Intracellular drug-drug interaction between nucleoside analogs leads to early virologic failure in HIV patients receiving triple nucleoside combinations of tenofovir, lamivudine and abacavir or d...

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INTRACELLULAR DRUG-DRUG INTERACTION BETWEEN NUCLEOSIDE
ANALOGS LEADS TO EARLY VIROLOGIC FAILURE IN HIV PATIENTS
RECEIVING TRIPLE NUCLEOSIDE COMBINATIONS OF TENOFOVIR,
LAMIVUDINE AND ABACAVIR OR DIDANOSINE

by

Lucun Bi

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHARMACEUTICAL SCIENCES)

May 2008

Copyright 2008 Lucun Bi
ii

Dedication

To Liz Li and Alexander Bi
iii
Acknowledgements

The work on this thesis has been an inspiring, challenging and always
interesting experience. It would have been impossible without the support of
many people around me.
First of all, I would like to express my deepest gratitude to my advisor,
Prof. Stan Louie for giving me this valuable opportunity to take on these
challenging projects and his patient guidance, encouragement and excellent
advice throughout this study. I am especially thankful for the continuous
training he gave me in the area of cellular biology and HIV clinical
pharmacology.
Next, I would like extend my sincere gratitude to the advisory
committee members, Professors Gilbert Burckart, David D’Argenio, Wei-
chiang Shen and Clay Wang. This study would not have been successful
without their valuable advice and guidance. I am especially grateful to Dr.
D’Argenio for his advice during the initial phase of my graduate study and his
continuous teaching in pharmacokinetic modeling.
I owe a special thank to Dr. Miguel Goicoechea for his advice and
instrumental role in the planning and execution of the clinical study. I warmly
thank the twenty-one HIV patients for their participation in this clinical study
and for their goodwill for advancing the science.
iv
I would also like to thank GlaxoSmithKline and Bristol Myer Squibb for
their financial support of my graduate study; and thank Gilead Science and
Taylor Lab for their contribution towards the clinical study.
Special thanks to Dr. Jerika Lam and Dr. Haejung An for their generous
help during my graduate study; to my lab mates, Dr. Run Tian, Shanshan Liu,
Dr. Nick Mordwinkin and Jared Russell for sharing the harmonious
environment with me.
I want to thank my parents for raising me through the difficult time, for
teaching me to work hard, care for the needy and be positive towards life. I
hope that they will share my success and happiness in heaven.
Last, but not the least, I want to thank my dear wife Liz Li and my son
Alexander Bi for their continuous support and understanding throughout my
study. I would also want to thank my parent-in-law for their moral support and
understanding. Without their caring heart and unselfish sacrifice in the past
few years, I would never have been able to accomplish this mission. I am
grateful for their unconditional love.
v
Table of Contents

Dedication ................................................................................................. ii
Acknowledgements......................................................................................... iii
List of Tables ................................................................................................ ix
List of Figures ................................................................................................ xi
Abbreviations ...............................................................................................xiii
Abstract ..............................................................................................xvii
Prologue: Significance .......................................................................... xix
Chapter 1: Introductions ............................................................................1
Success of Trizivir.....................................................................................2
Early virologic failure
Tenofovir, Lamivudine and Abacavir combination ...................3
Tenofovir, Lamivudine and Didanosine combination ...............4
Tenofovir, Didanosine and Efavirenz; Tenofovir,
Lamivudine and Nevirapine .....................................................5
Nucleoside pharmacology ........................................................................6
Pharmacological mechanism for early virologic failure .............................8
Low genetic barrier ..................................................................8
Pharmacodynamic interactions................................................9
Intracellular drug-drug interactions ........................................10
Intracellular Tenofovir and Didanosine interaction..................................10
TFV and ddI interaction in clinical setting ..............................10
TFV and ddI interaction in in vitro studies..............................11
Interaction of TFV and ddI at purine nucleoside
phosphorylase .......................................................................11
Involvement of endogenous nucleotide pools.........................................13
Role of cellular adaptive mechanism in cellular resistance to
nucleoside analogs................................................................16
Expression of metabolic enzymes .........................................17
Expression of efflux transporters ...........................................17
vi
Chapter 1 References.............................................................................21
Chapter 2: Determination of Triphosphate of Nucleoside Analogs
and Endogenous Nucleotide Pools by Liquid
Chromatography Coupled to Tandem Mass Spectrometry....27
Chapter 2 Abstract..................................................................................27
Chapter 2 Introduction ...........................................................................28
Chapter 2 Experimental methods ...........................................................31
Chemicals..............................................................................31
Cells.......................................................................................32
Instrumentation .....................................................................32
Preparation of calibration standards ......................................33
Preparation of quality control samples...................................34
Sample processing ...............................................................35
LC/MS/MS analysis ..............................................................36
Validation of procedures.........................................................................38
Extraction efficiency...............................................................39
Linearity.................................................................................39
Accuracy and precision..........................................................40
Extraction recovery................................................................40
Incubation time ......................................................................41
Stability .................................................................................41
Chapter 2 Results...................................................................................42
Extraction efficiency...............................................................42
Linearity, accuracy and precision...........................................43
Intraday variation ..................................................................47
Extraction recovery................................................................49
Incubation time ......................................................................49
Stability..................................................................................50
Chapter 2 Discussion..............................................................................52
Chapter 2 Conclusions ...........................................................................55
Chapter 2 References.............................................................................56
Chapter 3: Intracellular Drug-Drug Interaction Among Nucleoside
Analogs Lead to Early Virologic Failure in HIV Patients
Receiving Combination of Abacavir, Tenofovir and
Lamivudine ............................................................................59
Chapter 3 Abstract..................................................................................59
Chapter 3 Introduction ...........................................................................61
Chapter 3 Materials and methods...........................................................65
Chemicals..............................................................................65
Cell lines ................................................................................66
Intracellular drug-drug interaction study.................................66
24 hours treatment.................................................................66
vii
7 days treatment....................................................................67
Concentration escalation study..............................................67
Cellular viability assay ...........................................................68
Western blot analysis of MRP2 and MRP4............................68
Determination of triphosphate of nucleoside analogs ............69
Sample preparation and processing .....................................69
LC/MS/MS analysis ..............................................................71
Data analysis .........................................................................72
Statistical analysis ................................................................73
Chapter 3 Results ..................................................................................74
24 hours treatment.................................................................74
7 days treatment ...................................................................75
Concentration escalation study..............................................77
Effect of long-term treatment with nucleoside analogs on
U937
WT
.................................................................................77
ddNTP formation in U937
WT
and U937
TFV
cells .....................82
Chapter 3 Discussion..............................................................................83
Chapter 3 References.............................................................................94
Chapter 4: Pharmacologic Mechanism Leading to Early Virologic
Failure of Didanosine, Lamivudine and Tenofovir
Combinaiton .........................................................................99
Chapter 4 Abstract..................................................................................99
Chapter 4 Introduction ..........................................................................100
Chapter 4 Materials and methods.........................................................103
Chemicals............................................................................103
Cells.....................................................................................103
Cellular viability assays........................................................104
Intracellular interaction among tenofovir, didanosine and
lamivudine ...........................................................................105
Concentration escalation study............................................105
Cell processing for triphosphate determination....................106
LC/MS/MS analysis .............................................................107
Western analysis of MRP2 and MRP4.................................108
Data analysis ......................................................................109
Statistical analysis ...............................................................110
Chapter 4 Results.................................................................................110
Chapter 4 Discussion............................................................................118
Chapter 4 References...........................................................................128
Chapter 5: Intracellular and Plasma Interaction Between Abacavir
and Tenofovir in Treatment Naïve HIV Patients...................132
Chapter 5 Abstract................................................................................132
Chapter 5 Introduction .........................................................................133
viii
Chapter 5 Materials and methods.........................................................136
Chemicals............................................................................136
Study design........................................................................137
Sample size .........................................................................138
Study population .................................................................138
Drug administration and sample collection .........................139
Determination of plasma abacavir and tenofovir
concentration .......................................................................140
Plasma ABC analysis ..........................................................140
Plasma TFV analysis ..........................................................141
Determination of TFV-DP and CBV-TP concentration in
PBMC ..................................................................................142
Data analysis .......................................................................143
Pharmacokinetic analysis ....................................................143
Statistic analysis ..................................................................144
Chapter 5 Results.................................................................................145
Chapter 5 Discussion............................................................................149
Chapter 5 References...........................................................................157
Chapter 6: Summary and Future Work..................................................160
Chapter 6 References ..........................................................................166
Bibliography ............................................................................................168
ix
List of Tables
Table 1.1 Half life of nucleoside analogs in plasma and their
corresponding active moieties (ddNTP) in PBMC cells. ..........7
Table 1.2 Intracellular interaction between Tenofovir disoproxil
fumarate (TDF) and didanosine (ddI) in HIV infected
patients..................................................................................12
Table 2.1 Working standard solution of ddNTP in distilled water ..........34
Table 2.2 The mass transition and retention time for each
nucleoside analogs corresponding to ddNTP and dNTP .......38
Table 2.3A Linear calibration curves for ddNTP ......................................45
Table 2.3B Linear calibration curves for dNTP ........................................46
Table 2.4 Intraday precision and accuracy of ddNTP and dNTP...........48
Table 2.5 Extration recovery of ddNTP and internal stnadards ............48
Table 3.1 Change in ddNTP in U937 cells after 24 hour treatment ......74
Table 3.2 Change in ddNTP/dNTP ratio in U937 cells after 24 hour
treatment .............................................................................. 74
Table 3.3 Change in ddNTP in U937 cells after 7 day treatment...........76
Table 3.4 ddNTP/dNTP ratio changes in U937 cells after 7 day
treatment ..............................................................................76
Table 4.1 Intracellular interaction between TFV, 3TC and ddI in
CEMss or U937 cells after 24 hour treatment......................111
x
Table 5.1 Patient demographics for CCTG584 clinical study ..............144
Table 5.2 Intracellular TFV-DP PK parameters difference between
two sequences.....................................................................145
Table 5.3 The ratio changes in plasma PK parameters of TFV
between two sequences .....................................................148
Table 5.4 The ratio changes in plasma PK parameters of ABC
between two sequences .....................................................148
xi
List of Figures
Figure 1.1 Intracellular interaction between TFV and ddI in HIV
patients ....................................................................................12
Figure 1.2 Inhibition of purine nucleoside phosphorylase (PNP) by
anabolites of tenofovir ..............................................................14
Figure 1.3 Intracellular activation of nucleoside analogs ...........................19
Figure 2.1 LC/MS/MS chromatogram of system suitability .......................37
Figure 2.2 Response of CBV-TP after solid phase extraction on WAX
and MAX...................................................................................42
Figure 2.3 Extraction efficency in removing ddNMP..................................44
Figure 2.4 Presence of dGMP and dGDP has no impact on dGTP and
dATP after extraction on WAX..................................................44
Figure 2.5 Enzyme incubation time ...........................................................49
Figure 2.6 Storage stability of dNTP stock solution under -80ºC ...............51
Figure 2.7 Storage stability of dNTP in HPLC vial over 24 hours ..............51
Figure 3.1 Impact of TFV concentration on intracellular interaction
between ABC and TFV in CEM cells........................................78
Figure 3.2 Impact of ABC concentration on intracellular interaction
between ABC and TFV in CEM cells .......................................79
Figure 3.3 Cellular viability of U937
WT
and U937
TFV
variant cells
treated with ABC and TFV for 48 hours....................................80
xii
Figure 3.4 Formation of ddNTP in U937
WT
and U937
TFV
variant cells .......81
Figure 4.1 Intracellular interaction between TFV and ddI in CEMss
treated with TFV and escalting concentration of ddI for 24
hours .....................................................................................113
Figure 4.2 Intracellular changes in dNTP level in CEMss treated with
TFV and escalating concentration of ddI for 24 hours............114
Figure 4.3 Expression of MRP2 and MRP4 in U937 and CEMss cells
treated with increasing concentration of TFV for 24 hours ....115
Figure 4.4 Relationship between cell viability and expression of MRP4
in CEMss and its variant cells ................................................117
Figure 4.5 Schematic diagram of physiological mechanisms for
intracellular drug interaction between TFV and ddI ................121
Figure 5.1 CCTG584 clinical study design for the investigation of
intracellular and plasms drug-drug interaction between
TDF and ABC in treatment naïve HIV patients.......................138
Figure 5.2 Intracellular CBV-TP PK parameters between the two
sequences..............................................................................147
xiii
Abbreviations

2 Cl A 2-Chloro adenosine
2-Cl ATP 2-Chloro adenosine triphosphate
3TC Lamivudine
3TC-MP Lamivudine-monophosphate
3TC-TP Lamivudine-triphosphate
6N-ME 6-methyl analog of abacavir
ABC Abacavir
ABC-MP Abacavir-monophosphate
AK2 Adenylate kinase 2
ADV Adefovir
ADV-DP Adefovir-diphosphate
ANP Acyclic nucleoside phosphonate
AUCτ Area under the curve during dosing interval
AZT Zidovudine
AZT-MP Zidovudine monophosphate
CBV Carbovir
CBV-MP Carbovir-monophosphate
CBV-TP Carbovir-triphosphate
CLss/F Apparent steady state clearance
Cmax Maximum concentration
xiv
Ctrough Trough concentration
dA 2’deoxyadenosine
dATP 2’-deoxyadenosine triphosphate
dAMP 2’-deoxyadenosine monophosphate
dC 2’-deoxycytidine
dCK Deoxycytidine kinase
dCTP 2’-deoxycytidine triphosphate
ddA 2’,3’-didoexyadenosine
ddAMP 2’,3’-dideoxydanosine monophosphate
ddATP 2’,3’-dideoxyadenosine triphosphate
ddC 2’,3’-didoexycytidine
ddCTP 2’,3’-dideoxycytidine triphosphate
ddG 2’,3’-dideoxyguanosine
ddGTP 2’,3’dideoxyguanosine triphosphate
ddI Didanosine
ddIMP 2’,3’-dideoxyinosine monophosphate
ddN Nucleoside analog
ddNDP 2’,3’ dideoxynucleoside analog diphosphate
ddNMP 2’,3’ dideoxynucleoside analog monophosphate
ddNTP 2’,3’ dideoxynucleoside analog triphosphate
dGTP 2’-deoxyguanosine triphosphate
dGDP 2’-deoxyguanosine diphosphate
xv
dGMP 2’-deoxyguanosine monophosphate
dNTP Endogenous nucleotide pools
dT 2’ thymidine
dTTP 2’-deoxythymidine triphosphate
EFV Efavirenz
GK Guanylate kinase
Geo. Mean Geometric mean
HAART Highly active antiretroviral therapy
HGPRT Hypoxanthine-guanine phosphoribose transferase
HIV Human immunodeficiency virus
HPLC High pressure liquid chromatography
IC Intracellular
IMPDH Inosine monophosphate dehydrogenase
LC/MS/MS Liquid chromatography coupled to tandem mass
spectrometry

MAX Strong anion extraction
MRP2 Multidrug resistant protein 2
MRP4 Multidrug resistant protein 4
MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide

NA Nucleoside analog
NCA Non-compartmental analysis
NNRTI Non-nucleoside reverse transcriptase inhibitor
xvi
NRTI Nucleoside reverse transcriptase inhibitor
NVP Nevirapine
PK Pharmacokinetics
PBMC Peripheral blood mononuclear cells
PI Protease inhibitor
PNP Purine nucleoside phosphorylase
RT Reverse transcriptase
SPE Solid phase extraction
TAM Thymidine analog mutation
TDF Tenofovir disoproxil fumarate
TMF Tenofovir mono-isoproxil fumarate
TFV Tenofovir
TFV-MP Tenofovir monophosphate
TFV-DP Tenofovir diphosphate
RR Ribonucleotide reductase
Vss/F Apparent steady state volume of distribution
WAX Weak anion extraction
xvii
Abstract

High level of virologic failure was observed in HIV patients receiving
combinations of tenofovir (TFV), lamivudine (3TC) combined with either
abacavir (ABC) or didanosine (ddI). To investigate the pharmacologic
mechanisms involve with the virologic failures, a comprehensive study was
undertaken to evaluate the intracellular concentration of the active moiety
(ddNTP) of these respective nucleoside analogs.
Triple nucleoside combinations were tested at physiological
concentrations revealed reductions of 5 to 33% of the respective ddNTP. In
vitro studies evaluating TFV and ABC in a concentration dependent manner
showed a reduction of 40% and 30% in intracellular CBV-TP and TFV-DP.
Similar findings were demonstrated with TFV and ddI, where 40% and 25%
reduction in ddATP and TFV-DP, respectively, was detected. The level of
dATP and dGTP, endogenous nucleotides, were increased by 2.4- and 2.7-
fold, respectively, when cells were treated with 20 μM TFV, which can dilute
the effect of ddNTP.
The expression of MRP2 and MRP4 were increased in cells serially
passaged in either ABC or TFV, which correlated with cellular viability in the
presence of high concentrations of nucleosides. Moreover, these cells had
significantly lower level of ddNTP accumulation as compared to the wild type
counterparts.
xviii
A clinical study was undertaken to evaluate whether the in vitro findings
were also observed in humans. The combination of ABC and TDF resulted in
a 2-fold increase in intracellular TFV-DP in patients receiving the combination
of ABC and TFV as compared to TDF alone. Increases in TFV-DP may be
attributed to higher plasma level of TFV, which was detected in the second
phase of this study. A 4-fold increase in CBV-TP was detected three of nine
patients, while the other six (6/9) had a 41% reduction. The degree of CBV-TP
reduction is consistent with what is seen in vitro.
These findings suggest that cellular adaptive are critical in reducing
intracellular levels of nucleoside analogs and their corresponding ddNTP,
which may increase risk of virologic failures. The underlying pharmacological
mechanisms may include but are not limited to, the competitive inhibition of
anabolic enzymes, increase expression of efflux transporters and increase
dATP and dGTP, where the resultant effect is reduced antiviral activity.
xix
Prologue: Significance

This study comprehensively investigated the mechanism(s) that may
lead to intracellular drug-drug interaction between nucleoside analogs in CD4
positive lymphocytes. The intracellular interaction between TFV in
combination with either ABC or ddI was evaluated over various concentration
ranges and incubation times, where the end result was a significant reduction
in CBV-TP and ddATP formation, respectively. Furthermore, these results
were confirmed by a clinical study, in which intracellular interaction between
ABC and TFV in treatment naïve HIV patients resulted in 28% median
reduction of CBV-TP in ABC group and a 41% reduction in CBV-TP in a
subset of six out of 9 HIV patients. These results suggested intracellular drug-
drug interaction may be a major cause of early virologic failure observed in
HIV patients receiving combinations containing 3TC, TFV and either ABC or
ddI.
In addition, this study found that the presence of TFV could increase
endogenous deoxyadenosine triphosphate (dATP) and deoxyguanosine
triphosphate (dGTP) by two-fold. This study further found that long-term
exposure of CD4 cells to nucleoside analogs can trigger cellular adaptive
mechanisms resulting in increased expression of MRP2 and MRP4.
Increased expression of these efflux transporters could reduce levels of the
phosphorylated nucleoside analogs and thus increase the risk of viral
resistance to emerge as a consequence.
xx
Reduced levels of nucleotide analogs can lead to viral resistance,
where the viral mutations observed in patients are consistent with the
reduction of triphosphorylated anabolites, and the ultimate consequence is
virologic failure. This study has produced compelling data suggesting a
multitude of pharmacologic mechanisms that may give rise to clinical virologic
failures. The paradigm used in this series of studies may be a good template
for evaluating nucleoside combinations, and in the drug development process
to optimize effective nucleoside-based combination treatment for HIV and
cancer patients.
1
Chapter I. Introduction

Highly active antiretroviral therapy (HAART) has achieved great
success in inhibiting viral replication and delaying the disease progression of
human immunodeficiency syndromes (FDA, 2005; Palella et al, 1998).
Despite this, HAART continues to be unable to completely eradicate the virus
from the infected host, thus HIV infected patients will require life long antiviral
suppressive therapy to prevent the progression of disease (FDA, 2005; Finzi
et al, 1997; Siliciano et al, 2003).
Success of long-term treatment not only hinges on the presence of
potent components in the drug regimens, but also on the ability to minimize
drug-related side effects and thus increase tolerability, which will lead to an
increase in drug adherence. HAART is usually a combination of two
nucleoside reverse transcriptase inhibitors (NRTIs) and either a non-
nucleoside reverse transcriptase inhibitor (NNRTI) or protease inhibitor (PI).
However, prolonged treatment with PIs often cause metabolic and
morphologic abnormalities, such as fat accumulation, abnormalities of lipid
and glucose metabolism, and increased risk of cardiovascular disease
(Winston et al, 2005). In addition, combinations containing NNRTIs and PIs
often have complex dosing schedules and food restrictions and many
adverse effects, which may reduce patient’s adherence to the treatment and
result in an increase in the risk of clinical failure. As NRTIs are relatively
efficacious and the combinations generally have more convenient dosing
2
schedules, thereby the triple nucleoside combination therapy evolves
naturally as a good first choice.

Success of Trizivir
Trizivir
® , which is a fixed dose combination of zidovudine (AZT), ABC
and lamivudine (3TC) given as a twice daily regimen, was the first triple
nucleoside combination being evaluated in HIV clinical studies (Gulick et al,
2004). Gulick et al conducted a clinical study to compare the efficacy of triple
nucleoside therapy to the standard of treatment, which included two or three
nucleosides plus efavirenz (EFV). The results showed that 61% and 74% of
HIV patients in the triple nucleoside group were able to achieve viral
suppression and had an HIV-1 RNA level less than 50 and 200 copies per
milliliter (copies/mL) at week 48, respectively, as compared to 83% and 89%
in the EFV group at week 48, respectively. Although not as efficacious as the
EFV-containing combination, Trizivir
® had significant improvements in drug
adherence as it had a simplified dosing schedule, reduced the need for food
requirement and had fewer drug-drug interactions. Because of these
advantages, Trizivir
® was used extensively in some special populations such
as adolescents or as a maintenance regimen for patients who achieved
durable viral suppression after receiving PI or NNRTI-containing treatment
(Cuzin et al, 2005; Gulick et al, 2004; Handforth and Sharland, 2004).
3
Triple nucleoside combination of TFV, ABC and 3TC
In an effort to further refine and improve this pharmacologic strategy,
the triple nucleoside combination of TDF, ABC and 3TC in a once-daily
regimen was developed. In this combination, the newly approved drug TDF
was used to replace AZT, which had been used extensively over the past
twenty years, developing many resistant viruses selecting against it. TDF has
more potent antiviral activity as compared to AZT, and TDF maintains its
antiviral activity in AZT resistant viral strains (Kearney et al, 2004).
Unfortunately, HIV patients receiving combinations containing TDF
experienced early virologic failure in a number of clinical studies when it was
part of a triple nucleoside regimen (Farthing C, 2003; Gallant JE, 2003;
Landman R, 2004).
In the ESS30009 study, the triple nucleoside combination of TDF,
ABC and 3TC was compared to EFV, ABC and 3TC in HIV-positive
treatment-naïve patients. In the unplanned analysis of the 192 patients in this
study, it was found that 49% of patients receiving triple nucleoside regimens
failed to achieve viral suppression or had viral rebound after achieving viral
suppression, as compared to 5% in the EFV group within treatment weeks 8
through 12. In light of these findings, the TDF arm was halted and patients in
the triple nucleoside group were allowed to switch to other regimens. The
genotypic analysis was performed for patients who had virologic failure. In
this study, 98% of patients had HIV with the M184V/I mutation in viral reverse
transcriptase gene, while 54% of non-responders had both M184V/I and
4
K65R mutations. While 3TC and ABC selected for the M184V/I mutation; the
K65R mutation was shown to reduce the susceptibility to both ABC and TDF.
These results were confirmed by two other small clinical studies (Khanlou et
al, 2005, Landman et al, 2004).

Early virologic failure of combination of TDF, ddI and 3TC
Similarly, HIV patients receiving the combination of TDF, 3TC and ddI
had early virologic failures as well (Jemsek J, 2004). In a small clinical study,
twenty-two HIV-positive treatment-naïve patients were treated with the once
daily combination of ddI, 3TC and TDF. Twenty patients experienced early
virological failure with the median time to failure of 16 weeks. The virological
failure was defined as either failure to achieve viral suppression (defined as
viral load < 50 copies per mL of plasma) or viral rebound after achieving
initial viral suppression. Viral resistance testing of viruses derived from the
plasma of these patients showed that all twenty patients had the M184V/I
mutation, while 10 patients (50%) had an additional K65R viral mutation
which was consistent with resistance towards 3TC, and ddI and TDF,
respectively.
Virologic failures were also observed in patients switching to these two
triple nucleoside combinations after achieving viral suppression with a
previous regimen (Hoogewerf et al, 2003; Wirden et al, 2004). In these
patients, viral load was undetectable (<200 copies/mL), however 7 of 11
patients in this group developed virologic failure after six months, as
5
compared to none in the group of patients receiving PI and NNRTI-containing
regimens. The common mutations detected were either K65R or M184V
alone in five patients, while one patient had both K65R and M184V (Wirden
et al, 2004).

Early virologic failure in ddI, TFV and EFV; 3TC, TFV and Nevirapine
Early virologic failure was also observed in a clinical study involving
the nucleoside combination of ddI, TFV, and EFV. More recently the
combination of 3TC, TFV and nevirapine (NVP) also produced a large
number of patients with incomplete virologic suppression (Maitland et al,
2005; Podzamczer et al, 2005; Rey D, 2007).
These two combinations were composed of two NRTIs along with
either EFV or NVP. When 15 treatment-naïve HIV patients were treated with
the combination of TFV, ddI and EFV, 8 patients experienced virologic failure
as early as 2-3 weeks after initiation of the treatment. The K65R mutation
selected by ddI and TFV was detected in two out of eight patients, while the
L74V mutation selected by ddI was detected in four of eight non-responders
(Podzamczer et al, 2005). In addition, the combination of ddI, TFV and EFV
was also associated with lymphopenia (a reduction in lymphocytes, or T cells)
in a significant number of patients after achieving viral suppression. This
paradoxical decline in CD4+ T-cells was incongruous to a normal rise in CD4
count following successful suppression of viral replication as noted by a viral
6
load below detection (<400 copies/mL or <50 copies/mL) (Barrios et al, 2005;
Negredo et al, 2004).
More recently, treatment-naïve HIV patients who were treated with the
combination of 3TC, TFV and NVP revealed a large number of early virologic
failures (Rey D, 2007). In the DAUFIN study, 9 of out 36 treatment-naïve
patients receiving the once-daily combination of TFV, 3TC and NVP either
never achieved viral suppression or had their plasma viral load rebound after
initial suppression. Viral resistance testing revealed the presence of K65R
and M184V/I mutations which was present in 6 and 3 non-responders,
respectively. There were also more than one viral mutation associated with
reduced susceptibility towards NVP. Noticeably, these patients had higher
baseline plasma viral load and lower CD4 counts prior to treatment as
compared to the patients achieving viral suppression in the same group. In
the study, plasma trough concentration of NVP was found to be adequate for
viral suppression.

Nucleoside pharmacology
NRTIs are nucleoside analogs (NA) lacking a 3’ hydroxyl group as
compared to endogenous nucleosides involved in the DNA replication. NAs
are sequentially phosphorylated in host cells and form triphosphate of
nucleoside analogs (ddNTP), which compete with endogenous nucleotide
(dNTP) for binding to viral reverse transcriptase (RT) for incorporation into
the viral DNA chain, resulting in the termination of its elongation. It is critical
7
to achieve sufficient drug levels, especially at the trough concentration at the
targeted sites to achieve the intended antiviral activities. Insufficient
intracellular drug levels may allow the HIV virus to select for viral mutations
under the selective pressure. Because HIV viral replication occurs in CD4+
cells, intracellular concentration of ddNTPs would be a critical factor for the
success of treatment.
The pharmacokinetics of these NRTIs in plasma and their active
moieties in PBMCs has been investigated to verify whether these drugs were
adequate for once daily treatment. Most of the NRTIs have short half-lives in
plasma, where ABC, ddI and 3TC have half-lives of 1.5, 1.5 and 6 hours in
plasma, respectively (Table 1.1). TFV is a nucleotide analog that differs
markedly from other NRTIs where the plasma half life is 17 hours. The half-
lives of active moieties corresponding to these NRTIs in PBMCs were
significantly different as compared to their plasma half-lives. The active
8
moieties for these four NRTIs, carbovir-triphosphate (CBV-TP),
dideoxyadenosine triphosphate (ddATP), 3TC-TP and TFV-DP had half-lives
of 20, 24, 15 and 150 hours in PBMCs, respectively. The long intracellular
half-lives for these NRTIs in PBMC allow them to be dosed once-daily.

Pharmacological mechanisms for early virologic failure
In light of the early virologic failures in HIV patients receiving the triple
nucleoside combination of 3TC, TDF and either ABC or ddI, a number of
mechanisms have been postulated. There were mainly two viral mutations
found in HIV patients who experienced virologic failure, therefore the triple
nucleoside combination may have a low genetic barrier for viral resistance.
Pharmacodynamic interaction between these nucleoside analogs could also
lead to antagonism. Lastly, intracellular drug-drug interaction between NRTIs
could be another plausible mechanism leading to early virologic failure.

Low genetic barrier
There were only two common viral RT mutations (M184V/I and K65R)
observed in patients experienced early virologic failure after receiving one of
the two triple nucleoside combinations, which led to the perception of a low
genetic barrier with treatment. In the Tonus study, viral resistance dynamics
were studied for the treatment-naïve HIV patients who experienced early
virologic failure when treated with the combination of ABC, TFV and 3TC.
M184V viral mutants at the viral RT were observed as early as week 4 of the
9
study, and the K65R mutation appeared later (Delaunay et al, 2005).
Although ABC and TDF maintained antiviral activity against M184V mutants,
the low genetic barrier hypothesis alone did not address the rapid rate
leading to early virologic failure of the triple nucleoside combination of ABC,
TFV and 3TC. There could be some other pharmacological mechanisms
involved in the early virologic failure in HIV patients.

Pharmacodynamic interaction
It is also noteworthy that all these agents (ABC, TFV, 3TC and ddI)
have a favorable intracellular pharmacokinetic profile which warrants once
daily dosing (Table 1.1) More importantly these agents have been previously
evaluated in combination therapy as dual NRTIs in combination with either a
PI or NNRTI, and have achieved exceptional viral suppression with an
excellent tolerability profile, except for the TFV and ABC combination.
Thereby, the combination of TDF and ABC was investigated to verify if
the combination could cause any pharmacodynamic antagonism against
each other

(Havlir et al, 2005; Moyle et al, 2005). Two independent studies
demonstrated that the combination of ABC and TFV was virologically
synergistic in both wild type and viruses with K65R and M184V mutations
(Lanier et al., 2005, Ray et al., 2005).

10
Intracellular drug-drug interaction
The intracellular drug-drug interaction between nucleoside analogs
has been scrutinized in light of the clinical failure of the triple nucleoside
combination of TFV, 3TC and ABC or ddI. Plasma drug-drug interaction
between these NRTIs was investigated to verify whether this would lead to
early virologic failure. Out of all the optional combinations, there was no
plasma drug-drug interaction between nucleoside analogs, except ddI and
TFV. It was found that patients treated with combination of ddI and TFV had
40-300% elevation in ddI plasma levels as compared to ddI treatment alone.
Consequently, ddI dose reduction by 40% was proposed to patients receiving
ddI and TFV combination (Kearney et al, 2004).

Intracellular TFV and ddI interaction
TFV and ddI interaction in the clinical setting
Pruvost et al investigated the intracellular drug-drug interaction
between TFV and ddI in HIV patients using a sensitive mass spectrometry
assay. In the study, intracellular concentrations of ddATP and TFV-DP were
determined in the PBMC of HIV patients receiving drug treatment containing
TFV or ddI alone and in combination (Pruvost et al, 2005). Their results
indicated that median ddATP formation was 30% lower in HIV patients
receiving the combination of TFV and ddI after ddI dose adjustment as
compared to patients receiving ddI alone (see Table 1.2). In addition, there
was a strong correlation between the intracellular concentrations of ddATP
11
and TFV-DP in HIV patients receiving the combination of ddI and TFV, which
suggested a competitive inhibition between the two agents, as the two NRTIs
share two critical anabolic enzymes in their activation pathway (Figure 1.1).

Intracellular TFV and ddI interaction in in vitro study
Rodman et al investigated intracellular drug-drug interaction between
TFV and ddI in both quiescent and activated PBMCs (Robbins et al, 2003). In
this study, cells were treated with ddI at physiological achievable
concentrations from 2 to 20 μM alone and in combination with 5 μM of TFV.
Radiolabeled drugs were utilized in this study and phosphorylated anabolites
were separated by HPLC and detected by scintillation counter. The study
showed a 4-10% reduction in ddATP in cells treated with ddI and TFV
combination as compared to cells treated with ddI alone. In addition, results
demonstrated that the activations state of PBMC cells did not have significant
impact on intracellular interaction between ddI and TFV. In light of the
disagreements between this in vitro study to the aforementioned clinical
study, additional experiments with dose escalation of TFV and ddI or with a
relatively longer treatment time may be needed in order to determine the
intracellular interaction between TFV and ddI.

Interaction of TFV and ddI at purine nucleoside phosphorylase
The intracellular drug-drug interaction between ddI and TFV may also
involve a catabolic enzyme for purine nucleosides. In a cell culture study,
12
13
Ray et al found that anabolites of TFV, TFV-MP and TFV-DP, were potent
inhibitors of purine nucleoside phosphorylase (PNP) with IC
50
in the low μM
range. PNP is a catabolic enzyme that mediates the breakdown of purine
nucleosides to form the respective base and the ribose moiety. In the
presence of PNP, ddI is catabolized into inosine and dideoxyribose sugar.
TFV phosphorylated metabolite mediated inhibition of PNP can lead to an IC
ddI accumulation (Ray et al, 2004). Unfortunately, Ray et al did not evaluate
the intracellular concentrations of TFV-DP or ddATP to further investigate
whether the interaction between ddI and TFV was only restricted to ddI
metabolism. In addition, results also demonstrated that PNP had much
higher catalytical efficiency towards metabolizing endogenous nucleosides
when compared to ddI. This study illustrated the importance of PNP inhibition
with regard to the impact on endogenous nucleotide pools (dNTP). It is
conceivable that inhibition of PNP by TFV anabolites may be an important
factor that significantly impact levels of dNTP, where elevated level can out
compete with ddATP for viral RNA incorporation.

Involvement of endogenous nucleotide pools
Inhibition of PNP by phosphorylated anabolites of TFV could have a huge
impact on both the efficacy and toxicity of the treatments for HIV patients.
PNP breaks down purines like adenosine, guanosine, inosine and their
respective analogs into their corresponding nucleoside base and ribose
moiety to prevent excessive accumulation of purines (see Figure 1.2).
14
Inhibition of PNP may lead to an accumulation of endogenous nucleoside
bases and eventually lead to elevation of endogenous dNTP pools, which
compete with ddNTPs for the binding and incorporation into the active site of
HIV RT for proviral DNA elongation. Elevation of dNTPs may lead to a
reduction in the ddNTP/dNTP ratio, thereby reducing the antiviral activity of
the corresponding NRTI and may result in an incomplete viral suppression
and allow drug resistant viruses to emerge.
Inhibition of PNP mediated by TFV and its anabolites may increase
drug-related toxicity. A significant number of HIV patients receiving the
combination of ddI, TFV and EFV developed lymphopenia (reduction in
lymphocytes) after achieving viral suppression, approximately six month after
initiation of antiviral therapy (Barrios et al, 2005; Negredo et al, 2004). This
15
paradoxical decline in CD4+ T-cells was incongruous to a normal rise in CD4
count following successful suppression of viral replication as noted by a viral
load below detection (<50 copies/mL).
Children with PNP deficiency syndrome, a congenital disorder, often
suffer from severe combined immunodeficiency disease (SCID). In the
afflicted children, elevated levels of deoxyguanosine triphosphate (dGTP)
were observed in PBMCs, where an accumulation of dGTP was thought to
promote cellular apoptosis through the activation of caspase enzymes
(Bzowska et al, 2000). Thus, TFV anabolite-mediated inhibition of PNP may
elevate dGTP levels and lead to apoptosis of T-cells.
Ray et al investigated the role of endogenous nucleotide pools in
CEMss cells using 10 μM of TFV and 30 μM of ABC. No change in dATP or
dGTP was noted. Ray et al claimed that the intracellular concentration of
TFV and its anabolites were not high enough to achieve complete inhibition
of PNP; hence TFV treatment did not have significant impact on purine
endogenous nucleotide pools.
However, in real life HIV patients were treated with TDF, an esterified
prodrug of TFV. TDF is more lipophilic and can be readily absorbed in
intestine and converted into TFV by carboxyl esterases found in plasma. As
TDF can more readily diffuse into PBMCs by passive diffusion than TFV, it is
conceivable that some of the PBMCs may absorb TDF and have higher
intracellular concentrations of TFV and its anabolites when compared to
treatment with TFV (Mallant et al, 2005). Therefore, in some of the cells,
16
there might be a high concentration of TFV and its anabolites, which could
sufficiently result in PNP inhibition and eventually lead to lymphopenia. In
addition, the elevated levels of dATP and dGTP would likely be consumed by
the rapid viral replication at the early stage of treatment. However, with the
HIV viral replication being inhibited by the combination treatment, there could
be accumulation of dATP or dGTP within host cells and lead to lymphopenia
via activation of apoptotic pathways.

Role of cellular adaptive mechanisms in cellular resistance towards
nucleoside analogs
Pharmacological agents have both desirable and undesirable
properties. To reduce the impact of undesirable pharmacological properties,
humans have developed “biosensors” that alter the transport and metabolism
of endobiotics and xenobiotics leading to a reduced exposure to these toxic
agents. These cellular adaptive changes reduce cellular exposure to the
cytotoxic agents and increase the host cells chances of survival. This cellular
adaptive mechanism is exemplified in the cellular resistance against NAs in
the treatment of malignant tumors (Mansson et al, 2003), NAs were used
extensively in the treatment of tumors. NAs are activated in cells to form the
active moiety ddNTPs, which are then incorporated into the tumor DNA and
terminate viral DNA replication. Long-term treatment of malignant tumors
with NAs prompts cellular adaptive changes in the way of expression of
enzymes and efflux transporters. By adapting, cells reduce their exposure to
17
these cytotoxic agents and the formation of ddNTPs, which may eventually
develop into drug resistance.

Expression of metabolic enzymes
A notable cellular adaptive mechanism is the down regulation of RNA
and protein expression of critical anabolic enzymes required for the formation
of ddNTP. Increased expression of catabolic enzymes is another strategy
that leads to reduced formation of ddNTP in chemoresistant tumors.
Catabolic enzymes are responsible for the degradation of nucleosides and
ultimately reduce the precursor levels that form the active ddNTP (Bergman,
2005; Bergman, 2000). This suggests that the cells are able to use
alternative pathways for nucleic acid biosynthesis while preventing the
processing of the pharmaceutical agent. Similarly, normal cells may use
similar mechanisms to evade potentially toxic xenobiotic exposures.

Expression of efflux transporters
Another notable mechanism for cellular adaptive changes involves
efflux transporters. Mammalian cells express many efflux transporters to
protect cellular organisms from the assault of xenobiotics. Among the
prominent efflux transporters for NRTIs is the multidrug resistant protein 4
(MRP4). Expression of MRP4 in CEMss cells was found to be inducible in a
dose dependent fashion after treatment with adefovir (ADV), an analog of
TFV which differs by a methyl group. In one study, ADV resistant CEM-r1
18
cells significantly over-expressed MRP4. The CEM-r1 cells demonstrated a
250-fold increase in cellular viability against ADV treatment; in addition,
antiviral activity of ADV in CEM-r1 cells had more than a 150-fold reduction
compared to that in CEMss. In addition, cellular resistance toward ADV can
render resistance against a broad spectrum of nucleosides, which is
exemplified by 6- and 5-fold increases in resistance towards AZT and 3TC,
respectively (Schuetz et al., 1999). MRP4 can effectively efflux NRTI
monophosphates (e.g. AZT-MP) and ANPs (e.g. ADV and TFV) out of cells,
thereby reducing the amount of precursors leading to the formation of ddNTP,
and conferring cross-resistance against multiple NRTIs. As higher levels of
ddNTP may correlate to higher viral suppression, conversely, low levels of
ddNTP may reduce drug exposure and result in incomplete viral suppression
and clinical failure (Fletcher et al., 2002, Fletcher et al., 2000). Over-
expression of MRP4 may reduce the intracellular level of ddNTP and
increase the risk of incomplete viral suppression, allowing drug-resistant viral
mutations to emerge.
In addition, ddNTP may also cause some cytotoxicity by inhibition of
mitochondrial DNA in host cells (Lewis et al., 2003). Thereby, cellular
adaptive changes in host cells may reduce their exposure to the cytotoxic
nucleoside analogs. In CEM-r1 cells, there was a two-fold decrease in the
activity of adenylate kinase 2 (AK2), and over-expression of the efflux
transporter MRP4. AK2 is a critical anabolic enzyme responsible for the
phosphorylation of ADV to form ADV-MP (Robbins et al., 1995). The
19
combination of these cellular adaptive changes resulted in a 50% and 97%
reduction in ADV and ADV-DP, respectively.
As previously mentioned, up-regulation of catabolic enzymes is
another mechanism utilized by host cells to reduce the exposure to these
cytotoxic agents. In cells resistant to anti-tumor agent, 9-β-D-
arabinofuranosyl-2-fluoroadenine (fludarabine, Fara-A), a 43% increase in
activity of the small subunit R2 of ribonucleotide reductase (RR) was
detected when compared to parental CEMss cells. This increase in R2
activity will increase the formation of endogenous dNTPs, thereby resulting in
a decrease in the ratio of ddNTP/dNTP and a reduction in the efficacy of the
nucleoside analog (Mansson et al., 1999).
20
To summarize, cellular activation of nucleoside analogs depends on a
few critical cellular factors, which include the protein level of anabolic
(activation) enzymes, catabolic (deactivation) enzymes, and efflux
transporters. When treated with cytotoxic agents, such as nucleoside
analogs, the expression of cellular factors could be altered to reduce the host
cells exposure towards these cytotoxic agents. These cellular adaptive
changes can have an impact on the activation of NRTIs, and could cause
intracellular drug-drug interactions in patients receiving triple nucleoside
combinations of TFV, 3TC and ABC or ddI. Moreover, TFV-mediated
inhibition of PNP may also have a significant impact on endogenous
nucleotide pools and compromise the efficacy of NRTIs leading to early
virologic failure.
In this study, the intracellular drug-drug interaction among nucleoside
analogs TFV, 3TC and ABC or ddI will be investigated. Cellular adaptive
changes with in response to the nucleoside treatment will be evaluated, with
emphasis on cellular resistance against nucleoside analogs after long-term
treatment.

21
Chapter 1 References:

Barreiro, P. & Soriano, V. (2006) Suboptimal CD4 gains in HIV-infected
patients receiving didanosine plus tenofovir, J Antimicrob Chemother,
57(5), pp. 806-809.
Barrios, A., Rendon, A., Negredo, E., Barreiro, P., Garcia-Benayas, T.,
Labarga, P., Santos, J., Domingo, P., Sanchez-Conde, M., Maida, I.,
Martin-Carbonero, L., Nunez, M., Blanco, F., Clotet, B., Sambeat,
M.A., Gil, P., Gonzalez-Lahoz, J., Cooper, D. & Soriano, V. (2005)
Paradoxical CD4+ T-cell decline in HIV-infected patients with
complete virus suppression taking tenofovir and didanosine, AIDS,
19(6), pp. 569-575.
Bergman, A.M., Eijk, P.P., Ruiz van Haperen, V.W., Smid, K., Veerman, G.,
Hubeek, I., van den Ijssel, P., Ylstra, B. & Peters, G.J. (2005) In vivo
induction of resistance to gemcitabine results in increased expression
of ribonucleotide reductase subunit M1 as the major determinant,
Cancer Res, 65(20), pp. 9510-9516.
Bergman, A.M., Giaccone, G., van Moorsel, C.J., Mauritz, R., Noordhuis, P.,
Pinedo, H.M. & Peters, G.J. (2000) Cross-resistance in the 2',2'-
difluorodeoxycytidine (gemcitabine)-resistant human ovarian cancer
cell line AG6000 to standard and investigational drugs, Eur J Cancer,
36(15), pp. 1974-1983.
Bzowska, A., Kulikowska, E. & Shugar, D. (2000) Purine nucleoside
phosphorylases: properties, functions, and clinical aspects, Pharmacol
Ther, 88(3), pp. 349-425.
Carr, A., Samaras, K., Burton, S., Law, M., Freund, J., Chisholm, D.J. &
Cooper, D.A. (1998) A syndrome of peripheral lipodystrophy,
hyperlipidaemia and insulin resistance in patients receiving HIV
protease inhibitors, AIDS, 12(7), pp. F51-58.

22
Cuzin, L., Pugliese, P., Bugnon, F., Delpierre, C., Cua, E., Billaud, E.,
Massip, P., Raffi, F. & Dellamonica, P. (2005) Triple nucleoside
reverse transcriptase inhibitor- vs. nonnucleoside reverse
transcriptase inhibitor-containing regimens as first-line therapy:
efficacy and durability in a prospective cohort of French HIV-infected
patients, HIV Med, 6(6), pp. 388-395.
Delaunay, C., Brun-Vezinet, F., Landman, R., Collin, G., Peytavin, G.,
Trylesinski, A., Flandre, P., Miller, M. & Descamps, D. (2005)
Comparative selection of the K65R and M184V/I mutations in human
immunodeficiency virus type 1-infected patients enrolled in a trial of
first-line triple-nucleoside analog therapy (Tonus IMEA 021), J Virol,
79(15), pp. 9572-9578.
Farthing C, K.H., Yeh V. (2003) Early virologic failure in a pilot study
evaluating the efficacy of once daily abacavir (ABC), lamivudine (3TC)
and tenofovir DF (TDF) in treatment naive HIV-infected patients.
Presented at: Second International AIDS Society Conference on HIV
Pathogenesis and Treatment; x; Paris. (Paris,
FDA (2005) Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected
Adults and Adolescents.
Finzi, D., Hermankova, M., Pierson, T., Carruth, L.M., Buck, C., Chaisson,
R.E., Quinn, T.C., Chadwick, K., Margolick, J., Brookmeyer, R.,
Gallant, J., Markowitz, M., Ho, D.D., Richman, D.D. & Siliciano, R.F.
(1997) Identification of a reservoir for HIV-1 in patients on highly active
antiretroviral therapy, Science, 278(5341), pp. 1295-1300.
Fletcher, C.V., Anderson, P.L., Kakuda, T.N., Schacker, T.W., Henry, K.,
Gross, C.R. & Brundage, R.C. (2002) Concentration-controlled
compared with conventional antiretroviral therapy for HIV infection,
AIDS, 16(4), pp. 551-560.
Fletcher, C.V., Kawle, S.P., Kakuda, T.N., Anderson, P.L., Weller, D.,
Bushman, L.R., Brundage, R.C. & Remmel, R.P. (2000) Zidovudine
triphosphate and lamivudine triphosphate concentration-response
relationships in HIV-infected persons, AIDS, 14(14), pp. 2137-2144.
23
Gallant, J.E., Rodriguez, A.E., Weinberg, W.G., Young, B., Berger, D.S., Lim,
M.L., Liao, Q., Ross, L., Johnson, J. & Shaefer, M.S. (2005) Early
virologic nonresponse to tenofovir, abacavir, and lamivudine in HIV-
infected antiretroviral-naive subjects, J Infect Dis, 192(11), pp. 1921-
1930.
Galmarini, C.M., Clarke, M.L., Jordheim, L., Santos, C.L., Cros, E., Mackey,
J.R. & Dumontet, C. (2004) Resistance to gemcitabine in a human
follicular lymphoma cell line is due to partial deletion of the
deoxycytidine kinase gene, BMC Pharmacol, 4, pp. 8.
Gulick, R.M., Ribaudo, H.J., Shikuma, C.M., Lustgarten, S., Squires, K.E.,
Meyer, W.A., 3rd, Acosta, E.P., Schackman, B.R., Pilcher, C.D.,
Murphy, R.L., Maher, W.E., Witt, M.D., Reichman, R.C., Snyder, S.,
Klingman, K.L. & Kuritzkes, D.R. (2004) Triple-nucleoside regimens
versus efavirenz-containing regimens for the initial treatment of HIV-1
infection, N Engl J Med, 350(18), pp. 1850-1861.
Handforth, J. & Sharland, M. (2004) Triple nucleoside reverse transcriptase
inhibitor therapy in children, Paediatr Drugs, 6(3), pp. 147-159.
Havlir, D.V., Tierney, C., Friedland, G.H., Pollard, R.B., Smeaton, L.,
Sommadossi, J.P., Fox, L., Kessler, H., Fife, K.H. & Richman, D.D.
(2000) In vivo antagonism with zidovudine plus stavudine combination
therapy, J Infect Dis, 182(1), pp. 321-325.
Hoogewerf, M., Regez, R.M., Schouten, W.E., Weigel, H.M., Frissen, P.H. &
Brinkman, K. (2003) Change to abacavir-lamivudine-tenofovir
combination treatment in patients with HIV-1 who had complete
virological suppression, Lancet, 362(9400), pp. 1979-1980.
Jemsek J, H.P., Harper E. (2004) Poor virologic response and early
emergence of resistance in treatment naive, HIV-infected patients
receiving a once daily triple nucleoside regimen of didanosine,
lamivudine, and tenofovir DF11th Conference on Retroviruses and
Opportunistic Infections; (San Francisco,
Kearney, B.P., Flaherty, J.F. & Shah, J. (2004) Tenofovir disoproxil fumarate:
clinical pharmacology and pharmacokinetics, Clin Pharmacokinet,
43(9), pp. 595-612.
24
Khanlou, H., Yeh, V., Guyer, B. & Farthing, C. (2005) Early virologic failure in
a pilot study evaluating the efficacy of therapy containing once-daily
abacavir, lamivudine, and tenofovir DF in treatment-naive HIV-infected
patients, AIDS Patient Care STDS, 19(3), pp. 135-140.
Landman R, P.G., Descamps D, et al. (2004) Low genetic barrier to
resistance is a possible cause of early virologic failures in once-daily
regimen of abacavir, lamivudine, and tenofovir: the Tonus study11th
Conference on Retroviruses and Opportunistic Infections (San
Francisco,
Lewis, W., Day, B.J. & Copeland, W.C. (2003) Mitochondrial toxicity of NRTI
antiviral drugs: an integrated cellular perspective, Nat Rev Drug
Discov, 2(10), pp. 812-822.
Maitland, D., Moyle, G., Hand, J., Mandalia, S., Boffito, M., Nelson, M. &
Gazzard, B. (2005) Early virologic failure in HIV-1 infected subjects on
didanosine/tenofovir/efavirenz: 12-week results from a randomized
trial, AIDS, 19(11), pp. 1183-1188.
Mallants, R., Van Oosterwyck, K., Van Vaeck, L., Mols, R., De Clercq, E. &
Augustijns, P. (2005) Multidrug resistance-associated protein 2
(MRP2) affects hepatobiliary elimination but not the intestinal
disposition of tenofovir disoproxil fumarate and its metabolites,
Xenobiotica, 35(10-11), pp. 1055-1066.
Mansson, E., Flordal, E., Liliemark, J., Spasokoukotskaja, T., Elford, H.,
Lagercrantz, S., Eriksson, S. & Albertioni, F. (2003) Down-regulation
of deoxycytidine kinase in human leukemic cell lines resistant to
cladribine and clofarabine and increased ribonucleotide reductase
activity contributes to fludarabine resistance, Biochem Pharmacol,
65(2), pp. 237-247.
Mansson, E., Spasokoukotskaja, T., Sallstrom, J., Eriksson, S. & Albertioni,
F. (1999) Molecular and biochemical mechanisms of fludarabine and
cladribine resistance in a human promyelocytic cell line, Cancer Res,
59(23), pp. 5956-5963.

25
Negredo, E., Molto, J., Burger, D., Viciana, P., Ribera, E., Paredes, R., Juan,
M., Ruiz, L., Puig, J., Pruvost, A., Grassi, J., Masmitja, E. & Clotet, B.
(2004) Unexpected CD4 cell count decline in patients receiving
didanosine and tenofovir-based regimens despite undetectable viral
load, AIDS, 18(3), pp. 459-463.
Palella, F.J., Jr., Delaney, K.M., Moorman, A.C., Loveless, M.O., Fuhrer, J.,
Satten, G.A., Aschman, D.J. & Holmberg, S.D. (1998) Declining
morbidity and mortality among patients with advanced human
immunodeficiency virus infection. HIV Outpatient Study Investigators,
N Engl J Med, 338(13), pp. 853-860.
Podzamczer, D., Ferrer, E., Gatell, J.M., Niubo, J., Dalmau, D., Leon, A.,
Knobel, H., Polo, C., Iniguez, D. & Ruiz, I. (2005) Early virological
failure with a combination of tenofovir, didanosine and efavirenz,
Antivir Ther, 10(1), pp. 171-177.
Pruvost, A., Negredo, E., Benech, H., Theodoro, F., Puig, J., Grau, E.,
Garcia, E., Molto, J., Grassi, J. & Clotet, B. (2005) Measurement of
intracellular didanosine and tenofovir phosphorylated metabolites and
possible interaction of the two drugs in human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 49(5), pp.
1907-1914.
Ray, A.S., Olson, L. & Fridland, A. (2004) Role of purine nucleoside
phosphorylase in interactions between 2',3'-dideoxyinosine and
allopurinol, ganciclovir, or tenofovir, Antimicrob Agents Chemother,
48(4), pp. 1089-1095.
Rey D, S.M., Hoisey G, Meyer P, Chavanet P, Allavena C, Diemer M, May T,
Hoen B and Lang J. (2007) Early Virologic Non-response to Once
Daily Combination of Lamivudine, Tenofovir and Nevirapine in
Antiretroviral Naive HIV-infected Patients: Preliminary Results of the
Daufin Study14th Conference of Retrovirus and Opportunistic
Infections (Los Angeles, CA).
Robbins, B.L., Connelly, M.C., Marshall, D.R., Srinivas, R.V. & Fridland, A.
(1995) A human T lymphoid cell variant resistant to the acyclic
nucleoside phosphonate 9-(2-phosphonylmethoxyethyl)adenine
shows a unique combination of a phosphorylation defect and
increased efflux of the agent, Mol Pharmacol, 47(2), pp. 391-397.
26
Robbins, B.L., Wilcox, C.K., Fridland, A. & Rodman, J.H. (2003) Metabolism
of tenofovir and didanosine in quiescent or stimulated human
peripheral blood mononuclear cells, Pharmacotherapy, 23(6), pp. 695-
701.
Schuetz, J.D., Connelly, M.C., Sun, D., Paibir, S.G., Flynn, P.M., Srinivas,
R.V., Kumar, A. & Fridland, A. (1999) MRP4: A previously unidentified
factor in resistance to nucleoside-based antiviral drugs, Nat Med, 5(9),
pp. 1048-1051.
Siliciano, J.D., Kajdas, J., Finzi, D., Quinn, T.C., Chadwick, K., Margolick,
J.B., Kovacs, C., Gange, S.J. & Siliciano, R.F. (2003) Long-term
follow-up studies confirm the stability of the latent reservoir for HIV-1
in resting CD4+ T cells, Nat Med, 9(6), pp. 727-728.
Wirden, M., Marcelin, A.G., Tubiana, R., Valantin, M.A., Ghosn, J., Duvivier,
C., Dominguez, S., Paris, L., Agher, R., Peytavin, G., Katlama, C. &
Calvez, V. (2004) Virologic outcome after switching from a nucleoside
reverse transcriptase inhibitor to tenofovir in patients with
undetectable HIV-1 RNA plasma level, J Acquir Immune Defic Syndr,
36(3), pp. 876-878.

27
Chapter 2: Determination of Triphosphate of Nucleoside Analogs and
Endogenous Nucleotide Pools by Liquid Chromatography Coupled to
Tandem Mass Spectrometry

Chapter 2 Abstract
A sensitive multiplex liquid chromatography tandem mass
spectrometry (LC/MS/MS) assay was developed and validated for the
determination of intracellular concentrations of dideoxynucleoside
triphosphates (ddNTPs) corresponding to nucleoside analogs and
endogenous nucleotide pools (dNTP). The cell extracts were fractionated,
dephosphorylated and quantified using LC/MS/MS via corresponding
nucleoside/nucleotide to ddNTP and dNTP.
The method was validated in terms of linearity, specificity, precision,
accuracy, and extraction recovery in CEMss cells and demonstrated to be
effective using U937, CEM and PBMC cells. The assay was demonstrated
to be linear in the range of 0.1 to 20 pmole/10
6
cells for 3TC-TP and TFV-DP,
0.0105 to 2.1 pmole/10
6
for CBV-TP, and 0.005 to 1.0 pmole/10
6
for ddATP
when using 1×10
7
cells. The assay has a lower limit of quantification of 0.1
pmole/10
6
cells for both 3TC-TP and TFV-DP, 0.0105 pmole/10
6
cells for
CBV-TP, and 0.005 pmole/10
6
for ddATP. The linear range for endogenous
nucleotide pools was 1.02 to 102 pmole/10
6
, 0.986 to 98.6 pmole/10
6
and
1.07 to 107 pmole/10
6
cells for dATP, dGTP, and dCTP, respectively. It has
variation within and between batches of less than 25%.
28
This assay has been successfully applied in the analysis of cell
extracts derived from cell culture study and PBMC cells from a clinical study
for the intracellular drug-drug interaction.

Chapter 2 Introduction

Intracellular drug-drug interaction between nucleoside analogs has
been suggested as a leading cause for early virologic failure in HIV patients
treated with the once-daily combination of nucleoside reverse transcriptase
inhibitors (NRTIs) with either a non-nucleoside reverse transcriptase inhibitor
(NNRTI) or a protease inhibitor (PI) (Gallant et al., 2005, Jemsek J, 2004). To
prove this hypothesis, it is essential to quantify the intracellular
concentrations of triphosphate moieties of the corresponding ddNTP, and
determine whether the ddNTP level for each NRTI will be altered by the
presence of other nucleoside analogs.
Determination of triphosphate ddNTP was critical in understanding the
clinical pharmacology of this class of drugs. Nucleoside analogs (NA) are
prodrugs that are sequentially phosphorylated into monophosphate (ddNMP),
diphosphate (ddNDP) and triphosphate (ddNTP) moieties by intracellular
enzymes. The ddNTP competes with endogenous nucleotide pools for the
binding to viral reverse transcriptase (RT). The incorporation of ddNTP into
proviral DNA could inhibit viral DNA replication. Thereby, the ratio of
ddNTP/dNTP ratio is a critical biomarker for both efficacy and toxicity of NA
29
(Anderson et al., 2004, Groschel et al., 1997). Previous clinical studies
suggested that the intracellular ddNTP levels had better correlation to
treatment outcomes compared to plasma pharmacokinetics (Fletcher et al.,
2002, Fletcher et al., 2000).
Intracellular drug-drug interaction between nucleoside analogs may
occur since NA phosphorylation enzymes may have overlapping activity
(Anderson et al., 2004). The competitive inhibition was detected when either
two cytosine or two thymidine analogs were combined. This type of
interaction also led to poor clinical outcomes (Havlir et al., 2000, Kewn et al.,
2000).
There are two major challenges with ddNTP analysis: these
compounds are extremely hydrophilic and not retained on C18 HPLC
columns; in addition, these compounds, in their natural forms, carry three
negative charges and mass spectrometers are not as sensitive when
detecting negative ion carrying molecules. To circumvent these challenges,
ion pairing reagents are commonly utilized in the nucleotide separation
(Becher et al., 2002, Pruvost et al., 2005). In an elegantly designed method,
Ray et al was able to separate tenofovir (TFV) and its phosphorylated
anabolites, monophosphate and diphosphate of TFV (TFV-MP and TFV-DP)
using ion-pairing separation on a microbore C18 column (Ray et al., 2005,
Vela et al., 2007). An API 4000 mass spectrometer equipped with an
orthogonal turbo ion spray was used to quantify the phosphorylated
nucleotides. Using an orthogonal turbo spray source, the ion pairing agents
30
and buffers are not directly sprayed into the mass spectrometers and thereby
limit their ion suppression capacity. However, direct analysis of cell lysates
without extraction made it technically more challenging.
An indirect approach was also applied extensively before the
emergence of LC/MS/MS technology. In the indirect approach, cell lysates
were fractionated using a strong anion exchange (SAX) column, fractions of
ddNTP were collected, and radiolabelling was employed to determine the
presence of ddNTP in biological samples. Alternatively, after fractionation
with a salt gradient, the phosphate groups of ddNTP could be removed by
acid or alkaline phosphatase. After enzyme hydrolysis, salts in the samples
were removed by solid phase extraction (SPE) and the final nucleosides
corresponding to ddNTP were then quantified by measuring the amount of
radioactivity in the sample (Robbins et al., 1995, Robbins et al., 2003).
Though labor intensive, this approach was very effective in producing clean
cellular samples, and high sensitivity can be achieved using radiolabeled
drug. However, large cellular extracts are often required to determine
nucleotide levels.
Another method to quantify intracellular nucleotide levels is the primer
extension assay (Harris et al., 2002, Kewn et al., 2002). This approach for
the determination of ddNTP in cells is based on the competitive inhibition of
ddNTP with natural occurring endogenous nucleotide, after RT was added
with a carefully designed template. This approach has been used extensively
in the determination of endogenous nucleotide pools in cell culture. Though
31
very sensitive, this approach also requires radiolabeled nucleosides and
suffers from non-specific binding of nucleotide pools to RT. In addition, this
method also lacks the capability for simultaneous detection of multiple
compounds.
The present method utilizes a hybrid method designed to employ the
strengths of clean cellular preparation, and pairs it with high sensitivity and
specificity offered by LC/MS/MS detection. The extensive sample extraction
preceding mass spectrometry analysis provides the potential to analyze a
number of substrates in the same sample, and thus permit the quantification
of both ddNTP and dNTP simultaneously. By using this approach, a highly
sensitive method has been developed for the simultaneous determination of
3TC-TP, CBV-TP, TFV-DP, and ddATP in a small volume of cellular lysates.
In addition, the endogenous nucleotide pools can also be determined using
the same method; therefore the ddNTP/dNTP ratio can be determined in a
single analytical run. This approach was validated for specificity, selectivity,
linearity, stability and extraction recovery.

Chapter 2 Experimental methods
Chemicals
The nucleotide triphosphates, carbovir-triphosphate (CBV-TP) and
lamivudine-triphosphate (3TC-TP) were a generous gift from
GlaxoSmithKline (Dr. Katrina Oie), while ddATP was purchased from GE
Amersham (Piscataway, USA). TFV, 3TC-MP, TFV-DP and its internal
32
standard ADV-DP were purchased from Movareck, CA. Other triphosphate
internal standards such as dideoxycytidine triphosphate (ddCTP),
dideoxyguanosine triphosphate (ddGTP) and 2-chloroadenosine
triphosphate (2Cl-ATP) were purchased from Sigma Biologicals, St. Louis,
MO. Deoxyguanosine monophosphate (dGMP), deoxyguanosine
diphosphate (dGDP), deoxyadenosine triphosphate (dATP), deoxyguanosine
triphosphate (dGTP), deoxycytidine triphosphate (dCTP) and deoxythymidine
triphosphate (dTTP) were purchased from Sigma.

Cells
CEMss and U937
WT
were maintained in RPMI medium, which was
supplemented with 10% FBS, 1X pyruvate, 100 μg/mL of penicillin and
streptomycin. Cells were passaged every 5 days and experiments generally
used cells from passage 5 to 25. Cells were harvested and lysed with ice
cold methanol and these cell lysates were used as blanks for the calibration
standards. PBMC cells were obtained from HIV patients from University
Hospital at the University of California San Diego. (UCSD, CA).

Instrumentation
An API3000 (Applied Biosystems, Foster City, CA) LC/MS/MS
instrument was used throughout the study. The system is composed of an
Agilent 1100 (Agilent, San Jose, CA) high pressure liquid chromatography
(HPLC) system coupled to an API 3000 triple quadrupole tandem mass
33
spectrometer. The operating software is Analyst version 1.4.2. The weak
anion extraction (WAX) cartridge was manufactured by Waters Corporation
(Milford, MA). The sample dryer was a Thermo vacuum dryer and Alltech
cell culture sample dryer. The 24-port solid extraction manifold was
purchased from Alltech associates (Deerfield, IL). The ACE C18 column 2.0
x 50 mm and 3 μM packing was manufactured by Advanced
Chromatography Technologies, Aberdeen, Scotland and distributed by Mac-
Mod analytical Inc (Chadds Ford, PA).

Preparation of calibration standards
Stock solutions of 3TC-TP at 22.2 μg/mL, CBV-TP at 240 μg/mL,
ddATP at 1 μmol/mL and TFV-DP at 1 μmol/mL were prepared by diluting
their reference standards in distilled water. These stock solutions were
further diluted with distilled water to obtain working standard solutions A, B
and C, with the concentrations listed in the Table 2.1. Six calibration
standards were prepared by spiking blank cell lysates from CEMss, U937 or
PBMC cells with comparable cell numbers to the samples. Each standard
was then spiked with 50 μL of 500 ng/mL ddCTP, 1000 ng/mL of ddGTP, 500
ng/mL of 2-ClA and 500 ng/mL of ADV-DP. Finally the cellular calibration
standards were spiked with working standard solutions A, B, and C. The odd
and even numbered standards were spiked with 50 μL and 100 μL of
standard solutions A, B and C. The double blank sample was not spiked with
either standard or internal standard, while the blank was spiked with internal
34
standards only. These standards and blanks were processed in parallel to
the samples.
For the analysis of endogenous nucleotide pools, stock solutions
containing each nucleotide were prepared in distilled water. The stock
solutions were sequentially diluted to prepare working standard solutions at
50, 500 and 5000 ng/mL. Six neat dNTP calibration standards with odd and
even numbers were spiked with 50 and 100 μL of above working solutions,
respectively. Each neat standard solution was also spiked with 50 μL of
internal standards. These dNTP neat calibration standards were extracted
and processed in parallel to study samples.

Preparation of quality control samples
CEMss cells proliferating at logarithmic growth phase were treated
with 10 μM of ABC, 3TC, and ddI, and 30 μM of TFV for 24 hours. After the
treatment, cells were washed, enumerated and lysed using ice cold
35
methanol. Aliquots of cell lysates (15 million cells) were transferred into nunc
tubes and vacuum dried. The concentration of ddNTP corresponding to each
NRTI was determined and these samples served as quality control for the
subsequent assay.

Sample processing
Cell extracts previously treated with NRTI(s) were transferred into
polypropylene tubes and spiked with 50 μL each of 500 ng/mL 2-
chloroadenosine triphosphate (2-Cl ATP), 250 ng/mL of ddC-TP, 250 ng/mL
of ddGTP and 250 ng/mL of ADV-DP. Samples were then acidified in 1 mL
of 0.1% formic acid and solid phase extraction was started immediately.
Calibration standards for ddNTP, neat dNTP standards, and quality control
samples were prepared according to previous section and preprocessed in
parallel to the cell extract samples.
All cell extracts samples were fractionated using WAX SPE cartridges
(Waters Oasis WAX, 3 mL, 60 mg capacity), which were preconditioned with
3 mL methanol and 3 mL of distilled water, followed by 3 mL of 0.1% formic
acid. The cellular extracts and calibration standards were transferred into
WAX cartridges and were washed with acidified 100 mM potassium chloride
(KCl) to elute the monophosphate and diphosphate forms of the nucleoside
analogs. Samples were next washed with 0.1% formic acid. The nucleoside
triphosphate (ddNTP) and endogenous nucleotide pools (dNTP) were finally
eluted by 1 mL of alkalized methanol into clean polypropylene tubes. The
36
eluents contained ddNTPs, dNTP and internal standards. Samples were
neutralized, vacuum dried and then suspended in 0.5 mL of 0.01% formic
acid, which contained 1 unit of acid phosphatase (Sigma Biologicals, St.
Louis, MO). Following 30 minutes incubation at 37ºC, all of the phosphates
were removed to yield the corresponding nucleoside analogs or phosphonate
nucleotide analogs (i.e. TFV and ADV). 50 μL of samples were transferred
into HPLC vials for dNTP analysis. The rest of dephosphorylated samples
were dried and reconstituted with 50 μL of 5% methanol in deionized (DI)
water, and 30 μL of the samples were injected into a LC/MS/MS system for
analysis.

LC/MS/MS analysis
Intracellular concentrations of nucleosides corresponding to the
ddNTPs were analyzed on a LC/MS/MS system, composed of an Agilent
1100 HPLC system and a Sciex API 3000 mass detector. The analytes were
separated using a reversed phase ACE C18 column (Advanced
Chromatography Technologies) with the dimensions of 2.0 x 50 mm and 3
μM packing. A step gradient program was applied to separate all the
analytes. The mobile phase consisted of methanol as component A and 20
mM ammonia acetate buffer at pH 4.5 as component B. The initial mobile
phase was set at 7% methanol for three minutes, escalated to 15% methanol
rapidly and maintained for 7 minutes. The mobile phase was changed to 80%
methanol for two minutes and subsequently reduced to 7% methanol and
37
equilibrated for another 11 minutes. After separation, the analytes in HPLC
efferent was introduced into the mass spectrometer through a turbo ion spray
interface, in which analytes and internal standards were ionized and carried a
positive charge. In addition, a heated turbo nitrogen stream was used to
evaporate solvents and increase ionization efficiency. The mass
spectrometer operated in three periods. The first period had an 8 minute
scan time for the detection of deoxycytidine (dC), ADV, TFV, dideoxycytidine
(ddC), deoxyguanosine (dG), 3TC and dT; the second period lasted 9
minutes and deoxyadenosine (dA), dideoxyguanosine (ddG), carbovir (CBV)
dideoxyadenosine (ddA), and 2 chloroadenosine (2 Cl A) were detected.
The third period was used to re-equilibrate the column. A LC/MS/MS
38
chromatogram is shown in Figure 2.1 and the retention time and mass
transition for each analyte is shown in table 2.2.

Validation procedure
To demonstrate that the method is reliable and reproducible for the
intended use, it was validated with regards to the following parameters:
extraction efficiency, linearity, accuracy and precision, extraction recovery,
carry over, specificity, stability. Because of the complexity of this assay, the
method development and validation focused on a few critical processes and
is discussed in the following sections.

39
Extraction efficiency
The fractionation of cell extracts was evaluated on WAX cartridges
and SAX cartridges, which is the updated version of traditional silica-based
QMA. CBV-TP was used as a model drug for this evaluation.
The extraction efficiency of removing monophosphate and
diphosphate while maintaining the triphosphate is critical for the success of
this approach, as the residual ddNMP and ddNDP in cell lysates will also
lead to the formation of nucleoside entities corresponding to ddNTP after
enzyme hydrolysis. In the first experiment, dGTP at 5 ng alone and with 10
ng of dGMP and dGDP was used. A similar study was conducted for ddNTP.
1.0×10
7
U937 cells were spiked with 10 pmole of TFV-DP and 3TC-TP alone
and in combination with 20 pmole of lamivudine monophosphate (3TC-MP).
Samples were prepared in duplicates and went through the same extraction
process. The area counts of TFV-DP and 3TC-TP were compared to
determine the impact of the presence of TFV and 3TC-MP.

Linearity
Six calibration standards were freshly prepared in blank CEMss or
U937 cell extracts. Calibration standards were processed and analyzed on
the API3000. Least-square linear regression using a weighting of
1/concentration
2
was performed to establish a linear calibration curve
between the area ratios of analyte to internal standard and the
concentrations of analyte. The linearity was established by back calculating
40
the concentration for each calibration standard. The comparison of the actual
concentration to the expected theoretical value established the precision and
accuracy of the assay. The actual value should be within 20% of the target
concentration and only 30% of standards could be rejected as outliers.

Accuracy and precision
The accuracy and precision of the assay was established by analyzing
four replicates of quality control samples and a middle standard over three
analytical runs. The intra-day precision and accuracy was established by
comparing the mean of these replicates to the target concentration. The
coefficient of variation of these replicates established the intra-day variation.
The inter-day precision was established by comparing the mean to the
targeted concentration. The inter-day accuracy was based on the variation of
all the replicates over the three analytical runs. The precision and accuracy
should all be within 25%.

Extraction recovery
Extraction recoveries of ddNTPs were performed to evaluate sample
loss and reproducibility of the extraction process. Both extracted and non-
extracted samples were carried out using three replicates of the third highest
calibration standard. The extracted samples were prepared by spiking
working standard and internal standards into blank CEMss cells and went
through SPE on WAX cartridges and dried. The non-extracted samples were
41
prepared by spiking the same working standards and internal standards into
blank CEMss cells which had already gone through WAX extraction. Both
extracted and non-extracted samples went through enzyme hydrolysis and
the rest of sample processing. Extraction recovery was estimated by
comparing the mean area counts of the extracted samples to the non-
extracted ones. It is preferable to have a relative high extraction recovery;
however, reproducibility is equally critical.

Incubation time
The impact of incubation time of enzyme hydrolysis for ddNTP was
investigated using TFV-DP and dNTP as model drugs. Two replicates of the
third highest TFV calibration standard and its internal standard ADV went
through WAX extraction. After vacuum drying, samples were incubated for
30 minutes and overnight for TFV. The area count of TFV and ADV were
compared to verify if longer incubation times enhanced the signal.
In the second study, a set of dNTP calibration standards were
incubated for 30, 60 and 90 minutes. Fraction of samples were transferred
into HPLC vials and analyzed immediately. The area counts of dNTP were
compared to obtain the optimized incubation time.

Stability
The stability of ddNTP and dNTP samples was investigated with
regards to its storage stability in a -80 °C freezer and in HPLC vials at room
42
temperature. Also stability of dNTP was investigated under regard to the pH
of the medium used after WAX extraction and the final reconstitution solution.

Chapter 2 Results
Extraction efficiency
The application of WAX HPLC separation of nucleotides by pH
gradient was reported recently (Shi et al., 2002, Veltkamp et al., 2006). Cell
extract fractionation on WAX cartridges was evaluated first during method
development and compared to the traditionally utilized strong anion
exchange approach (MAX, Waters Corp, Milford MA). The mean area count
43
of CBV was 7696 ± 1690 from WAX extraction as compared to 844 ± 156
from MAX extraction. These results suggested that fractionation of cell
extracts on WAX achieved higher extraction efficiency and higher sensitivity.
Thereby, WAX extraction was selected for the study.
The quality of the extraction process was evaluated with regards to
the potential impact of residual ddNMP and ddNDP in cell lysates. The
deoxyguanosine area counts from samples containing 5 ng dGTP and 10 ng
each of dGMP and dGDP were only 5% higher as compared to samples
containing dGTP only (Figure 2.3). In addition, TFV-DP and 3TC-TP area
counts in samples spiked with TFV and 3TC-MP were 0.2% lower and 5.2%
higher compared to those samples without TFV and 3TC-MP, respectively
(Figure 2.4). A two-sided student’s t-test was performed between the two
groups and no statistical significant difference was detected. These results
suggested that the fractionation process using WAX cartridges was efficient
in removing the mono- and diphosphate species of nucleoside analogs and
endogenous nucleotides, and the area counts were specific to the dNTP or
ddNTP in the cell extracts.

Linearity, precision and accuracy
Sensitivity of this assay with regards to each ddNTP was determined
from analytical batches across six different days (Table 2.3A). The
established linear dynamic ranges for ddNTP were as follows: 0.1 to 20
44

45
46
47
pmol/10
6
cells for TFV-DP and 3TC-TP, 0.0105 to 2.1 pmol/10
6
cells for
CBV-TP and 0.005 to 1 pmol/10
6
for ddATP. The limits of quantification were
at 0.1 pmole/10
6
cells for TFV-DP and 3TC-TP, 0.0105 and 0.005 pmole/10
6

cells for CBV-TP and ddATP, respectively.
The linear dynamic ranges for dATP, dGTP and dCTP were from 1.02
to 102, 0.986 to 98.6 and 1.07 to 107 pmol/10
6
cells, respectively. The limit of
quantification was determined to be 1.02, 0.986 and 1.07 pmole/10
6
cells for
dATP, dGTP and dCTP, respectively. Descriptive analysis was performed for
these batches to establish the inter-day precision and accuracy of this assay.
All ddNTPs at each level were within 15% of variation and ±15% of the error
as compared to the target as shown in Table 2.3A. The dNTP results were
based on five experiments and reported in Table 2.3B.

Intraday variation
Five replicates of quality control samples were assayed to assess the
intraday variation and precision of the assay for ddNTP and dNTP (Table
2.4). Results demonstrated that all analytes had less than 15% variation. In
addition, dNTP in six calibration standards in 1.5X10
7
CEMss cell were also
determined and the intraday variation was less than 15%. These results
demonstrated that this method was highly reproducible.
48

49
Extraction recovery
Extraction recoveries of ddNTPs were studied and results were listed
in Table 2.5. The extraction recovery was dependent on the compound.
While TFV-DP and ddATP had 95% recovery, extraction recovery of CBV-TP
and 3TC-TP were 51% and 40%, respectively. The recovery of internal
standards was 40-50%. In addition, the variation in the area counts was less
than 16%, while extraction of ddATP and ddCTP had a variation of 25% and
22%, respectively. However, when the area counts were normalized to dATP
in the cells, ddATP variation was less than 9% (data not shown).

Incubation time
The enzyme incubation time was established experimentally. Both
TFV-DP and ADV-DP had higher area counts in the samples going through
50
overnight incubation as compared to the 1.5 hour incubation (Table 2.6).
The area ratios of TFV-DP to ADV-DP were not altered. In the second
experiment, dNTPs were incubated for 30, 60, 90 and 120 minutes (data not
shown). The longer incubation times demonstrated decreased stability of the
endogenous nucleosides; thereby the enzyme incubation time was set at 30
minutes (data not shown).

Stability
The comparison study of dNTP stock solutions prepared three years
apart showed that the two standard solutions differed by -4% to 1%. This
demonstrated that dNTP standards in DI water were stable for at least three
years at -80 ºC. Relative stability of endogenous samples in reconstitution
solution was also studied and the results indicated these samples were
stable in HPLC vials for 24 hours at room temperature (Figure2.6).
Nucleoside bases corresponding to endogenous nucleotide pools were found
to be stable for at least 24 hours in acidified deionized water in HPLC vials.
The peak area counts of analytes analyzed immediately after sample
processing were comparable to these re-injected into LC/MS/MS 24 hours
later (Figure 2.7).
51

52
Chapter 2 Discussion
SAX cartridges were commonly used to fractionate cell lysates using
high salt concentration to separate ddNMP and ddNDP, and thus allow to
ddNTP to be isolated. However, high salt concentration found in the ddNTP
fraction after could cause ion suppress that is detriment to mass
spectrometry detection. In recent years, the WAX HPLC method has been
employed in the direct analysis of nucleotides in cellular extracts (Shi et al.,
2002, Veltkamp et al., 2006). We successfully adapted this concept and
developed a sensitive assay for ddNTP and dNTP quantification.
This approach uses a high concentration of KCl to fractionate cell
extracts, more specifically to remove ddNMPs and ddNDPs from cellular
extracts. Instead of using a high concentration KCl solution to elute ddNTPs,
we washed these samples with 0.1% formic acid solution to eliminate KCl
salt and then harvest ddNTP with alkalized methanol. The packing of the
WAX cartridge contains a weak base with pKa of 6, therefore under acidic
conditions it carries positive charge which can interact with the negative
charges found on ddNTP. However, when the pH is switched to an alkalized
condition, this ionic interaction is suppressed and thus allows ddNTPs to be
eluted out of WAX cartridges. By taking advantage of this feature, we are
able to harvest ddNTP fraction without using a high concentration KCl salt
solution.
Direct analysis of ddNTP by LC/MS/MS is a very attractive option and
there were quite a few recently published methods (Shi et al., 2002,
53
Veltkamp et al., 2006). Most publications for the direct analysis of ddNTP
used ion pairing agents for HPLC separation and a triple quadrupole tandem
mass spectrometer for detection. While this approach is very attractive, it
was not feasible when this method was under development where only the
API 3+ instrument was available. The sensitivity of this instrument was not
adequate in the negative ion mode and the utilization of ion pairing agents
may further complicate MS detection via ion suppression. An alternative
method was to develop an indirect approach, where the ddNTP fraction is
hydrolyzed using acid phosphatase to remove phosphate group. The
corresponding nucleoside to their respective ddNTP and dNTP were
quantified by LC/MS/MS.
SPE using WAX cartridges yields very clean extracts and reduces the
extent of ion suppression, which is commonly observed in the analysis of
samples in biological matrix. The SPE methodology can be easily adapted
into direct analysis by taking advantages clean extracts and thereby less ion
suppression.
During solid phase extraction on WAX cartridges, a more extensive
wash with 100 mM KCl was applied in order to remove all the ddNMPs and
ddNDPs from the samples and ensure that only ddNTPs were harvested.
The extensive wash was able to remove ddNMP and ddNDP (Figures 2.3
and 2.4). This measure also consistently had extraction recoveries in the
range of 40-60% for most compounds, except for ddATP and TFV-DP (Table
2.6). However, the relative low recovery did not have negative impact, as the
54
sensitivity of the assay was sufficient for the analysis of ddNTP in cell extract
derived from both cell culture and clinical studies. In addition, the assay was
reproducible and had intraday variation in the range of less than 16%.
Enzyme incubation time was optimized in this study. Endogenous
nucleosides that were formed after enzymatic hydrolysis were stable in acidic
condition for at least three hours (results not shown). Longer enzymatic
incubation time was found to have a negative impact on stability of dN
(Figure 2.5). These experiments suggested the optimal enzyme incubation
time was 3 hours; however a 30 minutes enzyme incubation time was
adapted throughout the study to prevent the potential stability issues.
The stability of dNTP was investigated and was found that the
endogenous nucleotides were stable for at least three years when stored at
-80 ºC freezer (Figure 2.6). In addition, endogenous nucleosides are stable
under room temperature for at least 24 hours when reconstituted in acidic
solutions.
This method described here has the capability to quantify multiple
nucleoside analogs and endogenous nucleotide pools simultaneously. A
limitation of this method is that the processing is lengthy and thus maybe
prone to human error. Though sample extraction procedures could be
adapted into a direct analysis approach using ion pairing agents or an WAX
HPLC column, it still remains to be investigated whether the direct approach
could determine multiple ddNTP and dNTP in a single run with the same type
of sensitivity.
55
Chapter 2 Conclusions
A sensitive and versatile LC/MS/MS has been developed for the
determination of intracellular concentrations of ddNTP and dNTP
simultaneously. This method has been used successfully in the analysis for
ddNTP and dNTP in cell extracts for both cell culture study and clinical study.
The method demonstrated good linear dynamic range, reproducible
extraction process and extraction efficiency. The method was sensitive and
within the linear dynamic range covering the levels observed in the HIV
patients.
This method is versatile and can be adapted for the determination of
ddNTPs corresponding to any other nucleoside analogs. In addition, this
methodology can be utilized in the drug discovery stage for the screening of
compounds to evaluate possible intracellular drug-drug interaction. It could
also be used in the clinical setting for the optimization of combination
treatment involving two nucleoside analogs.

56
Chapter 2 References

Anderson, P.L., Kakuda, T.N. & Lichtenstein, K.A. (2004) The cellular
pharmacology of nucleoside- and nucleotide-analogue reverse-
transcriptase inhibitors and its relationship to clinical toxicities, Clin
Infect Dis, 38(5), pp. 743-753.
Becher, F., Pruvost, A., Goujard, C., Guerreiro, C., Delfraissy, J.F., Grassi, J.
& Benech, H. (2002) Improved method for the simultaneous
determination of d4T, 3TC and ddl intracellular phosphorylated
anabolites in human peripheral-blood mononuclear cells using high-
performance liquid chromatography/tandem mass spectrometry,
Rapid Commun Mass Spectrom, 16(6), pp. 555-565.
Fletcher, C.V., Anderson, P.L., Kakuda, T.N., Schacker, T.W., Henry, K.,
Gross, C.R. & Brundage, R.C. (2002) Concentration-controlled
compared with conventional antiretroviral therapy for HIV infection,
AIDS, 16(4), pp. 551-560.
Fletcher, C.V., Kawle, S.P., Kakuda, T.N., Anderson, P.L., Weller, D.,
Bushman, L.R., Brundage, R.C. & Remmel, R.P. (2000) Zidovudine
triphosphate and lamivudine triphosphate concentration-response
relationships in HIV-infected persons, AIDS, 14(14), pp. 2137-2144.
Gallant, J.E., Rodriguez, A.E., Weinberg, W.G., Young, B., Berger, D.S., Lim,
M.L., Liao, Q., Ross, L., Johnson, J. & Shaefer, M.S. (2005) Early
virologic nonresponse to tenofovir, abacavir, and lamivudine in HIV-
infected antiretroviral-naive subjects, J Infect Dis, 192(11), pp. 1921-
1930.
Groschel, B., Cinatl, J. & Cinatl, J., Jr. (1997) Viral and cellular factors for
resistance against antiretroviral agents, Intervirology, 40(5-6), pp. 400-
407.
Harris, M., Back, D., Kewn, S., Jutha, S., Marina, R. & Montaner, J.S. (2002)
Intracellular carbovir triphosphate levels in patients taking abacavir
once a day, AIDS, 16(8), pp. 1196-1197.
57
Havlir, D.V., Tierney, C., Friedland, G.H., Pollard, R.B., Smeaton, L.,
Sommadossi, J.P., Fox, L., Kessler, H., Fife, K.H. & Richman, D.D.
(2000) In vivo antagonism with zidovudine plus stavudine combination
therapy, J Infect Dis, 182(1), pp. 321-325.
Jemsek J, H.P., Harper E. (2004) Poor virologic response and early
emergence of resistance in treatment naive, HIV-infected patients
receiving a once daily triple nucleoside regimen of didanosine,
lamivudine, and tenofovir DF11th Conference on Retroviruses and
Opportunistic Infections; (San Francisco,
Kewn, S., Hoggard, P.G., Sales, S.D., Johnson, M.A. & Back, D.J. (2000)
The intracellular activation of lamivudine (3TC) and determination of
2'-deoxycytidine-5'-triphosphate (dCTP) pools in the presence and
absence of various drugs in HepG2 cells, Br J Clin Pharmacol, 50(6),
pp. 597-604.
Kewn, S., Hoggard, P.G., Sales, S.D., Jones, K., Maher, B., Khoo, S.H. &
Back, D.J. (2002) Development of enzymatic assays for quantification
of intracellular lamivudine and carbovir triphosphate levels in
peripheral blood mononuclear cells from human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 46(1), pp. 135-
143.
Pruvost, A., Negredo, E., Benech, H., Theodoro, F., Puig, J., Grau, E.,
Garcia, E., Molto, J., Grassi, J. & Clotet, B. (2005) Measurement of
intracellular didanosine and tenofovir phosphorylated metabolites and
possible interaction of the two drugs in human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 49(5), pp.
1907-1914.
Ray, A.S., Myrick, F., Vela, J.E., Olson, L.Y., Eisenberg, E.J., Borroto-Esodo,
K., Miller, M.D. & Fridland, A. (2005) Lack of a metabolic and antiviral
drug interaction between tenofovir, abacavir and lamivudine, Antivir
Ther, 10(3), pp. 451-457.
Robbins, B.L., Greenhaw, J., Connelly, M.C. & Fridland, A. (1995) Metabolic
pathways for activation of the antiviral agent 9-(2-
phosphonylmethoxyethyl)adenine in human lymphoid cells, Antimicrob
Agents Chemother, 39(10), pp. 2304-2308.
58
Robbins, B.L., Wilcox, C.K., Fridland, A. & Rodman, J.H. (2003) Metabolism
of tenofovir and didanosine in quiescent or stimulated human
peripheral blood mononuclear cells, Pharmacotherapy, 23(6), pp. 695-
701.
Shi, G., Wu, J.T., Li, Y., Geleziunas, R., Gallagher, K., Emm, T., Olah, T. &
Unger, S. (2002) Novel direct detection method for quantitative
determination of intracellular nucleoside triphosphates using weak
anion exchange liquid chromatography/tandem mass spectrometry,
Rapid Commun Mass Spectrom, 16(11), pp. 1092-1099.
Vela, J.E., Olson, L.Y., Huang, A., Fridland, A. & Ray, A.S. (2007)
Simultaneous quantitation of the nucleotide analog adefovir, its
phosphorylated anabolites and 2'-deoxyadenosine triphosphate by
ion-pairing LC/MS/MS, J Chromatogr B Analyt Technol Biomed Life
Sci, 848(2), pp. 335-343.
Veltkamp, S.A., Hillebrand, M.J., Rosing, H., Jansen, R.S., Wickremsinhe,
E.R., Perkins, E.J., Schellens, J.H. & Beijnen, J.H. (2006) Quantitative
analysis of gemcitabine triphosphate in human peripheral blood
mononuclear cells using weak anion-exchange liquid chromatography
coupled with tandem mass spectrometry, J Mass Spectrom, 41(12),
pp. 1633-1642.

59
Chapter 3: Intracellular Drug-Drug Interaction among Nucleoside
Analogs Lead to Early Virologic Failure in HIV Patients Receiving
Combination of Abacavir, Tenofovir and Lamivudine

Chapter 3 Abstract
Triple nucleoside analog (NA) combination as the sole components in
HIV therapy has produced mixed clinical outcomes. One of these regimens
using tenofovir (TDF), abacavir (ABC), and lamivudine (3TC) resulted in a
significant number of early virologic failures in clinical trials. Virus emanating
from these patients revealed M184V/1 and/or K65R resistance mutations.
Our study comprehensively evaluated the pharmacologic factors leading to
virologic failures, which include intracellular drug-drug interaction and the role
of efflux transporters.
U937 and CEMss were used to determine whether an intracellular (IC)
drug-drug interaction occurred between TFV, ABC, and 3TC., Nucleoside
resistant variants of both cell lines were developed using serial passage over
6 months. These cells were treated alone, in dual or triple combination at 5
and 20 μM for 24 hours and 7 days. The IC triphosphate anabolites of NAs
and their respective endogenous triphosphate levels were determined using
a validated LC-MS-MS method. In addition, the impact of nucleoside
exposure on efflux transporters was evaluated using Western blot analysis.
When U937
WT
cells were treated with 5 μM of ABC, TFV and 3TC
over 24 hours, intracellular ratios of CBV-TP/dGTP, TFV-DP/dATP and 3TC-
TP/dCTP were 82.5%, 88.3% and 79.6%, respectively, as compared to
60
control. After 7 days of treatment, intracellular ratios of CBV-TP/dGTP, TFV-
DP/dATP and 3TC-TP/dCTP were 84.8%, 95.7%, and 67.3%, respectively,
as compared to control. In cells treated with combination of TFV and ABC,
both IC TFV-DP and CBV-TP concentration were reduced significantly in a
concentration dependent manner.
When the IC levels 3TC-TP, CBV-TP and TFV-DP were measured in
wild type cells as compared to TFV resistant U937 cells (U937
TFV
), the IC
levels were 65%, 44% and 74% lower than levels achieved in U937 cells,
respectively. MRP2 and MRP4 expression were significantly enhanced in
U937
TFV
and U937
ABC
cells as compared to U937
WT
. The over-expression of
both MRP2 and MRP4 enhanced cellular viability by reducing their
exposures to both TFV and ABC at supra physiological concentrations.
These data suggest that there may be a competitive inhibition
between TFV and ABC within lymphocytes especially at 20 μM
concentrations. When these agents are combined in a triple nucleoside
combination, 3TC-TP/CTP is 33% lower after 7 days of treatment. More
over, after receiving long-term treatment with TFV and ABC, both CEM and
U937 can increase expression of both MRP2 and MRP4 and confer cross
resistance toward multiple nucleoside analogs. These findings may provide
significant insights leading to virologic failures seen in clinical trials.
61
Chapter 3 Introduction
Combinations of triple nucleoside reverse transcriptase inhibitors
(NRTIs) with either non-nucleoside reverse transcriptase inhibitors (NRTIs)
or protease inhibitors (PIs) have been evaluated in the treatment of HIV
positive patients with mixed outcomes (Gallant et al., 2005, Jemsek et al.,
2004, Khanlou et al., 2005, Maitland et al., 2005, Moyle & Gazzard, 2002,
Staszewski et al., 2001). In clinical trials, the combination of zidovudine
(AZT), lamivudine (3TC), and abacavir (ABC), which is also known as the
fixed dose combination Trizivir, achieved reasonable success in treating
HIV patients (Moyle & Gazzard, 2002, Staszewski et al., 2001). Since
Trizivir has twice daily dosing, strategies to improve triple nucleoside
therapy using nucleoside components with pharmacokinetic profiles
supporting once daily dosing were developed. This includes the combination
of 3TC and tenofovir (TFV) with either ABC or didanosine (ddI).
TFV, ABC and ddI are adenosine or inosine analogs, which are
classified as purine analogs. Despite the theoretical promise of using this
once-daily regimen of triple nucleosides, the combination of TFV, 3TC and
ABC did not perform well in clinical trials, as a significant number of patients
were either unable to suppress viral loads below 50 copies/mL or developed
virologic rebounds after achieving viral suppression (Gallant et al., 2005,
Khanlou et al., 2005). Clinical failures were also seen when patients were
given the triple nucleoside combination containing TFV, 3TC and ddI
(Jemsek et al., 2004, Maitland et al., 2005).
62
The combination of TFV and ABC has undergone rigorous scrutiny, as
this combination has never been evaluated in a formal clinical study.
Virologic studies suggested that the combination of ABC with TFV is not
antagonistic (Delaunay et al., 2005, Lanier et al., 2005), where resistant
viruses detected in HIV patients who experienced early virologic failure often
harbored K65R and M184V mutations at viral reverse transcriptase. In
addition, no plasma drug interactions between the two nucleosides analogs
were detected (Kearney BP, 2003). Intracellular drug-drug interaction
between TFV and ABC continued to be a plausible explanation for the early
virological failures observed in HIV patients receiving the combination of
TFV, 3TC and ABC.
Intracellular drug-drug interaction between ABC and TFV combination
were evaluated in an in vitro study, however, this study did not able to detect
significant changes of triphosphate nucleotides of the corresponding
nucleoside analogs (Ray et al., 2005). Although a moderate reduction of
triphosphate anabolites was detected, they were neither statistically nor
clinically relevant since the patient population had high variability.
Results from a clinical study reported by Hawkins et al also did not
demonstrate any drug interaction with regards to intracellular triphosphate
changes in TFV-DP and carbovir-triphosphate (CBV-TP) (Hawkins et al.,
2005). This clinical finding was further supported by in vitro studies
evaluating the active triphosphate anabolites of both TFV-DP and CBV-TP
(Ray et al., 2005). The in vitro study also evaluated the levels of
63
endogenous nucleotides, 2’-deoxyguanosine and 2’-deoxyadenosine, where
at the concentration range evaluated, no significant changes in endogenous
nucleotide pools were found. These two studies suggested intracellular
interaction between ABC and TFV was not the cause of early virological
failure seen in the aforementioned clinical studies. However, these studies
were performed over short time frame and at a concentration range were
limited and were physiologically achievable.
Despite these findings, intracellular drug-drug interactions involving a
number of nucleoside analogs have been identified, such as ddC with 3TC,
AZT and d4T (Havlir et al., 2000, Kewn et al., 1997). NRTIs are prodrugs and
require sequential cellular activation to form their respective active 5’-
triphosphate (ddNTP). There are multiple anabolic and catabolic enzymes
involved in these activation processes and often these enzymes have cross
activities towards multiple NRTIs, thus forming overlapping pathways during
their respective activation processes. This could constitute a potential
mechanism of their intracellular interaction (Prueksaritanont et al., 2002, Ray
et al., 2004).
This present study evaluated whether a change in the triphosphate
anabolites could be detected with increased concentrations and with longer
exposure time. In addition, this study evaluated the role of efflux transporters
and their possible contribution to a potential intracellular drug-drug
interaction. Multidrug resistant protein-2 (MRP2) and MRP4 in mammalian
cells are major efflux transporters with moderate affinity for nucleoside
64
analogs (Miller, 2001, Schuetz et al., 1999). The expression of these efflux
transporters is inducible; where over expression of these transporters can
confer cellular resistance towards a wide spectrum of nucleoside analogs.
One typical example is the over expression of MRP4 in CEM-r1 cells, which
was selected by adefovir (ADV, PMEA) and resulted in resistance towards
not only ADV but also many other nucleoside analogs (Schuetz et al., 1999,
Weiss et al., 2007).
To add more complexity, these cellular anabolic and catabolic
enzymes and transporters can undergo cellular adaptive changes with
prolonged treatment with nucleoside analogs (NAs). Long-term treatment
with NAs in cancer treatment often leads to resistance towards these
nucleoside-based therapies. A notable mechanism is the alteration of RNA
and protein expression of metabolic enzymes required for the
biotransformation of NAs to form the active ddNTP (Galmarini et al., 2003,
Mansson et al., 1999). Similarly, normal cells may also deploy these
mechanisms to reduce exposure to potentially toxic xenobiotics and result in
cellular resistance and reduced formation of ddNTP.
These cellular factors may reduce intracellular levels of nucleoside
analogs, like ABC, TFV and 3TC. Lowered levels of the nucleoside analogs
may lead to reduction of phosphorylated anabolites. Lastly, reduced
formation of ddNTP may compromise the antiviral activities of NAs and
contribute to the failure of the triple nucleoside combination of ABC, TFV and
3TC.
65
In this study, we determined the cellular activation of NRTI in U937
and CEMss cells when they were treated with ABC, TFV and 3TC alone or in
combination. We detected significant intracellular drug-drug interactions
among these NRTIs and evaluated their potential impact on early virological
failure in HIV patients receiving the triple nucleoside combination of ABC,
TFV and 3TC. This study estimated the ratio of triphosphate of NA to their
corresponding endogenous nucleoside pool (ddNTP/dNTP ratio), and further
determined the fold changes in these ratios in U937 or CEMss cells treated
with ABC, TFV and 3TC in combination. The change in these ratios
suggested that intracellular interaction among ABC, TFV and 3TC may have
huge impact on the antiviral activity of these NRTIs.

Chapter 3 Materials and methods
Chemicals
ABC, 3TC and the corresponding triphosphates carbovir-triphosphate
(CBV-TP) and lamivudine-triphosphate (3TC-TP) were generous gifts from
GlaxoSmithKline (Dr. Katrina Oie). TFV, tenofovir diphosphate (TFV-DP) and
adefovir diphosphate (ADV-DP) were purchased from Movareck, CA. Other
triphosphate internal standards ddCTP, ddGTP, 2Cl-ATP and endogenous
nucleotide pools, dATP, dGTP, dCTP and dTTP were purchased from Sigma.

66
Cell lines
The human promonocytic (U937
WT
) and lymphoblastic (CEMss)
leukemia T-cell lines were used in this study. The ABC and TFV resistant
variants, U937
ABC
, CEM
ABC
and U937
TFV
, CEM
TFV

were selected from serial
passage of the parental U937
WT
and CEMss cells in the presence of
escalating concentrations of ABC or TFV from 1 to 20 μg/mL over six months,
respectively. Cellular viability assay and Western blot analysis were applied
to characterize these variant cells. Cells were grown in RPMI supplemented
with 10% fetal bovine serum, 1X sodium pyruvate, and 100 μg/mL of
penicillin and streptomycin. Cells were maintained at 37° C in 5% CO
2
at 95%
humidity and were serially passaged weekly.

Intracellular drug-drug interaction study
24 hour treatment
Medium was changed 24 hours prior to treatment. On the day of
treatment, cells were washed twice with PBS, enumerated and then seeded
in T-75 flasks, where cellular density was set at 5X 10
5
/mL for U937
WT
and
its variant cells. In contrast,

cellular density was set at 7.5 X 10
5
cells/mL for
CEMss. NRTI stock solution was spiked into the cell suspension so that the
final RPMI medium contained the designated agent(s) at 5 μM or 20 μM
either alone or in combination. After 24 hour treatment, cells were washed
with 10 mL of ice cold PBS twice and lysed with 3 mL of ice cold methanol.
The cellular extracts were centrifuged for 5 minutes at 4500 RPM at 4°C. The
67
supernatants were transferred into Nunc tubes and stored at -80°C until
analysis.

7 days treatment
To mimic the drug-drug interaction among nucleoside analogs in HIV
patients after receiving treatment for a relatively longer term, drug interaction
studies were performed in which CEMss or U937 cells were treated with
nucleoside analogs for 7 days. In the 7-day treatment study, cells were
seeded at 1 X 10
5
cells per mL of RPMI in T-75 flasks and fed with fresh
medium containing a specified concentration of NRTI every other day until
day 7 when cells were harvested. At the end of the study, cells were
harvested and cell lysates were stored at -80°C until analysis.

Concentration escalation study
To explore the effect of concentration on intracellular drug-drug
interaction between ABC and TFV, CEMss cells were treated for 24 hours
with ABC at concentrations from 2 to 20 μM alone and in combination with
TFV at either 5 or 20 μM. At the end of the study, cells were harvested and
cell lysates were stored at -80°C until analysis.

68
Cellular viability assay
Cells were suspended at a density of 1 X 10
5
cells/well. Cells were
then treated with the respective nucleosides at concentrations from 2 to 1000
μg/mL and incubated for 24 hours at 37
o
C in 5% CO
2
. The cell viability was
determined using MTT [3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium
bromide] (Sigma Chemicals, MO, USA) or alomar blue assays (Biosource,
Thousand Oaks, CA). After incubation for 24 hours, an aliquot of 100 μL of
250 μg/mL MTT was added and incubated for 2 hours. The dark blue
formazan crystals were solubilized and liberated by the addition of 200 μL
isopropanol. The absorbance was measured using a Tecan 96-well
spectrophotometer set at a wavelength of 570 nm.
These results were compared with cell viability assays using alomar
blue (Biosource), which is a soluble non-toxic dye that does not alter the
viability of cultured cells. Cells were treated with various levels of
nucleosides, and after the treatment, 10 μL of alomar blue dye was added
into each well and incubated for three hours before UV absorbance was
measured at 570 and 620 nm. Cellular viability of cells receiving treatment of
NRTI was compared to the untreated cells.

Western analysis of MRP2 and MRP4
CEMss and U937
WT
cells were exposed to 10 and 50 μM of TFV for
24 hours to determine the expression of MRP2 and MRP4. After treatment,
cells were collected for Western Blot analysis. In addition, studies were
69
performed to investigate the cellular adaptive changes in regards to the
expression of efflux transporters after long-term treatment with nucleoside
analogs. Variant cell lines of CEMss and U937
WT
were collected for Western
Blot analysis.
To perform the Western Blot analysis, cellular membranes were
isolated, and 75 μg of extracted membrane proteins were separated using
7.5% SDS-PAGE. Proteins were then transferred overnight onto
nitrocellulose PVDF membranes, which were blocked with 1x PBS-Tween
plus 5% non-fat dry milk for 1 hour at room temperature. Afterwards, the
membranes were washed twice with 1x PBS-Tween and once with 1x PBS-
Tween plus 5% dry milk. They were probed with antibodies specific for MRP2
(Upstate) and MRP4 (Santa Cruz Biotechnology, CA, USA). Subsequently,
the membranes were further probed with horseradish peroxidase-conjugated
(HRP) secondary antibodies against goat IgG (Santa Cruz Biotechnology,
CA, USA). Immunoblots were developed using a chemiluminescent detection
system (Bio-Rad Labs, CA, USA). MRP2 and MRP4 expression by Western
Blot analysis was then compared to that of untreated cells or wild type cells.

Determination of triphosphate of nucleoside analogs
Sample preparation and processing
Cell lysates were spiked with 100 μL of 500 ng/mL 2-chloroadenosine
triphosphate (2-Cl ATP), 250 ng/mL of ddC-TP, and 250 ng/mL of ADV-DP
and dried under forced air. 2-Cl ATP, ddC-TP and ADV-DP were used as
70
internal standards for CBV-TP, 3TC-TP, and TFV-DP, respectively. Samples
were then dissolved in 1 mL of 0.1% formic acid and preceded to solid phase
extraction. Calibration standards were prepared by spiking the
aforementioned internal standards with 100 μL of standard solution of each
compound of interest into 1.5 X10
7
CEMss or 1.0 X10
7
U937 cell extracts,
and processed in parallel to cell extract samples.
All cell extracts were fractionated using a weak anion exchange SPE
cartridges (Waters Oasis WAX, 3 mL, 60 mg capacity). SPE cartridges were
preconditioned with 3 mL methanol, followed by 3 mL of 0.1% formic acid.
Then the cell extracts and calibration standards were loaded onto WAX
cartridges. Samples were washed with 100 mM KCl, then washed 0.1%
formic acid, and finally eluted into clean test tubes with 1 mL of methanol
solution containing 3% ammonia hydroxide.
The eluents were dried and suspended in 0.5 mL of 0.01% formic acid
(pH of 2.7). To this solution, 1 unit of acid phosphatase (Sigma) was added
and samples were incubated for 30 minutes at 37ºC. After enzyme
hydrolysis, 50 μL of samples were transferred into a HPLC vial for
endogenous nucleotide analysis. The remaining samples were dried and
reconstituted with 50 μL of 5% methanol solution. An aliquot of 30 μL of
solution were injected into a LC/MS/MS system for analysis.

71
LC/MS/MS analysis
Intracellular concentrations of nucleosides corresponding to the
ddNTPs were analyzed using a validated liquid chromatography-tandem
mass spectrometry system (LC/MS/MS), which consisted of an Agilent 1100
high performance liquid chromatography (HPLC) system coupled to a triple
quadruple mass spectrometer (Sciex API 3000). A reversed phase ACE C18
column (Advanced Chromatography Technologies, 3.0 X 50 mm, 3 μm
packing) was used at a flow rate of 0.3 mL per minute under a step gradient
to separate all nucleosides relating to dNTP and ddNTPs. The mobile phase
consisted of methanol as component A and 20 mM ammonia acetate buffer
at pH 4.5 as component B. The initial mobile phase was set at 93% B for
three minutes, decreased to 85% B and maintained for 6 minutes, washed
with 20% B for two minutes and increased to 93% B and equilibrated for
another 10 minutes. After separation, the analytes in HPLC efferent was
introduced into the mass spectrometer for detection. The mass spectrometer
operated in three periods. The first period had 8 minutes scan time for the
detection of deoxycytidine (dC), ADV, TFV, dideoxycytidine (ddC),
deoxyguanosine (dG), deoxythymidine (dT) and 3TC; the second period
lasted 8 minutes and deoxyadenosine (dA), dideoxyguanosine (ddG),
carbovir (CBV), dideoxyadenosine (ddA) and 2 chloroadenosine (2 Cl A)
were detected The third period was used to re-equilibrate the column.

72
Data analysis
The ratios of area counts of analyte to area counts of internal standard
were linearly regressed to the concentrations of analyte, using a weighting
factor of 1/concentration
2
. The established linear calibration curve was used
to back calculate the actual concentration for each calibration standard. The
comparison of the actual concentration to the expected theoretical value
established the precision and accuracy of the assay. Inter-day precision and
accuracy of the assay were determined by pooling the results of six
calibration curves obtained on different days. The calibration curve was also
used in the determination of ddNTP in cell lysates and the results were then
normalized to the cell counts and expressed as femtomole/10
6
cells.
To determine the intracellular drug-drug interactions, the cells treated
with single NRTI were used as controls. Intracellular ddNTP concentration of
cells receiving combination treatment was compared to that in control and
fold change were determined as follows:
Fold change = ddNTP in combination NRTIs/ddNTP in single NRTI

The endogenous nucleotide pools in the cell lysates were also
detected by the LC/MS/MS system. The ratios of area counts of dNTP pools
to their respective internal standards were determined. These ratios were
also normalized to the cell counts and incorporated in the estimation of
ddNTP/dNTP ratios. The ddNTP/dNTP ratio and its fold change were
calculated using the same equation as above. In the later part of the study,
73
the neat standard solutions containing endogenous nucleotide at various
concentrations were prepared and extracted in parallel with cell lysates.
Calibration curves for dNTP were established using linear regression, in
which the ratio of area counts of dNTP to area counts of internal standard
was regressed to the concentration. The weighting factor was
1/concentration
2
. The dNTP concentrations in cell lysates were determined
using these calibration curves. The final results were normalized to the cell
counts and reported as pmol/10
6
cells. The calibration range for 3TC-TP,
CBV-TP and TFV-DP in CEMss or U937 cells were at 0.10-20 pmole/10
6

cells, 0.0105-2.1 pmole/10
6
cells and 0.1-20 pmole/10
6
cells using 1×10
7

cells, respectively.

Statistical analysis
Descriptive statistic analysis was applied to the fold change in ddNTP
and ddNTP/dNTP using cells treated with single nucleoside analogs as
control. In addition, Wilcoxon rank test and one sample t-test were used to
determine fold changes against unity to identify any drug interactions.
ANOVA was applied to test the effect of dose on interaction between TFV
and ddI. The significance level was set a priori with significance α at 0.05.
All analyses were performed using GraphPad Prism version 4.0 for Windows
(GraphPad Software, San Diego, CA).

74
Chapter 3 Results
24 hours treatment
Intracellular triphosphorylated nucleotide levels were determined in
U937
WT
cells after 24 hours of incubation in the presence of 5 μM of ABC,
TFV and 3TC alone or in combination. Average and median fold-change in
five experiments is summarized in Table 3.1 and 3.2. When U937
WT
cells
were treated with the combination of ABC and TFV, IC concentrations of
CBV-TP and TFV-DP were 74.5% and 84.1% and CBV-TP/dGTP and TFV-
DP/dATP ratios were 80.9% and 84.5%, respectively, as compared to that in
75
control. When cells were treated with combination of TFV and 3TC for 24
hours, IC formation of 3TC-TP and 3TC-TP/dCTP were 75.2% and 95.9% of
the control, respectively, however, there were no changes in TFV-DP and
TFV-DP/dATP ratio. In the series of 24 hour treatment with ABC and 3TC,
the IC formation of CBV-TP and 3TC-TP were 111% and 95.1% and CBV-
TP/dGTP and 3TC-TP/dCTP ratios were 81.2% and 87.9%, respectively, as
compared to control. When U937
WT
cells were treated with combination of
ABC, 3TC and TFV for 24 hours, intracellular CBV-TP, TFV-DP and 3TC-TP
formation were 95.2%, 86.5% and 78.9% and CBV-TP/dGTP, TFV-DP/dATP
and 3TC-TP/dCTP were 97.2%, 88.3% and 79.2%, respectively, as
compared to that in control.

7 days treatment
U937
WT
cells were treated with 5 μM of ABC, TFV and 3TC alone or in
combination for 7 days to assess the intracellular drug-drug interaction
among nucleoside analogs over a longer period. Intracellular formation of
triphosphate nucleotide (ddNTP) corresponding to each NRTI was
determined and fold-change in ddNTP and ddNTP/dNTP ratios were
summarized in Table 3.3 and 3.4. After U937
WT
cells were treated with ABC
and TFV for one week, intracellular CBV-TP and TFV-DP concentrations
were at 78.8% and 82.4% and CBV-TP/dGTP and TFV-DP/dATP were at
74.1% and 83.9%, respectively, as compared to cells treated with single
agent only. When cells were treated with ABC and 3TC for one week,
76
intracellular CBV-TP and 3TC-TP concentrations were 83.6% and 96.2% and
CBV-TP/dGTP and 3TC-TP/dCTP was at 96.9% and 81.7%, respectively, as
compared to control. When treated with the combination of 3TC and TFV for
one week, intracellular 3TC-TP and TFV-DP concentrations were 91.5% and
88.9% and 3TC-TP/dCTP and TFV-DP/dATP ratios were at 77.7% and
88.0%, respectively, as compared to that in control. When U937
WT
cells
were treated with combination of ABC, TFV and 3TC, the intracellular
concentrations of CBV-TP, TFV-DP and 3TC-TP were 73.5% 94.2% and
77
81.3% and the ratios of CBV-TP/dGTP, TFV-DP/dATP and 3TC-TP/dCTP
were 84.8%, 95.7% and 67.3%, respectively, as compared to that in control.

Concentration escalation study
To further investigate the intracellular interaction between ABC and
TFV, CEMss were incubated for 24 hours with ABC at 2, 10, and 20 μM and
TFV at either 5 or 20 μM. The intracellular concentrations of CBV-TP and
TFV-DP were determined and intracellular interaction between the two drugs
was evaluated across the entire concentration as shown in Figure 3.1. When
2 μM of ABC was combined with either 5 or 20 μM of TFV, the IC level of
CBV-TP decreased in a concentration dependent manner (Figure 3.1, Panel
A). Similar findings were found when 10 and 20 μM of ABC were used,
where the IC CBV-TP concentration decreased significantly when 20 μM of
TFV was used (Figure 3.1, Panel B and C).
The intracellular concentration of TFV-DP was evaluated using 20 μM
of TFV and escalating levels of ABC (Figure 3.2). The decline in the level of
TFV-DP was detected even at 2 μM of ABC, where the drop in TFV-DP was
approximately 40% of that found when no ABC was added.

Effect of long-term treatment of NRTIs on U937
WT

To mimic the impact of long-term nucleoside therapy in HIV patients
with NRTIs, U937
TFV
cells were used to investigate cellular adaptive
mechanisms associated with NRTI therapy. U937
WT
cells were incubated in
78
79
RPMI medium in the presence of escalating concentrations of TFV or ABC
over a 6 month period to develop TFV-resistant and ABC-resistant variant
cells, U937
TFV
and U937
ABC
,

respectively. To characterize these cells, cellular
viability assays of U937
WT
were compared with its variant cells U937
ABC
and
U937
TFV
, which were resistant to ABC and TFV, respectively.
Cellular viability assays in U937
WT
and its variant cells U937
ABC
and
U937
TFV
using increasing concentrations of ABC and TFV were evaluated
and summarized in Figure 3.3. The results were based on four experiments.
All three cell variants were treated for 48 hours with escalating
concentrations of ABC or TFV, and cellular viability was assessed with an
80
81
82
Alomar blue assay (panel A and B). Cellular viability assays were also
performed using a MTT assay, and the results were similar (results not
presented) The IC
50
for U937
WT
cells was 150 and 1000 μg/mL after 48
hours treatment with TFV and ABC, respectively. In contrast, the IC
50
for
U937
ABC
and U937
TFV
was not achievable in these cells even at the highest
concentration, which was at 1000 μg/mL (Figure 3.3, panel C).
Concentrations above 1000 μg/mL were not used in this study because drug
solubility could not be assured. Two sided student t-tests were performed to
compare the three treatment groups, and both U937
TFV
and U937
ABC
were
significantly different from U937
WT
cells with P<0.05. The cellular viability
assay was also performed in CEMss, CEM
TFV
and CEM
ABC
cells and similar
results were obtained (results not presented).
To investigate the molecular mechanism of cellular resistance towards
TFV and ABC, Western blot analysis was performed to assess the
expression of efflux transporters in these cells. In U937
TFV
and U937
ABC
cells,
expression of both MRP2 and MRP4 were enhanced as compared to that in
U937
WT
(Figure 3.3 panel B).

ddNTP formation in U937
WT
and U937
TFV

U937
TFV
and U937
WT
cells were treated for 24 hours with 20 μM of
ABC, 3TC and TFV alone and the intracellular concentrations of CBV-TP,
3TC-TP and TFV-DP were evaluated. Results were based on three
independent experiments in U937
WT
and U937
TFV
. The intracellular levels of
83
CBV-TP, 3TC-TP and TFV-DP were found to be 1313 ± 517, 3614 ± 753 and
2193 ± 627 fmole/10
6
in U937
TFV
cells, respectively. However, the IC
concentration of CBV-TP, 3TC-TP and TFV-DP in the U937
WT
cell receiving
identical treatment were 2343 ± 609, 10,371 ± 4178 and 8505 ± 2713
fmole/10
6
cells, respectively (Figure 3.3, Panel A). In comparison with the
level of CBV-TP, 3TC-TP and TFV-DP in U937
WT
cells, a reduction of 44%,
65% and 74% was observed in U937
TFV
, respectively. A two-sided student t-
test was performed to compare the two groups and the results showed
significant level for TFV-DP with P<0.05.

Chapter 3 Discussion:
The interaction between ABC and TFV has been previously evaluated,
however no pharmacologic finding can currently explain the lack of clinical
activity (Hawkins et al., 2005, Ray et al., 2005). These findings were
supported by virologic assessments, where virologic antagonism was also
not detected (Lanier et al., 2005, Ray et al., 2005). Despite the absence of
adequate data explaining the lack of antiviral activity with 3TC, ABC and
TFV, caution should be taken when employing these three nucleosides
without the presence of a non-nucleoside reverse transcriptase inhibitor
(NNRTI) or a protease inhibitor (PI).
When evaluating the above studies, a commonality exists in that only
physiological achievable concentrations were used in assessing the
84
interaction (Hawkins et al., 2005, Lanier et al., 2005, Ray et al., 2005). In an
elegantly designed clinical trial, Hawkins et al evaluated patients who were
viral controlled and receiving nucleoside therapy. The intracellular
concentrations of CBV-TP and TFV-DP were determined, where no
significant alteration of either triphosphate anabolites were detected before
and after the cessation of either one of the corresponding nucleosides
(Hawkins et al., 2005). This clinical finding was supported by an in vitro
study, in which CEMss were used to evaluate intracellular interaction
between ABC and TFV. In this study, the endogenous nucleotide pools were
also determined to study whether treatment with ABC or TFV had any impact
on the level of naturally nucleotides and on the capacity of ddNTP to be
incorporated into the elongating viral DNA (Ray et al., 2005). When CEMss
were treated with 30 μM of ABC or 10 μM of TFV for 24 hours, no notable
changes in the endogenous nucleotides (e.g. dGTP, dATP) were observed.
When U937
WT
cells were treated with 5 μM of ABC, TFV and 3TC,
intracellular concentrations of CBV-TP, TFV-DP and 3TC-TP were 95.2%,
99.7% and 78.9%, respectively, as compared to the control (Table 3.1).
When incorporating endogenous nucleotide pools, the ratios of CBV-
TP/dGTP, TFV-DP/dATP and 3TC-TP/dCTP were 82.5%, 88.3% and 79.6%,
respectively, as compared to that in control. The endogenous nucleotide was
incorporated into the ratio by setting up a ddNTP/dNTP ratio. CBV-TP/dGTP,
TFV-DP/dATP and 3TC-TP/dCTP dropped by 17.5%, 11.7% and 20.4%,
respectively after 24 hours of treatment (Table 3.2). The results from this
85
study using the combination of 5 μM of 3TC, ABC, and TFV combined
produced a modest reduction in the corresponding 3TC-TP, CBV-TP and
TFV-DP after 24 hour treatment, which corresponded to the findings
presented by others (Ray et al., 2005). When these combinations were used
in 7 days treatments, moderate reductions were seen across all
combinations. After 7 days of treatment with the combination of 5 μM of ABC,
TFV and 3TC, IC concentration of CBV-TP, TFV-DP and 3TC-TP were
76.6%, 97.8% and 79.3%, respectively, as compared to control levels. The
ratios of CBV-TP/dGTP, TFV-DP/dATP and 3TC-TP/dCTP were 84.8% ±
12.1%, 95.7% ± 38.6% and 67.3% ± 25.1%, respectively, as compared to
control (Table 1). After one week of incubation, the reduction of 3TC-TP was
found to be the greatest. A reduction in 3TC-TP is consistent with the
virologic resistance seen in virus isolated from patients who fail this therapy.
The majority of the viruses have M184V mutations which is consistent with
lowered exposure to 3TC-TP. However a drop in 3TC-TP alone cannot alone
explain the frequency of K65R mutations alone and in combination with
M184V (Lanier et al., 2005, Parikh et al., 2007).
To further evaluate whether an interaction occurs with ABC and TFV,
this study evaluated the intracellular concentration of triphosphate
nucleotides corresponding to the nucleoside analogs under an escalating
concentration condition. TFV was evaluated at 0, 5, and 20 μM in
combination with ABC concentrations from 2 to 20 μM (Figure 1A-C). At all
ABC concentrations, the intracellular CBV-TP concentration was significantly
86
lower when combined with TFV. When the level of TFV-DP was evaluated
using 20 μM with varying concentrations of ABC, the addition of ABC
reduced the intracellular levels of TFV-DP (Figure 1D).
The data suggest that there may be a competitive inhibition between
these two NAs. TFV and ABC may share the same activation enzyme during
their respective activation pathways, and additional studies will be needed to
determine which enzymes were involved (Faletto et al., 1997). The results
from this study demonstrated that the intracellular interaction between ABC
and TFV led to a statistically significant reduction in CBV-TP formation and
CBV-TP/dGTP ratio. The reduced formation of CVB-TP indicated that the
antiviral activity of ABC was compromised; thereby HIV viral mutation K65R
would likely be selected against ABC.
An in vitro viral resistant study of the HIV patients who experienced
early virologic failure suggested that the M184V and K65R mutation strains
started from two independent pathways and merged together later (Delaunay
et al., 2005). In HIV patients who experienced early virologic failure after
receiving the combination of ABC, TFV and 3TC, viral resistant strains of
M184V were detected in many colonies as early as week 4, while K65R
alone was rare. At week 12, a number of patients had viruses with both
M184V and K65R. Our results correlated well with this clinical finding and
offered a glimpse of the potential pharmacological mechanism. The
emergence of M184V/I early on correlated well with 33% reduction in 3TC-
TP/dCTP after 7 day treatment with ABC, TFV and 3TC combination. In
87
addition, even though ABC retained its antiviral activity against M184V viral
mutants, the two-three fold reduction in viral susceptibility to ABC (Daluge et
al., 1997), coupled with the reduced formation of CBV-TP could allow
M184V/I to further select the second viral mutation, K65R, which was
resistant against both ABC and TFV.
In vitro testing by two independent research groups did not find any
antagonistic pharmacodynamic interactions between TFV and ABC (Lanier et
al., 2005, Ray et al., 2005). Both studies were conducted at concentrations
around the EC50 of these nucleoside analogs, which were in the low μM
range, under which the intracellular drug–drug interaction may not occur and
the pharmacokinetic impact was not incorporated.
In addition, a recent study showed that the in vitro antagonistic study
between NAs did not have good correlation to clinical outcomes. In vitro
testing showed that SPD754 was found to be synergistic when combined
with 3TC or FTC (Gu et al., 2006). However, SPD754 had significant
intracellular drug-drug interaction with 3TC in clinical studies. The steady
state area under the curve (AUC) and C
max
of SPD754-TP in patients treated
with SPD754 and 3TC was 4- to 6-fold lower as compared to that in patients
treated with SPD754 alone. More importantly, the IC
50
value of SPD754
against M184V HIV-1 was increased 2- to 4-fold with the presence of 3TC. At
the same time, intracellular levels of 3TC-TP were not affected by the
presence of SPD754 (Bethell et al., 2007). The results illustrated the
limitation of an in vitro antagonism pharmacodynamic interaction study using
88
plasma concentration to detect changes. Further, it illustrated that extreme
caution should be taken in the evaluation of the combination of NAs in the
treatment of HIV viral infection.
Most of the known intracellular drug-drug interactions between NAs
happened when they had competitive inhibition toward each other at their
activation enzymes. In the case of combination treatment of dual thymidine
or dual cytosine analogs, intracellular drug-drug interaction occurred and the
antiviral activities of the combination treatment was greatly compromised
(Bethell et al., 2007, Havlir et al., 2000, Hernandez-Santiago et al., 2007,
Kewn et al., 1997).
Combination therapy of AZT and stavudine (d4T) in HIV patients
revealed a drug-drug interaction between the two NRTIs, which also
compromised antiviral activities. The viral decay in patients receiving AZT
and d4T in combination was worse than patients receiving single drug
treatment; in addition, CD4 counts in patients receiving this combination
continued to decrease from baseline as compared to the CD4 cell count
rebound in patients receiving AZT or d4T alone (Havlir et al., 2000).
Noticeably, intracellular concentrations of d4T-TP in patients receiving
combination of AZT and d4T were 83% lower as compared to that in patients
receiving d4T alone.
The combination of TFV and ddI was found to have a systemic drug-
drug interaction; resulting in ddI plasma concentrations to increase 40-60% in
HIV patients receiving the combination of ddI and TFV as compared to
89
receiving ddI alone. Subsequently, ddI dose reduction was adapted,
however, the drug-drug interaction at the intracellular level still occurred as
the two may cause competitive inhibition of the anabolic enzymes during
their activation processes (Bi L., 2007, Pruvost et al., 2005).
Efflux transporters were found to play an important role in drug-drug
interaction as well, as in the case of interaction between TFV and 3TC.
Although they do not share the same anabolic enzyme during their
respective activation processes, they may share same efflux transporters
(Schuetz et al., 1999). To further determine the impact of efflux transporters,
TFV and ABC resistant variant cells were created. These cells were
characterized to assess the adaptive cellular responses when U937 and
CEMss were under constant exposure to NAs for long periods. Our results
demonstrated that long-term treatment with NAs led to cellular resistance.
The variant cells were resistant towards TFV and ABC as compared to the
wild type U937 (Figure 2). Western blot analysis further illustrated both
MRP2 and MRP4 were inducible when wild type U937 was treated with
increasing concentration of TFV. In addition, prolonged exposure to NAs also
increased expression of both MRP2 and MRP4 in both U937
ABC
and
U937
TFV
, as compared to the wild type cells (Figure 3A). These cellular
adaptive changes led to a significant reduction in intracellular ddNTP
formation, as a 40-77% reduction in 3TC-TP, TFV-DP, and CBV-TP level
was seen in the U937
TFV
cells when compared to wild-type cells (Figure 3B).
90
MRP4 is a wide spectrum nucleoside transporter and can reduce
intracellular levels of nucleosides and nucleotides (Miller, 2001, Schuetz et
al., 1999). This cellular resistance mechanism works by transporting the
monophosphate of NRTIs (ddNMP) out of cells, thereby reducing the
precursors needed for the formation of ddNTP. The over-expression of
MRP4 contributed to the cellular resistance toward treatment with NAs. This
was demonstrated in adefovir (ADV) resistant CEM (CEMr1), which has an
increased expression of MRP4 and a down regulation of adenosine kinase 2
(AK2) (Robbins et al., 1995, Schuetz et al., 1999). An analog to ADV with the
structural difference of a methyl group, TFV lacks a ribose sugar moiety and
can not be catabolized by purine nucleoside phosphorylase (PNP) (Ray et
al., 2004). In addition, cells lack an efficient way to remove TFV-DP; thereby
the intracellular half life of TFV-DP is estimated to be 160 hours, which is
about the half-life of white blood cells. Therefore, cells have to rely on efflux
transporters to remove TFV out of cells, and cellular response to the
treatment of TFV often involves an over-expression of efflux transporters. In
this study, both MRP4 and MRP2 were over-expressed in U937
TFV
, and over
expression of these efflux transporters conferred cross resistance to multiple
NAs.
The role of MRP2 with regards to nucleoside efflux activity has been
controversial. NAs were actively secreted in the pathway involving the
organic anion transporter OAT1 and OAT3 at the basolateral side, and MRP2
and MRP4 at the apical side of the proximal tubular cells. A number of
91
studies suggested that MRP2 played a major role in renal active secretion by
mediating the efflux of TFV, ADV and CDV out of the renal proximal tubular
cells (Miller, 2001). Inhibition of MRP2 and MRP4 by PIs, especially lopinavir
and ritonavir could lead to accumulation of TFV in tubular cells, thereby
caused kidney toxicity (Lam et al., 2007, Louie et al, 2005).
The correlations between intracellular levels of NAs and increased
expression of MRP2 and MRP4 cannot be directly attributed, as anabolic
enzymes may also contribute to this result. In addition, the relative function of
these two efflux transporters could not determined easily as there lacked
specific inhibitor and/or probes for each transporter.
PNP is an important enzyme regulating the balance of endogenous
nucleotide pools in cell. It breaks the glycosidic bond in guanosine and
inosine to keep the balance of endogenous nucleotide pool (Bzowska et al.,
2000). The anabolites of TFV, especially the monophosphate of TFV (TFV-
MP) are potent inhibitors of PNP (Ray et al., 2004). It was postulated that
inhibition of PNP may lead to an elevation of the endogenous purine
nucleotides dATP and dGTP, which would reduce the efficacy of TFV and
ABC and lead to early virologic failure in HIV patients receiving ABC, TFV
and 3TC in combination (Kakuda et al., 2004). In this study, no significant
change in endogenous dATP and dGTP was detected. Inhibition of PNP by
anabolites of TFV was not a critical factor here as ABC and carbovir (CBV)
were not good substrates of PNP (Faletto et al., 1997). Although, intracellular
concentrations of TFV-DP were not high enough to cause complete inhibition
92
of PNP within cells, TFV-MP was a more potent inhibitor of PNP and its
contribution to PNP inhibition could be significant, assuming TFV-MP was
one-third the level of TFV-DP in cells. The results from this suggested that
this might not cause complete inhibition of PNP; however, it may shift the
balance of endogenous nucleotide pools.
In this cell culture study, we investigated the intracellular drug-drug
interaction between NAs. The concentration of ABC and 3TC were in the
physiologically achievable range, while TFV was 5 times above the
physiological level. However, the concentration was still appropriate for the
study of an intracellular drug-drug interaction. In another study, HIV patients
were treated with tenofovir disoproxil fumarate (TDF), the esterified prodrug
of TFV. TDF was converted into TFV in plasma by esterase. In vivo study
results demonstrated that TDF had 35% bioavailability in humans as
compared to less than 3% for TFV (Kearney et al., 2004). As TDF is readily
absorbed at the intestine, it could also passively diffuse into PBMC cells.
Thereby, some of the PBMC cells may have a higher intracellular TFV
concentration.
To summarize, in this study we investigated intracellular drug-drug
interaction between ABC, TFV and 3TC in U937
WT
and CEMss cells. Our
results showed that there were intracellular drug interactions between these
three NAs, with CBV-TP and 3TC-TP formation being reduced when U937
WT

cells were treated with the triple NRTI combination, while TFV-DP formation
was less affected. The interaction became more evident with an increase in
93
treatment duration. We also investigated the underlying pharmacological
mechanism of the intracellular drug-drug interaction and found that efflux
transporters played an important role. In the TFV resistant U937
TFV
cells,
both MRP2 and MRP4 were over-expressed and their expression was
dependent on the dose and length of TFV treatment. In addition, U937
TFV

cells conferred cross-resistance towards ABC and 3TC, which contributed
significantly to the intracellular drug-drug interaction. We also estimated the
endogenous nucleotide pools and the ddNTP/dNTP ratio for each one of the
NAs. Our results suggested that intracellular drug-drug interactions
compromised the efficacy of ABC and 3TC and may directly lead to early
virologic failure in HIV patients receiving this regimen.

94
Chapter 3 References:

Bethell, R., De Muys, J., Lippens, J., Richard, A., Hamelin, B., Ren, C. &
Collins, P. (2007) In vitro interactions between apricitabine and other
deoxycytidine analogues, Antimicrob Agents Chemother, 51(8), pp.
2948-2953.
Bzowska, A., Kulikowska, E. & Shugar, D. (2000) Purine nucleoside
phosphorylases: properties, functions, and clinical aspects, Pharmacol
Ther, 88(3), pp. 349-425.
Daluge, S.M., Good, S.S., Faletto, M.B., Miller, W.H., St Clair, M.H., Boone,
L.R., Tisdale, M., Parry, N.R., Reardon, J.E., Dornsife, R.E., Averett,
D.R. & Krenitsky, T.A. (1997) 1592U89, a novel carbocyclic
nucleoside analog with potent, selective anti-human immunodeficiency
virus activity, Antimicrob Agents Chemother, 41(5), pp. 1082-1093.
Delaunay, C., Brun-Vezinet, F., Landman, R., Collin, G., Peytavin, G.,
Trylesinski, A., Flandre, P., Miller, M. & Descamps, D. (2005)
Comparative selection of the K65R and M184V/I mutations in human
immunodeficiency virus type 1-infected patients enrolled in a trial of
first-line triple-nucleoside analog therapy (Tonus IMEA 021), J Virol,
79(15), pp. 9572-9578.
Faletto, M.B., Miller, W.H., Garvey, E.P., St Clair, M.H., Daluge, S.M. &
Good, S.S. (1997) Unique intracellular activation of the potent anti-
human immunodeficiency virus agent 1592U89, Antimicrob Agents
Chemother, 41(5), pp. 1099-1107.
Gallant, J.E., Rodriguez, A.E., Weinberg, W.G., Young, B., Berger, D.S., Lim,
M.L., Liao, Q., Ross, L., Johnson, J. & Shaefer, M.S. (2005) Early
virologic nonresponse to tenofovir, abacavir, and lamivudine in HIV-
infected antiretroviral-naive subjects, J Infect Dis, 192(11), pp. 1921-
1930.

95
Galmarini, C.M., Thomas, X., Graham, K., El Jafaari, A., Cros, E., Jordheim,
L., Mackey, J.R. & Dumontet, C. (2003) Deoxycytidine kinase and cN-
II nucleotidase expression in blast cells predict survival in acute
myeloid leukaemia patients treated with cytarabine, Br J Haematol,
122(1), pp. 53-60.
Gu, Z., Allard, B., de Muys, J.M., Lippens, J., Rando, R.F., Nguyen-Ba, N.,
Ren, C., McKenna, P., Taylor, D.L. & Bethell, R.C. (2006) In vitro
antiretroviral activity and in vitro toxicity profile of SPD754, a new
deoxycytidine nucleoside reverse transcriptase inhibitor for treatment
of human immunodeficiency virus infection, Antimicrob Agents
Chemother, 50(2), pp. 625-631.
Havlir, D.V., Tierney, C., Friedland, G.H., Pollard, R.B., Smeaton, L.,
Sommadossi, J.P., Fox, L., Kessler, H., Fife, K.H. & Richman, D.D.
(2000) In vivo antagonism with zidovudine plus stavudine combination
therapy, J Infect Dis, 182(1), pp. 321-325.
Hawkins, T., Veikley, W., St Claire, R.L., 3rd, Guyer, B., Clark, N. & Kearney,
B.P. (2005) Intracellular pharmacokinetics of tenofovir diphosphate,
carbovir triphosphate, and lamivudine triphosphate in patients
receiving triple-nucleoside regimens, J Acquir Immune Defic Syndr,
39(4), pp. 406-411.
Hernandez-Santiago, B.I., Mathew, J.S., Rapp, K.L., Grier, J.P. & Schinazi,
R.F. (2007) Antiviral and cellular metabolism interactions between
Dexelvucitabine and lamivudine, Antimicrob Agents Chemother, 51(6),
pp. 2130-2135.
Jemsek, J., Hutcherson, P. & Harper, E. (February 2004) Poor virologic
responses and early emergence of resistance in treatment naive, HIV-
1 infected patients receiving an once daily triple nucleoside regimen of
didanosine, lamivudine and tenofovir [abstract 51], Presented at: The
11th Conference on Retroviruses and Opportunistic Infections, San
Francisco, CA.
Jerika T. Lam, L.T., Lucun Bi, Lee Ming Boo, Paul Beringer, David K. Ann,
Michael N. Neely and Stan G. Louie (2007) Inhibition of Multidrug
Resistance-Associated Protein 2 (MRP2) by Ritonavir Leads to
Tenofovir-Mediated Renal Dysfunction (manuscripts submitted).
96
Kakuda, T.N., Anderson, P.L. & Becker, S.L. (2004) CD4 cell decline with
didanosine and tenofovir and failure of triple nucleoside/nucleotide
regimens may be related, AIDS, 18(18), pp. 2442-2444.
Kearney, B.P., Flaherty, J.F. & Shah, J. (2004) Tenofovir disoproxil fumarate:
clinical pharmacology and pharmacokinetics, Clin Pharmacokinet,
43(9), pp. 595-612.
Kearney BP, I.E., Ebrahim, R and Cheng AK (2003) The Pharmacokinetics
of Abacavir, a purine nucleoside analog, are not Afftected by Tenofovir
DF43rd Annual Interscience Conf on Antimicrobial Agents and
Chemotherapy (Chicago, Illinois,
Kewn, S., Veal, G.J., Hoggard, P.G., Barry, M.G. & Back, D.J. (1997)
Lamivudine (3TC) phosphorylation and drug interactions in vitro,
Biochem Pharmacol, 54(5), pp. 589-595.
Khanlou, H., Yeh, V., Guyer, B. & Farthing, C. (2005) Early virologic failure in
a pilot study evaluating the efficacy of therapy containing once-daily
abacavir, lamivudine, and tenofovir DF in treatment-naive HIV-infected
patients, AIDS Patient Care STDS, 19(3), pp. 135-140.
Lanier, E.R., Hazen, R., Ross, L., Freeman, A. & Harvey, R. (2005) Lack of
antagonism between abacavir, lamivudine, and tenofovir against wild-
type and drug-resistant HIV-1, J Acquir Immune Defic Syndr, 39(5),
pp. 519-522.
Louie, SG, Lam, J., Neely, MN, Beringer, P. (2005) Multidrug resistance-
associated protein 2 (MRP2) inhibition by ritonavir increases tenofovir-
associated cytotoxicity3rd IAS Conference on HIV Pathogenesis and
Treatment. (Rio de Janiero, Brazil
Maitland, D., Moyle, G., Hand, J., Mandalia, S., Boffito, M., Nelson, M. &
Gazzard, B. (2005) Early virologic failure in HIV-1 infected subjects on
didanosine/tenofovir/efavirenz: 12-week results from a randomized
trial, AIDS, 19(11), pp. 1183-1188.

97
Mansson, E., Spasokoukotskaja, T., Sallstrom, J., Eriksson, S. & Albertioni,
F. (1999) Molecular and biochemical mechanisms of fludarabine and
cladribine resistance in a human promyelocytic cell line, Cancer Res,
59(23), pp. 5956-5963.
Miller, D.S. (2001) Nucleoside phosphonate interactions with multiple organic
anion transporters in renal proximal tubule, J Pharmacol Exp Ther,
299(2), pp. 567-574.
Moyle, G.J. & Gazzard, B.G. (2002) Switching to zidovudine plus lamivudine
plus abacavir maintains viral suppression in patients with high viral
load before antiretroviral therapy: a retrospective clinical cohort
analysis, AIDS, 16(7), pp. 1086-1087.
Parikh, U.M., Zelina, S., Sluis-Cremer, N. & Mellors, J.W. (2007) Molecular
mechanisms of bidirectional antagonism between K65R and thymidine
analog mutations in HIV-1 reverse transcriptase, AIDS, 21(11), pp.
1405-1414.
Prueksaritanont, T., Zhao, J.J., Ma, B., Roadcap, B.A., Tang, C., Qiu, Y., Liu,
L., Lin, J.H., Pearson, P.G. & Baillie, T.A. (2002) Mechanistic studies
on metabolic interactions between gemfibrozil and statins, J
Pharmacol Exp Ther, 301(3), pp. 1042-1051.
Pruvost, A., Negredo, E., Benech, H., Theodoro, F., Puig, J., Grau, E.,
Garcia, E., Molto, J., Grassi, J. & Clotet, B. (2005) Measurement of
intracellular didanosine and tenofovir phosphorylated metabolites and
possible interaction of the two drugs in human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 49(5), pp.
1907-1914.
Ray, A.S., Myrick, F., Vela, J.E., Olson, L.Y., Eisenberg, E.J., Borroto-Esodo,
K., Miller, M.D. & Fridland, A. (2005) Lack of a metabolic and antiviral
drug interaction between tenofovir, abacavir and lamivudine, Antivir
Ther, 10(3), pp. 451-457.
Ray, A.S., Olson, L. & Fridland, A. (2004) Role of purine nucleoside
phosphorylase in interactions between 2',3'-dideoxyinosine and
allopurinol, ganciclovir, or tenofovir, Antimicrob Agents Chemother,
48(4), pp. 1089-1095.
98
Robbins, B.L., Connelly, M.C., Marshall, D.R., Srinivas, R.V. & Fridland, A.
(1995) A human T lymphoid cell variant resistant to the acyclic
nucleoside phosphonate 9-(2-phosphonylmethoxyethyl)adenine
shows a unique combination of a phosphorylation defect and
increased efflux of the agent, Mol Pharmacol, 47(2), pp. 391-397.
Schuetz, J.D., Connelly, M.C., Sun, D., Paibir, S.G., Flynn, P.M., Srinivas,
R.V., Kumar, A. & Fridland, A. (1999) MRP4: A previously unidentified
factor in resistance to nucleoside-based antiviral drugs, Nat Med, 5(9),
pp. 1048-1051.
Staszewski, S., Keiser, P., Montaner, J., Raffi, F., Gathe, J., Brotas, V.,
Hicks, C., Hammer, S.M., Cooper, D., Johnson, M., Tortell, S., Cutrell,
A., Thorborn, D., Isaacs, R., Hetherington, S., Steel, H. & Spreen, W.
(2001) Abacavir-lamivudine-zidovudine vs indinavir-lamivudine-
zidovudine in antiretroviral-naive HIV-infected adults: A randomized
equivalence trial, JAMA, 285(9), pp. 1155-1163.
Weiss, J., Theile, D., Ketabi-Kiyanvash, N., Lindenmaier, H. & Haefeli, W.E.
(2007) Inhibition of MRP1/ABCC1, MRP2/ABCC2, and MRP3/ABCC3
by nucleoside, nucleotide, and non-nucleoside reverse transcriptase
inhibitors, Drug Metab Dispos, 35(3), pp. 340-344.

99
Chapter 4: Pharmacologic Mechanisms Leading to Early Virologic
Failure of Didanosine, Lamivudine, and Tenofovir Combination

Chapter 4 Abstract
The pharmacologic mechanism(s) associated with early virologic
failures in treatment-naïve HIV patients receiving triple nucleoside therapy, in
particular, the combination of didanosine (ddI), lamivudine (3TC) and
tenofovir (TFV) was investigated in this in vitro study. The intracellular
nucleotide triphosphate of the corresponding nucleoside analogs and
corresponding endogenous nucleotides were measured in CEMss cells
treated with the nucleoside(s) alone or in combination. A reduction of 10% to
22% was noted when cells were treated with the nucleosides in combination
as compared to the individual nucleosides at clinically achievable levels (5
μM). When cells were treated with higher concentrations (20 μM) of TFV and
ddI, the reduction in TFV-diphosphate (TFV-DP) and dideoxyadenosine-
triphosphate (ddATP) was approximately 25% and 40%, respectively.
The endogenous purine nucleotide levels were evaluated when cells
were treated with TFV and ddI in combination. CEMss treated with 2 to 20 μ
M of ddI did not impact endogenous levels of deoxyguanosine-triphosphate
(dGTP) or deoxyadenosine-triphosphate (dATP). However, the addition of 20
μM of TFV increased dGTP and dATP 2.4- and 2.7-fold as compared to
untreated CEMss.

100
This study also evaluated the expression of multidrug resistant
protein-2 (MRP2) and MRP4 and their possible role in relations to virologic
failures. An increased expression in MRP2 and MRP4 was seen when cells
were treated with TFV in a concentration dependent manner. The level of
MRP4 was found to correlate with tolerance toward nucleoside-induced
cytotoxicity.
Taken together this study suggested that the combination of ddI, 3TC
and TFV caused modest changes in the corresponding triphosphate
anabolites. However concentration escalation of TFV and ddI to 20 μM could
produce a 25% and 40% reduction of TFV-DP and ddATP, respectively.
When the reduction of triphosphate anabolites are coupled with increased
level in endogenous nucleotides, these changes may point to critical factors
that may lead to early virologic failures in patients receiving this triple
nucleoside regimen.

Chapter 4 Introduction
Viral resistance is a major obstacle in the management of HIV. The
focus of HIV resistance has concentrated on viral evasion mechanisms that
include viral capacity to overcome selective pressure through adaptive
responses which often evolve as viral mutations (Goudsmit et al., 1996,
Larder et al., 1995, Richman et al., 1991). Resistant viruses isolated from
patients who were treated with highly active antiretroviral therapy (HAART)
had mutations along the viral genes encoding for critical proteins or enzymes
101
such as reverse transcriptase and viral protease. However, the impact of
incomplete viral suppression as a consequence of host cell adaptation
towards xenobiotic exposure has not been fully explored in the HIV arena.
Pharmaceutical agents have both desirable and undesirable
properties. The desired pharmacological activity is normally the intended
activity, however, there are adverse effects associated with
pharmacotherapy. To reduce the impact of undesirable xenobiotics or
chemicals, humans have evolved and developed “biosensors” such as
pregnane xenobiotics receptor (PXR) and constitutive androstene receptor
(CAR), which are able to increase the transcription of efflux transporters
and/or alter metabolism of endobiotics and xenobiotics, where the ultimate
cellular response is to reduce toxic exposure (Kliewer et al., 2002, Kliewer et
al., 1999, Xie & Evans, 2001). Nowhere is this better described than the use
of nucleoside/nucleotide analogs (NAs). Although the understanding of
cellular resistance towards NAs in the treatment of HIV is still in its infancy,
there is a wealth of knowledge of cellular resistance in relationship to tumor
resistance towards chemotherapy, where nucleosides are also highly utilized.
Tumor resistance towards cytotoxic chemotherapy can emerge as a
consequence of a number of cellular adaptive mechanisms, which alone or in
combination with other processes can lead to resistance towards NA-based
therapy. One notable cellular mechanism is to alter RNA and protein
expression of metabolic enzymes required for NAs biotransformation to the
active moiety, the triphosphate (TP) metabolite (Groschel et al., 2001,
102
Groschel et al., 2002, Groschel et al., 2000). Reduction in the expression of
anabolic intracellular enzymes correlated well with reduced levels of
phosphorylated nucleotide analogs (Groschel et al., 2002, Han et al., 2004).
Another mechanism associated with drug resistance towards
nucleoside analogs has been attributed to the increased expression of efflux
transporters like MRP2 and MRP4 (Louie SG, 2005, Miller, 2001, Schuetz et
al., 1999). Cells that were grown in the presence of acyclic nucleoside
phosphonates (ANPs) resulted in an increased cellular resistance toward a
wide spectrum of antiviral agents. Resistance towards ANPs, such as PMEG
and adefovir (PMEA) was proportional to levels of MRP4 expression
(Schuetz et al., 1999).
The role of these efflux transporters in relationship to nucleoside
interactions has not been sufficiently explored. This is critical in light of rapid
clinical failure associated with triple nucleoside combinations like lamivudine
(3TC), abacavir (ABC) and tenofovir (TFV), and 3TC, didanosine (ddI) and
TFV. In this study, we explore the possible drug-drug interaction between
3TC, ddI
and TFV that may lead to an increase risk for clinical failure (Farthing
C, 2003, Gallant JE, 2003, Jemsek J, 2004). We determined the intracellular
levels of nucleotide triphosphates for the corresponding NAs in both CEM
and U937 cells when treated either alone or in two or three drug combination.
This study also evaluated the cellular response towards long-term nucleoside
treatment and the impact of efflux transport expression.
103
Chapter 4 Materials and Methods
Chemicals:
Didanosine and efavirenz were a generous gift from Bristle Myers-
Squibb (Dr. Daniel Seekins). Lamivudine was a gift from GlaxoSmithKline
(Dr. Katrina Oie), and tenofovir was obtained from NIH AIDS Research &
Reference Reagent Program.
The corresponding triphosphates: ddATP was purchased from GE
Amersham Bioscience. Adefovir-diphosphate (ADV-DP) and TFV-DP were
purchased from Movareck, CA. Carbovir-triphosphate (CBV-TP) and
lamivudine-triphosphate (3TC-TP) were a generous gift from
GlaxoSmithKline (Dr. Katrina Oie). Other triphosphate internal standards
such as ddCTP, ddGTP, 2Cl-ATP, dATP, dGTP, dCTP and dTTP were
purchased from Sigma Biologicals, St. Louis, MO.

Cells
CEMss and U937
WT
were maintained in RPMI medium, which was
supplemented with 10% FBS, 1X pyruvate, 100 μg/mL of penicillin and
streptomycin. Cells were passaged every 5 days and experiments generally
used cells from passage 5 to 25.
U937
ABC
, CEM
ABC
and U937
TFV
, CEM
TFV

are ABC and TFV resistant
variant cells, respectively. The variant cells were selected by serial passage
of CEMss or U937
WT
cells in RPMI medium containing ABC or TFV
concentration, where concentration escalated from 1 to 20 μg/mL over six
104
months. Cells were maintained in medium containing either ABC or TFV, at
37° C in 5% CO
2
at 95% humidity. Cell viability assay and Western blot
analysis were applied to characterize these variant cell lines.

Cell viability assays
Cellular viability of U937
WT
, U937
ABC
, U937
TFV
, CEMss, CEM
ABC
and,
CEM
TFV

were evaluated by incubating the cells in increasing concentration of
TFV or ABC from 0 to 1000 ng/mL for 48 hours. Cellular viability was
determined using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide (MTT) (Sigma Chemicals, MO, USA) and alomar blue dye
(Biosource Diagnostics, Sunnyvale, CA).
Following cellular treatment, 100 μL of 250 μg/mL MTT was added
into each well and incubated for 2 hours. The dark blue formazan crystals
were solubilized and liberated by the addition of 200 μL isopropanol. The
absorbance was measured using a Tecan 96-well spectrophotometer set at a
wavelength of 570 nm. These results were compared with cell viability
assays using alomar blue assays, which is a soluble non-toxic dye that does
not alter the viability of cells cultured. Cells were treated with various levels of
nucleosides and their viability was compared to untreated cells, which served
as control.

105
Intracellular interaction among TFV, ddI and 3TC
U937 cells (1.0 X 10
7
cells) or CEMss (1.5 X 10
7
cells) were treated
with TFV, ddI and 3TC alone or in combination for 24 hours. NRTIs were
dissolved in PBS first, then sterile filtered and diluted into 1000 μM in RPMI
medium. Predetermined amount of stock solution were spiked into each
flask so that the final RPMI medium contained 5 μM of NRTI alone or in
combination. Cells were placed at 37
o
C in 5% CO
2
for 24 hour treatment.
After treatment, cells were washed with ice cold PBS twice, enumerated. The
cells were centrifuged for 5 minutes at 4500 RPM; the supernatants were
discarded and 3 mL of ice cold methanol was added. The cellular pellets
were mixed with methanol to ensure complete protein precipitation. The cell
extracts were centrifuge at 4500 RPM for 15 minutes at 4 ºC. The
supernatants were transferred into clean polypropylene tubes and stored at
-80 ºC.

Concentration escalation study
CEMss (1.5 X 10
7
cells) were treated with 20 μM of TFV alone or in
combination with ddI at 2, 5, 10 and 20 μM for 24 hours at 37
o
C in 5% CO
2
in
to study the effect of dose on the drug-drug interaction. The cells were then
harvested via centrifugation at 1500 rpm for 10 minutes, and the supernatant
was aspirated. PBS was added and the cells were washed. The cellular
pellet was dissociated using ice-cold methanol and the entire cellular extract
was vigorously mixed. The cellular extract was then centrifuged at 5,000
106
rpm for 15 minutes at 4
o
C, and the supernatant was then transferred into a
clean polypropylene tube.

Cell processing for triphosphate determination
To each cellular lysate, internal standards for each of the nucleotide
analogs were added as follows: 100 μL of each of 250 ng/mL of adefovir-
diphosphate (ADV-DP), 500 ng/mL of 2-chloroadenosine triphosphate (2-Cl-
ATP), 250 ng/mL of dideoxycytidine triphosphate (ddCTP) and 250 ng/mL of
dideoxyguanosine triphosphate (ddGTP), respectively. Cellular extracts
containing all of the triphosphorylated anabolites and respective internal
standards were dried under a steady steam of filtered dry air. The residuals
were reconstituted with 1 mL of 0.1% formic acid, and further processed
using solid phase extraction (SPE).
The reconstituted cellular samples were applied onto a weak anion
exchange (WAX) SPE cartridge (Waters Oasis WAX cartridges, 3 mL, 60 mg
capacity) that was previously conditioned with 3 mL of methanol, 3 mL of
deionized water and 3 mL of 0.1% formic acid. The samples were washed
with 100 mM KCl acidified with 0.1% formic acid to elute monophosphate
and diphosphate forms of nucleoside analogs (ddNMP and ddNDP). The
columns were then washed with 0.1% formic acid solution. The triphosphate
anabolites were then eluted into collection tubes with 1 mL of alkalinized
methanol.

107
The eluents were then dried under a steady stream of compressed dry
air, and then reconstituted with 0.5 mL of 0.01% formic acid (pH 2.7)
containing 1 units of sweet potato acid phosphatase (Sigma Biologicals, St.
Louis, MO). Following 30 minutes incubation at 37ºC, all of the phosphates
were removed to yield the corresponding nucleoside analogs or phosphonate
nucleotide analogs (i.e. TFV and ADV). The dephosphorylated samples
were evaporated to dryness and reconstituted with 50 μL of 5% methanol in
deionized (DI) water and 30 μL of the samples were injected into a
LC/MS/MS system for analysis.

LC/MS/MS analysis
Intracellular concentrations of corresponding nucleosides or
nucleotides were analyzed using liquid chromatography tandem mass
spectrometry (LC-MS/MS), which consisted of an Agilent 1100 (Agilent, San
Jose, CA) HPLC system coupled to an API 3000 triple quadrupole tandem
mass spectrometer (Applied Biosystems, Foster City, CA). The analytes
were separated using a C18 reversed phase ACE (Advanced
Chromatography Technologies, Aberdeen, Scotland) with the following
dimensions, 3.0 X 50 mm, 3 μm packing. A step gradient program was used
to separate and elute the analytes. The mobile phase consisted of two
components, A and B. Component A was 100% methanol while component
B consisted of 20 mM ammonium acetate with pH adjusted to 4.5 by acetic
acid. The mobile phase program was initially set at 7% of methanol for 4
108
minutes, which was increased to 15% and held there for an additional 6
minutes. The concentration of methanol was subsequently increased to 80%
and maintained for 2 minutes, at the end of which the program returned to
initial setting of 7% methanol and allowed to equilibrate for 11 minutes. The
analytes were eluted using a flow rate was 300 μL/min.
After the HPLC separation, the efferent was introduced into the mass
spectrometer through a Turbo ion spray interface, in which all analytes and
internal standards were ionized and carried a positive charge. In addition, a
heated turbo nitrogen stream was used to evaporate solvents to increase
ionization efficiency. The mass transition ions were monitored at 248→152,
288→176, 230→112, 236→136, 274→162, 212→112, 252→152 and
302→136 for CBV, TFV, 3TC, ddA, ADV, ddC, ddG and 2-Cl A, respectively.
The mass transition ions used to determine the concentration of endogenous
nucleosides deoxycytidine (dC), deoxyadenosine (dA), deoxyguanosine (dG)
and deoxythymidine (dT) were at 228→112, 252→136, 268→152, 243→127,
respectively.

Western analysis of MRP2 and MRP4
CEMss and U937
WT
cells were exposed to 10 and 50 μM of TFV for
24 hours and cells were collected to determine protein expression of MRP2
and MRP4. In addition, protein expression of MRP2 ad MRP4 was also
determined in U937
WT
, U937
TFV
and U937
ABC
.
109
Western Blot analysis was performed following cellular membrane
isolation, and 75 μg of extracted membrane proteins were separated using
7.5% SDS-PAGE. Proteins were then transferred overnight onto
nitrocellulose PVDF membranes, which were blocked with 1x PBS-Tween
plus 5% non-fat dry milk for 1 hour at room temperature. Afterwards, the
membranes were washed twice with 1x PBS-Tween and once with 1x PBS-
Tween plus 5% dry milk. They were probed with antibodies specific for
MRP2 (Upstate, MA, USA) and MRP4 (Santa Cruz Biotechnology, CA, USA).
Subsequently, the membranes were further probed with horseradish
peroxidase-conjugated (HRP) secondary antibodies against goat IgG (Santa
Cruz Biotechnology, CA, USA). Immunoblots were developed using a
chemiluminescent detection system (Bio-Rad Labs, CA, USA).

Data analysis:
The peak area ratios of analyte to internal standard were linearly
regressed to the concentrations of analytes, using a weighting factor of
1/concentration
2
. The established linear calibration curve was used in the
determination of ddNTP in cell lysates and the results were then normalized
to the cell counts and expressed as fmole/10
6
cells.
To determine the intracellular drug-drug interactions, the cells treated
with single NRTI were used as controls. Intracellular ddNTP concentration of
110
cells receiving combination treatment was compared to that in control and
fold change were determined as follows:
Fold change = ddNTP in combination NRTIs/ddNTP in single NRTI

Statistic analysis:
Descriptive statistic analysis was applied to the fold change in ddNTP.
In addition, Wilcoxon rank test and one sample t-test were used to determine
fold changes against parity to identify any drug interactions. ANOVA were
applied to test the effect of dose on interaction between TFV and ddI. The
significance level was 0.05. All analyses were performed using GraphPad
Prism version 4.0 for Windows (GraphPad Software, San Diego, CA).

Chapter 4 Results:
Intracellular triphosphate nucleotides were quantified in both CEMss
and U937, which are lymphoblastic and monocytic leukemia cell lines,
respectively. Both cell lines were treated with 5 μM nucleoside analogs
alone or in combination for 24 hours. No statistical difference (p = 0.82) was
detectable when the intracellular triphosphate anabolite ratios were
compared between the two cells lines (data not presented). The formation of
triphosphate anabolites in cells treated with more than one nucleoside
combination were compared to cells treated with only one nucleoside. Table
4.1 is a summary of 6 separate studies performed in CEMss. The ratio of the
111
combination was compared to cells treated alone with TFV, 3TC or ddI at the
corresponding concentrations.
The dual combination treatment of TFV and 3TC demonstrated a
reduction of TFV-DP and 3TC-TP at 21.7% and 14.8% respectively, as
compared to the same cells treated with either nucleoside analogs used
alone. When ddI was combined with 3TC, there was an 18.8% reduction in
3TC-TP and a 16.6% reduction in ddATP as compared to control. In these
series of experiments where TFV was combined with ddI, there was 20.2%
and 8% reduction in TFV-DP and ddATP, respectively, as compared to the
triphosphate levels achieved when cells treated with TFV or ddI alone. The
combination of all three nucleosides at 5 μM was also analyzed, where the
average reduction of 3TC-TP, TFV-DP and ddATP was 21.6%, 9.3% and
9.6%, respectively. The median changes were also evaluated where their
112
values were similar to the mean values (Table 4.1). Since clinical failures
were observed for patients receiving TFV, ddI and EFV or nevirapine,
additional studies were undertaken to elucidate whether the combination of
TFV and ddI was a factor leading to early virologic failures in treatment naïve
HIV patients (Leon et al., 2005, Rey D, 2007). Additional experiments
explored whether higher concentrations could induce cellular adaptive
mechanisms that were not observed at lower concentrations. In these
experiments, escalating concentrations of ddI from 2 to 20 μM were
combined with 20 μM of TFV to determine whether interactions were
concentration dependent. The level of ddATP was reduced in a dose-
dependent manner, when compared to cells treated with ddI alone at the
corresponding concentrations (Figure 4.1).
In cells treated with 20 μ of TFV and 20 μM ddI, the intracellular
ddATP level was 60% of ddATP achieved in cells treated with 20 μM ddI
alone (Figure 4.1A). The intracellular TFV-DP was also reduced in
concentration dependent manner of ddI, where the drop was approximately
25% as compared to cells treated with 20 μM of TFV alone (Figure 4.1B).
The largest reduction was seen with cells that were treated with both 20 μM
TFV and 20 μM ddI. These studies demonstrated that formation of ddATP
was saturable which may be due to the possible sharing of the same
anabolic enzymes that are required to form the triphosphate moiety (Pruvost
et al., 2005).
113
114
115
Other factors can lead to resistance towards nucleoside treatment which
include cellular increases in the levels of endogenous, naturally occurring
nucleotides that may compete with nucleoside analogs for viral DNA
incorporation (Back & Berkhout, 1997). Therefore, the fold-change of
endogenous dATP and dGTP were analyzed in cells treated with ddI or TFV
alone and in combination (Figure 4.2). The addition of ddI alone did not
increase endogenous levels of dATP or dGTP as compared to the untreated
cells, even when 20 μM ddI was used. However, 20 μM TFV was able to
increase both dGTP and dATP by approximately 2.4- and 2.7-fold for dGTP
and dATP as compared to endogenous nucleotide levels in untreated cells,
respectively (Figure 4.2). When increasing concentrations of ddI were
116
combined with TFV, the fold change in both dATP and dGTP was reduced in
a concentration dependent manner as compared to 20 μM of TFV alone.
Resistance towards nucleoside therapy has been attributed to other
cellular adaptive response, which include increased expression of nucleoside
transporters (Groschel et al., 2001, Groschel et al., 2002, Groschel et al.,
2000, Han et al., 2004, Schuetz et al., 1999). To determine whether the
efflux of nucleosides may be responsible for reducing xenobiotic exposure
such as nucleoside exposure, both CEMss and U937 cells were treated with
10 and 50 μM of TFV for 24 hours, after which the levels of MRP2 and MRP4
were determined and compared to controls (Figure 4.3).
TFV is an analog of adefovir (ADV) with structure difference of a
methyl group. ADV was previously found to induce the expression of MRP4
that correlated with resistance towards to ADV and multiple nucleoside
analogs (Schuetz et al., 1999). The expression of MRP4 was increased in
both U937 and CEMss cells when exposed to TFV in a concentration-
dependent manner. Since MRP2 was recently described to be a possible
efflux transporter of TFV, this study also evaluated MRP2 expression and
response towards increased concentrations of TFV(Louie SG, 2005). Similar
to MRP4, the expression of MRP2 in CEM and U937 cells treated with TFV
was increased in a concentration dependent manner (Figure 4.3).
Since the concentration of nucleosides used in the aforementioned
experiments may be higher than those seen in humans, where the length of
exposure is short, CEM
TFV
and CEM
ABC
variant cells were used to determine
117
118
cellular response in cells chronically exposed to clinically achievable levels of
nucleosides. CEMss were maintained in growth medium where initial
concentrations of 1 μg/mL of either TFV or ABC were added to mimic
continuous exposures, as in the case of patients who receive these agents
as part of HAART. CEMss were serially passaged with either TFV or ABC to
produce CEM variants that were resistant towards TFV (CEM
TFV
) or ABC
(CEM
ABC
). After 6 months of treatment these variants were grown in the
presence of 20 μg/mL of either TFV or ABC. To test their resistance towards
nucleoside therapy, cellular viability of CEMss, CEM
TFV
and CEM
ABC
was
determined in the presence of increasing concentrations of ABC (Figure
4.4A).
The IC
50
for CEMss, CEM
ABC
and CEM
TFV
were determined to be 210,
275, and 450 μg/mL, respectively. In this study, cellular mechanisms
associated with increased resistance towards nucleoside treatment were also
evaluated using Western analysis, which correlated with increased
expression of MRP4 with increased cellular viability towards ABC treatments.
CEM
ABC
and CEM
TFV
had a 1.46- and 4.5 fold enhancement in MRP4
expression as compared to CEMss (Figure 4.4B).

Chapter 4 Discussion:
Drug-drug interactions between antiretroviral agents can increase or
reduce plasma concentrations of these antiviral compounds. Clinical drug-
drug interactions between NRTIs have been described previously, where the
119
combination of zidovudine (AZT) and stavudine (d4T) had significantly lower
antiviral activity as measured by viral decline. The viral decline in patients
receiving AZT or d4T alone was better than patients receiving the
combination of the two nucleosides (Havlir et al., 2000). The lack of viral
decline was accompanied by a progressive decline in CD4 counts when
patients were given the combination. In the same study, the comparable
treatment group consisting of the combination ddI and d4T produced viral
decline with an accompanying CD4 increase. In vitro data supported the
clinical findings, where the combination of d4T and AZT resulted in
competitive inhibition at the critical enzymes required for phosphorylation to
form the active triphosphate moiety, which reduced phosphorylation of d4T to
form d4T-TP (Hoggard et al., 1995).
Similar in vitro findings were observed in the combination of 3TC and
zalcitabine (ddC). The interaction between these two cytosine analogs
occurred in a concentration-dependent manner (Kewn et al., 1999).
Increasing concentrations of ddC, from 10 to 100 μM, demonstrated a
progressive decline in the production of 3TC-TP. When 100 μM ddC was
combined with 3TC, a 64% reduction of intracellular 3TC-TP was observed
when compared to cells treated with 3TC alone. Other nucleoside analogs
were tested, and while the addition of deoxycytosine resulted in a decrease
in 3TC-TP formation as well, other nucleosides such as d4T did not have a
statistically significant impact on the formation of 3TC-TP. This is consistent
with the data presented in this study (Table 4.1), which suggest that 3TC
120
does not significantly interact with ddI. In this study, the combination of 3TC
and TFV was also evaluated, and resulted in a reduction of 3TC-TP and
TFV-DP by 21.7% and 14.8%, respectively. The combination of 3TC, TFV
and ddI nucleosides reduced levels of their corresponding triphosphate
nucleotide by 21.6%, 9.3%, and 9.6%, respectively. This reduction by itself
may not be adequate to increase the risk of viral resistance; however other
clinical studies have demonstrated treatment-naïve patients receiving this
triple nucleoside combination have an increased risk of early virologic
failures.
A plasma pharmacokinetic drug-drug interaction was seen when ddI
and tenofovir difumarate (TDF) were combined, where ddI plasma
concentration was increased by 40-300% when TDF was added (Kearney et
al., 2005). Ray et al further evaluated the interaction between ddI and TFV in
an in vitro system, where an increase in the intracellular accumulation of ddI
in the presence of TFV was reported (Ray et al., 2004). The addition of TFV
reduced the formation of dideoxyribose and dideoxyribose-phosphate, which
are catabolites resulting from purine nucleoside phosphorylase (PNP)
mediated catabolism of ddI (Ray et al., 2004). The phosphorylated
anabolites of acyclic nucleoside like TFV-monophosphate (TFV-MP), TFV-
diphosphate (TFV-DP) and ganciclovir-monophosphate (GCV-MP) were all
found to be inhibitors of PNP which will lead to accumulation of ddI.(Ray et
al., 2004)
121
Despite an increase in intracellular ddI as a consequence of inhibition
of PNP mediated by anabolites of TFV, this study suggested that intracellular
levels of ddATP were not altered when combined with 5 μM of TFV (Table
4.1). This is consistent with data presented by Robbins et al, who had
evaluated the intracellular accumulation of the triphosphate anabolites of ddI
and TFV (Robbins et al., 2003). When the concentration of ddI was
increased to 20 μM, a slight drop of 6% to 10% of ddATP was noted when
122
combined with 5 μM TFV. Unlike the study performed by Robbins et al,
which used 5 μM TFV, this study also increased the concentration of TFV to
20 μM. In addition, increasing concentrations of ddI from 2 to 20 μM were
also used in combination with 20 μM of TFV. As shown in Figure 4.1,
intracellular concentrations of ddATP were reduced by 40% when treated
with combination of 20 μM of ddI and TFV as compared to the corresponding
treatment with ddI alone; in addition, TFV-DP were moderately reduced by
25% as compared to cells treated with 20 μM of TFV alone. This occurred
despite increased intracellular ddI concentrations as a consequence of PNP
inhibition (Figure 4.5).
One possible explanation may be decreased expression of anabolic
enzymes, such as deoxycytosine kinase and thymidine kinase (Groschel et
al., 2001, Groschel et al., 2002, Groschel et al., 2000, Han et al., 2004). In
previous studies, reduction of anabolic enzymes correlated with reduced
anabolic activity (Groschel et al., 2002, Han et al., 2004). When intracellular
anabolites were quantified, the levels of the phosphorylated nucleosides
were decreased (Han et al., 2004). Pruvost et al suggested that the
reduction of the triphosphate nucleotides may be due to competitive inhibition
of the same anabolic enzymes by ddI and TFV (Pruvost et al., 2005). The
data presented in this study supports the hypothesis that increased
concentrations of ddI result in reduction of both ddATP and TFV-DP (Figure
4.2), and further support the hypothesis that ddI and TFV anabolites use the
same anabolic enzymes such as adenylate kinase and nucleoside
123
diphosphate kinase to form the triphosphate moiety. The reduction of ddATP
and TFV-DP may be adequate to explain why virologic failures occur in some
patients receiving the triple nucleoside combination of 3TC, TFV and ddI.
However, inhibition of PNP may have other consequences with regards to
the level of natural substrates of this catabolic enzyme, adenosine and
guanosine (Bantia et al., 2003, Ray et al., 2004). The inhibition of PNP may
also alter the endogenous level of dATP and dGTP, which was confirmed in
Figure 4.2. The addition of ddI even up to 20 μM did not alter the level of
dGTP or dATP, however 20 μM TFV was able to increase endogenous
dGTP and dATP by 2.4- and 2.7-fold when endogenous levels were
compared to untreated cells. Other known PNP inhibitors such as BCX1777
have also been shown to increase endogenous purines like dGTP and dATP;
suggesting the increase of the endogenous nucleotides may be a
consequence of PNP inhibition (Bantia et al., 2003). When the concentration
of the TFV-DP were calculated, the levels were adequate to inhibit PNP as
described by Ray et al (Ray et al., 2004). The alteration of endogenous
nucleotides increased their competition with TFV-DP and ddATP for binding
to the viral reverse transcriptase, which may constitute a second mechanism
for increasing risk for the emergence of viral resistance. In addition, high
levels of endogenous nucleotides are associated with apoptosis.
Accumulation of dATP and/or dGTP may increase the risk of activating pro-
apoptotic pathways, which may explain why a drop in CD4 cells have been
observed in patients receiving ddI and TFV (Kakuda et al., 2004).

124
It is important to note that 20 μM of TFV may not be achievable in
humans without toxicity. The data presented here differs from that presented
by others who did not find that TFV altered endogenous levels of dATP and
dGTP (Ray et al., 2004), however others used 10 μM TFV instead of 20 μM
TFV used in this study. The higher TFV concentration used in this study may
allow for intracellular accumulation to reach the concentration required to
inhibit PNP, which was reported to have a Ki of 0.126 μM for TFV-MP (Ray
et al., 2004). When 5 μM of TFV was used, dATP and dGTP levels were not
significantly altered compared to untreated cells, thus suggesting the level of
intracellular of TFV-MP and TFV-DP is critical. One limitation to this study is
that the type of assay used by Ray et al differs from the one used in this
study, which utilized an indirect assay, where the isolated triphosphorylated
nucleotides are dephosphorylated and the level of the corresponding
nucleosides quantified.
This study further evaluated other mechanism(s) that may reduce the
levels of triphosphorylated nucleotides, which include increased expression
of efflux transporters, where an increase in expression will reduce the level of
precursor anabolites like ddNMPs. Schuetz et al demonstrated that CEM
cells incubated in the presence of adefovir (ADV) or PMEA were able to
develop resistance towards a wide spectrum of nucleosides (Robbins et al.,
1995, Schuetz et al., 1999). The mechanism of the resistance was in part
attributed to a down regulation of adenosine kinase 2 (AK2) activity and an
increased expression of an efflux transporter, MRP4 (Robbins et al., 1995).
125
AK2 is a critical enzyme that phosphorylates monophosphate of adenosine
analogs to form the diphosphate anabolite. Thus a down regulation of AK2
will reduce the formation of ddADP and TFV-MP
Resistance towards adefovir was also attributed to an increased
expression of MRP4, which correlated with an increased ability to tolerate
nucleosides (Figure 4.4A). This mechanism of resistance may be attributed
to the efflux of monophosphate of nucleoside analogs (ddNMPs) out of the
cell, and thus another mechanism to reduce the formation of the ddNTP. In
this study, we reported that resistance towards ABC and TFV in CEMss and
U937 cells involved with increased levels of MRP4 expression as well;
however the level of expression differs between the two-nucleoside analogs.
The expression of the MRP4 correlated well with the ability of cells to resist
the cytotoxic activity of ABC and TFV (Figure 4.4B).
More recently, the role of MRP2 was found to be important in
exporting ANPs, in particular TFV and ADV. ADV utilizes cMOAT or MRP2
as an efflux transporter, thus it would not be surprising if TFV, an analog of
ADV with structural difference of a methyl group, also utilized MRP2. There
were a number of studies that support the notion that TFV also utilizes MRP2
as a transporter of TFV (Louie SG, 2005, Mallants et al., 2005, Miller, 2001).
In this study, MRP2 expression was enhanced in the presence of TFV for
both U937 and CEMss cells (Figure 4.3). Its expression was found to be
dependent on TFV concentration, which further supported that the
expression of both MRP2 and MRP4 were important in response toward
126
nucleoside exposure. Increased expression of these efflux transporters may
also contribute to the reduced level of precursors, leading to an increased
tolerance towards nucleoside analogs (Figure 4.5). This is likely due to the
ability of MRP2 and MRP4 to reduce the levels of nucleosides and their
corresponding phosphorylated anabolite (Figure 4.5) (Han et al., 2004). Thus
the mechanism associated with viral resistance towards nucleoside
combinations in HAART may be a combination of competition for anabolic
enzymes, decreased expression of anabolic enzymes, and increased
expression of efflux transporters such as MRP2 and MRP4.
The findings with regards to changes in the active moiety of these
nucleosides were evaluated alone and in two-drug and three-drug
combinations. The reduction in the triphosphorylated nucleotide analogs was
modest when the concentrations of these nucleosides were at clinically
achievable levels (Table 4.1). However, higher concentrations of 20 μM TFV
were able to produce more dramatic changes in the both ddATP and TFV-DP
alteration (Figure 4.1). This study also investigated the role of altered
endogenous nucleotides such as dGTP and dATP when these cells were
treated with ddI alone or in combination with TFV (Figure 4.2). The increase
of dGTP and dATP by 2.4- and 2.8-fold may further reduce the effectiveness
of the nucleoside through competitive inhibition. This study presents
compelling data that a modest reduction in the active nucleoside moiety
coupled with increase in endogenous nucleotides may allow viral resistance
to emerge clinically. The reduction of the nucleotide triphosphate analogs
127
may be attributed to increased expression of both MRP2 and MRP4, and
reduced expression of the anabolic enzymes. This report suggests that
when evaluating possible nucleoside interactions, it may be necessary to
comprehensively evaluate both ddNTP and endogenous nucleotide levels.

128
Chapter 4 References:

Back, N.K. & Berkhout, B. (1997) Limiting deoxynucleoside triphosphate
concentrations emphasize the processivity defect of lamivudine-
resistant variants of human immunodeficiency virus type 1 reverse
transcriptase, Antimicrob Agents Chemother, 41(11), pp. 2484-2491.
Bantia, S., Ananth, S.L., Parker, C.D., Horn, L.L. & Upshaw, R. (2003)
Mechanism of inhibition of T-acute lymphoblastic leukemia cells by
PNP inhibitor--BCX-1777, Int Immunopharmacol, 3(6), pp. 879-887.
Farthing C, K.H., Yeh V. (2003) Early virologic failure in a pilot study
evaluating the efficacy of once daily abacavir (ABC), lamivudine (3TC)
and tenofovir DF (TDF) in treatment naive HIV-infected patients.
Presented at: Second International AIDS Society Conference on HIV
Pathogenesis and Treatment; x; Paris.
Gallant JE, R.A., Weinberg W, et al. (2003) Early non-response to tenofovir
DF (TDF) + abacavir (ABC) and lamivudine (3TC) in a randomized
trial compared to efavirenz (EFV) + ABC and 3TC: ESS30009
unplanned interim analysis43rd Annual Interscience Conference on
Antimicrobial Agents and Chemotherapy (Chicago, IL).
Goudsmit, J., De Ronde, A., Ho, D.D. & Perelson, A.S. (1996) Human
immunodeficiency virus fitness in vivo: calculations based on a single
zidovudine resistance mutation at codon 215 of reverse transcriptase,
J Virol, 70(8), pp. 5662-5664.
Groschel, B., Hover, G., Doerr, H.W. & Cinatl, J., Jr. (2001) Zidovudine (AZT)
resistance in H9 cells due to decreased TK expression is associated
with hypermethylation of TK gene, Nucleosides Nucleotides Nucleic
Acids, 20(4-7), pp. 487-492.
Groschel, B., Kaufmann, A., Hover, G., Cinatl, J., Doerr, H.W., Noordhuis, P.,
Loves, W.J., Peters, G.J. & Cinatl, J., Jr. (2002) 3'-Azido-2',3'-
dideoxythymidine induced deficiency of thymidine kinases 1, 2 and
deoxycytidine kinase in H9 T-lymphoid cells, Biochem Pharmacol,
64(2), pp. 239-246.
129
Groschel, B., Miller, V., Doerr, H.W. & Cinatl, J., Jr. (2000) Activity of cellular
thymidine kinase 1 in PBMC of HIV-1-infected patients: novel therapy
marker, Infection, 28(4), pp. 209-213.
Han, T., Fernandez, M., Sarkar, M. & Agarwal, R.P. (2004) 2', 3'-
Dideoxycytidine represses thymidine kinases 1 and 2 expression in T-
lymphoid cells, Life Sci, 74(7), pp. 835-842.
Havlir, D.V., Tierney, C., Friedland, G.H., Pollard, R.B., Smeaton, L.,
Sommadossi, J.P., Fox, L., Kessler, H., Fife, K.H. & Richman, D.D.
(2000) In vivo antagonism with zidovudine plus stavudine combination
therapy, J Infect Dis, 182(1), pp. 321-325.
Hoggard, P.G., Veal, G.J., Wild, M.J., Barry, M.G. & Back, D.J. (1995) Drug
interactions with zidovudine phosphorylation in vitro, Antimicrob
Agents Chemother, 39(6), pp. 1376-1378.
Jemsek J, H.P., Harper E. (2004) Poor virologic response and early
emergence of resistance in treatment naive, HIV-infected patients
receiving a once daily triple nucleoside regimen of didanosine,
lamivudine, and tenofovir DF11th Conference on Retroviruses and
Opportunistic Infections; (San Francisco, CA).
Kakuda, T.N., Anderson, P.L. & Becker, S.L. (2004) CD4 cell decline with
didanosine and tenofovir and failure of triple nucleoside/nucleotide
regimens may be related, AIDS, 18(18), pp. 2442-2444.
Kearney, B.P., Sayre, J.R., Flaherty, J.F., Chen, S.S., Kaul, S. & Cheng, A.K.
(2005) Drug-drug and drug-food interactions between tenofovir
disoproxil fumarate and didanosine, J Clin Pharmacol, 45(12), pp.
1360-1367.
Kewn, S., Hoggard, P.G., Henry-Mowatt, J.S., Veal, G.J., Sales, S.D., Barry,
M.G. & Back, D.J. (1999) Intracellular activation of 2',3'-dideoxyinosine
and drug interactions in vitro, AIDS Res Hum Retroviruses, 15(9), pp.
793-802.
Kliewer, S.A., Goodwin, B. & Willson, T.M. (2002) The nuclear pregnane X
receptor: a key regulator of xenobiotic metabolism, Endocr Rev, 23(5),
pp. 687-702.
130
Kliewer, S.A., Lehmann, J.M. & Willson, T.M. (1999) Orphan nuclear
receptors: shifting endocrinology into reverse, Science, 284(5415), pp.
757-760.
Larder, B.A., Kemp, S.D. & Harrigan, P.R. (1995) Potential mechanism for
sustained antiretroviral efficacy of AZT-3TC combination therapy,
Science, 269(5224), pp. 696-699.
Leon, A., Martinez, E., Mallolas, J., Laguno, M., Blanco, J.L., Pumarola, T. &
Gatell, J.M. (2005) Early virological failure in treatment-naive HIV-
infected adults receiving didanosine and tenofovir plus efavirenz or
nevirapine, AIDS, 19(2), pp. 213-215.
Louie, SG., Lam, J., Neely, MN and Beringer, P. (2005) Multidrug resistance-
associated protein2 (MRP2) inhibition by ritonavir increases tenofovir-
associated cytotoxicity3rd IAS Conference on HIV Pathogenesis and
Treatment. (Rio de Janiero, Brazil).
Mallants, R., Van Oosterwyck, K., Van Vaeck, L., Mols, R., De Clercq, E. &
Augustijns, P. (2005) Multidrug resistance-associated protein 2
(MRP2) affects hepatobiliary elimination but not the intestinal
disposition of tenofovir disoproxil fumarate and its metabolites,
Xenobiotica, 35(10-11), pp. 1055-1066.
Miller, D.S. (2001) Nucleoside phosphonate interactions with multiple organic
anion transporters in renal proximal tubule, J Pharmacol Exp Ther,
299(2), pp. 567-574.
Pruvost, A., Negredo, E., Benech, H., Theodoro, F., Puig, J., Grau, E.,
Garcia, E., Molto, J., Grassi, J. & Clotet, B. (2005) Measurement of
intracellular didanosine and tenofovir phosphorylated metabolites and
possible interaction of the two drugs in human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 49(5), pp.
1907-1914.
Ray, A.S., Olson, L. & Fridland, A. (2004) Role of purine nucleoside
phosphorylase in interactions between 2',3'-dideoxyinosine and
allopurinol, ganciclovir, or tenofovir, Antimicrob Agents Chemother,
48(4), pp. 1089-1095.
131
Rey D, S.M., Hoisey G, Meyer P, Chavanet P, Allavena C, Diemer M, May T,
Hoen B and Lang J. (2007) Early Virologic Non-response to Once
Daily Combination of Lamivudine, Tenofovir and Nevirapine in
Antiretroviral Naive HIV-infected Patients: Preliminary Results of the
Daufin Study14th Conference of Retrovirus and Opportunistic
Infections (Los Angeles, CA).
Richman, D., Shih, C.K., Lowy, I., Rose, J., Prodanovich, P., Goff, S. &
Griffin, J. (1991) Human immunodeficiency virus type 1 mutants
resistant to nonnucleoside inhibitors of reverse transcriptase arise in
tissue culture, Proc Natl Acad Sci U S A, 88(24), pp. 11241-11245.
Robbins, B.L., Connelly, M.C., Marshall, D.R., Srinivas, R.V. & Fridland, A.
(1995) A human T lymphoid cell variant resistant to the acyclic
nucleoside phosphonate 9-(2-phosphonylmethoxyethyl)adenine
shows a unique combination of a phosphorylation defect and
increased efflux of the agent, Mol Pharmacol, 47(2), pp. 391-397.
Robbins, B.L., Wilcox, C.K., Fridland, A. & Rodman, J.H. (2003) Metabolism
of tenofovir and didanosine in quiescent or stimulated human
peripheral blood mononuclear cells, Pharmacotherapy, 23(6), pp. 695-
701.
Schuetz, J.D., Connelly, M.C., Sun, D., Paibir, S.G., Flynn, P.M., Srinivas,
R.V., Kumar, A. & Fridland, A. (1999) MRP4: A previously unidentified
factor in resistance to nucleoside-based antiviral drugs, Nat Med, 5(9),
pp. 1048-1051.
Xie, W. & Evans, R.M. (2001) Orphan nuclear receptors: the exotics of
xenobiotics, J Biol Chem, 276(41), pp. 37739-37742.

132
Chapter 5: Intracellular and Plasma Interaction between Abacavir and
Tenofovir in Treatment Naïve HIV Patients

Chapter 5 Abstract
Intracellular interaction between abacavir (ABC) and tenofovir (TFV) in
treatment naïve HIV patients was investigated in this clinical study, which is
referred to as California Clinical Trials Group 584 (CCTG584). In this study,
HIV treatment naïve patients were randomized to receive either ABC or TDF
monotherapy in the first sequence, then the combination of ABC and TDF in
the second sequence. The pharmacokinetic parameters for plasma and
PBMCs were evaluated using geometric mean ratio comparing values from
sequences two with sequence one.
The results revealed an intracellular interaction between ABC and
TDF, where a two-fold increase in the steady state area under the curve
during dosing interval (AUCτ) of TFV-DP was detected. Concomitantly, a
28% reduction in CBV-TP was detected in the same cells. Individual analysis
of intracellular CBV-TP revealed two types of responses. In three out of nine
patients there were 2-10 fold increase in CBV-TP; in contrast, six patients
demonstrated a 41% (28-65%) reduction in CBV-TP.
The results of the clinical study were supported by in vitro cell culture
studies. In vitro studies combining ABC and TFV reveals a 35% reduction in
CBV-TP, while showing a 5% increase in TFV-DP in TFV resistant variant
U937
TFV
cells. The ABC and TFV interaction was found to be dependent on
the concentration of ABC and TFV. The intracellular interaction resulted in a
133
30-50% reduction in CBV-TP formation and a 30% reduction in TFV-DP. The
reduction in CBV-TP coincided with the clinical studies, however increases in
TFV-DP were not generally observed in cell culture studies.
Cell culture studies also suggested that both ABC and TFV resistant
variant cells over-express multi-drug resistant efflux transporter MRP2 and
MRP4. While the exact pharmacological mechanism is still unknown, the
interactions between ABC and TDF may involve competitive inhibition at
certain anabolic enzyme(s) resulting in reduced formation in CBV-TP.
Competitive competition could also occur between ABC and TFV at the efflux
transporter, resulting in reduced efflux of TFV, which leads to an increase in
TFV-DP.
The same competitive inhibition between ABC and TFV could also
occur in kidney cells, which would result in reduced active secretion of TFV
from plasma into urine, thereby resulting in an increase in plasma TFV
concentration in patients treated with ABC and TFV as compared to TFV
alone.

Chapter 5 Introduction
A higher than usual percentage of HIV patients experienced early
virologic failure after receiving the triple combination of tenofovir disoproxil
fumarate (TDF), abacavir (ABC) and lamivudine (3TC) in a once-daily
regimen in multiple clinical studies (Farthing C, 2003, Gallant et al., 2005).
Proposed mechanisms leading to clinical failures included weak genetic
134
barrier associated drugs for viral resistance, pharmacodynamic antagonism
and intracellular (IC) dug-drug interactions between ABC and TDF (Kakuda
et al., 2004, Landman R, 2004).
The combination of ABC and TDF has received additional scrutiny as
this combination of dual nucleoside analogs (NA) has never been evaluated
in a clinical setting, while the combinations of TDF/3TC and ABC/3TC have
demonstrated potent antiviral activities in clinical studies when co-
administered with efavirenz (EFV) (Cassetti et al., 2007, Gallant et al., 2005,
Moyle et al., 2005, Torti et al., 2005).
Results from two independent studies showed the lack of detectable
pharmacodynamic antagonism between ABC and TDF which would explain
the clinical failures (Lanier et al., 2005, Ray et al., 2005). In addition, results
from clinical pharmacokinetics studies also indicated that absence of
interaction between ABC and TDF in plasma (Kearney BP, 2003).
Intracellular interaction between ABC and TDF was also investigated
in clinical and in vitro studies (Hawkins et al., 2005, Ray et al., 2005). The
clinical study utilized a cohort of HIV patients who were viral suppressed
when treated with a nucleoside combination consisting of ABC, TDF and
3TC. ABC was withdrawn and switched to either a protease inhibitor (PI) or a
non-nucleoside reverse transcriptase inhibitor (NNRTI), to evaluate whether
the pharmacokinetics of IC CBV-TP were altered. Likewise, TDF was
withdrawn in the other cohort of patients and was started on either PI or
NNRTI to evaluate the pharmacokinetic of IC TFV-DP. The results from this
135
study did not reveal increase intracellular elimination of either TFV-DP or
CBV-TP, rather the intracellular half-life of both triphosphates were prolonged
(Hawkins et al., 2005). In addition, intracellular interaction between TDF and
ABC was also investigated in CEMss cells, where no significant interaction
between the two NAs were detected at the concentrations explored (Ray et
al., 2005).
The role of low genetic barrier towards viral resistance also did not
adequately address the clinical observation, since the combination of the two
agents demonstrated either additive or synergistic antiviral activity (Ray et
al., 2005, Lanier et al., 2005). Patients who experienced early virologic
failure often had the common viral mutations of M184V/I, K65R, or both.
M184V/I and K65R are selected by 3TC and ABC, and ABC and TFV,
respectively. M184V/I mutation was detected as early as 4 weeks after
initiation of treatment, while K65R was detected much later and often found
in viruses with M184V/I (Delaunay et al., 2005). When viruses with M184V/I
mutation were treated with the combination of 3TC, ABC and TDF, the
treatment was able to suppress the proliferation of this resistant virus
(Daluge et al., 1997, Miller, 2004); thus the suggestion that this combination
has low genetic barrier could not adequately explain what was seen clinically.
In addition, a viral dynamic study found M184V and K65R viral mutations in
different colonies and suggested that the two viral mutations emerged from
different selection pathways and thus suggest virologic failures is not a
consequence of low genetic barrier (Delaunay et al., 2005).
136
To clearly determine whether intracellular interaction between ABC
and TDF occurred, a clinical study was conducted in 20 treatment naïve HIV
patients. A pair wise two phase study design was applied to minimize the
variations between HIV patients. The primary endpoint of the study was to
determine the intracellular drug-drug interaction between ABC and TDF in
relation to plasma virologic decline. The second primary endpoint was to
detect relative antiviral activity of the two single nucleoside treatments as
compared to the combination treatment of ABC and TDF. The phase one
viral decay rate over seven day was used as biomarker (Notermans et al,
1998). The other objective of the clinical study was to determine whether
there was any drug-drug interaction at the systemic level; in addition, results
from this clinical study were compared to the in vitro cell culture studies to
verify if there was any in vitro and in vivo correlation.

Chapter 5 Materials and methods
Chemicals
Abacavir was a generous gift from GlaxoSmithKline (Dr. Katrina Oie).
The 6-methyl analog of abacavir analog (6N-ME) was a gift from Dr. Robert
Vince at University of Minnesota. Tenofovir and adefovir were purchased
from purchased from Movareck Biochemicals, Inc. (Brea, CA).
Deoxyadenosine monophosphate (dAMP), endogenous nucleotide pools,
dATP, dGTP, dCTP, dTTP and their internal standards dideoxycytidine
triphosphate (ddCTP), dideoxyguanosine triphosphate (ddGTP) and 2 chloro
137
adenosine triphosphate (2Cl-ATP) were purchased from Sigma Biologicals,
St. Louis, MO.

Study design
This was a randomized, open-label, multiple dose drug-drug
interaction study of ABC and TDF monotherapy as compared to ABC and
TDF combination treatment in treatment naïve HIV patients. After
randomization, each naïve patient started with sequence one, in which the
patient was treated with a once-daily dose of either ABC or TDF for one
week to reach steady state. On day 7 and 8, plasma and peripheral blood
mononuclear cells (PBMC) were collected. Patients had a 5 week wash out
period to clear all drugs in the system. The second sequence was initiated on
day 42, patients were treated with a once-daily regimen of TDF and ABC in
combination for one week to reach steady state. On day 48 and 49, plasma
and PBMC samples were collected. After completion of sequence two, EFV
was added to the combination treatment and the patient continued on a
once-daily regimen of EFV, ABC and TDF for an additional two weeks. On
day 63, TDF was stopped and replaced by 3TC. Patients continued on a
once-daily regimen of EFV, ABC and 3TC for an additional 46 weeks.

138
Sample size
Based on a 25% variation in the phase one viral decay rate observed
in previous clinical studies, a sample size of 20 (10 in each TDF and ABC
group) was chosen to have 80% power to detect 25% difference in the phase
one viral decay rate difference between the monotherapy and the dual
treatment group with the significance level of 0.05.
Study population
Eligible patients included men and women at least 18 years old with
documented HIV-1 infection and naïve to antiretroviral treatment. Additional
inclusion criteria included a CD4 count of ≥200 cells/ mm
3
, a HIV-1 RNA level
in plasma of > 5000 copies/mL and a Karnofsky performance score of ≥ 70
within 90 days of screening. Eligible patients were required to have an
139
absolute neutrophil count (ANC) of ≥ 750/mm
3
, a hemoglobin concentration
of ≥ 8.0 g/dL, platelet count of ≥ 50,000/mm
3
; calculated creatinine
clearance (CrCl) of > 50 mL/min (estimated by the Cockcroft-Gault equation),
AST (SGOT), ALT (SGPT), and alkaline phosphatase of ≤ 5 times of the
upper limit of the normal (ULN), and total bilirubin of ≤ 2.5 x ULN. A negative
pregnancy test was required within 30 days of study entry for women with
childbearing potential. Patients were also required to have signed informed
consent.

Drug administration and sample collection
In sequence one, HIV patients were treated with a once-daily dose of
600 mg ABC or 300 mg TDF every morning for one week; on days 7 and 8, 3
ml of plasma samples at predose, 0.5, 1, 2, 3, 6 and 24 hour post dose were
collected. In addition, 20 mL of blood were be collected at predose, 3, 6 and
24 hours post dose, and PBMC cells were harvested by applying Ficoll
hypaque technique. Patients had a 5 week wash out period to clear all drug
from system. On day 42, patients were treated with the once-daily
combination of 600 mg ABC and 300 mg TDF every morning for one week;
on days 48 and 49, plasma and PBMC samples were collected at the same
times as in sequence 1.
PBMC cells were harvested according to the sample processing
protocol. Blood samples for intracellular PK measurements were collected in
8.0 mL CPT tubes. The contents were mixed with anticoagulant thoroughly
140
by inverting the CPT tubes gently. The CPT tubes were centrifuged for 20
minutes at 1500 to 1800 RCF (Relative Centrifugal Force) at room
temperature (18-25
o
C). PBMC cells were transferred into a 50 mL conical
tube, washed with saline and enumerated. PBMC cells were then centrifuged
and supernatants were discarded. The cell pellets were lysed with 1.5 mL of
ice cold 70% methanol in water, vortexed for 3-5 minutes to ensure complete
lyses of cells. Cell extracts were stored under -80 ºC.
Blood samples in EDTA tubes were spun at room temperature (18-25

o
C) for 10 minutes at 800 RCF. Plasma samples from the top layer of each
tube were dispensed into two fresh 1.8 mL cryogenic sample tubes and
stored at -80 ºC.

Determination of plasma ABC and TFV concentration
Plasma ABC analysis
To 100 μL of plasma, 100 μL of methanol, 100 μL of 6-amino methyl
analog of abacavir (6N-ME) in methanol at 500 ng/mL and 0.7 mL methanol
were added. The samples were vortexed for a few seconds, then centrifuged
for 15 minutes at 14,000 rpm to remove any debris and proteins. 20 μL of
the supernatants were transferred into HPLC vials and mixed with 100 μL of
20 mM ammonia acetate buffer (pH 4.5). 5 μL of sample was injected into a
LC/MS/MS system for ABC analysis.

141
The LC/MS/MS was composed of an Agilent 1100 high performance
liquid chromatography (HPLC) system coupled to an Applied Biosystem API
3000 triple quadruple mass spectrometer. A reversed phase ACE C18
column (Advanced Chromatography Technologies) 3.0 X 50 mm, 3 μm
packing was used to separate the analytes. The mobile phase was
composed of 25% of methanol and 75% of 0.1% formic acid; the flow rate
was 0.3 mL/min. The mass transitions and retention times were 287→191
and 261→165, and 2.2 and 1.45 minutes for ABC and its internal standard
(6N-Me), respectively. The total analysis time for each sample was 4
minutes. The calibration curve range for ABC in human plasma was 5 to
5000 ng/mL, with a lower limit of quantification of 5 ng/mL.

Plasma TFV analysis
To 100 μL of plasma, 100 μL of methanol, 100 μL of deoxyadenosine
monophosphate (dAMP) and adefovir at 500 ng/mL and 700 mL of methanol
were added. The entire mixture was vortexed for a few seconds to mix, then
centrifuged for 15 minutes at 14000 rpm. Supernatants were transferred into
another clean Eppendorf tubes and dried under compressed air. Samples
were reconstituted with 100 μL of 5% methanol solution and 30 μL was
injected into a LC/MS/MS system for TFV analysis. Analytes were separated
on an ACE C18 column 3.0 X 50 mm, 3 μm packing. The mobile phase was
composed of 5 % of methanol and 95% of 20 mM ammonia acetate with its
pH adjusted to 4.5 by adding acetic acid. The flow-rate was set at 0.3
142
mL/min. The retention times of tenofovir, adefovir and dAMP were 4.0, 2.6
and 3.1 minutes, respectively. The mass transitions were 288→176,
274→162 and 332→156 for tenofovir, adefovir and dAMP, respectively. The
calibration curve range for tenofovir in human plasma was 5-1000 ng/mL,
with a lower limit of quantification of 5 ng/ml. The assay was linear over this
range (r
2
>0.99) and demonstrated excellent inter-day accuracy and precision
with the mean bias less than 15% and coefficients of variation less than 15%.

Determination of PBMC TFV-DP and CBV-TP concentration
PBMC TFV-DP analysis was completed by Gilead Pharmaceuticals
(Foster City, CA) using a validated direct LC/MS/MS assay. The standard
curve range was from 20 to 5000 fmole/10
6
cells, which was prepared by
using 10 μL of standard solutions into 300 μL PBMC cell lysate containing
about 5x10
6
cells along with an internal standard. Quality control (QC)
samples at 40, 400 and 4000 fmole/10
6
cells were placed at the beginning
and the end of the run. All standards and QCs were within 15% of their target
values.
IC CBV-TP concentrations in PBMC cell extracts were determined by
using a validated LC/MS/MS method. CBV-TP and the stable isotope internal
standard were extracted from 200 μL of human PBMC extracts by anion
exchange using a Waters Accell
TM
QMA solid phase extraction plate to
isolate the CBV-TP from the mono- (CBV-MP) and di-phosphorylated (CBV-
DP) metabolites. This was followed by dephosphorylation to carbovir using
143
alkaline phosphatase. Finally, high salt and enzyme levels were removed
from the sample using a Varian C
18
plate prior to LC/MS/MS analysis.
Gradient HPLC analysis was accomplished using a YMC ODS AQ 2.0 x 50
mm column with tandem mass spectrometry (MS/MS) detection utilizing
positive ion Turbo IonSpray ionization. The method has good specificity for
the quantitation of carbovir (dephosphorylated carbovir triphosphate). The
linear calibration range was from 0.05 to 10 ng/mL with correlation
coefficients better than 0.996; the lower limit of quantitation being 0.05
ng/mL. The within-run accuracy (% bias) was better than + 6%; the within-run
and between-run precision (%CV) was better than 15% and 6%, respectively.

Data analysis
Pharmacokinetics analysis
Plasma concentration of ABC and TFV were analyzed using the
noncompartmental analysis (NCA) approach in WinNonlin 5.0 (Pharsight,
Cary, NC). The linear trapezoidal rule was used to estimate area under the
curve during 24 hour dosing interval (AUCτ). The elimination rate constant
was estimated using the last three plasma concentrations at the terminal
elimination phase and the elimination half-life (t
1/2
) was calculated
accordingly in NCA. The other pharmacokinetic (PK) parameters estimated
by NCA analysis were the maximum plasma concentration (C
max
), steady
state oral apparent clearance (CLss/F) and apparent volume of distribution
(Vss/F).
144
Pharmacokinetics of TFV-DP and CBV-TP in PBMC cells were
analyzed by NCA as well. The AUCτ of TFV-DP and CBV-TP were
estimated using the linear trapezoidal rule; C
trough
were estimated by taking
the average of concentration at predose and 24 hour post dose; C
max
was the
maximum concentration in PBMC cells during the 24 hour intervals

Statistic analysis:
Descriptive statistic analysis was applied to all PK parameters of
plasma and PBMC. In addition, the ratios of these PK parameters of
sequence 2 over sequence 1 were derived and tested by using Mann-
Whitney or Wilcoxon signed rank test, when appropriate. The significance
level was 0.05. All analyses were performed using GraphPad Prism version
4.00 for Windows (GraphPad Software, San Diego, CA).
145
Chapter 5 Results:
The demographics and clinical characteristics of HIV patients in ABC
and TDF group are summarized in Table 5.1. Patients in the two groups had
similar demographics, except the ABC group was statistically younger.
Intracellular PK parameters of TFV-DP of HIV patients in the TDF group are
summarized in Table 5.2. The ratios of PK parameters in the two sequences
were compared to investigate whether the addition of ABC in the second
sequence would affect the TFV-DP formation in PBMC cells. The trough
concentration (C
trough
), C
max
and AUCτ of TFV-DP in the sequence 2 were
more likely to be higher than that in sequence 1 during the monotherapy
146
phase. The median change in TFV-DP C
trough
, C
max
and AUCτ in HIV patients
receiving ABC and TDF combination were 2-fold higher as compared to the
same patients receiving only TDF. In addition, the ratio of IC PK parameters
in sequence 2 over sequence 1 were all significantly higher than 1,
suggesting that intracellular interaction between ABC and TDF led to an
increase in TFV-DP formation in PBMC cells. Although a large variation in
every PK parameter was observed across the patient population, with the
coefficient of variation was 109% and 74.6% in C
trough
concentration at
sequence 1 and 2, respectively, the variation in ratio of IC C
trough
was
reduced to 38%.
The IC CBV-TP pharmacokinetics profile of HIV patients in the ABC
group is summarized in Figure 5.2. There were variations in the patient
population with regards to their C
trough
, C
max
and AUCτ of CBV-TP in PBMC
cells. Steady state AUCτ of CBV-TP in HIV patients receiving ABC alone
varied from 458 to 5343 fmole/10
6
cells*hr, while AUCτ of the same patients
group receiving ABC and TDF combination ranged from 675 to 6517
fmole/10
6
cells*hr. When the AUCτ ratio between the two sequences was
compared, a median 28% reduction in CBV-TP was detected when ABC was
combined with TDF, however, no statistical significance was found. Further
analysis demonstrated that there were two types of response when TDF was
added in sequence 2. One group of patients (3/9) had a geometric mean
272% increase in CBV-TP in the second sequence. In this group, one patient
had a ten-fold increase in CBV-TP. In the second group, six patients (6/9)
147
revealed a mean reduction of 41% in AUCτ of CBV-TP. The intracellular
C
max
and C
trough
of CBV-TP also supported this divergent trend, suggesting
that multiple factors contributed to the intracellular interaction between ABC
and TDF.
The plasma drug-drug interaction between ABC and TDF was also
investigated in this clinical study. The ratios of plasma pharmacokinetics
parameters of TFV in sequence 2 over sequence 1 were listed in Table 5.3.
Steady state AUCτ of TFV in plasma in the second sequence had a 22.7%
increase and the apparent oral clearance had a 14.7% reduction. Both the t-
148
test and Wilcoxon signed rank test showed that these changes were
statistically significant with p values less than 0.05. At the same time, plasma
TFV elimination half life, C
max
and apparent volume of distribution Vss/F did
not show much difference between the two sequences.
The ratios of PK parameters of ABC in sequence 2 over sequence 1
are summarized in Table 5.4. There was more variation in the PK parameter
among these patients; however, the ratios of the PK parameters of ABC in
149
sequence 2 over sequence 1 were not statistically different between the two
sequences. These results suggested absence of interaction between ABC
and TFV with regards to plasma ABC pharmacokinetics.

Chapter 5 Discussion
In this study, we have evaluated drug-drug interaction between ABC
and TDF in plasma and in PBMC cells. Twenty HIV treatment naïve patients
were randomized to receive either ABC or TDF as monotherapy. In the
second phase all of the patients were given both ABC and TDF. By
comparing the PK parameters over these two sequences, we were able to
use each patient as his or her own control and thereby effectively reduced
the inter-patient variability.
The results from this study indicated that intracellular interaction
between ABC and TDF affected both intracellular TFV-DP and CBV-TP
levels in PBMC cells. The interaction consistently increased the TFV-DP
levels in PBMC cells for all patients in TDF group except one. These
increases were detected in C
trough
, C
max
and AUCτ of TFV-DP where the
change was statistically significant with a p value less than 0.05 when tested
with Wilcoxon signed rank test and one sample t-test (Table 5.2).
The intracellular interaction of ABC and TDF also had an impact on
CBV-TP formation. After evaluating the ratios of intracellular
pharmacokinetics parameters in the two sequences, it was found that there
was a 28% and 19% reduction in the median level of AUCτ and C
max
of CBV-
150
TP, respectively, while C
trough
of IC CBV-TP had 9% increase. These
changes were not statistically significant when tested with the Wilcoxon
signed rank test. Further data analysis revealed a divergent trend in
intracellular CBV-TP. Among the nine patients completed both sequences in
the clinical study, three patients had an increase of AUCτ of CBV-TP in the
range of 2-10 fold, while six patients had an average of a 41% reduction in
AUCτ of CBV-TP. In addition, there was a 31% and 33% reduction in the
C
trough
and C
max
of CBV-TP, respectively (Figure 5.2C and D). These results
suggested that intracellular reduction of CBV-TP as a consequence of TDF
addition may increase risk for emergence of viral resistance in a large
proportion of the patients receiving this combination. Since IC CBV-TP
reduction was not found in all patients, this portion is consistent in what is
observed in clinical trials. It is important to mention that TDF-DP was
consistently elevated in patients receiving ABC and TDF.
The results in this study differ from another clinical study evaluating
whether a drug-drug interaction was seen patients receiving ABC and TDF.
However this study had a clinical bias with regards to entry criteria, where
only HIV patients treated with ABC, TDF and 3TC who were able to achieve
viral suppression were allowed to be included. In this study, the intracellular
concentration of CBV-TP and TFV-DP and their terminal elimination were
evaluated following either TDF or ABC was withdrawn and the patients were
allowed to add either NNRTIs or PIs. The results showed that the median
concentration of TFV-DP and CBV-TP did not differ significantly before and
151
after the withdrawal of TDF and ABC. However, the comparison of the mean
and median of the two sequences may not detect the intracellular interaction
between ABC and TDF, as the variation between patients may disguise the
real difference. The patients enrolled in the study had all achieved viral
suppression before the drug switch, and thereby were not representative of
patients for the intended research. This selection bias identify patients who
may have elevated level of nucleosides and thus produce the adequate
levels of ddNTPs. Lastly, the presence of a third drug in their study may
have introduced additional confounding factors, since both NNRTIs and PIs
have been shown to block efflux transporters like Pgp and MRPs.
This study was designed as a two sequences study with a washout
period in order to minimize variation between patients. When comparing
either the means or medians of IC C
max
and AUCτ of CBV-TP between the
two sequences, the slight difference between the two sequences would not
be very meaningful. When utilizing pair-wise sequential analysis, the ratios of
PK parameters were determined for HIV patients receiving ABC or TDF
mono-therapy and the combination of ABC and TDF. It was found that a
large proportion of the patients (6/9) had a significant reduction of CBV-TP in
PBMC cells. Intracellular reduction in CBV-TP of 41% was detected. This
level of reduction may be adequate to increased risk for the emergence of
viral resistance, especially viruses with M184V, K65R or both. Because of
the small size of the study, we were not able to draw a statistically significant
152
conclusion, however, we were able to detect the intricate intracellular
interaction between ABC and TDF and its potential clinical impact.
The intracellular interaction between ABC and TFV found in this study
further support what was seen the in vitro cell culture studies (Chapter 3 and
4). When CEMss cells were treated for 24 hours with the combination of ABC
from 2 to 20 μM and TFV at 5 and 20 μM, intracellular CBV-TP formation
was reduced in a TFV concentration dependent fashion, and maximum
reduction was 45% in cells treated with ABC and TFV combination as
compared to cells treated with ABC alone (Table 3.1). In addition, TFV-DP
had a 30% reduction (Table 3.2). These in vitro study results confirmed the
clinical findings and suggested a competitive interaction between ABC and
TFV. The interaction of ABC and TFV in U937
TFV
cells occurred in a similar
fashion and resulted in a 5% increase in TFV-DP and 35% reduction in CBV-
TP when treated with combination of ABC and TFV as compared to cells
treated with ABC or TFV alone (data not shown).
In vitro cell culture studies also suggested that cells developed cellular
resistance towards TFV by making certain cellular adaptive changes in the
expression of transporters and metabolic enzymes. It was found that multi-
drug resistant efflux transporters MRP2 and MRP4 were induced by TFV in a
dose and time dependent manner. These cellular adaptive changes resulted
in a 77% reduction of TFV-DP in TFV resistant variant U937
TFV
cells as
compared to wild type cells (Table 3.4A). Thereby, the increase in TFV-DP in
153
PBMCs after treatment with ABC and TFV could be offset by the cellular
resistance towards TFV, and should not be reviewed optimistically.
The over-expression of efflux transporters is one of the cellular
adaptive responses to treatment with nucleoside analogs in order to minimize
their exposure to cytotoxic agents. MRP4 was found to be part of cellular
resistance in CEMss cells towards adefovir, a phosphonate nucleoside
analog differing from TFV by a methyl group. Not surprisingly, MRP4 was
also a key efflux transporter for the cellular adaptive changes towards TFV.
In addition, results from cell culture studies demonstrated that both MRP2
and MRP4 were over-expressed in a concentration- and time-dependent
manner to the treatment of TFV (Figure 3.4).
The role of efflux transporters in cellular resistance or tolerance
toward TFV treatment was not that all surprising. TFV-DP had an elimination
half-life of 150 hours, which resembled that of lymphocytes. This suggested
that cells lacked efficient mechanisms to break down TFV-DP, and had to
rely on efflux transporters to remove TFV out of cells in order to reduce their
exposure to TFV. However, when this efflux transport pathway was
compromised, it would result in an increase of TFV-DP in PBMC cells.
Therefore, it is conceivable that ABC or its anabolites may have competitive
inhibition with TFV at the efflux transporters.
The same competitive inhibition at the efflux pathway may also result
in the 20% elevation of TFV in plasma. The competitive inhibition of efflux
transport of TFV by ABC or its anabolites may reduce the active secretion in
154
kidney, therefore causing a 17% reduction in the geometric mean of apparent
oral steady state clearance of TFV, and resulting in an increased
concentration of TFV in plasma and higher AUCτ. This study only had 10
patients in the TDF group, however it revealed a complex and intricate
interaction between ABC and TDF and possible molecular mechanisms for
this interaction.
This study also provided critical information and explanation of the
viral mutations observed in HIV patients who experienced early virologic
failure after receiving the triple combination of ABC, TDF and 3TC. In these
patients, half harbored M184V/I viral mutations and the other half had
M184V/I and K65R mutations. A study of viral dynamics also showed that
the two mutations had independent selection pathways, with M184V/I being
selected at week 4 and K65R emerging later. The M184V virus resulted in
three-fold reductions in viral susceptibility to ABC, however ABC still
maintained its antiviral activity. The reduced formation of CBV-TP due to the
intracellular interaction between ABC and TDF may provide additional
pressure on ABC and allow the HIV virus to select for the K65R mutation.
The combination of these factors gave the M184V/I virus an advantage of
selecting for the second mutation of K65R, thereby resulting in the
occurrence of M184V/I and K65R dual mutations in HIV patients.
When HIV patients were treated with the quadruple nucleoside
combination of TFV, 3TC, ABC and zidovudine (AZT), the common viral
mutations found were thymidine analog mutations (TAM) selected by AZT or
155
K65R selected by ABC and TDF. However, viral mutations carrying both
TAM and K65R were almost never observed. Viral dynamic and kinetic
studies found that the two mutations worked under completely different
mechanisms (Parikh et al., 2006, Parikh et al., 2007). The K65R mutation
reduced the incorporation of triphosphate of nucleoside analogs (ddNTP) into
viral reverse transcriptase, while TAM increased the excision rate of
monophosphate of nucleoside analogs from proviral DNA chains. The early
selection of TAM by reverse transcriptase prevented it from further protein
structural changes for the selection of K65R without compromising its own
viral replication capacity.
Taken together, intracellular interaction between nucleoside analogs
reduced CBV-TP level within PBMCs and contributed to the emergence of
viral resistance M184V, K65R and the combination of both. This interaction
was a critical factor for the development of early virologic failure in patients
receiving the combination of TFV, 3TC and ABC. Its impact was diminished
in the combination containing AZT, as the early selection of TAM prevented
the emergence of K65R as the two mutations had conflicting pathways.
These results suggested the intracellular interaction between TDF and
ABC also affected the plasma pharmacokinetics of TFV and resulted in a
modest yet statistically significant increase in TFV exposure, which may
explain why there is a significant increase in IC TFV-DP which was not
detected in the in vitro studies. This study employed a two sequence pair
wise study design, using the patients as their own control, thereby minimizing
156
the variation between patients. This also made the statistical testing and
comparison less affected by these confounding factors. These results were
derived from limited patients; thereby the significance of this finding will need
to be validated from samples from previous well-controlled, large clinical
studies.
Unfortunately, this clinical study did was not able to add lamivudine
and its role into the interactions were observed in this clinical trial. The
addition of lamivudine may cause significant changes from intracellular
ddNTP concentrations. The addition of lamivudine had changed intracellular
concentrations ddNTP in one study (Bantia et al., 2003).
This study provides some descriptive insights as what the addition of
TDF impact intracellular concentrations of CBV-TP. The mechanism(s)
associated with this finding has not been elucidated. Additional studies using
real time polymerase chain reactions to determine the expression of various
targeted genes may be informative. Other studies should include the impact
of TDF on plasma elimination of ABC and intracellular concentrations of
CBV-TP are required to provide clarity on this issue.

157
Chapter 5 References:

Bantia, S., Ananth, S.L., Parker, C.D., Horn, L.L. & Upshaw, R. (2003)
Mechanism of inhibition of T-acute lymphoblastic leukemia cells by
PNP inhibitor--BCX-1777, Int Immunopharmacol, 3(6), pp. 879-887.
Cassetti, I., Madruga, J.V., Suleiman, J.M., Etzel, A., Zhong, L., Cheng, A.K.
& Enejosa, J. (2007) The safety and efficacy of tenofovir DF in
combination with lamivudine and efavirenz through 6 years in
antiretroviral-naive HIV-1-infected patients, HIV Clin Trials, 8(3), pp.
164-172.
Daluge, S.M., Good, S.S., Faletto, M.B., Miller, W.H., St Clair, M.H., Boone,
L.R., Tisdale, M., Parry, N.R., Reardon, J.E., Dornsife, R.E., Averett,
D.R. & Krenitsky, T.A. (1997) 1592U89, a novel carbocyclic
nucleoside analog with potent, selective anti-human immunodeficiency
virus activity, Antimicrob Agents Chemother, 41(5), pp. 1082-1093.
Delaunay, C., Brun-Vezinet, F., Landman, R., Collin, G., Peytavin, G.,
Trylesinski, A., Flandre, P., Miller, M. & Descamps, D. (2005)
Comparative selection of the K65R and M184V/I mutations in human
immunodeficiency virus type 1-infected patients enrolled in a trial of
first-line triple-nucleoside analog therapy (Tonus IMEA 021), J Virol,
79(15), pp. 9572-9578.
Farthing C, K.H., Yeh V. (2003) Early virologic failure in a pilot study
evaluating the efficacy of once daily abacavir (ABC), lamivudine (3TC)
and tenofovir DF (TDF) in treatment naive HIV-infected patients.
Presented at: Second International AIDS Society Conference on HIV
Pathogenesis and Treatment; x; Paris.
Gallant, J.E., Rodriguez, A.E., Weinberg, W.G., Young, B., Berger, D.S., Lim,
M.L., Liao, Q., Ross, L., Johnson, J. & Shaefer, M.S. (2005) Early
virologic nonresponse to tenofovir, abacavir, and lamivudine in HIV-
infected antiretroviral-naive subjects, J Infect Dis, 192(11), pp. 1921-
1930.

158
Hawkins, T., Veikley, W., St Claire, R.L., 3rd, Guyer, B., Clark, N. & Kearney,
B.P. (2005) Intracellular pharmacokinetics of tenofovir diphosphate,
carbovir triphosphate, and lamivudine triphosphate in patients
receiving triple-nucleoside regimens, J Acquir Immune Defic Syndr,
39(4), pp. 406-411.
Kakuda, T.N., Anderson, P.L. & Becker, S.L. (2004) CD4 cell decline with
didanosine and tenofovir and failure of triple nucleoside/nucleotide
regimens may be related, AIDS, 18(18), pp. 2442-2444.
Kearney BP, I.E., Ebrahim, R and Cheng AK (2003) The Pharmacokinetics
of Abacavir, a purine nucleoside analog, are not Affected by Tenofovir
DF43rd Annual Interscience Conf on Antimicrobial Agents and
Chemotherapy (Chicago, Illinois,
Landman R, P.G., Descamps D, et al. (2004) Low genetic barrier to
resistance is a possible cause of early virologic failures in once-daily
regimen of abacavir, lamivudine, and tenofovir: the Tonus study11th
Conference on Retroviruses and Opportunistic Infections (San
Francisco,
Lanier, E.R., Hazen, R., Ross, L., Freeman, A. & Harvey, R. (2005) Lack of
antagonism between abacavir, lamivudine, and tenofovir against wild-
type and drug-resistant HIV-1, J Acquir Immune Defic Syndr, 39(5),
pp. 519-522.
Miller, M.D. (2004) K65R, TAMs and tenofovir, AIDS Rev, 6(1), pp. 22-33.
Moyle, G.J., DeJesus, E., Cahn, P., Castillo, S.A., Zhao, H., Gordon, D.N.,
Craig, C. & Scott, T.R. (2005) Abacavir once or twice daily combined
with once-daily lamivudine and efavirenz for the treatment of
antiretroviral-naive HIV-infected adults: results of the Ziagen Once
Daily in Antiretroviral Combination Study, J Acquir Immune Defic
Syndr, 38(4), pp. 417-425.
Parikh, U.M., Barnas, D.C., Faruki, H. & Mellors, J.W. (2006) Antagonism
between the HIV-1 reverse-transcriptase mutation K65R and
thymidine-analogue mutations at the genomic level, J Infect Dis,
194(5), pp. 651-660.
159
Parikh, U.M., Zelina, S., Sluis-Cremer, N. & Mellors, J.W. (2007) Molecular
mechanisms of bidirectional antagonism between K65R and thymidine
analog mutations in HIV-1 reverse transcriptase, AIDS, 21(11), pp.
1405-1414.
Ray, A.S., Myrick, F., Vela, J.E., Olson, L.Y., Eisenberg, E.J., Borroto-Esodo,
K., Miller, M.D. & Fridland, A. (2005) Lack of a metabolic and antiviral
drug interaction between tenofovir, abacavir and lamivudine, Antivir
Ther, 10(3), pp. 451-457.
Torti, C., Quiros-Roldon, E., Regazzi, M., Antinori, A., Patroni, A., Villani, P.,
Tirelli, V., Cologni, G., Zinzi, D., Lo Caputo, S., Perini, P. & Carosi, G.
(2005) Early virological failure after tenofovir + didanosine + efavirenz
combination in HIV-positive patients upon starting antiretroviral
therapy, Antivir Ther, 10(4), pp. 505-513.

160
Chapter 6: Summary and Future Work

Intracellular drug-drug interaction between TFV, 3TC with either ABC
or ddI, and their underlying pharmacological mechanisms have been
extensively investigated in the current study using CD4 positive lymphocytes.
The intracellular interactions between 20 μM TFV with either ABC or ddI
resulted in approximately 40% reduction in intracellular concentration of
CBV-TP or ddATP. In addition, these interactions were also dependent on
the length and concentration range of the treatment, suggesting a nature of
competitive inhibition at certain critical activation enzymes.
Intracellular drug-drug interaction between ABC and TDF in treatment
naïve HIV patients was further confirmed by a clinical study. In this clinical
study, patients receiving combination of ABC and TDF had a median of 28%
reduction in AUCτ of CBV-TP at steady state. In six of nine patients, a
geometric mean reduction of 41% in IC CBV-TP was detected in between
sequence 1 and 2. In contrast three of nine (3/9) had a more than two fold
increase in the IC CBV-TP. It is conceivable that pharmacogenomics
analysis may be necessary to determine if genetic difference may impact on
intracellular disposition of CBV-TP. The trough concentration of CBV-TP in
the subset of six patients also had a 31% reduction. The reduction in CBV-
TP may be sufficient to allow for emergence of viral resistance.
This study further investigated the underlying pharmacological
mechanism(s) that may lead to the intracellular interaction between
161
nucleoside analogs. The results indicated that long-term exposure with TFV
caused cellular adaptive changes where increased expression of efflux
transporters, MRP2 and MRP4, were detected. These cellular adaptive
changes have been previously described to reduce cellular exposure to
cytotoxic agents and resulted in significant reductions in ddNTP formation for
multiple nucleoside analogs. These transporters can efflux monophosphate
of nucleoside analogs (ddNMP), which can reduce the level of precursors to
form the triphosphate anabolites.
Cellular adaptive mechanisms can reduce nucleoside exposure by
increasing endogenous nucleotide pools. Results from this study
demonstrated in CEMss treated with 20 μM TFV was able to significantly
increase in the formation of endogenous purine, dATP and dGTP, nucleotide
pools. The increase in endogenous purines is likely due to inhibition of PNP
by anabolites of TFV (Ray et al., 2004). Elevation of dGTP and dATP in cells
treated with combination of TFV and ddI could increase competition with
ddNTP for the binding into viral DNA replication. Also, elevation of dGTP has
been shown to inhibit ribonucleotide reductase and lead T cells into
apoptosis (Arpaia et al., 2000a, Arpaia et al., 2000b). It is conceivable that
the same mechanism may lead to T-cell depletion (lymphopenia) in HIV
patients receiving the combination of TDF and ddI, even at the reduced ddI
dose.
This study illustrated the complexity of the intracellular drug-drug
interaction between nucleoside analogs and its potential clinical significance
162
on the treatment outcomes. In addition, cellular adaptive changes in
response to long-term treatment with nucleoside analogs should be carefully
and comprehensively evaluated in either HIV or cancer patients, where
nucleoside analogs have been heavily utilized.
Viral resistance is a major obstacle in the management of HIV. The
focus of HIV resistance in the past is on viral evasion mechanisms that
includes viral capacity to overcome selective pressure through adaptive
responses which often evolves as viral mutations (Goudsmit et al., 1996,
Larder et al., 1995, Parikh et al., 2007, Richman et al., 1991). Resistant
viruses isolated from patients who were treated with HAART had mutations
along the viral genes encoding for critical proteins or enzymes such as
reverse transcriptase and viral protease. However, the impact of incomplete
viral suppression as a consequence of host cell adaptation towards
xenobiotic exposure has not been fully explored in the HIV arena.
This study evaluated a few cellular factors with regards to their impact
on the combination treatment involving with nucleoside analogs. Our results
demonstrated that cellular adaptive changes in expression of enzymes
and/or transporters had a significant impact on cellular activation of
nucleoside analogs (i.e. the intracellular ddNTP concentration). In addition,
treatment with TFV may also have a significant impact on endogenous
nucleotide pools. These cellular factors and cellular adaptive changes may
have a critical impact on viral resistance to combination treatment.
163
This study established a model as how to evaluate combination
treatment involving nucleoside analogs and provided a paradigm shift in the
treatment of HIV or cancer patients, where nucleoside analogs are the
cornerstone of the combination therapy. The results in this study illustrated
the pitfalls of the conventional pharmacodynamic (PD) interaction study for
the antiviral activities study of combination of nucleoside analogs (Bethell et
al., 2007). First, the concentrations of pharmacological agents in the above
mentioned PD interaction study were significantly lower than the levels at
which significant pharmacokinetic interactions may occur between these
agents; thereby the results of PD interaction did not entail the potential
pharmacokinetic interactions between these agents (Lanier et al., 2005, Ray
et al., 2005). Alternatively, a PD interaction study in chronically exposed cells
may be better suited for this purpose as it will incorporate the cellular
adaptive changes occurring after long-term treatment with nucleoside
analogs.
The intracellular interaction between nucleoside analogs should be
evaluated in healthy volunteers prior to the initiation of treatment on HIV
patients. Since the pharmacological mechanisms for intracellular drug
interaction are likely consistent in these two populations; conducting
intracellular drug interaction study in healthy volunteers will reduce the risk of
exposure of HIV patients to suboptimal antiviral combination therapies all
together.
164
One question that was not adequately answered was that
supraphysiological concentration of TFV concentration was required produce
significant drug-drug interaction, yet patients receiving TDF, 3TC and ABC or
ddI experienced virologic failure a few weeks after initiation of the treatment.
One explanation is that TDF is the esterified prodrug of TFV, which is more
lipophillic and can be absorbed quickly through the GI tract and passively
diffuse through the cell membrane. However, TDF is converted into tenofovir
mono-isoproxil fumarate (TMF), then into TFV in endothelial cells and
hepatocytes (Kearney et al., 2004). There is currently no data with regards to
plasma concentrations of TDF. Proving the presence of TDF and/or TMF in
plasma will add tremendous value into the understanding of the
pharmacological mechanism of drug interaction between TFV and ABC or
ddI. Since TDF can penetrate cells 10 to 100 fold more efficiently than TFV
(Robbins et al., 1998).
To determine the presence of TDF, it may be necessary to collect
blood in tubes containing carboxyl esterase inhibitor during sample collection
for plasma and PBMC cells. Inhibitors such as PMSF have been shown to
inhibit carboxyl esterase, and thus prevent de-esterification in stored blood
samples.
Other critical studies may include evaluating the possible competitive
inhibition between TFV and anabolites of ABC or ddI at both guanylate
kinase and adenylate kinase, respectively. A methodology capable of
165
analyzing the diphosphate of nucleoside analogs should be developed for
this study.
On a separate front, cellular adaptive changes in enzymes and
transporters in cells receiving treatment with nucleoside analogs alone and in
combination should be investigated with real time PCR analysis. Expression
of MRP2 and MRP4 by treatment with TFV could be confirmed with this
experiment. In addition, inhibition of PNP by anabolites of TFV could trigger
cells to increase transcription of mRNA for PNP.
Additionally, it may be necessary to evaluate the impact of
pharmacogenetics on IC CBV-TP when ABC is co-administered with TFV.
Since there were two divergent CBV-TP dispositions in this study which is
consistent with what was seen in the ES20009, it may be necessary to
evaluate the impact of pharmacogenetics in relations to virologic failures.
This can be accomplished by evaluating targeted genes in patients who were
part of the ES20009.
In summary, this dissertation revealed some intriguing intracellular
drug-drug interaction between nucleoside analogs. The underlying
pharmacological mechanisms entailed the complex nature of cellular
interaction and cellular adaptive changes towards the treatment with
nucleoside analogs. This study demonstrated that the host cells played a
significant role with regard to the clinical response to combination treatment
for HIV. It may also have profound impact on the combination treatment for
cancers, in which nucleoside analogs have been applied extensively also.
166
Chapter 6 References:

Arpaia, E., Benveniste, P., Di Cristofano, A., Gu, Y., Dalal, I., Kelly, S.,
Hershfield, M., Pandolfi, P.P., Roifman, C.M. & Cohen, A. (2000a)
Mitochondrial basis for immune deficiency. Evidence from purine
nucleoside phosphorylase-deficient mice, J Exp Med, 191(12), pp.
2197-2208.
Arpaia, E., Gu, Y., Dalal, I., Kelly, S., Hershfield, M., Roifman, C.M. & Cohen,
A. (2000b) Biochemical and immunological abnormalities in purine
nucleoside phosphorylase deficient mice, Adv Exp Med Biol, 486, pp.
41-45.
Bethell, R., De Muys, J., Lippens, J., Richard, A., Hamelin, B., Ren, C. &
Collins, P. (2007) In vitro interactions between apricitabine and other
deoxycytidine analogues, Antimicrob Agents Chemother, 51(8), pp.
2948-2953.
Goudsmit, J., De Ronde, A., Ho, D.D. & Perelson, A.S. (1996) Human
immunodeficiency virus fitness in vivo: calculations based on a single
zidovudine resistance mutation at codon 215 of reverse transcriptase,
J Virol, 70(8), pp. 5662-5664.
Kearney, B.P., Flaherty, J.F. & Shah, J. (2004) Tenofovir disoproxil fumarate:
clinical pharmacology and pharmacokinetics, Clin Pharmacokinet,
43(9), pp. 595-612.

Lanier, E.R., Hazen, R., Ross, L., Freeman, A. & Harvey, R. (2005) Lack of
antagonism between abacavir, lamivudine, and tenofovir against wild-
type and drug-resistant HIV-1, J Acquir Immune Defic Syndr, 39(5),
pp. 519-522.
Larder, B.A., Kemp, S.D. & Harrigan, P.R. (1995) Potential mechanism for
sustained antiretroviral efficacy of AZT-3TC combination therapy,
Science, 269(5224), pp. 696-699.

167
Parikh, U.M., Zelina, S., Sluis-Cremer, N. & Mellors, J.W. (2007) Molecular
mechanisms of bidirectional antagonism between K65R and thymidine
analog mutations in HIV-1 reverse transcriptase, AIDS, 21(11), pp.
1405-1414.
Ray, A.S., Myrick, F., Vela, J.E., Olson, L.Y., Eisenberg, E.J., Borroto-Esodo,
K., Miller, M.D. & Fridland, A. (2005) Lack of a metabolic and antiviral
drug interaction between tenofovir, abacavir and lamivudine, Antivir
Ther, 10(3), pp. 451-457.
Ray, A.S., Olson, L. & Fridland, A. (2004) Role of purine nucleoside
phosphorylase in interactions between 2',3'-dideoxyinosine and
allopurinol, ganciclovir, or tenofovir, Antimicrob Agents Chemother,
48(4), pp. 1089-1095.
Richman, D., Shih, C.K., Lowy, I., Rose, J., Prodanovich, P., Goff, S. &
Griffin, J. (1991) Human immunodeficiency virus type 1 mutants
resistant to nonnucleoside inhibitors of reverse transcriptase arise in
tissue culture, Proc Natl Acad Sci U S A, 88(24), pp. 11241-11245.

Robbins, B.L., Srinivas, R.V., Kim, C., Bischofberger, N. & Fridland, A.
(1998) Anti-human immunodeficiency virus activity and cellular
metabolism of a potential prodrug of the acyclic nucleoside
phosphonate 9-R-(2-phosphonomethoxypropyl)adenine (PMPA),
Bis(isopropyloxymethylcarbonyl)PMPA, Antimicrob Agents
Chemother, 42(3), pp. 612-617.

168
Bibliography

Anderson, P.L., Kakuda, T.N. & Lichtenstein, K.A. (2004) The cellular
pharmacology of nucleoside- and nucleotide-analogue reverse-
transcriptase inhibitors and its relationship to clinical toxicities, Clin
Infect Dis, 38(5), pp. 743-753.
Arpaia, E., Benveniste, P., Di Cristofano, A., Gu, Y., Dalal, I., Kelly, S.,
Hershfield, M., Pandolfi, P.P., Roifman, C.M. & Cohen, A. (2000a)
Mitochondrial basis for immune deficiency. Evidence from purine
nucleoside phosphorylase-deficient mice, J Exp Med, 191(12), pp.
2197-2208.
Arpaia, E., Gu, Y., Dalal, I., Kelly, S., Hershfield, M., Roifman, C.M. & Cohen,
A. (2000b) Biochemical and immunological abnormalities in purine
nucleoside phosphorylase deficient mice, Adv Exp Med Biol, 486, pp.
41-45.
Back, N.K. & Berkhout, B. (1997) Limiting deoxynucleoside triphosphate
concentrations emphasize the processivity defect of lamivudine-
resistant variants of human immunodeficiency virus type 1 reverse
transcriptase, Antimicrob Agents Chemother, 41(11), pp. 2484-2491.
Bantia, S., Ananth, S.L., Parker, C.D., Horn, L.L. & Upshaw, R. (2003)
Mechanism of inhibition of T-acute lymphoblastic leukemia cells by
PNP inhibitor--BCX-1777, Int Immunopharmacol, 3(6), pp. 879-887.
Barreiro, P., Garcia-Benayas, T., Soriano, V. & Gallant, J. (2002)
Simplification of antiretroviral treatment--how to sustain success,
reduce toxicity and ensure adherence avoiding PI use, AIDS Rev,
4(4), pp. 233-241.
Barreiro, P. & Soriano, V. (2006) Suboptimal CD4 gains in HIV-infected
patients receiving didanosine plus tenofovir, J Antimicrob Chemother,
57(5), pp. 806-809.

169
Barrios, A., Rendon, A., Negredo, E., Barreiro, P., Garcia-Benayas, T.,
Labarga, P., Santos, J., Domingo, P., Sanchez-Conde, M., Maida, I.,
Martin-Carbonero, L., Nunez, M., Blanco, F., Clotet, B., Sambeat,
M.A., Gil, P., Gonzalez-Lahoz, J., Cooper, D. & Soriano, V. (2005)
Paradoxical CD4+ T-cell decline in HIV-infected patients with
complete virus suppression taking tenofovir and didanosine, AIDS,
19(6), pp. 569-575.
Becher, F., Pruvost, A., Goujard, C., Guerreiro, C., Delfraissy, J.F., Grassi, J.
& Benech, H. (2002) Improved method for the simultaneous
determination of d4T, 3TC and ddl intracellular phosphorylated
anabolites in human peripheral-blood mononuclear cells using high-
performance liquid chromatography/tandem mass spectrometry,
Rapid Commun Mass Spectrom, 16(6), pp. 555-565.
Bethell, R., De Muys, J., Lippens, J., Richard, A., Hamelin, B., Ren, C. &
Collins, P. (2007) In vitro interactions between apricitabine and other
deoxycytidine analogues, Antimicrob Agents Chemother, 51(8), pp.
2948-2953.
Bi L., L.J., Mordwinkin N., Russel J. and Louie S. (2007) Pharmacologic
Mechanisms Leading to Early Virologic Failure of Didanosine,
Tenofovir and Lamivudine Combination. (in preparation)
Bzowska, A., Kulikowska, E. & Shugar, D. (2000) Purine nucleoside
phosphorylases: properties, functions, and clinical aspects, Pharmacol
Ther, 88(3), pp. 349-425.
Calmy, A., Petoumenos, K., Lewden, C., Law, M., Bocquentin, F., Hesse, K.,
Cooper, D., Carr, A. & Bonnet, F. (2007) Combination antiretroviral
therapy without a nucleoside reverse transcriptase inhibitor:
experience from 334 patients in three cohorts, HIV Med, 8(3), pp. 171-
180.
Carr, A., Samaras, K., Burton, S., Law, M., Freund, J., Chisholm, D.J. &
Cooper, D.A. (1998) A syndrome of peripheral lipodystrophy,
hyperlipidaemia and insulin resistance in patients receiving HIV
protease inhibitors, AIDS, 12(7), pp. F51-58.
170
Cassetti, I., Madruga, J.V., Suleiman, J.M., Etzel, A., Zhong, L., Cheng, A.K.
& Enejosa, J. (2007) The safety and efficacy of tenofovir DF in
combination with lamivudine and efavirenz through 6 years in
antiretroviral-naive HIV-1-infected patients, HIV Clin Trials, 8(3), pp.
164-172.
Cihlar, T., Birkus, G., Greenwalt, D.E. & Hitchcock, M.J. (2002) Tenofovir
exhibits low cytotoxicity in various human cell types: comparison with
other nucleoside reverse transcriptase inhibitors, Antiviral Res, 54(1),
pp. 37-45.
Claire, R.L., 3rd (2000) Positive ion electrospray ionization tandem mass
spectrometry coupled to ion-pairing high-performance liquid
chromatography with a phosphate buffer for the quantitative analysis
of intracellular nucleotides, Rapid Commun Mass Spectrom, 14(17),
pp. 1625-1634.
Cuzin, L., Pugliese, P., Bugnon, F., Delpierre, C., Cua, E., Billaud, E.,
Massip, P., Raffi, F. & Dellamonica, P. (2005) Triple nucleoside
reverse transcriptase inhibitor- vs. nonnucleoside reverse
transcriptase inhibitor-containing regimens as first-line therapy:
efficacy and durability in a prospective cohort of French HIV-infected
patients, HIV Med, 6(6), pp. 388-395.
Daluge, S.M., Good, S.S., Faletto, M.B., Miller, W.H., St Clair, M.H., Boone,
L.R., Tisdale, M., Parry, N.R., Reardon, J.E., Dornsife, R.E., Averett,
D.R. & Krenitsky, T.A. (1997) 1592U89, a novel carbocyclic
nucleoside analog with potent, selective anti-human immunodeficiency
virus activity, Antimicrob Agents Chemother, 41(5), pp. 1082-1093.
Delahunty, T., Bushman, L. & Fletcher, C.V. (2006) Sensitive assay for
determining plasma tenofovir concentrations by LC/MS/MS, J
Chromatogr B Analyt Technol Biomed Life Sci, 830(1), pp. 6-12.
Delaunay, C., Brun-Vezinet, F., Landman, R., Collin, G., Peytavin, G.,
Trylesinski, A., Flandre, P., Miller, M. & Descamps, D. (2005)
Comparative selection of the K65R and M184V/I mutations in human
immunodeficiency virus type 1-infected patients enrolled in a trial of
first-line triple-nucleoside analog therapy (Tonus IMEA 021), J Virol,
79(15), pp. 9572-9578.
171
Faletto, M.B., Miller, W.H., Garvey, E.P., St Clair, M.H., Daluge, S.M. &
Good, S.S. (1997) Unique intracellular activation of the potent anti-
human immunodeficiency virus agent 1592U89, Antimicrob Agents
Chemother, 41(5), pp. 1099-1107.
Farthing C, K.H., Yeh V. (2003) Early virologic failure in a pilot study
evaluating the efficacy of once daily abacavir (ABC), lamivudine (3TC)
and tenofovir DF (TDF) in treatment naive HIV-infected patients,
Presented at: Second International AIDS Society Conference on HIV
Pathogenesis and Treatment; x; Paris.
FDA (2005) Guidelines for the Use of Antiretroviral Agents in HIV-1-
Infected Adults and Adolescents.
Finzi, D., Hermankova, M., Pierson, T., Carruth, L.M., Buck, C., Chaisson,
R.E., Quinn, T.C., Chadwick, K., Margolick, J., Brookmeyer, R.,
Gallant, J., Markowitz, M., Ho, D.D., Richman, D.D. & Siliciano, R.F.
(1997) Identification of a reservoir for HIV-1 in patients on highly active
antiretroviral therapy, Science, 278(5341), pp. 1295-1300.
Fletcher, C.V., Anderson, P.L., Kakuda, T.N., Schacker, T.W., Henry, K.,
Gross, C.R. & Brundage, R.C. (2002) Concentration-controlled
compared with conventional antiretroviral therapy for HIV infection,
AIDS, 16(4), pp. 551-560.
Fletcher, C.V., Kawle, S.P., Kakuda, T.N., Anderson, P.L., Weller, D.,
Bushman, L.R., Brundage, R.C. & Remmel, R.P. (2000) Zidovudine
triphosphate and lamivudine triphosphate concentration-response
relationships in HIV-infected persons, AIDS, 14(14), pp. 2137-2144.
Gallant JE, R.A., Weinberg W, et al. (2003) Early non-response to tenofovir
DF (TDF) + abacavir (ABC) and lamivudine (3TC) in a randomized
trial compared to efavirenz (EFV) + ABC and 3TC: ESS30009
unplanned interim analysis43rd Annual Interscience Conference on
Antimicrobial Agents and Chemotherapy (Chicago,

172
Gallant, J.E., Rodriguez, A.E., Weinberg, W.G., Young, B., Berger, D.S., Lim,
M.L., Liao, Q., Ross, L., Johnson, J. & Shaefer, M.S. (2005) Early
virologic nonresponse to tenofovir, abacavir, and lamivudine in HIV-
infected antiretroviral-naive subjects, J Infect Dis, 192(11), pp. 1921-
1930.
Galmarini, C.M., Thomas, X., Graham, K., El Jafaari, A., Cros, E., Jordheim,
L., Mackey, J.R. & Dumontet, C. (2003) Deoxycytidine kinase and cN-
II nucleotidase expression in blast cells predict survival in acute
myeloid leukaemia patients treated with cytarabine, Br J Haematol,
122(1), pp. 53-60.
Garcia-Benayas, T., Rendon, A.L., Rodriguez-Novoa, S., Barrios, A., Maida,
I., Blanco, F., Barreiro, P., Rivas, P., Gonzalez-Lahoz, J. & Soriano, V.
(2006) Higher risk of hyperglycemia in HIV-infected patients treated
with didanosine plus tenofovir, AIDS Res Hum Retroviruses, 22(4), pp.
333-337.
Goudsmit, J., De Ronde, A., Ho, D.D. & Perelson, A.S. (1996) Human
immunodeficiency virus fitness in vivo: calculations based on a single
zidovudine resistance mutation at codon 215 of reverse transcriptase,
J Virol, 70(8), pp. 5662-5664.
Groschel, B., Cinatl, J. & Cinatl, J., Jr. (1997) Viral and cellular factors for
resistance against antiretroviral agents, Intervirology, 40(5-6), pp. 400-
407.
Groschel, B., Hover, G., Doerr, H.W. & Cinatl, J., Jr. (2001) Zidovudine (AZT)
resistance in H9 cells due to decreased TK expression is associated
with hypermethylation of TK gene, Nucleosides Nucleotides Nucleic
Acids, 20(4-7), pp. 487-492.
Groschel, B., Kaufmann, A., Hover, G., Cinatl, J., Doerr, H.W., Noordhuis, P.,
Loves, W.J., Peters, G.J. & Cinatl, J., Jr. (2002) 3'-Azido-2',3'-
dideoxythymidine induced deficiency of thymidine kinases 1, 2 and
deoxycytidine kinase in H9 T-lymphoid cells, Biochem Pharmacol,
64(2), pp. 239-246.
173
Groschel, B., Miller, V., Doerr, H.W. & Cinatl, J., Jr. (2000) Activity of cellular
thymidine kinase 1 in PBMC of HIV-1-infected patients: novel therapy
marker, Infection, 28(4), pp. 209-213.
Gu, Z., Allard, B., de Muys, J.M., Lippens, J., Rando, R.F., Nguyen-Ba, N.,
Ren, C., McKenna, P., Taylor, D.L. & Bethell, R.C. (2006) In vitro
antiretroviral activity and in vitro toxicity profile of SPD754, a new
deoxycytidine nucleoside reverse transcriptase inhibitor for treatment
of human immunodeficiency virus infection, Antimicrob Agents
Chemother, 50(2), pp. 625-631.
Gulick, R.M., Ribaudo, H.J., Shikuma, C.M., Lustgarten, S., Squires, K.E.,
Meyer, W.A., 3rd, Acosta, E.P., Schackman, B.R., Pilcher, C.D.,
Murphy, R.L., Maher, W.E., Witt, M.D., Reichman, R.C., Snyder, S.,
Klingman, K.L. & Kuritzkes, D.R. (2004) Triple-nucleoside regimens
versus efavirenz-containing regimens for the initial treatment of HIV-1
infection, N Engl J Med, 350(18), pp. 1850-1861.
Han, T., Fernandez, M., Sarkar, M. & Agarwal, R.P. (2004) 2', 3'-
Dideoxycytidine represses thymidine kinases 1 and 2 expression in T-
lymphoid cells, Life Sci, 74(7), pp. 835-842.
Handforth, J. & Sharland, M. (2004) Triple nucleoside reverse transcriptase
inhibitor therapy in children, Paediatr Drugs, 6(3), pp. 147-159.
Harris, M., Back, D., Kewn, S., Jutha, S., Marina, R. & Montaner, J.S. (2002)
Intracellular carbovir triphosphate levels in patients taking abacavir
once a day, Aids, 16(8), pp. 1196-1197.
Havlir, D.V., Tierney, C., Friedland, G.H., Pollard, R.B., Smeaton, L.,
Sommadossi, J.P., Fox, L., Kessler, H., Fife, K.H. & Richman, D.D.
(2000) In vivo antagonism with zidovudine plus stavudine combination
therapy, J Infect Dis, 182(1), pp. 321-325.
Hawkins, T., Veikley, W., St Claire, R.L., 3rd, Guyer, B., Clark, N. & Kearney,
B.P. (2005) Intracellular pharmacokinetics of tenofovir diphosphate,
carbovir triphosphate, and lamivudine triphosphate in patients
receiving triple-nucleoside regimens, J Acquir Immune Defic Syndr,
39(4), pp. 406-411.
174
Hernandez-Santiago, B.I., Mathew, J.S., Rapp, K.L., Grier, J.P. & Schinazi,
R.F. (2007) Antiviral and cellular metabolism interactions between
Dexelvucitabine and lamivudine, Antimicrob Agents Chemother, 51(6),
pp. 2130-2135.
Hoggard, P.G., Veal, G.J., Wild, M.J., Barry, M.G. & Back, D.J. (1995) Drug
interactions with zidovudine phosphorylation in vitro, Antimicrob
Agents Chemother, 39(6), pp. 1376-1378.
Hoogewerf, M., Regez, R.M., Schouten, W.E., Weigel, H.M., Frissen, P.H. &
Brinkman, K. (2003) Change to abacavir-lamivudine-tenofovir
combination treatment in patients with HIV-1 who had complete
virological suppression, Lancet, 362(9400), pp. 1979-1980.
Jemsek, J., Hutcherson, P. & Harper, E. (February 2004) Poor virologic
responses and early emergence of resistance in treatment naive, HIV-
1 infected patients receiving a once daily triple nucleoside regimen of
didanosine, lamivudine and tenofovir [abstract 51], Presented at: The
11th Conference on Retroviruses and Opportunistic Infections, San
Francisco, CA.
Jullien, V., Treluyer, J.M., Rey, E., Jaffray, P., Krivine, A., Moachon, L., Lillo-
Le Louet, A., Lescoat, A., Dupin, N., Salmon, D., Pons, G. & Urien, S.
(2005) Population pharmacokinetics of tenofovir in human
immunodeficiency virus-infected patients taking highly active
antiretroviral therapy, Antimicrob Agents Chemother, 49(8), pp. 3361-
3366.
Kakuda, T.N., Anderson, P.L. & Becker, S.L. (2004) CD4 cell decline with
didanosine and tenofovir and failure of triple nucleoside/nucleotide
regimens may be related, AIDS, 18(18), pp. 2442-2444.
Kearney, B.P., Flaherty, J.F. & Shah, J. (2004) Tenofovir disoproxil fumarate:
clinical pharmacology and pharmacokinetics, Clin Pharmacokinet,
43(9), pp. 595-612.
Kearney BP, I.E., Ebrahim, R and Cheng AK (2003) The Pharmacokinetics
of Abacavir, a purine nucleoside analog, are not Affected by Tenofovir
DF43rd Annual Interscience Conf on Antimicrobial Agents and
Chemotherapy (Chicago, Illinois).
175
Kearney, B.P., Sayre, J.R., Flaherty, J.F., Chen, S.S., Kaul, S. & Cheng, A.K.
(2005) Drug-drug and drug-food interactions between tenofovir
disoproxil fumarate and didanosine, J Clin Pharmacol, 45(12), pp.
1360-1367.
Kewn, S., Hoggard, P.G., Henry-Mowatt, J.S., Veal, G.J., Sales, S.D., Barry,
M.G. & Back, D.J. (1999) Intracellular activation of 2',3'-dideoxyinosine
and drug interactions in vitro, AIDS Res Hum Retroviruses, 15(9), pp.
793-802.
Kewn, S., Hoggard, P.G., Sales, S.D., Johnson, M.A. & Back, D.J. (2000)
The intracellular activation of lamivudine (3TC) and determination of
2'-deoxycytidine-5'-triphosphate (dCTP) pools in the presence and
absence of various drugs in HepG2 cells, Br J Clin Pharmacol, 50(6),
pp. 597-604.
Kewn, S., Hoggard, P.G., Sales, S.D., Jones, K., Maher, B., Khoo, S.H. &
Back, D.J. (2002) Development of enzymatic assays for quantification
of intracellular lamivudine and carbovir triphosphate levels in
peripheral blood mononuclear cells from human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 46(1), pp. 135-
143.
Kewn, S., Veal, G.J., Hoggard, P.G., Barry, M.G. & Back, D.J. (1997)
Lamivudine (3TC) phosphorylation and drug interactions in vitro,
Biochem Pharmacol, 54(5), pp. 589-595.
Khanlou, H., Yeh, V., Guyer, B. & Farthing, C. (2005) Early virologic failure in
a pilot study evaluating the efficacy of therapy containing once-daily
abacavir, lamivudine, and tenofovir DF in treatment-naive HIV-infected
patients, AIDS Patient Care STDS, 19(3), pp. 135-140.
King, T., Bushman, L., Kiser, J., Anderson, P.L., Ray, M., Delahunty, T. &
Fletcher, C.V. (2006) Liquid chromatography-tandem mass
spectrometric determination of tenofovir-diphosphate in human
peripheral blood mononuclear cells, J Chromatogr B Analyt Technol
Biomed Life Sci, 843(2), pp. 147-156.
176
Kliewer, S.A., Goodwin, B. & Willson, T.M. (2002) The nuclear pregnane X
receptor: a key regulator of xenobiotic metabolism, Endocr Rev, 23(5),
pp. 687-702.
Kliewer, S.A., Lehmann, J.M. & Willson, T.M. (1999) Orphan nuclear
receptors: shifting endocrinology into reverse, Science, 284(5415), pp.
757-760.
Lam JT, T.L., Bi L, Lee MB, Beringer P, Ann D, Neely MN and Louie SG
(2007) Inhibition of Multidrug Resistance-Associated Protein 2 (MRP2)
by Ritonavir Leads to Tenofovir-Mediated Renal Dysfunction.
(submitted)

Landman R, P.G., Descamps D, et al. (2004) Low genetic barrier to
resistance is a possible cause of early virologic failures in once-daily
regimen of abacavir, lamivudine, and tenofovir: the Tonus study11th
Conference on Retroviruses and Opportunistic Infections (San
Francisco,
Lanier, E.R., Hazen, R., Ross, L., Freeman, A. & Harvey, R. (2005) Lack of
antagonism between abacavir, lamivudine, and tenofovir against wild-
type and drug-resistant HIV-1, J Acquir Immune Defic Syndr, 39(5),
pp. 519-522.
Larder, B.A., Kemp, S.D. & Harrigan, P.R. (1995) Potential mechanism for
sustained antiretroviral efficacy of AZT-3TC combination therapy,
Science, 269(5224), pp. 696-699.
Leon, A., Martinez, E., Mallolas, J., Laguno, M., Blanco, J.L., Pumarola, T. &
Gatell, J.M. (2005) Early virological failure in treatment-naive HIV-
infected adults receiving didanosine and tenofovir plus efavirenz or
nevirapine, AIDS, 19(2), pp. 213-215.
Louie SG, Lam J., Neely MN, Beringer P. (2005) Multidrug resistance-
associated protein 2 (MRP2) inhibition by ritonavir increases tenofovir-
associated cytotoxicity3rd IAS Conference on HIV Pathogenesis and
Treatment. (Rio de Janiero, Brazil).
177
Maitland, D., Moyle, G., Hand, J., Mandalia, S., Boffito, M., Nelson, M. &
Gazzard, B. (2005) Early virologic failure in HIV-1 infected subjects on
didanosine/tenofovir/efavirenz: 12-week results from a randomized
trial, AIDS, 19(11), pp. 1183-1188.
Mallants, R., Van Oosterwyck, K., Van Vaeck, L., Mols, R., De Clercq, E. &
Augustijns, P. (2005) Multidrug resistance-associated protein 2
(MRP2) affects hepatobiliary elimination but not the intestinal
disposition of tenofovir disoproxil fumarate and its metabolites,
Xenobiotica, 35(10-11), pp. 1055-1066.
Mansson, E., Spasokoukotskaja, T., Sallstrom, J., Eriksson, S. & Albertioni,
F. (1999) Molecular and biochemical mechanisms of fludarabine and
cladribine resistance in a human promyelocytic cell line, Cancer Res,
59(23), pp. 5956-5963.
Margolis, D.M., Kewn, S., Coull, J.J., Ylisastigui, L., Turner, D., Wise, H.,
Hossain, M.M., Lanier, E.R., Shaw, L.M. & Back, D. (2002) The
addition of mycophenolate mofetil to antiretroviral therapy including
abacavir is associated with depletion of intracellular deoxyguanosine
triphosphate and a decrease in plasma HIV-1 RNA, J Acquir Immune
Defic Syndr, 31(1), pp. 45-49.
Markowitz, M., Hill-Zabala, C., Lang, J., DeJesus, E., Liao, Q., Lanier, E.R.,
Davis, E.A. & Shaefer, M. (2005) Induction with
abacavir/lamivudine/zidovudine plus efavirenz for 48 weeks followed
by 48-week maintenance with abacavir/lamivudine/zidovudine alone in
antiretroviral-naive HIV-1-infected patients, J Acquir Immune Defic
Syndr, 39(3), pp. 257-264.
Miller, D.S. (2001) Nucleoside phosphonate interactions with multiple organic
anion transporters in renal proximal tubule, J Pharmacol Exp Ther,
299(2), pp. 567-574.
Miller, M.D. (2004) K65R, TAMs and tenofovir, AIDS Rev, 6(1), pp. 22-33.

178
Moyle, G.J., DeJesus, E., Cahn, P., Castillo, S.A., Zhao, H., Gordon, D.N.,
Craig, C. & Scott, T.R. (2005) Abacavir once or twice daily combined
with once-daily lamivudine and efavirenz for the treatment of
antiretroviral-naive HIV-infected adults: results of the Ziagen Once
Daily in Antiretroviral Combination Study, J Acquir Immune Defic
Syndr, 38(4), pp. 417-425.
Moyle, G.J. & Gazzard, B.G. (2002) Switching to zidovudine plus lamivudine
plus abacavir maintains viral suppression in patients with high viral
load before antiretroviral therapy: a retrospective clinical cohort
analysis, AIDS, 16(7), pp. 1086-1087.
Negredo, E., Molto, J., Burger, D., Viciana, P., Ribera, E., Paredes, R., Juan,
M., Ruiz, L., Puig, J., Pruvost, A., Grassi, J., Masmitja, E. & Clotet, B.
(2004) Unexpected CD4 cell count decline in patients receiving
didanosine and tenofovir-based regimens despite undetectable viral
load, AIDS, 18(3), pp. 459-463.
Palella, F.J., Jr., Delaney, K.M., Moorman, A.C., Loveless, M.O., Fuhrer, J.,
Satten, G.A., Aschman, D.J. & Holmberg, S.D. (1998) Declining
morbidity and mortality among patients with advanced human
immunodeficiency virus infection. HIV Outpatient Study Investigators,
N Engl J Med, 338(13), pp. 853-860.
Parikh, U.M., Barnas, D.C., Faruki, H. & Mellors, J.W. (2006) Antagonism
between the HIV-1 reverse-transcriptase mutation K65R and
thymidine-analogue mutations at the genomic level, J Infect Dis,
194(5), pp. 651-660.
Parikh, U.M., Zelina, S., Sluis-Cremer, N. & Mellors, J.W. (2007) Molecular
mechanisms of bidirectional antagonism between K65R and thymidine
analog mutations in HIV-1 reverse transcriptase, AIDS, 21(11), pp.
1405-1414.
Podzamczer, D., Ferrer, E., Gatell, J.M., Niubo, J., Dalmau, D., Leon, A.,
Knobel, H., Polo, C., Iniguez, D. & Ruiz, I. (2005) Early virological
failure with a combination of tenofovir, didanosine and efavirenz,
Antivir Ther, 10(1), pp. 171-177.
179
Prueksaritanont, T., Lin, J.H. & Baillie, T.A. (2006) Complicating factors in
safety testing of drug metabolites: kinetic differences between
generated and preformed metabolites, Toxicol Appl Pharmacol,
217(2), pp. 143-152.
Prueksaritanont, T., Zhao, J.J., Ma, B., Roadcap, B.A., Tang, C., Qiu, Y., Liu,
L., Lin, J.H., Pearson, P.G. & Baillie, T.A. (2002) Mechanistic studies
on metabolic interactions between gemfibrozil and statins, J
Pharmacol Exp Ther, 301(3), pp. 1042-1051.
Pruvost, A., Negredo, E., Benech, H., Theodoro, F., Puig, J., Grau, E.,
Garcia, E., Molto, J., Grassi, J. & Clotet, B. (2005) Measurement of
intracellular didanosine and tenofovir phosphorylated metabolites and
possible interaction of the two drugs in human immunodeficiency
virus-infected patients, Antimicrob Agents Chemother, 49(5), pp.
1907-1914.
Quercia, R., Amiel, C., Marechal Eda, S., Lebrette, M.G., Thevenet, S.,
Renaud, S., Schneider, V., Kara, A., Guessant, S. & Pialoux, G.
(2006) Virologic and immunologic efficacy of the
tenofovir/didanosine/lamivudine regimen, J Acquir Immune Defic
Syndr, 42(3), pp. 389-390.
Ray, A.S., Cihlar, T., Robinson, K.L., Tong, L., Vela, J.E., Fuller, M.D.,
Wieman, L.M., Eisenberg, E.J. & Rhodes, G.R. (2006) Mechanism of
active renal tubular efflux of tenofovir, Antimicrob Agents Chemother,
50(10), pp. 3297-3304.
Ray, A.S., Myrick, F., Vela, J.E., Olson, L.Y., Eisenberg, E.J., Borroto-Esodo,
K., Miller, M.D. & Fridland, A. (2005) Lack of a metabolic and antiviral
drug interaction between tenofovir, abacavir and lamivudine, Antivir
Ther, 10(3), pp. 451-457.
Ray, A.S., Olson, L. & Fridland, A. (2004) Role of purine nucleoside
phosphorylase in interactions between 2',3'-dideoxyinosine and
allopurinol, ganciclovir, or tenofovir, Antimicrob Agents Chemother,
48(4), pp. 1089-1095.

180
Reid, G., Wielinga, P., Zelcer, N., De Haas, M., Van Deemter, L., Wijnholds,
J., Balzarini, J. & Borst, P. (2003) Characterization of the transport of
nucleoside analog drugs by the human multidrug resistance proteins
MRP4 and MRP5, Mol Pharmacol, 63(5), pp. 1094-1103.
Rey, D., Krebs, M., Partisani, M., Hess, G., Cheneau, C., Priester, M.,
Bernard-Henry, C., de Mautort, E. & Lang, J.M. (2006) Virologic
response of zidovudine, lamivudine, and tenofovir disoproxil fumarate
combination in antiretroviral-naive HIV-1-infected patients, J Acquir
Immune Defic Syndr, 43(5), pp. 530-534.
Rey D, S.M., Hoisey G, Meyer P, Chavanet P, Allavena C, Diemer M, May T,
Hoen B and Lang J. (2007) Early Virologic Non-response to Once
Daily Combination of Lamivudine, Tenofovir and Nevirapine in
Antiretroviral Naive HIV-infected Patients: Preliminary Results of the
Daufin Study14th Conference of Retrovirus and Opportunistic
Infections (Los Angeles, CA,
Richman, D., Shih, C.K., Lowy, I., Rose, J., Prodanovich, P., Goff, S. &
Griffin, J. (1991) Human immunodeficiency virus type 1 mutants
resistant to nonnucleoside inhibitors of reverse transcriptase arise in
tissue culture, Proc Natl Acad Sci U S A, 88(24), pp. 11241-11245.
Robbins, B.L., Connelly, M.C., Marshall, D.R., Srinivas, R.V. & Fridland, A.
(1995a) A human T lymphoid cell variant resistant to the acyclic
nucleoside phosphonate 9-(2-phosphonylmethoxyethyl)adenine
shows a unique combination of a phosphorylation defect and
increased efflux of the agent, Mol Pharmacol, 47(2), pp. 391-397.
Robbins, B.L., Greenhaw, J., Connelly, M.C. & Fridland, A. (1995b)
Metabolic pathways for activation of the antiviral agent 9-(2-
phosphonylmethoxyethyl)adenine in human lymphoid cells, Antimicrob
Agents Chemother, 39(10), pp. 2304-2308.
Robbins, B.L., Srinivas, R.V., Kim, C., Bischofberger, N. & Fridland, A.
(1998) Anti-human immunodeficiency virus activity and cellular
metabolism of a potential prodrug of the acyclic nucleoside
phosphonate 9-R-(2-phosphonomethoxypropyl)adenine (PMPA),
Bis(isopropyloxymethylcarbonyl) PMPA, Antimicrob Agents
Chemother, 42(3), pp. 612-617.
181
Robbins, B.L., Wilcox, C.K., Fridland, A. & Rodman, J.H. (2003) Metabolism
of tenofovir and didanosine in quiescent or stimulated human
peripheral blood mononuclear cells, Pharmacotherapy, 23(6), pp. 695-
701.
Schuetz, J.D., Connelly, M.C., Sun, D., Paibir, S.G., Flynn, P.M., Srinivas,
R.V., Kumar, A. & Fridland, A. (1999) MRP4: A previously unidentified
factor in resistance to nucleoside-based antiviral drugs, Nat Med, 5(9),
pp. 1048-1051.
Shi, G., Wu, J.T., Li, Y., Geleziunas, R., Gallagher, K., Emm, T., Olah, T. &
Unger, S. (2002) Novel direct detection method for quantitative
determination of intracellular nucleoside triphosphates using weak
anion exchange liquid chromatography/tandem mass spectrometry,
Rapid Commun Mass Spectrom, 16(11), pp. 1092-1099.
Siliciano, J.D., Kajdas, J., Finzi, D., Quinn, T.C., Chadwick, K., Margolick,
J.B., Kovacs, C., Gange, S.J. & Siliciano, R.F. (2003) Long-term
follow-up studies confirm the stability of the latent reservoir for HIV-1
in resting CD4+ T cells, Nat Med, 9(6), pp. 727-728.
Staszewski, S., Keiser, P., Montaner, J., Raffi, F., Gathe, J., Brotas, V.,
Hicks, C., Hammer, S.M., Cooper, D., Johnson, M., Tortell, S., Cutrell,
A., Thorborn, D., Isaacs, R., Hetherington, S., Steel, H. & Spreen, W.
(2001) Abacavir-lamivudine-zidovudine vs indinavir-lamivudine-
zidovudine in antiretroviral-naive HIV-infected adults: A randomized
equivalence trial, JAMA, 285(9), pp. 1155-1163.
Torti, C., Quiros-Roldon, E., Regazzi, M., Antinori, A., Patroni, A., Villani, P.,
Tirelli, V., Cologni, G., Zinzi, D., Lo Caputo, S., Perini, P. & Carosi, G.
(2005) Early virological failure after tenofovir + didanosine + efavirenz
combination in HIV-positive patients upon starting antiretroviral
therapy, Antivir Ther, 10(4), pp. 505-513.
Trotta, M.P., Bonfigli, S., Ceccherini-Silberstein, F., Bellagamba, R., D'Arrigo,
R., Soldani, F., Zaccarelli, M., Concetta Bellocchi, M., Lorenzini, P.,
Marconi, P., Boumis, E., Forbici, F., Comandini, U.V., Tozzi, V.,
Narciso, P., Federico Perno, C. & Antinori, A. (2006) Clinical and
genotypic correlates of mutation K65R in HIV-infected patients failing
regimens not including tenofovir, J Med Virol, 78(5), pp. 535-541.
182
Vela, J.E., Olson, L.Y., Huang, A., Fridland, A. & Ray, A.S. (2007)
Simultaneous quantitation of the nucleotide analog adefovir, its
phosphorylated anabolites and 2'-deoxyadenosine triphosphate by
ion-pairing LC/MS/MS, J Chromatogr B Analyt Technol Biomed Life
Sci, 848(2), pp. 335-343.
Veltkamp, S.A., Hillebrand, M.J., Rosing, H., Jansen, R.S., Wickremsinhe,
E.R., Perkins, E.J., Schellens, J.H. & Beijnen, J.H. (2006) Quantitative
analysis of gemcitabine triphosphate in human peripheral blood
mononuclear cells using weak anion-exchange liquid chromatography
coupled with tandem mass spectrometry, J Mass Spectrom, 41(12),
pp. 1633-1642.
Weiss, J., Theile, D., Ketabi-Kiyanvash, N., Lindenmaier, H. & Haefeli, W.E.
(2007) Inhibition of MRP1/ABCC1, MRP2/ABCC2, and MRP3/ABCC3
by nucleoside, nucleotide, and non-nucleoside reverse transcriptase
inhibitors, Drug Metab Dispos, 35(3), pp. 340-344.
White, K.L., Chen, J.M., Feng, J.Y., Margot, N.A., Ly, J.K., Ray, A.S.,
Macarthur, H.L., McDermott, M.J., Swaminathan, S. & Miller, M.D.
(2006) The K65R reverse transcriptase mutation in HIV-1 reverses the
excision phenotype of zidovudine resistance mutations, Antivir Ther,
11(2), pp. 155-163.
Wirden, M., Marcelin, A.G., Tubiana, R., Valantin, M.A., Ghosn, J., Duvivier,
C., Dominguez, S., Paris, L., Agher, R., Peytavin, G., Katlama, C. &
Calvez, V. (2004) Virologic outcome after switching from a nucleoside
reverse transcriptase inhibitor to tenofovir in patients with
undetectable HIV-1 RNA plasma level, J Acquir Immune Defic Syndr,
36(3), pp. 876-878.
Xie, W. & Evans, R.M. (2001) Orphan nuclear receptors: the exotics of
xenobiotics, J Biol Chem, 276(41), pp. 37739-37742.

Abstract (if available)

Abstract High level of virologic failure was observed in HIV patients receiving combinations of tenofovir (TFV), lamivudine (3TC) combined with either abacavir (ABC) or didanosine (ddI). To investigate the pharmacologic mechanisms involve with the virologic failures, a comprehensive study was undertaken to evaluate the intracellular concentration of the active moiety (ddNTP) of these respective nucleoside analogs.

Linked assets

University of Southern California Dissertations and Theses

Asset Metadata

Creator Bi, Lucun (author)

Core Title Intracellular drug-drug interaction between nucleoside analogs leads to early virologic failure in HIV patients receiving triple nucleoside combinations of tenofovir, lamivudine and abacavir or d...

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Pharmaceutical Sciences

Publication Date 03/15/2008

Defense Date 11/29/2007

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

Tag ABC,cellular adaptive response,ddI,ddNTP,efflux transporters,endogenous nucleotide pools,intracellular drug-drug interaction,MRP2,MRP4,NRTI,nucleoside analog,OAI-PMH Harvest,PNP,TFV

Language English

Advisor Burckart, Gilbert J. (committee chair), D'Argenio, David (committee member), Louie, Stan (committee member), Shen, Wei-Chiang (committee member), Wang, Clay C. C. (committee member)

Creator Email lucunbi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1051

Unique identifier UC190360

Identifier etd-Bi-20080315 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-47592 (legacy record id),usctheses-m1051 (legacy record id)

Legacy Identifier etd-Bi-20080315.pdf

Dmrecord 47592

Document Type Dissertation

Rights Bi, Lucun

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu