in microbial life and occurs via natural
selection, removing all antibiotic resistance is unrealistic and not necessary. Therefore, a targeted
approach can significantly help with preventing the threat of antibiotic resistance spread to the
environment.
We showed that membrane fouling layer can significantly increase iARGs, while decreasing eARG
abundance in the effluent of AnMBRs. The fouling layer can also improve performance of AnMBRs in
regard to COD removal. According to our results, the abundance of most targeted ARGs in the inner layer
of the HF membrane was significantly higher than its outer layer, except for sul2 and ermB. This might
indicate that some ARG hosts were enriched in the biofilm, while some (e.g., hosts of sul2 and ermB) were
not enriched. The association between ARG hosts and attached/suspended growth remains understudied
and requires further investigation. Our results also revealed that iARGs and eARGs have different fates in
fouled membrane. It has been previously reported that disinfection also distinctly affects iARGs and
eARGs.
22, 27
Therefore, in order to effectively remove both iARGs and eARGs from waste streams, a
combination of different technologies may be an important strategy. In addition, comprehensive research
assessing the risk of iARGs versus eARGs in downstream environments is needed that considers
144
prevalence/persistence and horizontal transfer of different fraction of ARGs. With a better understanding
of the relative risk of iARGs and eARGs, we can begin to design and operate treatment systems to reduce
the overall risk of ARG proliferation from wastewater treatment systems.
References
1. WHO, No time to wait: securing the future from drug-resistant infections. Report to the Secretary-
General of the United Nations. World Health Organization, Geneva 2019.
2. Rizzo, L.; Manaia, C.; Merlin, C.; Schwartz, T.; Dagot, C.; Ploy, M.; Michael, I.; Fatta-Kassinos, D., Urban
wastewater treatment plants as hotspots for antibiotic resistant bacteria and genes spread into the
environment: a review. Sci. Total Environ. 2013, 447, 345-360.
3. Guo, J.; Li, J.; Chen, H.; Bond, P. L.; Yuan, Z., Metagenomic analysis reveals wastewater treatment plants
as hotspots of antibiotic resistance genes and mobile genetic elements. Water Res. 2017, 123, 468-478.
4. Zarei-Baygi, A.; Harb, M.; Wang, P.; Stadler, L. B.; Smith, A. L., Evaluating Antibiotic Resistance Gene
Correlations with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors. Environ. Sci.
Technol. 2019, 53, (7), 3599-3609.
5. Baygi, A. Z.; Harb, M.; Wang, P.; Stadler, L.; Smith, A. L., Microbial community and antibiotic resistance
profiles of biomass and effluent are distinctly affected by antibiotics addition to an anaerobic membrane
bioreactor. Environmental Science: Water Research & Technology 2020.
6. Zhang, Y.; Snow, D. D.; Parker, D.; Zhou, Z.; Li, X., Intracellular and extracellular antimicrobial resistance
genes in the sludge of livestock waste management structures. Environ. Sci. Technol. 2013, 47, (18),
10206-10213.
7. Mao, D.; Luo, Y.; Mathieu, J.; Wang, Q.; Feng, L.; Mu, Q.; Feng, C.; Alvarez, P., Persistence of extracellular
DNA in river sediment facilitates antibiotic resistance gene propagation. Environ. Sci. Technol. 2014, 48,
(1), 71-78.
8. Sui, Q.; Chen, Y.; Yu, D.; Wang, T.; Hai, Y.; Zhang, J.; Chen, M.; Wei, Y., Fates of intracellular and
extracellular antibiotic resistance genes and microbial community structures in typical swine wastewater
treatment processes. Environment international 2019, 133, 105183.
9. Wang, P.; Harb, M.; Zarei-Baygi, A.; Stadler, L. B.; Smith, A. L., Comparative analysis of intracellular and
extracellular antibiotic resistance gene abundance in anaerobic membrane bioreactor effluent. BioRxiv
2019, 702076.
10. Yuan, Q.-B.; Huang, Y.-M.; Wu, W.-B.; Zuo, P.; Hu, N.; Zhou, Y.-Z.; Alvarez, P. J., Redistribution of
intracellular and extracellular free & adsorbed antibiotic resistance genes through a wastewater
treatment plant by an enhanced extracellular DNA extraction method with magnetic beads. Environment
international 2019, 131, 104986.
11. Zhang, Y.; Niu, Z.; Zhang, Y.; Zhang, K., Occurrence of intracellular and extracellular antibiotic
resistance genes in coastal areas of Bohai Bay (China) and the factors affecting them. Environ. Pollut. 2018,
236, 126-136.
145
12. Zhao, Z.; Zhang, K.; Wu, N.; Li, W.; Xu, W.; Zhang, Y.; Niu, Z., Estuarine sediments are key hotspots of
intracellular and extracellular antibiotic resistance genes: A high-throughput analysis in Haihe Estuary in
China. Environment International 2020, 135, 105385.
13. Ohlsen, K.; Ternes, T.; Werner, G.; Wallner, U.; Löffler, D.; Ziebuhr, W.; Witte, W.; Hacker, J., Impact of
antibiotics on conjugational resistance gene transfer in Staphylococcus aureus in sewage. Environ.
Microbiol. 2003, 5, (8), 711-716.
14. Li, B.; Qiu, Y.; Zhang, J.; Liang, P.; Huang, X., Conjugative potential of antibiotic resistance plasmids to
activated sludge bacteria from wastewater treatment plants. International biodeterioration &
biodegradation 2019, 138, 33-40.
15. Inoue, D.; Sei, K.; Soda, S.; Ike, M.; Fujita, M., Potential of predominant activated sludge bacteria as
recipients in conjugative plasmid transfer. Journal of bioscience and bioengineering 2005, 100, (6), 600-
605.
16. Volkova, V. V.; Lu, Z.; Besser, T.; Gröhn, Y. T., Modeling the infection dynamics of bacteriophages in
enteric Escherichia coli: estimating the contribution of transduction to antimicrobial gene spread. Appl.
Environ. Microbiol. 2014, 80, (14), 4350-4362.
17. Varga, M.; Kuntová, L.; Pantůček, R.; Mašlaňová, I.; Růžičková, V.; Doškař, J., Efficient transfer of
antibiotic resistance plasmids by transduction within methicillin-resistant Staphylococcus aureus USA300
clone. FEMS Microbiol. Lett. 2012, 332, (2), 146-152.
18. Nielsen, K. M.; Van Weerelt, M.; Berg, T. N.; Bones, A. M.; Hagler, A. N.; Van Elsas, J., Natural
transformation and availability of transforming DNA to Acinetobacter calcoaceticus in soil microcosms.
Appl. Environ. Microbiol. 1997, 63, (5), 1945-1952.
19. Lu, N.; Zilles, J. L.; Nguyen, T. H., Adsorption of extracellular chromosomal DNA and its effects on
natural transformation of Azotobacter vinelandii. Appl. Environ. Microbiol. 2010, 76, (13), 4179-4184.
20. Bielaszewska, M.; Idelevich, E. A.; Zhang, W.; Bauwens, A.; Schaumburg, F.; Mellmann, A.; Peters, G.;
Karch, H., Effects of antibiotics on Shiga toxin 2 production and bacteriophage induction by epidemic
Escherichia coli O104: H4 strain. Antimicrob. Agents Chemother. 2012, 56, (6), 3277-3282.
21. Zhang, S.; Wang, Y.; Song, H.; Lu, J.; Yuan, Z.; Guo, J., Copper nanoparticles and copper ions promote
horizontal transfer of plasmid-mediated multi-antibiotic resistance genes across bacterial genera.
Environment international 2019, 129, 478-487.
22. Liu, S.-S.; Qu, H.-M.; Yang, D.; Hu, H.; Liu, W.-L.; Qiu, Z.-G.; Hou, A.-M.; Guo, J.; Li, J.-W.; Shen, Z.-Q.,
Chlorine disinfection increases both intracellular and extracellular antibiotic resistance genes in a full-
scale wastewater treatment plant. Water Res. 2018, 136, 131-136.
23. Cheng, H.; Hong, P.-Y., Removal of antibiotic-resistant bacteria and antibiotic resistance genes affected
by varying degrees of fouling on anaerobic microfiltration membranes. Environ. Sci. Technol. 2017, 51,
(21), 12200-12209.
24. Yoon, Y.; Chung, H. J.; Di, D. Y. W.; Dodd, M. C.; Hur, H.-G.; Lee, Y., Inactivation efficiency of plasmid-
encoded antibiotic resistance genes during water treatment with chlorine, UV, and UV/H2O2. Water Res.
2017, 123, 783-793.
25. CDC., Antibiotic Resistance Threats in the United States. U.S. Department of Health and Human
Services 2019.
26. Rice, E. W.; Wang, P.; Smith, A. L.; Stadler, L. B., Determining Hosts of Antibiotic Resistance Genes: A
Review of Methodological Advances. Environmental Science & Technology Letters 2020, 7, (5), 282-291.
146
27. McKinney, C. W.; Pruden, A., Ultraviolet disinfection of antibiotic resistant bacteria and their antibiotic
resistance genes in water and wastewater. Environ. Sci. Technol. 2012, 46, (24), 13393-13400.

147
Appendices
Appendix A: Supporting information for chapter 2
Table S1.1 Absolute abundance of iARGs and eARGs in different environments.
Absolute Abundance
Environment Targeted ARG iARG eARG Reference
Raw Municipal
Wastewater
(copies/L)

tetC 9.0E+09 3.0E+07
1

sul2 6.2E+10 9.0E+07
blaPSE-1 4.0E+08 6.1E+06
ermB 2.0E+10 7.3E+07

tetA 3.0E+11 2.5E+10
2

tetC 2.1E+09 1.8E+08
tetM 8.0E+09 2.2E+09
tetX 1.0E+09 2.8E+07
sul1 2.3E+09 2.7E+08
sul2 1.0E+09 1.1E+08
blaTEM 4.0E+08 1.0E+08
ermB 1.1E+09 4.0E+08

WWTPs Sludge
(copies/g)

tetC 1.2E+10 1.6E+06
1

sul2 1.8E+11 9.6E+05
blaPSE-1 1.80E+08 1.1E+05
ermB 1.6E+10 5.3E+06

tetA 3.0E+11 6.6E+10
2

tetC 3.0E+09 5.5E+08
tetM 1.6E+10 6.3E+09
tetX 2.0E+09 8.0E+07
sul1 4.0E+09 7.6E+08
sul2 9.4E+09 1.1E+09
blaTEM 5.0E+08 6.0E+08
148
ermB 3.1E+09 1.2E+09

tetW 3.0E+09 5.4E+07
3

tetX 4.0E+11 1.1E+09
sul1 4.0E+12 2.0E+10
sul2 2.1E+11 1.4E+09
ampC 1.0E+08 8.0E+06
blaTEM 5.2E+11 2.3E+08
ermB 3.4E+11 3.0E+08
intI1 1.0E+11 3.0E+09

Tap Water
(copies/L)

tetM 3.4E+05 6.0E+02
4

tetX 7.0E+03 9.2E+01
sul1 7.6E+06 1.3E+05
sul2 6.0E+05 2.0E+04
ampC 2.1E+04 1.4E+03
blaTEM 2.0E+05 1.0E+04

Manure
(copies/g)

tetO 1.0E+08 1.1E+04
5

tetQ 3.5E+09 1.0E+07
tetX 6.1E+08 0.0E+00
sul1 1.0E+09 5.4E+06
sul2 2.4E+09 1.0E+07

tetW 8.1E+10 1.0E+08
3

tetX 2.0E+11 2.4E+08
sul1 1.3E+11 7.0E+08
sul2 1.0E+10 9.5E+07
ampC 2.0E+09 4.0E+04
blaTEM 8.3E+11 8.0E+06
ermB 6.0E+09 2.3E+07
intI1 5.1E+10 2.0E+06

tetO 2.1E+08 3.0E+05
5

149
Manure Treatment
Lagoon
(copies/g)

tetQ 1.0E+09 2.3E+07
tetX 8.4E+08 6.4E+05
sul1 1.2E+09 7.2E+06
sul2 1.0E+09 1.0E+07

River Water
(copies/mL)

tetT 2.0E+05 4.1E+01
6

tetW 1.4E+05 3.0E+01
sul1 3.0E+09 1.3E+03
sul2 3.2E+08 6.0E+03

River Sediment
(copies/g)

tetT 7.0E+04 2.1E+05
6

tetW 1.3E+04 5.0E+04
sul1 5.0E+07 1.5E+08
sul2 1.2E+10 3.0E+10

Estuarine
Sediments
(copies/g)
Aminoglycoside 3.2E+06 6.3E+06
7

Lactamase 3.7E+06 2.0E+05
Chloramphenicol 3.5E+05 3.9E+05
Multidrug 2.8E+07 3.5E+07
Sulfonamide 1.9E+06 1.5E+07
MLSB 3.2E+06 1.0E+06
Tetracycline 1.1E+06 7.9E+05
Vancomycin 6.7E+05 0.0E+00

150
Table S1.2 Log removal of iARGs and eARGs in secondary clarifier effluent achieved by different
disinfection methods.

Log Removal
Disinfection Method Operation Conditions Targeted ARG iARG eARG Reference
Chlorination (ClO 2) 8-9 mg/L
30 min
tetA -0.3 -0.7
8

tetC -0.4 -0.2
tetM -0.3 0.3
tetQ -0.2 -0.4
tetX -0.2 -0.1
ampC -0.5 -0.3
blaTEM -0.4 -0.3
sul1 -0.3 -0.2
sul2 -0.5 -0.4
ermB -0.3 -1
qnrA -0.6 0
gyrA -0.1 0.7
vanA -0.4 -0.4

Chlorination (Cl 2) 4.2 mg/L
5 h
tetC 0.7 0.3
1

sul2 1 0.8
blaPSE-1 0.3 -0.1
ermB 0.7 -0.1

Chlorination (Cl 2) 10 mg/L
1 h
ampR 0.6 -0.3
9

kanR 1.2 0.3

Chlorination (Cl 2) 20 mg/L
1 h
ampR 4 4
9

kanR 4 4

Ultraviolet - tetA 0.2 -0.5
2

tetC 0.1 0.3
tetM 0.4 -0.3
tetX -0.5 -0.7
sul1 0.1 -0.1
151
sul2 0.3 0.1
blaTEM 0.3 0.5
ermB -0.1 -1.2

Ultraviolet 50 mJ/cm
2
tetA 0.9 0.7
10

ampC 0.9 0.6
mecA 1.8 1.6
vanA 1.6 1.4

Ultraviolet 100 mJ/cm
2
tetA 1.3 1.4
10

ampC 1.4 1.1
mecA 2.8 2.6
vanA 2.5 2.3

Ultraviolet 150 mJ/cm
2
tetA 1.8 1.7
10

ampC 1.7 1.4
mecA 3.1 3.5
vanA 3.2 2.9

Ultraviolet 20 mJ/cm
2
ampR 0.5 0.9
9

kanR - 1.4

Ultraviolet 70 mJ/cm
2
ampR 2.2 3.3
9

kanR 2.5 2.8

Ultraviolet 130 mJ/cm
2
ampR 3.6 4
9

kanR 3.7 4

Ozonation 3.5 mg/L
25 min
tetC 0 0.6
1

sul2 -0.2 0.2
blaPSE-1 0.4 0.6
ermB -0.6 0

blaTEM 0.5 1.3
11

152
Photo
electrocatalytic
28 mW/cm
2

2 h acc(3)II
0.5 1.3

Photo
electrocatalytic
28 mW/cm
2

4 h
blaTEM 1 2.6
11

acc(3)II 1.4 1.7

Photo
electrocatalytic
28 mW/cm
2

16 h
blaTEM 7.8 6
11

acc(3)II 7.6 6

References
1. Zhang, Y.; Li, A.; Dai, T.; Li, F.; Xie, H.; Chen, L.; Wen, D., Cell-free DNA: a neglected source for antibiotic
resistance genes spreading from WWTPs. Environ. Sci. Technol. 2018, 52, (1), 248-257.
2. Yuan, Q.-B.; Huang, Y.-M.; Wu, W.-B.; Zuo, P.; Hu, N.; Zhou, Y.-Z.; Alvarez, P. J., Redistribution of
intracellular and extracellular free & adsorbed antibiotic resistance genes through a wastewater
treatment plant by an enhanced extracellular DNA extraction method with magnetic beads. Environment
international 2019, 131, 104986.
3. Dong, P.; Wang, H.; Fang, T.; Wang, Y.; Ye, Q., Assessment of extracellular antibiotic resistance genes
(eARGs) in typical environmental samples and the transforming ability of eARG. Environment international
2019, 125, 90-96.
4. Hao, H.; Shi, D.-y.; Yang, D.; Yang, Z.-w.; Qiu, Z.-g.; Liu, W.-l.; Shen, Z.-q.; Yin, J.; Wang, H.-r.; Li, J.-w.,
Profiling of intracellular and extracellular antibiotic resistance genes in tap water. J. Hazard. Mater. 2019,
365, 340-345.
5. Zhang, Y.; Snow, D. D.; Parker, D.; Zhou, Z.; Li, X., Intracellular and extracellular antimicrobial resistance
genes in the sludge of livestock waste management structures. Environ. Sci. Technol. 2013, 47, (18),
10206-10213.
6. Mao, D.; Luo, Y.; Mathieu, J.; Wang, Q.; Feng, L.; Mu, Q.; Feng, C.; Alvarez, P., Persistence of extracellular
DNA in river sediment facilitates antibiotic resistance gene propagation. Environ. Sci. Technol. 2014, 48,
(1), 71-78.
7. Zhao, Z.; Zhang, K.; Wu, N.; Li, W.; Xu, W.; Zhang, Y.; Niu, Z., Estuarine sediments are key hotspots of
intracellular and extracellular antibiotic resistance genes: A high-throughput analysis in Haihe Estuary in
China. Environment International 2020, 135, 105385.
8. Liu, S.-S.; Qu, H.-M.; Yang, D.; Hu, H.; Liu, W.-L.; Qiu, Z.-G.; Hou, A.-M.; Guo, J.; Li, J.-W.; Shen, Z.-Q.,
Chlorine disinfection increases both intracellular and extracellular antibiotic resistance genes in a full-
scale wastewater treatment plant. Water Res. 2018, 136, 131-136.
9. Yoon, Y.; Chung, H. J.; Di, D. Y. W.; Dodd, M. C.; Hur, H.-G.; Lee, Y., Inactivation efficiency of plasmid-
encoded antibiotic resistance genes during water treatment with chlorine, UV, and UV/H2O2. Water Res.
2017, 123, 783-793.
153
10. McKinney, C. W.; Pruden, A., Ultraviolet disinfection of antibiotic resistant bacteria and their antibiotic
resistance genes in water and wastewater. Environ. Sci. Technol. 2012, 46, (24), 13393-13400.
11. Jiang, Q.; Yin, H.; Li, G.; Liu, H.; An, T.; Wong, P. K.; Zhao, H., Elimination of antibiotic-resistance
bacterium and its associated/dissociative blaTEM-1 and aac (3)-II antibiotic-resistance genes in aqueous
system via photoelectrocatalytic process. Water Res. 2017, 125, 219-226.

154

Appendix B: Supporting information for chapter 3
B.1 Bench-scale anaerobic membrane bioreactor operation
The reactor was inoculated with 5 L of anaerobic digester sludge. Mixed liquor suspended solids (MLSS)
and mixed liquor volatile suspended solids (MLVSS) of the primary sludge were 7.9  0.4 g/L and 6.1  0.5
g/L, respectively. The AnMBR was fed using a synthetic wastewater prepared twice a week to prevent
degradation. To provide additional measures against degradation, the synthetic wastewater was
comprised of a concentrate solution acidified to a pH of 3.5 and refrigerated and blended with a basic
dilution water prior to the AnMBR. Table S2.1 shows the composition of these two solutions and their
final concentration in the synthetic wastewater.
Membrane cleaning was done periodically by physically removing attached foulants and then
submerging the modules in 0.5% (v/v) NaOCl solution overnight. To increase the efficiency of chemical
cleaning, a peristaltic pump was employed to pump the 0.5% NaOCl solution through the membrane.
Next, DI water was pumped through the membrane modules until the pH of the permeate was neutral.
Permeate flux and TMP indicated no irreversible fouling occurred during AnMBR operation.
Total biogas produced was measured using a flowmeter (GFM17 Flow Meter, Aalborg, Orangeburg, NY).
Headspace biogas was recirculated through sparging tubes below the membrane modules at a rate of 30
mL/min to scour the surface of the membranes and prevent membrane fouling. The working volume of
the reactor was maintained using an automatic level float switch. Effluent permeate flow was controlled
at a rate of 8 min filtration and 2 min backwashing per 10 min period using a peristaltic pump (BT100-1L
Multi-channel Peristaltic Pump, Longer, China). Trans-membrane pressure (TMP) of each membrane
module was measured using a pressure transducer. Permeate flux was controlled at 7 LMH, resulting in a
hydraulic retention time (HRT) of 16 hours. Sludge was not wasted except for sampling, resulting in a
155

solids retention time (SRT) of 300 days. AnMBR operational parameters were monitored and recorded
using LabVIEW 2014 (Student Edition).
Table S2.1 Synthetic wastewater composition.
Concentrate solution Dilution water
Reagent Concentration
(mg/L)
Reagent Concentration
(mg/L)
Ammonium Chloride 11.5 Sodium Bicarbonate 369
Calcium Chloride 11.5 Magnesium Phosphate 30.8
Iron Sulfate 7.7 Potassium Phosphate 13.8
Sodium Sulfate 11.5 Sodium Hydroxide 18.5
Sodium Acetate 27
Urea 87
Peptone 11.5
Yeast 46
Milk Powder 115.4
Soy Oil 13.5
Hydrochloric Acid 0.2
Starch 115.4
Chromium Nitrate 3.7
Copper Chloride 2.5
Manganese Sulfate 4.9
Nickel Sulfate 1.2
Lead Chloride 0.5
Zinc Chloride 1.2

Chemical oxygen demand (COD) was measured in accordance with USEPA Method 410.4 using a HI801
Spectrophotometer (Hanna Instruments, Woonsocket, RI, USA). Volatile fatty acids (Acetate, Propionate,
Formate and Valerate), sulfate, phosphate and chloride were measured by ion chromatography on an ICS
2100 (Thermo Fisher Scientific, Waltham, MA) using methods described previously.
1
Biogas samples from
156

the reactor headspace and effluent were analyzed using a Trace 1310 GC system (Thermo Scientific, NY)
with flame ionization detection (FID) as described previously.
1

B.2 Antibiotic quantification
All glassware, including sampling, stock, and standard solution preparation vials, were baked at 400 ºC
for a minimum of four hours and washed with methanol prior to use. 10 mL Samples (influent and effluent)
and standard solutions were filtered through 0.2 µm PTFE syringe filters (Whatman) using 10 mL syringes
with Luer lock tips and stored in certified 2 mL amber LC vials (Agilent) and stored at 4 ºC for no more
than 3 days prior to analysis. Stock solutions of sulfamethoxazole and erythromycin were prepared in
HPLC-grade methanol at concentrations of 20 mg/L and stored at -20 ºC. Ampicillin stock solution was
prepared in HPLC-grade water at 20 mg/L due to its lack of solubility in methanol and stored at 4 ºC. Five-
point standard calibration curves were generated within the appropriate range for each of the
incremental antibiotic influent concentration phases (i.e., 0.1-15 µg/L for 10 µg/L target influent
concentration, 0.5-70 µg/L for 50 µg/L, and 1-300 µg/L for 250 µg/L). All calibration curve R
2
values were
above 99%. Both solvent-based and matrix-matched calibration curves were generated for all three
compounds to ensure that the solvent based standards were representative of influent and effluent
concentrations. Specifically, sulfamethoxazole, ampicillin, and erythromycin were dissolved in both
influent and effluent solutions at concentrations of 0.1, 1, 10, 100, and 1000 µg/L and processed using the
same procedure employed to prepare reactor samples described above. Calibration curves of the results
from these solutions were plotted against standard solutions of the same concentrations dissolved in
HPLC-grade water. Both matrices (influent and effluent) showed equivalent concentrations and scaling at
the ranges detected in the HPLC-grade water with R
2
values of above 99.9% for all three antibiotics. No
isotope-labelled internal standards were used, as the samples were analyzed by direct injection LC-MS on
the same day that samples were collected (no solid phase extraction was needed).
157

Table S2.2 Targeted antibiotic properties and MS data acquisition parameters.
Compound Molecular
Weight (MW)
Retention Time
(min)
MS Spectrum
(m/z)
Fragmentor
Voltage (V)
Sulfamethoxazole 253.052 2.17 254.059 400
Erythromycin 733.461 2.35 734.469 100
Ampicillin 349.110 1.44 350.117 400

Sulfamethoxazole, erythromycin and ampicillin were all targeted using positive ESI MS-Q-TOF mode.
The LC gradient program for detection of all three compounds utilized 0.1% formic acid in water as mobile
phase A and acetonitrile as mobile phase B as follows: t = 0.0 min, A = 90% and B = 10%; t = 3.0 min, A =
0% and B = 100%; t = 5.0 min, A = 0% and B = 100%; t = 5.10 min, A = 90% and B = 10%. LC conditions used
included a flow rate of 0.4 mL/min, maximum pressure of 600 bar, column temperature of 40 ºC, and
autosampler tray temperature of 8 ºC. A post-column switch was used to divert the first 0.5 min of column
elution to waste to avoid sending hydrophilic compounds from the effluent matrix through the MS.
Injection volumes ranged from 0.5-10 µL, depending on the target sample range for each operational
phase, to ensure that no compound extracted ion chromatogram peaks exceeded saturation detection
values. MS conditions used were as follows: sheath gas temp. of 400 ºC, sheath gas flow rate of 12 L/min,
gas temperature of 225 ºC, drying gas flow rate of 5 L/min, nebulizer pressure of 20 psi, capillary voltage
of 3500V, nozzle voltage of 500V, acquisition rate of 1.5 spectra/s, and acquisition time of 667
ms/spectrum. Targeted compound acquisition parameters are provided in Table S2. All compound
detection and quantification analyses were performed using the Agilent MassHunter Qualitative Analysis
Navigator program.
158

B.3 Quantification of ARGs using real time qPCR
Standards for qPCR quantification were obtained by amplifying all ARGs (listed in Table S2.2) from DNA
extracted from activated sludge samples using gradient PCR (Mastercycler nexus, Eppendorf, Hamburg,
Germany). 2% agarose gel electrophoresis was employed to verify the size of PCR products, which were
then purified using Wizard
®
SV Gel and PCR Clean-Up System (Promega, Madison, WI). Purified amplicons
were cloned into linearized pMiniT 2.0 vectora and transformed into NEB 10-beta Competent E. coli using
the NEB PCR Cloning Kit (New England Biolabs, Ipswich, MA). Ampicillin selection plates were used to grow
E. coli cells overnight. Plasmids were then extracted from E. coli cells using PureLink™ Quick Plasmid
Miniprep kit (Invitrogen, Carlsbad, CA) and sent for sanger sequencing (Laragen Sequencing & Genotyping,
Culver City, CA) to confirm the presence of target ARGs in plasmids. Plasmid DNA concentrations were
measured using the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA). Standard curves were
established using serial dilutions of the purified plasmids (10
-1
to 10
-8
). For all ARGs, the intI1 gene, and
the rpoB gene, qPCR efficiencies ranged from 92% to 102%. Melting curve and gel electrophoresis were
performed to assure the specificity of each qPCR reaction. Table S2.2 shows the forward and reverse
primers, annealing temperatures, and general thermal cycling conditions, of all ARGs, as well as the intI1
and rpoB genes.

159

Table S2.3 Forward and reverse primers and qPCR thermocycling conditions of all ARGs, intI1 and rpoB gene.
160

Table S2.4 The abundance of targeted ARGs (Copy/rpoB) in biomass and biofilm during the addition of increment concentrations of
sulfamethoxazole, erythromycin and ampicillin. Abundance and errors respectively represent the mean values and standard deviations calculated
according to the triplicate qPCR results.

161

162

Table S2.5 The abundance of targeted ARGs (Copy/mL) in effluent during the addition of increment concentrations of sulfamethoxazole,
erythromycin and ampicillin. Abundance and errors respectively represent the mean values and standard deviations calculated according to the
triplicate qPCR results

163

164

Figure S2.1 Abundance of targeted ARGs in the biomass, biofilm (Copy/rpoB) and effluent (Copy/mL)
during the addition of 250 (µg/L) (A) erythromycin; (B) ampicillin. ARGs are sorted by the biomass
abundance (highest to the lowest). Bar charts and error bars respectively represent the mean values and
standard deviations calculated according to the results from three samples collected on at different days
during the addition of 250 (µg/L) of antibiotics and each of their triplicate qPCR results. For biofilm results,
mean values and standard deviations were calculated according to the triplicate qPCR results.
A
B
165

Figure S2.2 (A) Absolute abundance of total and antibiotic resistant bacteria; (B) antibiotic resistant
bacteria normalized to total bacteria; in the effluent of AnMBR during the addition of sulfamethoxazole,
erythromycin and ampicillin at increment concentrations. Error bars represent the standard deviation of
the results obtained from replicate samples. SMX, ERY, AMP and TET stand for sulfamethoxazole,
erythromycin, ampicillin and tetracycline; respectively. RB also stands for resistant bacteria.
0
1
2
3
4
5
6
7
8
42 52 59 66 73 83 100 111 118 125 132 140 169 179 185 191 199
Log 10 (CFU/mL)
Operation Time (day)
Total Bacteria AMP RB TET RB SMX RB ERY RB
Sulfamethoxazole
250 50 10
Erythromycin
250 50 10
Ampicillin
250 50 10
0
0.2
0.4
0.6
0.8
1
42 52 59 66 73 83 100 111 118 125 132 140 169 179 185 191 199
Resistant Bacteria/Total Bacteria
Operation Time (day)
AMP RB TET RB SMX RB ERY RB
Sulfamethoxazole
250 50 10
Erythromycin
250 50 10
Ampicillin
250 50 10
A
B
166

References
1. Chen, S.; Smith, A. L., Methane-driven microbial fuel cells recover energy and mitigate dissolved
methane emissions from anaerobic effluents. Environmental science: water research & technology 2018,
4, (1), 67-79.
2. Chen, J.; Yu, Z. T.; Michel, F. C.; Wittum, T.; Morrison, M., Development and application of real-time
PCR assays for quantification of erm genes conferring resistance to macrolides-lincosamides-
streptogramin B in livestock manure and manure management systems. Appl. Environ. Microbiol. 2007,
73, (14), 4407-4416.
3. Pei, R.; Kim, S.-C.; Carlson, K. H.; Pruden, A., Effect of river landscape on the sediment concentrations
of antibiotics and corresponding antibiotic resistance genes (ARG). Water Res. 2006, 40, (12), 2427-2435.
4. Barlow, R. S.; Pemberton, J. M.; Desmarchelier, P. M.; Gobius, K. S., Isolation and characterization of
integron-containing bacteria without antibiotic selection. Antimicrobial Agents and Chemotherapy 2004,
48, (3), 838-842.
5. Feria, C.; Ferreira, E.; Correia, J. D.; Goncalves, J.; Canica, M., Patterns and mechanisms of resistance to
beta-lactams and beta-lactamase inhibitors in uropathogenic Escherichia coli isolated from dogs in
Portugal. J Antimicrob Chemother 2002, 49, (1), 77-85.
6. Szczepanowski, R.; Linke, B.; Krahn, I.; Gartemann, K. H.; Gutzkow, T.; Eichler, W.; Puhler, A.; Schluter,
A., Detection of 140 clinically relevant antibiotic-resistance genes in the plasmid metagenome of
wastewater treatment plant bacteria showing reduced susceptibility to selected antibiotics. Microbiology
2009, 155, (Pt 7), 2306-19.
7. Aminov, R. I.; Garrigues-Jeanjean, N.; Mackie, R. I., Molecular ecology of tetracycline resistance:
development and validation of primers for detection of tetracycline resistance genes encoding ribosomal
protection proteins. Appl Environ Microb 2001, 67, (1), 22-32.
8. Dahllöf, I.; Baillie, H.; Kjelleberg, S., rpoB-based microbial community analysis avoids limitations
inherent in 16S rRNA gene intraspecies heterogeneity. Appl. Environ. Microbiol. 2000, 66, (8), 3376-3380.

167

Appendix C: Supporting information for chapter 4
C.1 Quantification of antibiotics by LC-MS
For antibiotics quantification, 10 mL samples were collected for each sampling time point from the
influent and effluent of the AnMBR. Both collected samples and standard solutions were filtered through
0.2 µm PTFE syringe filters (Whatman) using 10 mL syringes with Luer lock tips and stored in certified 2
mL amber LC vials (Agilent) at 4 ºC refrigerator for no more than 3 days prior to analysis. Stock solutions
of sulfamethoxazole and erythromycin were prepared in HPLC-grade methanol at concentrations of 20
mg/L and stored at -20 ºC. Ampicillin stock solution was prepared in HPLC-grade water at 4 mg/L due to
its lack of solubility in methanol and stored at 4 ºC. For each antibiotic, a six-point standard calibration
curve was constructed within the appropriate range (i.e., 0.1-30 µg/L to target effluent antibiotics and 30-
400 µg/L to target influent antibiotics). All calibration curve R2 values were above 0.99. Both solvent-
based and matrix-matched calibration curves were generated for all three compounds to ensure that the
solvent based standards were representative of influent and effluent concentrations.
Positive ESI MS-Q-TOF mode was employed to target sulfamethoxazole, erythromycin, and ampicillin.
The LC gradient program for detection of all three compounds utilized 0.1% formic acid in water as mobile
phase A and acetonitrile as mobile phase B as follows: t=0.0 min A=90% B=10%, t=3.0 min A=0% B=100%,
t=5.0 min A=0% B=100%, t=5.10 min A=90% B=10%. LC conditions used included a flow rate of 0.4 mL/min,
maximum pressure of 600 bar, column temperature of 40 ºC, and autosampler tray temperature of 8 ºC.
A post-column switch was used to divert the first 0.5 min of column elution to waste to avoid sending
hydrophilic compounds from the effluent matrix through the MS. Injection volumes ranged from 0.5-10
µL, depending on the target sample range (influent and effluent) for each operational phase, to ensure
that no compound extracted ion chromatogram peaks exceeded saturation detection values. MS
168

conditions used were as follows: sheath gas temp. of 400 ºC, sheath gas flow rate of 12 L/min, gas
temperature of of 225 ºC, drying gas flow rate of 5 L/min, nebulizer pressure of 20 psi, capillary voltage
of 3500V, nozzle voltage of 500V, acquisition rate of 1.5 spectra/s, and acquisition time of 667
ms/spectrum. Targeted compound acquisition parameters are provided in Table S1. All compound
detection and quantification analyses were performed using the Agilent MassHunter Qualitative Analysis
Navigator program.
Table S3.1 Targeted antibiotic properties and MS data acquisition parameters.
Compound Molecular
Weight (MW)
Retention Time
(min)
MS Spectrum
(m/z)
Fragmentor
Voltage (V)
Sulfamethoxazole 253.052 2.17 254.059 400
Erythromycin 733.461 2.35 734.469 100
Ampicillin 349.110 1.44 350.117 400

169

Table S3.2 Synthetic wastewater composition.
Concentrate solution Dilution water
Reagent Concentration
(mg/L)
Reagent Concentration
(mg/L)
Ammonium Chloride 11.5 Sodium Bicarbonate 369
Calcium Chloride 11.5 Magnesium Phosphate 30.8
Iron Sulfate 7.7 Potassium Phosphate 13.8
Sodium Sulfate 11.5 Sodium Hydroxide 18.5
Sodium Acetate 27
Urea 87
Peptone 11.5
Yeast 46
Milk Powder 115.4
Soy Oil 13.5
Hydrochloric Acid 0.2
Starch 115.4
Chromium Nitrate 3.7
Copper Chloride 2.5
Manganese Sulfate 4.9
Nickel Sulfate 1.2
Lead Chloride 0.5
Zinc Chloride 1.2

170

Table S3.3 Forward and reverse primers and qPCR thermocycling conditions of all ARGs, intI1, and rpoB gene.

171

Table S3.4 Biomass ARG abundances (Copy/rpoB) in the pre-antibiotics, antibiotics loading, and post-antibiotics periods. Abundance and errors
respectively represent the mean values and standard deviations calculated from triplicate qPCR results.

Table S3.5 Effluent ARG abundances (Copy/mL) in the pre-antibiotics, antibiotics loading, and post-antibiotics periods. Abundance and errors
respectively represent the mean values and standard deviations calculated from triplicate qPCR results.

172

Table S3.6 Correlation analysis results between the first 100 most abundant OTUs and ARGs in biomass and effluent of the AnMBR. Correlated
ARGs had strong significant correlation (p < 0.05; and ρ > 0.7 or ρ < −0.7) with the assigned OTUs. Green rows show correlation in biomass
samples and the yellow rows show correlation in effluent samples.

173

174

175

Figure S3.1 Rarefaction curves for (a) biomass and (b) effluent microbial community samples of the
AnMBR.

(a)
(b)
176

Figure S3.2 Performance of the AnMBR in (a) COD removal and biogas production, and (b) antibiotic
removal efficiency.
0
10
20
30
40
50
60
70
80
90
100
0
50
100
150
200
250
300
7 10 14 18 22 26 7 10 14 18 22 26 7 10 14 18 22 26
Removal %
Antibiotic Concentration (ug/L)
Time (day)
SMX-Inf SMX-Eff ERY-Inf ERY-Eff AMP-Inf AMP-Eff SMX Rmv ERY Rmv AMP Rmv
(a)
(b)
177

Figure S3.3 PCA of ARG profiles in the biomass and effluent of the AnMBR

E1
E6
E14
E20
E27
E35
E46
B1
B6
B14
B20
B27
B35
B46
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1
F2 (19.87 %)
F1 (77.91 %)
Biomass Effluent
178

Figure S3.4 Relative abundance of (a) methanogens and (b) syntrophic bacteria in the biomass of the
AnMBR throughout the experimental period. Day 1 represents pre-antibiotics period, days 14, 20, 27 and
35 represent antibiotics loading period (area bordered by red dashed line) and day 46 represents post-
antibiotics period.

0
2
4
6
8
10
1 6 14 20 27 35 46
Relative abundance (%)
Time (day)
Methanoregula
Methanobacterium
Uncl.Methanobacteriaceae
Methanosarcina
Methanomassiliicoccus
Uncl.Euryarchaeota
Methanomethylovorans
Methanosaeta
Methanolinea
0
2
4
6
8
10
1 6 14 20 27 35 46
Relative abundance (%)
Time (day)
Syntrophobacter
Desulfomonile
Syntrophus
Smithella
Uncl.Syntrophobacteraceae
Uncl.Syntrophorhabdus
Uncl.Syntrophaceae
(a)
(b)
179

Figure S3.5 Relative abundance of the (a) Biomass and (b) effluent Microbial Community in genus Level
throughout the experimental period. Day 1 represents pre-antibiotics period, days 14, 20, 27 and 35
represent antibiotics loading period and day 46 represents post-antibiotics.

(a)
(b)
180

C.2 COD Mass Balance in the Reactor
Influent COD: 453  32 mg/L
Effluent COD: 45  10 mg/L
Mixed Liquor Volatile Suspended Solid: 9.6  0.3 g/L
Biogas Methane Production: 536  14 mL/d
Effluent Dissolved Methane: 112  5 mL/d
Wasted Biomass: 10 mL/d
Volume of the Reactor: 5 L
gCOD/gVSS: 1.42
8

Figure S3.6 COD mass balance in the reactor.
0
0.5
1
1.5
2
2.5
Inpot Output
g COD/d
Mass of COD in Influent
Mass of COD in Effluent
Mass of COD as Dissolved Methane in the Effluent
Mass of COD as Biogas Methane
Mass of COD in the Wasted Biomass
Mass of COD for Sulfate Reduction
181

Reference
1. Chen, J.; Yu, Z. T.; Michel, F. C.; Wittum, T.; Morrison, M., Development and application of real-time
PCR assays for quantification of erm genes conferring resistance to macrolides-lincosamides-
streptogramin B in livestock manure and manure management systems. Appl. Environ. Microbiol. 2007,
73, (14), 4407-4416.
2. Pei, R.; Kim, S.-C.; Carlson, K. H.; Pruden, A., Effect of river landscape on the sediment concentrations
of antibiotics and corresponding antibiotic resistance genes (ARG). Water Res. 2006, 40, (12), 2427-2435.
3. Barlow, R. S.; Pemberton, J. M.; Desmarchelier, P. M.; Gobius, K. S., Isolation and characterization of
integron-containing bacteria without antibiotic selection. Antimicrobial Agents and Chemotherapy 2004,
48, (3), 838-842.
4. Feria, C.; Ferreira, E.; Correia, J. D.; Goncalves, J.; Canica, M., Patterns and mechanisms of resistance to
beta-lactams and beta-lactamase inhibitors in uropathogenic Escherichia coli isolated from dogs in
Portugal. J Antimicrob Chemother 2002, 49, (1), 77-85.
5. Szczepanowski, R.; Linke, B.; Krahn, I.; Gartemann, K. H.; Gutzkow, T.; Eichler, W.; Puhler, A.; Schluter,
A., Detection of 140 clinically relevant antibiotic-resistance genes in the plasmid metagenome of
wastewater treatment plant bacteria showing reduced susceptibility to selected antibiotics. Microbiology
2009, 155, (Pt 7), 2306-19.
6. Aminov, R. I.; Garrigues-Jeanjean, N.; Mackie, R. I., Molecular ecology of tetracycline resistance:
development and validation of primers for detection of tetracycline resistance genes encoding ribosomal
protection proteins. Appl Environ Microb 2001, 67, (1), 22-32.
7. Dahllöf, I.; Baillie, H.; Kjelleberg, S., rpoB-based microbial community analysis avoids limitations
inherent in 16S rRNA gene intraspecies heterogeneity. Appl. Environ. Microbiol. 2000, 66, (8), 3376-3380.
8. Grady Jr, CP Leslie, Glen T. Daigger, Nancy G. Love, and Carlos DM Filipe. Biological wastewater
treatment. CRC press, 2011.

182

Appendix D: Supporting information for chapter 5
Table S4.1 Synthetic wastewater composition.
Concentrate solution Dilution water
Reagent Concentration
(mg/L)
Reagent Concentration
(mg/L)
Ammonium Chloride 11.5 Sodium Bicarbonate 369
Calcium Chloride 11.5 Magnesium Phosphate 30.8
Iron Sulfate 7.7 Potassium Phosphate 13.8
Sodium Sulfate 11.5 Sodium Hydroxide 18.5
Sodium Acetate 27
Urea 87
Peptone 11.5
Yeast 46
Milk Powder 115.4
Soy Oil 13.5
Hydrochloric Acid 0.2
Starch 115.4
Chromium Nitrate 3.7
Copper Chloride 2.5
Manganese Sulfate 4.9
Nickel Sulfate 1.2
Lead Chloride 0.5
Zinc Chloride 1.2
D.1 Anaerobic membrane bioreactor operation
Headspace biogas was recirculated through sparging tubes below the membrane modules at a rate of
30 mL/min (for each membrane module) to scour the surface of the membranes and control membrane
fouling. Effluent permeate flow was controlled using a peristaltic pump (BT100-1L Multi-channel
183

Peristaltic Pump, Longer, China) at a rate of 8 min filtration and 2 min backwashing. Silicon carbide
membrane modules were purchased from Cembrane (Denmark). The average effective area for each
module was 150 cm2, and the nominal pores size was 0.1 µm. The maximum temperature and pressure
tolerance of the membrane modules were 60 C and 6 bar, respectively.

Figure S4.1 Schematic diagram of the bench-scale anaerobic membrane bioreactor.
D.2 Membrane cleaning
First, membrane modules were cleaned physically by removing attached foulants from the surface of
the membranes. To chemically clean the membrane modules, they were submerged in a 0.5% (v/v) NaOCl
solution overnight. Further, a peristaltic pump was used to pump the 0.5% NaOCl solution through the
membrane modules to enhance chemical cleaning. Next, DI water was pumped through the membrane
modules until the pH of the permeate was neutral. DI water filtration test was employed for each
membrane modules, before and after each cleaning cycle, and results revealed total recovery of both
permeate flux and TMP.
184

D.3 Analysis methods
Chemical oxygen demand (COD) was measured in accordance with USEPA Method 410.4 using a HI801
Spectrophotometer (Hanna Instruments, Woonsocket, RI, USA). Mixed liquor suspended solid (MLSS) and
mixed liquor volatile suspended solid (MLVSS) were determined according to standard methods described
previously.
1
Biogas samples from the reactor headspace and effluent were analyzed using a Trace 1310
GC system (Thermo Scientific, NY) with flame ionization detection (FID) as described previously.
1

D.4 Antibiotic quantification by LC-MS
Quantification of antibiotics was carried out according to the method described in our previous study.
1

Briefly, 10 mL samples were collected from the influent and effluent of each membrane modules. 0.2 µm
PTFE syringe filters (Whatman) was employed to filter collected samples and standard solutions. The
filtrates, then, were stored at 4 C refrigerator for no more than 3 days prior to analysis. Sulfamethoxazole
(SMX) and erythromycin (ERY) stock solutions were prepared in HPLC-grade methanol at concentrations
of 20 mg/L and stored at -20 C. Due to the lack of solubility of ampicillin (AMP) in methanol, its stock
solution was prepared in HPLC-grade water at 4 mg/L and stored at 4 C. For each antibiotic, a six-point
standard calibration curve was constructed. All calibration curve R
2
values were above 0.99. To ensure
that the solvent-based standards were representative of influent and effluent concentrations, both
solvent-based and matrix-matched calibration curves were generated for all three antibiotics. Antibiotic
quantification was performed by direct injection liquid chromatography mass spectrometry with
electrospray ionization (LC-ESI-MS).
Positive ESI MS-Q-TOF mode was employed to target SMX, ERY and AMP. 0.1% formic acid in water as
mobile phase A and acetonitrile as mobile phase B were used for the LC gradient program as follows:
T = 0.0 min, A = 90%, and B = 10%,
185

T = 3.0 min, A = 0%, and B = 100%,
T = 5.0 min, A = 0%, and B = 100%,
T = 5.10 min, A = 90%, and B = 10%.
LC conditions were flow rate of 0.4 mL/min, maximum pressure of 600 bar, column temperature of 40
C, and autosampler tray temperature of 8 C. To avoid sending hydrophilic compounds from the effluent
matrix through the MS, a post-column switch was employed to send the first 0.5 min of column elution to
waste. Injection volumes ranged from 2-10 µL, depending on the target sample concentration range
(influent and effluent), to ensure that no compound extracted ion chromatogram peaks exceeded
saturation detection values. MS conditions were as follows: sheath gas temperature of 400 C, sheath gas
flow rate of 12 L/min, gas temperature of 225 C, drying gas flow rate of 5 L/min, nebulizer pressure of 20
psi, capillary voltage of 3500V, nozzle voltage of 500V, acquisition rate of 1.5 spectra/s, and acquisition
time of 667 ms/spectrum. Targeted compound acquisition parameters are provided in Table S4.2. All
compound detection and quantification analyses were performed using the Agilent MassHunter
Qualitative Analysis Navigator program.
Table S4.2 Targeted antibiotic properties and MS data acquisition parameters.
Compound Molecular
Weight (MW)
Retention Time
(min)
MS Spectrum
(m/z)
Fragmentor
Voltage (V)
Sulfamethoxazole 253.052 2.17 254.059 400
Erythromycin 733.461 2.35 734.469 100
Ampicillin 349.110 1.44 350.117 400
D.3 DNA extraction efficiency
DNA extraction efficiencies were 12.3 ± 1.9%, 27.1 ± 11.1%, and 24.5 ± 22.4% for the biomass/biofilm,
effluent intracellular, and effluent extracellular DNA, respectively. These efficiencies were similar to a
186

previous study by Mumy and Findlay which analyzed/extracted DNA from different samples through three
commercial DNA extraction kits.
9
Low extraction efficiency of biomass/biofilm samples in the present
study may be due to their lower purity compared to the effluent samples which resulted in lower DNA
recovery.
187

Table S4.3 Forward and reverse primers and qPCR thermocycling conditions of all ARGs, intI1, and rpoB gene.

188

D.4 Fouling reduced antibiotic concentrations in the effluent of the
AnMBR
During the control experiment (Phase 1), no significant changes were observed in antibiotic removal in
the permeates of three membrane modules. AMP had the highest removal rate (86 ± 4%), followed by
SMX (82 ± 3%) and ERY (72 ± 4%). These results were similar to the observations from previous studies on
AnMBR systems.
1,10
At the start of Phase 3, the concentrations of SMX and ERY in the permeate of the LF
membrane were higher than their concentrations in the MF and HF membrane permeates. However, SMX
and ERY concentrations decreased gradually in the permeate of the LF membrane and reached the same
value as the MF and HF membrane after 3 weeks operation of the AnMBR in Phase 3 (Figure S4.1). Given
that fouling in AnMBRs starts with pore blockage due to the deposition of organic matters on the surface
of the membrane and/or inside of the pores, and continuous with the formation of cake layer,
11,12
it is
likely that removal of SMX and ERY was impacted by pore blockage (size exclusion), and not cake layer
formation (adsorption into the cake layer). AMP removal followed a different trend, showing a significant
positive correlation with the extent of membrane fouling in Phase 3 (p<0.02, and ρ>0.7). The highest
removal rate of AMP was achieved by the HF membrane, followed by the MF, and then the LF membrane.
The concentration of AMP further decreased continuously in the permeates of all three membrane
modules throughout Phase 3. Cheng et al. also stated that fouling layers can directly contribute to
antibiotic removal rates in AnMBR-fouled membrane units.
13
Based on our results, it seems that the
extent of fouling had a direct impact on AMP removal capacity by the system, whereas the effect of fouling
on SMX and ERY effluent concentrations was less clear. This observation may be due to the higher
degradability of AMP compared to SMX and ERY,
14
which would have likely been further enhanced by the
type of bacteria enriched in the biofilm and/or increased activity of the more mature biofilm (HF).
189

Figure S4.2 Effluent concentration of (a) sulfamethoxazole (SMX), (b) erythromycin (ERY), and (c)
ampicillin (AMP) during Phase 3. LF, MF and HF represent low, medium and highly fouled membranes,
respectively.

Figure S4.3 Absolute abundance of intracellular and extracellular (a) tetO, (b) tetW, (c) intI1, and (d) rpoB,
in the effluent of LF, MF and HF membranes during Phase 3. LF, MF and HF represent low, medium and
highly fouled membranes in Phase 3, respectively.
0
4
8
12
16
20
46 52 59 65 70
SMX concentration ( µg/L)
Time (day)
0
4
8
12
16
20
46 52 59 65 70
AMP concentration ( µg/L)
Time (day)
0
4
8
12
16
20
46 52 59 65 70
ERY concentration ( µg/L)
Time (day)
0
4
8
12
16
20
46 52 59 65 70
ERY concentration ( µg/L)
Time (day)
LF effluent MF effluent HF effluent
(a) (c) (b)
190

References
1. Zarei-Baygi A, Harb M, Wang P, Stadler LB, Smith AL. Evaluating Antibiotic Resistance Gene Correlations
with Antibiotic Exposure Conditions in Anaerobic Membrane Bioreactors. Environmental science &
technology. 2019 Feb 27;53(7):3599-609.
2. Chen, J.; Yu, Z. T.; Michel, F. C.; Wittum, T.; Morrison, M., Development and application of real-time
PCR assays for quantification of erm genes conferring resistance to macrolides-lincosamides-
streptogramin B in livestock manure and manure management systems. Appl. Environ. Microbiol. 2007,
73, (14), 4407-4416.
3. Pei, R.; Kim, S.-C.; Carlson, K. H.; Pruden, A., Effect of river landscape on the sediment concentrations
of antibiotics and corresponding antibiotic resistance genes (ARG). Water Res. 2006, 40, (12), 2427-2435.
4. Barlow, R. S.; Pemberton, J. M.; Desmarchelier, P. M.; Gobius, K. S., Isolation and characterization of
integron-containing bacteria without antibiotic selection. Antimicrobial Agents and Chemotherapy 2004,
48, (3), 838-842.
5. Feria, C.; Ferreira, E.; Correia, J. D.; Goncalves, J.; Canica, M., Patterns and mechanisms of resistance to
beta-lactams and beta-lactamase inhibitors in uropathogenic Escherichia coli isolated from dogs in
Portugal. J Antimicrob Chemother 2002, 49, (1), 77-85.
6. Szczepanowski, R.; Linke, B.; Krahn, I.; Gartemann, K. H.; Gutzkow, T.; Eichler, W.; Puhler, A.; Schluter,
A., Detection of 140 clinically relevant antibiotic-resistance genes in the plasmid metagenome of
wastewater treatment plant bacteria showing reduced susceptibility to selected antibiotics. Microbiology
2009, 155, (Pt 7), 2306-19.
7. Aminov, R. I.; Garrigues-Jeanjean, N.; Mackie, R. I., Molecular ecology of tetracycline resistance:
development and validation of primers for detection of tetracycline resistance genes encoding ribosomal
protection proteins. Appl Environ Microb 2001, 67, (1), 22-32.
8. Dahllöf, I.; Baillie, H.; Kjelleberg, S., rpoB-based microbial community analysis avoids limitations
inherent in 16S rRNA gene intraspecies heterogeneity. Appl. Environ. Microbiol. 2000, 66, (8), 3376-3380.
9. Grady Jr, CP Leslie, Glen T. Daigger, Nancy G. Love, and Carlos DM Filipe. Biological wastewater
treatment. CRC press, 2011.
9. Mumy, K. L.; Findlay, R. H., Convenient determination of DNA extraction efficiency using an external
DNA recovery standard and quantitative-competitive PCR. J. Microbiol. Methods 2004, 57, (2), 259-268.
10. Baygi, A. Z.; Harb, M.; Wang, P.; Stadler, L.; Smith, A. L., Microbial community and antibiotic resistance
profiles of biomass and effluent are distinctly affected by antibiotics addition to an anaerobic membrane
bioreactor. Environmental Science: Water Research & Technology 2020.
11. Huang, Z.; Ong, S. L.; Ng, H. Y., Submerged anaerobic membrane bioreactor for low-strength
wastewater treatment: effect of HRT and SRT on treatment performance and membrane fouling. Water
Res. 2011, 45, (2), 705-713.
12. Lin, H.; Peng, W.; Zhang, M.; Chen, J.; Hong, H.; Zhang, Y., A review on anaerobic membrane
bioreactors: applications, membrane fouling and future perspectives. Desalination 2013, 314, 169-188.
191

13. Cheng, D.; Ngo, H. H.; Guo, W.; Liu, Y.; Chang, S. W.; Nguyen, D. D.; Nghiem, L. D.; Zhou, J.; Ni, B.,
Anaerobic membrane bioreactors for antibiotic wastewater treatment: performance and membrane
fouling issues. Bioresource technology 2018, 267, 714-724.
14. Alexy, R.; Kümpel, T.; Kümmerer, K., Assessment of degradation of 18 antibiotics in the closed bottle
test. Chemosphere 2004, 57, (6), 505-512.

Abstract (if available)

Abstract Antibiotic resistance infection is one of the biggest threats to human health, which is currently responsible for over 700,000 death annually worldwide. The world health organization (WHO) has called for urgent action to avert an antimicrobial crisis, and the U.S. Centers for Disease Control and Prevention (CDC) has identified addressing antibiotic resistance as a national priority. The CDC has proposed five core actions to better prepare the United States for an antibiotic resistance pandemic, one of which is keeping antibiotic resistance from entering the environment. Wastewater treatment plants (WWTPs) as the main interfaces between the build and natural environment, has been identified as a primary source of antibiotic resistance spread into the environment. Conventional wastewater treatment technologies have not been designed to mitigate release of antibiotic resistance, and more advanced technologies are required. Anaerobic membrane bioreactors (AnMBRs) are an emerging biotechnology that can provide similar treatment performance to aerobic processes, while promoting energy and nutrient recovery. AnMBRs have unique features that can also impact release of antibiotic resistance, however, they remained unexplored in this regard. Here, first we evaluated the role of influent antibiotics on antibiotic resistance gene (ARG) profile of the biomass and effluent of a bench-scale AnMBRs. A gradual increase in biomass ARG profile, and an initial increase followed by gradual decrease in effluent ARG profile were observed. ARGs can be transferred vertically and horizontally, however, it was unclear which transfer mechanisms was responsible for the variation of the biomass and effluent ARG profiles. Next, we designed another experiment, introducing three antibiotics to the influent of the AnMBR, and evaluated microbial community structures, ARG profiles, and their potential association. The gradual increase in biomass ARG profile in the presence of antibiotics, while biomass microbial community structure was not impacted by antibiotics addition, indicated the greater influence of horizontal gene transfer (HGT) compared to vertical gene transfer (VGT) in variation of ARG profile in the AnMBR. Effluent microbial community structure, however, shifted significantly upon initial exposure to antibiotics, probably due to its lower diversity (richness and evenness) compared to the biomass community. Last, we investigated role of membrane fouling layer on release of intracellular and extracellular ARGs (iARGs and eARGs) from an AnMBR. Results reveled that, compared to the biomass, fouling layer provided more conducive condition for HGT. Fouling layer also increased abundance of iARGs in the effluent, while it reduced eARG abundances. As an emerging biotechnology, AnMBRs are expected to be the future of waste management technologies, and this dissertation provides worthwhile information on operational strategies and design for full-scale AnMBR systems.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Wastewater-based epidemiology for emerging biological contaminants

PDF

Advancing energy recovery from food waste using anaerobic biotechnologies: performance and microbial ecology

PDF

Bioelectrochemical treatment of anaerobic process effluents: mitigation of dissolved methane and sulfide

PDF

Optimizing biomembrane reactor systems for water reclamation and reuse applications

Asset Metadata

Creator Zarei Baygi, Ali (author)

Core Title Fate of antibiotic resistance in anaerobic membrane bioreactors

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Engineering (Environmental Engineering)

Publication Date 08/10/2020

Defense Date 08/10/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag anaerobic membrane bioreactor,antibiotic resistance gene,biofilm,horizontal gene transfer,intracellular and extracellular,microbial community,OAI-PMH Harvest,resource recovery

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Smith, Adam (committee chair), Boedicker, James (committee member), Childress, Amy (committee member), McCurry, Daniel (committee member)

Creator Email ali.zarei.baygi@gmail.com,zareibay@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-363329

Unique identifier UC11666742

Identifier etd-ZareiBaygi-8911.pdf (filename),usctheses-c89-363329 (legacy record id)

Legacy Identifier etd-ZareiBaygi-8911.pdf

Dmrecord 363329

Document Type Dissertation

Rights Zarei Baygi, Ali

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA