University of Southern California Dissertations and Theses

The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes

(USC Thesis Other)

The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1
THE MOLECULAR ADAPTATION OF TRICHODESMIUM TO LONG-TERM CO
2
-
SELECTION UNDER MULTIPLE NUTRIENT LIMITATION REGIMES

Nathan G. Walworth

A Dissertation Submitted to the
University of Southern California in Partial Fulfillment of
the Requirements for the degree of Doctor of Philosophy (Biology)

Department of Biological Sciences
Marine Environmental Biology/Marine Biology and Biological Oceanography
University of Southern California

May 2016

Approved by

Advisory Committee

Chair
David A. Hutchins Eric A. Webb Kenneth H. Nealson

Ian M. Ehrenreich Moh Y. El-Naggar

2
Acknowledgements

First and foremost, I would like to acknowledge my parents, Stephen and Daisy
Walworth, who have shown me nothing but love and support throughout all my
endeavors while giving me the freedom to follow my heart. I would like to thank the rest
of my family, friends, and love (Karen Momper) who have empowered me every step of
the way without whom I could never have come this far. Thank you especially to my
advisors, Eric Webb and David Hutchins, for the (seemingly) endless debates and
discussions about essentially everything one can think of. Through their co-advising, I
not only grew as a scientist but just as importantly, as a person. Our collaboration taught
me patience and compromise, which I will take with me moving forward. Oh yea, and
thank you for sending me all over the world, that was really nice of you two. I would also
especially like to thank Fei-Xue Fu for her critical mentorship and unparalleled culturing
skills without which this dissertation would not have been possible. I believe our
collaboration together in the lab enabled us to take our research to heights not possible
without the other. It was truly a pleasure working with her. I would also like to extend
thanks to Ken Nealson who invested in me both as a scientist and as a person and
welcomed me into opportunities I had only dreamed of. His value system changed so
much for me as he always brought me back to the true reasons I set out for a career in the
environmental sciences. I will sorely miss my bi-monthly therapy session in his office.
Similarly, I’d like to thank the Hutchins and Webb lab personnel particularly Avery
Tatters, Mike Lee, and Zhi (Spider) Zhu for their attentive help in lab and insightful
discussions in my research. Thank you to the faculty of MBBO for always taking the time
to entertain my thoughts and questions and for treating me as a valued colleague. A
special thanks also goes out to the staff of MBBO in the front office who answered every
odd and logistical question I had about the whole dissertation process at USC. Thank you
to the universities and institutes who have given me these opportunities over the years
including UC Santa Cruz, Lund University, the J. Craig Venter Institute, and USC as well
as my primary funding body, the National Science Foundation. As I move forward to try
and better this world, my ultimate mission for myself will forever be to leave no
trace…thank you universe.
3
Abstract

The colony-forming photoautrophic nitrogen fixer, Trichodesmium, is among the
most important contributors of newly fixed nitrogen (N
2
) to the global oceans. Hence, its
evolutionary fate in the face of multiple, interacting global change factors will have large
impacts on both marine food webs and global biogeochemistry. In order to
comprehensively link global change-driven evolutionary mechanisms with future
biogeochemical impacts, Trichodesmium erythraeum IMS101 (IMS101) was adapted to
low and high CO
2
for ~550-900 generations under multiple nutrient limitation scenarios
and subjected to a battery of physiological, biogeochemical, and molecular analyses.
Analysis of genomes from multiple Trichodesmium isolates as well as
metagenomes and metatranscriptomes from natural populations revealed genus-wide
conserved, genome architecture unusually littered with numerous large intergenic
spacers, repetitive elements, and transposases. These findings not only defy streamlining
observations generally seen in free-living prokaryotes in the oligotrophic oceans but also
importantly confirm the genetic potential of the lab-based evolution of IMS101 with
evolutionary potential contained within natural populations. After 4.5 years of adaptation,
all 6 replicates in the high-CO
2
cell lines maintained significantly higher growth and N
2

fixation rates relative to the low-CO
2
cell lines under either replete nutrients or
phosphorus limitation. Surprisingly, a 44% fitness increase was observed relative to the
low-CO
2
cell lines when high-CO
2
cell lines were reciprocally transplanted back to the
ancestral CO
2
condition. This observation is extremely rare in microbial evolution
literature and has broad implications for nitrogen and carbon cycling in the future oceans.
Transcriptome analysis revealed CO
2
-specific transcriptional regulation of key upstream
processes including transposition and sigma factors, which exert control over vast gene
regulatory networks, implicating these regulatory mechanisms to be important in both
short- and long-term responses to CO
2
increases. Low- and high CO
2
-adapted cell lines
were also subjected to iron- (Fe), phosphorus- (P), and Fe/P co-limitation in which Fe/P
co-limited cell lines exhibited higher growth and N
2
fixation rates relative to either Fe- or
P-limitation alone accompanied by reductions in cell size. This unexpected increase
suggests Trichodesmium to be independently adapted to Fe/P co-limited regimes, which
redefines classical single-nutrient Liebig limitation typically invoked to describe what is
most limiting in different oceans. Global proteome restructuring in the Fe/P co-limited
cell lines included alterations to both core and precursor metabolic pathways, increased
abundance of proteins implicated in cell size regulation, and an increase in abundance of
a protein complement specific to Fe/P co-limitation. Additionally, high CO
2
induced
fundamental proteomic shifts from low to high CO
2
co-limitation, which revealed
specific metabolic pathways that may come under CO
2
selective pressure in the future
ocean. Accordingly, this thesis uncovered multiple levels of transcriptional and proteomic
responses underlying the adaptation of Trichodesmium to high CO
2
under multiple
nutrient limitation scenarios, which will further serve as invaluable data for future
modeling efforts assessing the impacts of interactive global change factors on the global
biosphere.

4
TABLE OF CONTENTS

Acknowledgements………………………………………………………..........2

Abstract……………………………………………………….............................3

Chapter 1………………………………………………………...........................5
The effects of global change on microbial biogeochemistry and metabolism

Chapter 2………………………………………………………...........................33
Trichodesmium genome maintains abundant, widespread noncoding DNA in situ,
despite oligotrophic lifestyle

Chapter 3………………………………………………………...........................40
The molecular transition underlying plasticity-mediated CO
2
adaptation in the
globally distributed marine cyanobacterial nitrogen-fixer Trichodesmium

Chapter 4………………………………………………………...........................64
Mechanisms of increased Trichodesmium fitness under iron and phosphorus co-
limitation in the present and future ocean

Conclusions………………………………………………………........................93

5

Chapter 1
The effects of global change on microbial biogeochemistry and metabolism
Nathan G. Walworth

6
Introduction
Global fossil fuel consumption has induced a broad range of anthropogenically-
mediated environmental impacts by releasing greenhouse gases (primarily carbon dioxide
(CO
2
)) into the atmosphere at an unprecedented rate
1
. This rise in atmospheric carbon
dioxide is increasing concentrations of dissolved inorganic carbon (DIC) in seawater,
thereby reducing both hydroxide and carbonate ion (OH
-
and CO
3
2-
) concentrations and
leading to decreased pH. This phenomenon is known as ocean acidification (OA)
2
.
Predicting future effects on marine ecosystems remains challenging due to the unknown
interactions between carbon dioxide and other abiotic and biotic factors such as nutrient
availability and biological metabolism and community structure, respectively.
Additionally, rising sea surface temperatures as a result of greenhouse gas emissions
1
are
predicted to impart both local and global effects on physical oceanic phenomena such as
intensified stratification and expanded ocean suboxia regimes
3
.
Increasing numbers of global change studies assessing both short- and long-term
responses of phototrophic (light-consuming) carbon- (photoautotrophs) and nitrogen-
fixing (diazotrophs) organisms to ocean acidification have been carried out. This is due to
their bottom-up control of global biogeochemical cycles and food webs
4,5
. Direct effects
of OA and increased temperature on these metabolisms can include, but are not limited
to, impacts on enzyme kinetics
6
, physiological processes (e.g. nitrogen and carbon
fixation)
7
, nutrient acquisition mechanisms
4
, and ultimately growth
8
.
Global change factors are also predicted to cause changes to nutrient
bioavailability and community composition as a result of physicochemical alterations like
decreased physical mixing
9
and/or altered ligand-trace metal complexation as a result of
reduced pH
10
. For example, there have been conflicting predictions concerning future
iron (Fe) bioavailability in an acidified ocean. Some reports predict increases
10,11
,
especially from anthropogenic sources
12
, and others predict general decreases
13
. In
addition, changing relationships between organisms
14
may further alter community
composition due to differing degrees of within-population evolution
8
. For instance,
marine diazotrophs that provide critical bioavailable nitrogen to oligotrophic (low
nutrient) ecosystems exhibit variable genus-specific physiological responses to elevated
CO
2
reflective of their biogeographic distributions. This suggests that the fitness of future
7
diazotrophic populations and other taxa relying on their fixed nitrogen may be partially
influenced by CO
2
levels
15
. Although not traditionally considered limiting
16
, these and
other studies have demonstrated CO
2
(DIC) to stimulate both short- and long-term growth
in diverse photoautotrophic microbial phyla, which will likely affect nutrient limited
metabolisms in the future oceans
17
.Hence, to comprehensively predict future effects on
marine ecosystems and global biogeochemistry, it will be of paramount importance to
reassess oceanic nutrient species and distributions in the context of global change factors.

Key determinants of nutrient cycling and bioavailability
Energy from sunlight imparts a critical constraint on open-ocean primary
productivity, as the rapid attenuation of light with depth confines the growth of
photoautotrophs to a thin euphotic layer. Here, nutrients are depleted at the surface
through biological uptake and assimilated into macromolecules resulting in the formation
of particulate organic matter (POM). POM is then transported downward through the
water column and microbially remineralized (i.e. the microbial loop), leading to enhanced
nutrient concentrations below the photic zone as surface layer nutrient concentrations
become limiting
17
.
Nutrient limitation itself can refer to different spatial scales of biological and
ecological complexity, from the cellular level to a regional biosphere. In terms of a single
cell, depletion of external nutrients can reduce the cellular nutrient pool (quota), cell
size
18,19
, and growth
4
. These constraints can eventually limit overall community biomass.
Ecological interactions such as microbial dependencies and predation can also indirectly
influence nutrient availability by altering species composition and abundance
20
.
Additionally, an increase in a limiting nutrient may not initially benefit dominant
community members that are adapted to low concentrations (e.g. small cells) but instead
may enhance growth of rarer, larger cells that are more resistant to predation
21
.
Accordingly, differing microbial capacities to vary intracellular nutrient pools depending
on external concentrations (i.e. plasticity) are key determinants of community structure
and ultimately biogeochemistry
22
. Alfred Redfield first noted the similarity in
quantitative elemental relationships (i.e. stoichiometry) between cell-derived POM and
dissolved nutrients via a nitrogen (N) to phosphorus (P) ratio (N:P) of 16:1, which was
8
ultimately extended to carbon (C) and termed the ‘Redfield ratio” at 106C:16N:1P
23
. This
conserved C:N:P stoichiometry derives from elemental composition early in microbial
evolutionary history
24
.
Since then, oceanographers have utilized deviations from this ratio to infer potential
nutrient limitation and ultimately biogeochemical cycling in global microbial
communities
22
. Additionally, large differences in stoichiometry observed across broad
microbial lineages like the green and red eukaryotic algal superfamilies, reflect different
abilities to store nutrients in internal pools, modify osmolyte composition, and substitute
molecules depending on nutrient availability
20
. For instance, varying Redfield ratios have
been observed due to the substitution of phospholipids by non-phosphorus containing
lipids in P-limited regimes
25
. Additionally, different cellular components retaining
different stoichiometric ratios in addition to diel-specific stoichiometric variability
26
can
impact general cellular ratios, depending on nutrient status. For example, resource
acquisition machinery (e.g. uptake proteins) exhibits high N:P ratios relative to growth
machinery (e.g. ribosomal RNA), which is typically high in both N and P
20
.
Hence, the Redfield ratio represents an average for a diverse oceanic
phytoplankton assemblage growing under different conditions with different growth
strategies. This metric serves as a starting point when investigating cellular nutrient
limitation in a given system, in which the nutrient(s) in shortest supply will limit the rate
and yield of new biomass production. However, varying degrees of stoichiometric
plasticity and surface recycling rates for different elements may change the ultimate
source of limitation from what might be expected based on the locally dissolved nutrient
pool
27
. Thus, stoichiometry exerts fundamental control over the coupling of nutrient
limitation and biogeochemical cycling and highlights the need to study future
evolutionary processes underlying nutrient limited metabolisms, which will determine
ultimate constraints on cell growth and population dynamics in an acidifying ocean.

Concepts of nutrient limitation
Low nutrient concentrations can limit cell growth rates, while the total available
stock of a nutrient(s) can set an upper bound on overall yield of a population. These two
modes of limitation are classically referred to as Blackman and Liebig limitation,
9
respectively
17,28
. Since multiple inorganic nutrients are close to Liebig limitation in the
oligotrophic oceans including nitrogen, phosphorus, and iron, both limitations can come
into play simultaneously depending on respective concentrations and uptake rates
29
. For
example, carbon limitation of microbial photoautotrophs is traditionally considered as
Blackman (rate) limitation since dissolved inorganic carbon (DIC) is quite high, yet only
~1% of it exists as CO
2
. Simultaneously, low concentrations of either P or Fe can result
in Liebig (yield) limitation
29
.
Traditionally, biomass limitation by the single nutrient in shortest supply has been
invoked to define the controlling mechanism limiting primary productivity and carbon
sequestration, which can then be secondarily limited by the next most limiting
nutrient
30,31
. However, several studies have demonstrated nutrient co-limitation whereby
two (or more) nutrients can be reduced to levels where addition of both are required to
stimulate growth
20,29,32
. Several forms of nutrient co-limitation have been demonstrated in
culture
19,33
and in situ
32,34
, thus redefining conventional concepts (e.g. Liebig limitation)
ascribing the ultimate constraints on productivity in different ocean biomes.
Independent nutrient co-limitation is defined as the simultaneous depletion of two
growth limiting nutrients that explicitly differ in biochemical function, as in nitrogen-
phosphorus co-limitation
29
. Biochemical substitution co-limitation describes a
circumstance where one limiting nutrient may be biologically substituted for another,
either directly within the same macromolecule (i.e. enzyme) as in zinc-cobalt co-
limitation
33
, or indirectly by substituting one macromolecule for another
17
. Biochemically
dependent co-limitation, refers to the limitation in the ability to take up one nutrient due
to the inability of a cell to acquire sufficient supplies of another, as in zinc-phosphorus
co-limitation
20
. Finally, at the community level, one subpopulation of the microbial
community may respond to additions of one nutrient whereas another may respond to that
of a different nutrient.
These definitions necessarily provide a hypothetical framework to attempt to explicitly
determine different nutrient statuses from the cell to the community level. However,
interpretations to delineate simultaneous versus secondary limitation scenarios remain
challenging. Nevertheless, the aforementioned studies provide evidence that marine
microbes experience periods of persistent nutrient co-limitation, which may have selected
10
for the evolution of independent metabolic responses to co-limiting conditions relative to
single limitation (e.g. Fe- or P-limitation). For instance, evidence for independent
selection under co-limitation has been shown in diazotrophic cyanobacteria that grow and
fix N
2
faster when co-limited for Fe and P (Fe/P co-limitation) than when limited by
either nutrient alone. This may indicate independent adaptations specific to co-limited,
oligotrophic regimes
19
(Fu et al. in prep)(see Ch. 4). These relatively recent co-limitation
studies together with rapidly surfacing global change factors compel reassessment of
global nutrient cycles as they interact with anthropogenically-driven evolutionary
processes.

Global patterns of oceanic nutrient limitation
In order to assess global change impacts to future nutrient cycles, it is necessary
to begin with known patterns of nutrient limitation and how they affect and are affected
by microbial processes. Ocean biome spatiotemporal patterns of upper-ocean micro- (e.g.
Fe, Cu, Ni, Cd, Zn, and Co) and macronutrient (e.g. N and P) limitation have been
inferred from multiple sources, including molecular techniques
35
, shipboard nutrient
manipulation experiments to natural microbial communities
32
, and in situ open-ocean
enrichment experiments
36,37
. For instance, productivity and biomass only increase with
experimental additions of nitrogen in the low-biomass (oligotrophic), low-latitude oceans
27,38
. The exception can be certain oceanic regions like the Mediterranean and subtropical
North Atlantic, where bioavailable phosphorus (P) can also become severely limiting in
addition to nitrogen (N)
25,39
, leading to N/P co-limitation in stratified conditions
27
.
However, there is a general tendency towards nitrogen limitation, which could be
explained in part by multiple forms of P-limited acclimatization mechanisms
25
, high
bioavailability of organic P compounds
40
, and/or selection towards higher N:P ratios
under low-nutrient conditions
41
. Conversely, photoautotrophs in areas with substantial
bioavailable nitrogen concentrations yet low photoautotrophic biomass (high-nitrate low-
chlorophyll regions, aka HNLC) are usually only stimulated with experimental iron
additions
36
. Other micronutrients (see above) can also potentially limit growth, but our
inadequate understanding of their physiological nature, interactions, and geographic
distributions stems from a scarcity of experiments due to methodological limitations and
11
logistical challenges from aspects such as metal-metal interactions (e.g. biochemical
substitution co-limitation)
17
. Hence, molecular diagnostics developed through laboratory
experimental manipulations
17
is promising for assessment of future nutrient interactions
with respect to microbes in situ
33,34,42
.

Global change interactions with microbial biogeochemistry

From ocean circulation to organism
The recent, rapid increase in single-variable CO
2
6,8

and temperature
43-45

experiments have set the stage for more realistic investigations of interacting global
change variables on microbial populations. There is evidence for large changes in
nutrient biogeochemistry coinciding with global shifts in atmospheric CO
2
concentrations
over glacial-interglacial cycles
46-48
. Thus, anthropogenically-mediated climate
interactions are predicted to affect future nutrient cycling, including changes to surface
ocean chemistry, external nutrient supply, nutrient composition (e.g. anthropogenic
sources), nutrient demands, and ocean circulation. When examining multiple interacting
drivers, it is useful to classify interactions in terms of their scale as well as their
abiotic/biotic nature.
Abiotic physicochemical interactions can be exclusively driven by the interplay of
physical processes. For instance, macronutrient-rich waters upwelled from the deep ocean
contribute to the physical resupply by the Southern Ocean. These waters are typically
depleted in micronutrients leading to iron-limited HNLC waters low in biomass, which
subsequently allows macronutrient-containing waters to travel northward within the
thermocline and eventually mix with smaller-scale upwelling to support production in N-
limited low latitudes
49
. However, the predicted strengthening of density stratification due
to warming and freshening (polar sea-ice melt) may restrict physical resupply of nutrients
to surface waters
9
, decrease export production of organic matter
50,51
, and expand N-
limited subtropical gyres
52
and oxygen minimum zones (OMZ)
53
. Conversely,
anthropogenic sources are increasing external nutrient supply via atmospheric and fluvial
inputs
54,55
, which further complicates predictions concerning nutrient bioavailability
depending on the chemical nature of the source, the geographic source of input, and its
12
unknown abiotic and biotic interactions. Anthropogenic nutrient sources are of
comparable magnitudes to natural ones, and are predicted to increase in the near-future.
Atmospheric deposition of bioavailable fixed nitrogen has already tripled since 1860, and
is deposited in low-latitude N-limited regions, which may stimulate primary production
17
.
However, future atmospheric deposition of iron and phosphorus are more difficult to
predict due to strong regional gradients and continued shifts in climate and land use
54
.
Overall though, atmospheric fluxes are predicted to be enriched in nitrogen relative to
both iron and phosphorus
17
, thereby potentially pushing once N-limited regimes towards
Fe- and/or P-limitation.
Biotic ecosystem-level interactions also have the potential to affect different
trophic levels contingent upon varying organismal sensitivities to different drivers and
their downstream effects on trophic relationships (e.g. predator-prey). For example,
environmental shifts can alter predatory-prey relationships by changing the palatability of
algae to marine amphipods
56
or by changing structural cellular elements via decreased net
calcification rates
57
. The effects of altered trophic relationships can subsequently result in
transmission through food webs, thereby forming explicit links between lower and upper
components. For instance, warming in polar waters may allow for range extension of
predators like crabs into new habitats
58
, which may be able to exploit OA-induced
thinning of calcite skeletons of prey
59
.
Organismal-level interactions with environmental drivers (e.g. metabolic and
physiological processes) have been the most intensively studied arena in the global
change field. Nutrient-derived physiological interplay has been largely summarized as
different forms of nutrient (co)-limitation (see above). Additionally, non-nutrient drivers
including temperature, light, pH, inorganic carbon availability (e.g. CO
2
or DIC), and
grazing can all interact with nutrient (co)-limitation to control physiological growth and
abundance of different microbial populations, which may lead to altered nutrient
demands at both the cellular and community level. For instance, short-term experiments
employing increased temperature and CO
2
have been shown to affect both primary
production and N
2
fixation
4,6,19,52,60
, suggesting potentially altered nitrogen inputs and/or
phosphorus or iron uptake in the future oceans. Accordingly, numerous short-term single-
and multivariate studies on photoautotrophic metabolism and assemblages have been
13
conducted, and serve as an appropriate foundation for subsequent long-term
investigations.

Short-term CO
2
experiments on photoautotrophs
The short-term response of a given genotype is characterized by its reaction norm,
which is a trait value range (e.g. photosynthesis, growth, etc.) as a function of two or
more environments
61
. These reaction norms are a measure of phenotypic plasticity, which
is defined as a rapid physiological (phenotypic) response to an environmental change.
Conversely, adaptive evolution is defined as a change in the underlying genetic (allelic)
composition of a population due to natural selection resulting in a change in phenotype
62
.
Differences in inter-clonal plasticity among photoautotrophs have been observed and
described as genotype x environment (G x E) interactions, in which genotypes that
rapidly exploit new environments will be favored by inter-genotypic selection on days to
months timescales
63
. Thus, studying the progression from plasticity to adaptation (see Ch.
3) is important because varying degrees of phenotypic plasticity can affect
adaptation
8,64,65
and a broad range of phenotypic plasticity can exist within a single
species
44,63
. Since global change may introduce conditions outside of the present-day
range of environments typically experienced by populations
2,66
, investigating acclimatory
capacity (i.e. phenotypic buffering) of biogeochemically-critical photoautotrophs will be
important to determine degrees of adaptive phenotypic plasticity as it affects
biogeochemistry.
The most abundant photosynthetic organisms on Earth are the unicellular
picocyanobacteria (<2 μm diameter) of the genera Prochlorococcus and Synechococcus,
which are estimated to contribute 40-80% of gross primary production in the tropical and
subtropical oceans
6
. Prochlorococcus strain MED4 showed no changes in growth or
photosynthesis when exposed to short-term elevated CO
2
and temperature in culture but
exhibited increased cell quotas, pigment concentration, and cell volume under increased
temperature alone
67
. Conversely, Fu et al. observed increased photosynthesis and growth
rates in Synechococcus strain WH7803 under elevated CO
2
and temperature
67
while
another study contrastingly demonstrated decreased growth with decreasing pH
(increasing CO
2
)
68
. Field studies of picocyanobacteria showed that natural populations
14
did not respond strongly to changes in CO
2
, with only a transient increase in
photosynthetic rates under high CO
2
followed by a decrease several days later.
Additionally, no changes to cell size and pigment content were observed
69
contrasting
with culture studies
67
. However, the ratio of Synechococcus to Prochlorococcus increased
under elevated CO
2
suggesting Synechococcus to have a competitive advantage over
Prochlorococcus that cannot be explained by the measured physiological parameters (e.g.
photosynthesis). Additionally, culture-independent molecular techniques have detected
different ecotypic clades of Prochlorococcus and Synechococcus to be significantly
correlated to different environmental levels of temperature, macronutrients, and iron
70,71
.
Taken together, these data suggest that Synechococcus is more sensitive to elevated
temperature and CO
2
relative to Prochlorococcus on short timescales, with ecotype-
specific responses possibly modulated by factors such as nutrient availability and
competition
72
. It remains to be seen whether these short-term studies are indicators of
long-term adaptation under global change factors in these picocyanobacteria.
Eukaryotic microbial photoautotrophs including diatoms and dinoflagellates
exhibit varying short-term responses to global change drivers. Some diatoms exhibit
increases in growth and photochemical yield
73,74
while others show decreased growth
75
in
response to increased CO
2
and light. Similarly, some dinoflagellates exhibit increased
photosynthetic rates while others increase growth but not photosynthetic rates under
elevated CO
2
and temperature
76
. Most OA short-term experiments like these have
typically co-varied CO
2
and either light or temperature, with fewer studies co-varying
nutrient availability.
Coupling nutrient limitation with OA has come under intensive study, particularly
for toxic algae due to their detrimental effects on commercial fisheries. For instance, the
toxic diatom Pseudo-nitzschia produces the neurotoxin domoic acid (DA), which is
regulated by pH and nutrient availability. DA production increases with increasing CO
2

concentrations under silica-limiting conditions for P. fraudulenta
77
and phosphorus (P)
limitation for P. multiseries
78
. Similarly, the toxic dinoflagellate Karlodinium veneficum
increases its toxicity under elevated CO
2
and P-limitation
79
. However, the dinoflagellate
Alexandrium catanella significantly increased its toxicity in high CO
2
irrespective of
nutrient limitation
80
. Hence, different diatom and dinoflagellate genera exhibit varying
15
short-term responses to OA and other environmental drivers, with implications for
enhanced toxicity under different nutrient regimes for certain taxa.
Nevertheless, different nutrient limitation permutations and long-term incubations
are largely absent from the breadth of available photoautotrophic studies. Thus,
conducting more long-term, multivariate OA studies will not only shed important new
insights on the relationship between different plastic responses and adaptive ones but also
on long-term, evolutionary effects to biogeochemistry. Together with OA experiments on
natural assemblages (see below), these short- and long-term data will enable better
predictions for future diatom and dinoflagellate community composition shifts in the
future oceans.
Coccolithophores are eukaryotic haptophytes that extrude calcium carbonate
plates (i.e. coccoliths) and appear to rely on aqueous CO
2
for photosynthesis and prefer
bicarbonate (HCO
3
-
) for calcification
81
. Certain species and strains have been shown to
calcify more under short-term increased CO
2
82
, but the vast majority tend to reduce
calcification under OA
6
consistent with fossil records
83
. Lower calcification may in turn
lead to greater carbon export, suggesting coccolithophores will continue to be a net CO
2

sink in the near future
84
. CO
2
-enhanced growth also depends on other factors such as
nitrogen source
85
and light
86
, but more studies are required to delineate among strain-
specific differences. Thus, coccolithophores are affected by both pH for calcification and
elevated CO
2
for photosynthetic physiology. Bach et al. (2013) varied these parameters
separately and found that high CO
2
reduced growth when pH was allowed to decrease
naturally, but this effect was not seen when pH was artificially held constant in Emiliania
huxleyi
81
. Calcification reductions and aberrant coccolith formation at high CO
2
was also
controlled by decreasing pH, as these effects were also undetected when pH was held
constant, thereby suggesting the effects of rising CO
2
to be dependent on pH changes and
shifts in carbonate equilibrium for E. huxleyi. This dependency has implications for long-
term competitive fitness by coccolithophores in natural assemblages. For example,
coccolithophore abundance increased relative to that of diatoms when natural populations
in mesocosms were enriched with predicted year 2100 CO
2
87
, implicating
coccolithophores to have a competitive advantage over diatoms. However, experimental
evolution studies subjecting coccolithophores to long-term OA (see below) suggest
16
general reductions in calcification
116
raising unknowns about pleiotropic effects that may
affect competitive fitness in the future oceans.
Other short-term mesocosm studies enriching natural assemblages with elevated
CO
2
have yielded varying results, with some demonstrating increases in total diatom
abundance but not total biomass or primary production
88
. Others showed increases in
phytoplankton productivity through the growth of larger chain-forming diatoms,
suggesting enhanced carbon export via larger, faster sinking cells
89
. Conversely, OA
conditions have also been shown to cause community shifts away from diatom
dominance
90,91
. Additional factors like temperature
45
and iron (Fe)
92
can also modulate
the effects of enhanced CO
2
on microbial communities. For instance, Hoppe et al.
demonstrated that phytoplankton productivity was not stimulated under Fe-limitation yet
floristic shifts and increased primary production were observed under Fe-replete
conditions
92
. Therefore, the effects of enhanced CO
2
on natural assemblages depend on
the initial community composition in addition to nutrient (e.g. iron) and non-nutrient (e.g.
light) drivers. Time of exposure to environmental drives may also play a large role in
determining eventual community shifts in the future oceans, but more long-term
community studies are needed to assess possible outcomes (see below).

Short-term CO
2
experiments on diazotrophs
Since much of the ocean’s biosphere is limited by nitrogen, primary productivity
and net export of organic matter (i.e. carbon sequestration) to the deep ocean (i.e. the
biological pump)
5
rely upon bioavailable nitrogen produced by cyanobacterial N
2
fixation. N
2
fixation is an energy-intensive, oxygen-sensitive process that reduces N
2
gas
into two molecules of bioavailable ammonia, and is conducted by several types of
diazotrophs in the euphotic water column. Due to oxygen (O
2
) inactivation of
nitrogenase, unicellular diazotrophs typically temporally separate photosynthesis and N
2

fixation, with the former process occurring during the day and the latter at night. In most
filamentous species, photosynthesis and N
2
fixation are spatially separated with
specialized cells that contain nitrogenase to carry out N
2
fixation (i.e. heterocysts).
Heterocysts are fully differentiated cells devoid of the O
2
-evolving photosystem II, and
are surrounded by thick cell walls acting as O
2
barriers. Fixed nitrogen is transported
17
from heterocysts in the form of amino acids, while heterocysts receive carbohydrates
from vegetative cells
93
. Conversely, the filamentous nonheterocystous, Trichodesmium,
fixes both N
2
and CO
2
during the photoperiod. Multiple studies employing different
techniques demonstrate temporal segregation of these processes, with peak CO
2
fixation
occurring around mid-morning and peak N
2
fixation in the afternoon
94
. However,
conflicting views on the spatial segregation of these processes have surfaced with some
immunochemical studies detecting nitrogenase in all cells along a filament
95
and others
detecting it in only ~15% of cells that reside in clusters along a filament
96

97,98
. These
clusters are referred to as diazocytes, and suggest spatial segregation of N
2
fixation from
other cells along a filament.
Trichodesmium, is estimated to carry out as much as half of total N
2
fixation in the vast
subtropical gyre biomes, which has made it the most intensively studied diazotroph in
OA research
60
. T. erythraeum IMS101 (hereafter IMS101) has been shown to increase
growth, biomass production, C acquisition, and/or N
2
fixation under high CO
2

conditions
7,99-101
. While Hutchins et al. observed growth stimulation by elevated CO
2
in
T. erythraeum and T. contortum, T. theibautii growth was already saturated under low
CO
2
, suggesting that marine N
2
-fixers may have experienced differential selection by
temporal fluctuations in CO
2
over Earth’s history
15
. Additionally, elevated CO
2
still
enhanced Trichodesmium productivity even with varying light, temperature, and
phosphorus availability
102-104
. However, coupling Fe-limitation with short-term elevated
CO
2
exposure neutralizes high CO
2
-growth and N
2
fixation in IMS101 (see Ch. 4)
105
.
Indeed, due to the high iron requirement of the N
2
-fixing (i.e. nitrogenase) and
photosynthetic protein complexes, dissolved Fe (dFe) concentrations are a major control
on Trichodesmium distributions. This is evidenced by elevated Trichodesmium spp.
abundance in the North Atlantic and Pacific near the Hawaiian Islands, where dFe is
relatively high, but not in the South Atlantic or Pacific central gyre where dFe is
extremely low
5
. Conversely, unicellular diazotrophs numerically dominate low dFe
regimes, potentially due to their smaller cell size and Fe quotas
106
.
The unicellular diazotroph Crocosphaera (UCYN-B) contributes substantial
amounts of fixed N to the oligotrophic Atlantic and Pacific Oceans, and fixes CO
2
during
the day and N
2
at night. Similar to Trichodesmium, elevated CO
2
in an Fe-replete regime
18
stimulates growth and N
2
fixation of C. watsonii WH8501, but the effect is neutralized in
Fe-limited conditions
107
. Additionally, enhanced N
2
fixation under OA and Fe-replete
conditions are strain-specific where WH0401 and WH8501 exhibit increases (especially
under low light for WH0401) while WH0003 and WH0402 remain unaffected
15,108
.
Interestingly, Garcia et al. found that WH0003 increased growth and N
2
fixation under
high CO
2
and high light, while elevated CO
2
only increased growth and not N
2
fixation
under low light in both low and high phosphorus regimes
109
. However, CO
2
- and N
2
-
fixation were both positively affected by high CO
2
in concert with low light and low P.
Accordingly, total P requirements declined as CO
2
increased demonstrating enhanced
phosphorus use efficiency, which may indicate that cells require less energy to fuel
carbon concentrating mechanisms (CCM) under OA conditions. Fewer OA studies have
been done on the closely related unicellular diazotroph, Cyanothece (UCYN-B), which
demonstrated increased production rates but not N
2
fixation under elevated CO
2
110
.
Interestingly, the single-cell, symbiotic diazotroph Calothrix showed no changes
in growth or primary production but strong increases in N
2
fixation under elevated CO
2
.
In contrast, mixed results have been obtained for the filamentous heterocyst-containing
diazotroph Nodularia spp. that forms blooms in the Baltic Sea. Under elevated CO
2
, N.
spumigena has been reported to reduce growth and increase N
2
fixation
110
in one study
and increase growth and CO
2
and N
2
fixation in another
111
.
Finally, the uncultured unicellular diazotrophic cyanobacteria group A (UCYN-
A) also contributes a significant fraction of marine N
2
fixation
106
in which their global
nitrogenase gene abundance is estimated to exceed that of Trichodesmium
112
. Members
belonging to UCYN-A lack photosystem II (PSII) and rubisco genes and acquire fixed
carbon symbiotically
113
. The single study examining short-term elevated CO
2
exposure
on natural communities containing UCYN-A found no stimulation in N
2
fixation
114
.
The above OA studies examining photoautotrophic and diazotroph physiology
represent the vast majority of marine microbial global change research, employing short-
term incubations that sometimes co-vary light and/or temperature and more rarely
limiting nutrients
6,60
. These studies thus only reveal short-term phenotypic responses to
projected future changes to ocean chemistry. Additionally, almost nothing is known
about how microbial evolutionary processes adapting to interacting factors will relate to
19
short-term plastic responses and subsequently affect future biogeochemical cycling and
ecosystems. Interestingly, several recent long-term studies on marine phytoplankton (see
below) have shown that short-term responses to elevated CO
2
(i.e. plastic response) can
reverse in sign and magnitude upon long-term exposures as part of an adaptive strategy
8
,
suggesting that short-term responses to global change factors are not necessarily
indicative of long-term adaptation in certain taxa
64
. Hence, due to the species- and strain-
specific variability in OA responses outlined above coupled to unknown effects to
nutrient availability from physicochemical forcings, it is paramount to conduct long-term
global change studies on both microbial assemblages and biogeochemically-important
microbes under a range of different scenarios predicted to occur in the future oceans.

Effects of global change-driven adaptation on microbial biogeochemistry

Short-term global change studies may only elucidate transient stress responses to
abrupt perturbations that may or may not be congruent with long-term adaptive
responses. Consequently, laboratory-based experimental evolution studies under well-
defined conditions are necessary to investigate both short- (plastic) and long-term
(adaptive) responses to global change factors and how they are related. One of the main
challenges in experimental evolution is extrapolating laboratory evolution to evolutionary
potential contained within natural populations due to limited knowledge of gene flow,
population sizes, mutation, and recombination in addition to the oversimplification of the
environment. However, substantial advances in next-generation sequencing technology
have enabled the study of global metabolic potential (e.g. genome, transcriptome,
proteome, epigenome) within individual organisms in response to environmental stimuli
in addition to global surveys of in situ genetic potential contained within microbial
populations
115
. Additionally, these methods have also enabled enhanced resolution into
horizontal gene transfer (i.e. the physical transfer of genetic material from one organism
to another) between cohabitating microbes that may aid in adaptation to new
environments and/or colonization of new niches
116
.
Therefore, connecting adaptive responses observed in the laboratory with genetic
potential observed in situ can give environmental context to laboratory-based evolution
20
and allow for a more comprehensive assessment of evolutionary potential to global
change. For example, we confirmed that the genome of the isolate, T. erythraeum
IMS101, used in this thesis is remarkably conserved with genomes of natural populations
(see Ch. 2), thus implying that changes to genetic potential of IMS101 observed in lab
experiments are also possible in Trichodesmium populations in situ. Hence, coupling
molecular techniques with experimental evolution helps to enable delineation between
plastic and adaptive responses to external changes through observations of underlying
genetic evolution
117
.

Long-term global change experiments on photoautotrophs
To date, very few experimental evolution studies have been conducted on marine
photoautotrophs, although these numbers are rapidly increasing. Several major challenges
in working with biogeochemically-important microbes include maintenance of those that
may not be amenable to the culturing environment, do not contain a genetic
transformation system, and cannot be cryopreserverd through classic techniques. These
issues compound the amount of work necessary to maintain adequate replicates per
treatment and require maintenance of numerous “backup” cultures should mortality occur
in experimental cell lines.
To date, no long-term studies on marine picocyanobacteria have been conducted
and only one examining the cyanobacterial diazotroph, IMS101, has been published (see
below). Several eukaryotic phytoplankton systems have been subjected to CO
2
-driven
experimental evolution, and have yielded remarkably consistent results in terms of both
their short- and long-term responses. In coccolithophores that experience fitness (i.e.
growth) loss and reduced calcification under short-term elevated CO
2
(see above),
adaptive evolution partially restores both traits after several hundred generations of
selection through sorting standing variation between clonal lineages and acquiring
heritable changes within clonal lineages attributed to novel mutations
118,119
. Interestingly,
Benner et al. (2013) co-varied high CO
2
and temperature but observed no changes in
calcification-related gene expression associated with changes in calcification
120
. Instead
they observed up-regulated genes involved in cellular processes and signaling,
information storage, core metabolism, and viral processes. Additionally, coccolithophore
21
populations adapted to high CO
2
exhibited phenotypic evidence of functional genetic
divergence between replicate populations suggesting that multiple genetic trajectories can
be selected for during adaptation to OA leading to several different genotypes
121
. Hence,
experimental evolution studies on coccolithophores have demonstrated that initial short-
term plastic responses to OA can reverse in sign upon adaptation, and can be achieved
through sorting standing variation, de novo mutation, and/or functional divergence. These
findings are important for highlighting an inverse relationship between plasticity and
adaptation, in addition to better informing future predictions on OA-influenced
biogeochemistry.
Experimental evolution on 16 physiologically distinct lineages of the globally
distributed chlorophyte, Ostreococcus, evolved in either constant or fluctuating, high and
low CO
2
regimes revealed short-term increases in growth as part of a plastic response.
This was followed by decreases in growth as part of an evolutionary response after 400
generations
64
. Populations demonstrating greater plasticity evolved more, and this
relationship was strongest in cell lines adapted to fluctuating environments favoring the
evolution and maintenance of plasticity. Hence, plasticity predicted the extent but not the
direction of phenotypic evolution. The evolved cell lines exhibiting slower growth
retained higher mitochondrial potential, and better withstood subsequent environmental
changes relative to faster growing cells. Additionally, cells evolved in fluctuating, high
CO
2
(FH) were smaller and had reduced C:N ratios relative to stable, high CO
2
(SH)
lines
65
. Taken together, the authors hypothesize that slower growth is adaptive under
long-term CO
2
enrichment when associated with production of higher quality daughter
cells that can maintain greater adaptive plasticity in the face of further environmental
changes. Here, the plastic and adaptive responses of Ostreococcus were also opposite
each other as in coccolithophores, even though each response was reversed in sign
relative to their corresponding responses in coccolithophores. Additionally, evolving cell
lines in fluctuating environments resulted in greater adaptive plasticity to further
environmental changes, which elucidates the importance of maintaining evolutionary
plasticity in dynamic environments. These results are important in that they relate
plasticity to adaptation, and highlight the need to consider evolving populations in
dynamic environments to better understand the role of maintaining plasticity for better
22
fitness. Additionally, increased growth is typically synonymous with higher fitness in
microbial evolution
117,122
, but these results suggest that adaptation leading to greater
fitness is not associated with increased growth but instead slower growth with
improvements in fitness-related traits (e.g. mitochondrial potential). These insights are
critical ones when predicting evolutionary responses of microbial populations to OA.
In the case of the freshwater chlorophyte, Chlamydomonas reinhardtii, cell
division rates did not increase following hundreds of generations of growth at high CO
2

relative to the plastic response of control populations
123
. Another study exposing 7
species from three major freshwater phytoplankton groups to high CO
2
for ~750
generations detected no evolutionary adaptation
124
.
Finally, in addition to determining direct adaptive responses to elevated CO
2
,
several studies have examined indirect effects following individual CO
2
selection namely
through competitive assays. These experiments addressed the question of whether
changes in competitive interactions between phytoplankton can be predicted from
evolutionary changes in individual lineages. One study using a freshwater community of
cyanobacteria, chlorophytes, and diatoms found that individual responses to long-term
(hundreds of generations) elevated CO
2
were good predictors of shifts in community
structure
125
. However, several other studies using more closely related taxa of either
freshwater chlorophytes or different species of marine diatoms or dinoflagellates found
that individual fitness of lineages evolved in isolation was a poor predictor of competitive
ability
45,126,127
. These discrepant results may be due to the fact that closely related
lineages have similar evolutionary constraints, while more distantly related genera retain
fundamental biological differences that are difficult or impossible to overcome, thereby
retaining more robust community structure
8
.

Long-term global change experiments on diazotrophs
All previous studies examining diazotroph physiology under interacting global-
change and nutrient stressors have only employed brief exposures (weeks to months; see
above). Only one study examining the long-term adaptation of Trichodesmium to high
CO
2
has been published. Briefly, the ancestral population was split into low (380 micro-
atmospheres (uatm)) and high (750 uatm) CO
2
treatments with 6 biological replicates
23
each
60
. At the onset of the experiment, ancestral cell lines placed in high CO
2
(750-
ancestral) rapidly increased both growth and N
2
fixation rates, whereas cell lines in low
CO
2
(380-ancestral) sustained lower physiological rates. This immediate increase in
growth and N
2
fixation demonstrates a plastic response to high CO
2
consistent with prior
observations (see above). Following 4.5 years of low and high CO
2
selection, no further
changes in fitness were observed for either the low or high CO
2
treatments. All 750-
selected replicate cell lines still maintained significantly higher growth and N
2
fixation
relative to both the 380-ancestral and 380-selected cell lines. Once subcultures of the
380-selected cell lines were placed in high CO
2
for 2 weeks following the 4.5-year
incubation at low CO
2
, both growth and N
2
fixation rapidly increased similar to the 750-
ancestral response.
Surprisingly, a 44% fitness increase was observed relative to both the 380-
selected and 380-ancestral cell lines when 750-selected cell lines were reciprocally
transplanted back to the ancestral CO
2
condition (380 uatm). Hence, the adaptive
response following long-term high CO
2
selection was characterized not by steady fitness
increases in the selection environment, but instead large fitness increases in the ancestral
environment, which is an extremely rare phenomenon in microbial evolution
literature
8,60,117
. Both the plastic and adaptive responses of IMS101 to CO
2
retain the
same phenotype in the selection environment but not the ancestral environment, and thus,
the transition from plasticity to adaptation remained phenotypically neutral in the
selection environment. 750-selected cell lines demonstrated an extended N
2
fixation diel
period, but shotgun proteomics revealed no major changes to detected protein abundances
including those involved in nitrogen, photosynthesis, and carbon. This suggests that
either undetected proteins or other regulatory elements such as DNA methylation may
cause phenotypic changes under high CO
2
conditions. Enhanced DNA methyltransferase
activity was indeed detected in the 750-to-380 reciprocal transplant, which may indicate
DNA methylation to play a role in maintaining increased growth and N
2
fixation in the
ancestral CO
2
environment. Additionally, 750-selected cell lines maintained higher
growth and N
2
fixation in P-limited conditions, relative to 380-selected cell lines under P-
limitation.
24
Although these adaptive responses reveal novel evolutionary phenomena
influencing an important biogeochemical process (i.e. N
2
fixation), it is clear from short-
term experiments that more interacting environmental drivers must be investigated to
determine evolutionary tradeoffs that may occur in response to an acidified ocean.
Additionally, more OA experiments investigating species interactions modulated by
environmental drivers are important to determine shifts in community and/or trophic
structure and their cascading effects through marine ecosystems. Finally, mostly all
experiments to date have traditionally compared two separate CO
2
regimes without
accounting for short-term CO
2
variability or more gradual increases in CO
2
known to
occur in situ
128
. Although these challenges are not trivial, they are surmountable
especially with the burgeoning global change research community. These experimental
complexities should encourage interdisciplinary collaboration as OA investigations begin
to incorporate more interacting variables. Hence, this dissertation aimed to provide a
framework going from single- to multivariate experiments and assessing organismal
responses on both short and long timescales to assess impacts to biogeochemistry and
evolution.

Dissertation Goals
The dissertation work described here examines the metabolic potential inherent in
both cultures and natural populations of Trichodesmium spp., and relates it to molecular
responses underlying phenotypic changes following both short- and long-term exposure
to elevated CO
2
in multiple nutrient limitation regimes. One of my primary goals in terms
of evolutionary processes was to investigate the effects of short-term plastic responses on
long-term adaptive ones, at both the molecular and physiological level. Several recent
studies in photoautotrophs have demonstrated the importance of short-term physiology in
determining evolutionary outcomes, but no studies to date have assessed these dynamics
on biogeochemically-critical marine N
2
fixers. The findings described here have
important implications for fundamental microbial evolution and physiology, as we
describe both novel experimental approaches as well as the molecular adaptation of
previously unidentified microbial evolutionary phenomena.
25
Additionally, this work holds broad implications for biogeochemical and nutrient
cycling including nitrogen, carbon, phosphorus, and iron. Since Trichodesmium is critical
to supporting primary production in the oligotrophic oceans, it is paramount to study its
evolutionary potential in the face of different global change scenarios and relate the
outcomes to future effects on food webs and export production of organic matter to the
deep ocean. With rapidly advancing technology, this type of global-change experimental
evolution study can yield empirical data that will be crucial to inform future climate
change modeling efforts. Only through integrating these disciplines within one
framework will we be able to comprehensively assess future uncertainties concerning
elemental cycling and the global biosphere.

References

1. Climate Change 2013 The Physical Science Basis. 1–14 (2014).
2. Caldeira, K. & Wickett, M. E. Oceanography: Anthropogenic carbon and ocean
pH. Nature 425, 365–365 (2003).
3. Gnanadesikan, A., Russell, J. L. & Zeng, F. How does ocean ventilation change
under global warming? Ocean Science 3, 43–53 (2007).
4. Hutchins, D. A., Mulholland, M. R. & Fu, F.-X. Nutrient cycles and Marine
Microbes in a CO
2
-Enriched Ocean. Oceanography 22, 128–145 (2009).
5. Sohm, J. A., Webb, E. A. & Capone, D. G. Emerging patterns of marine nitrogen
fixation. Nat Rev Micro 9, 499–508 (2011).
6. Mackey, K., Morris, J. J., Morel, F. & Kranz, S. Response of Photosynthesis to
Ocean Acidification. Oceanography 25, 74–91 (2015).
7. Hutchins, D. A. et al. CO
2
control of Trichodesmium N
2
fixation, photosynthesis,
growth rates, and elemental ratios. Limnol. Oceanogr. 52, 1293–1304 (2007).
8. Collins, S., Rost, B. & Rynearson, T. A. Evolutionary potential of marine
phytoplankton under ocean acidification. Evol Appl 7, 140–155 (2013).
9. Sarmiento, J. L., Hughes, T. M. C., Stouffer, R. J. & Manabe, S. Simulated
response of the ocean carbon cycle to anthropogenic climate warming. Nature
393, 245–249 (1998).
10. Millero, F. J., Woosley, R., Ditrolio, B. & Waters, J. Effect of ocean acidification
on the speciation of metals in seawater. Oceanography 22, 72–85 (2009).
11. Breitbarth, E. et al. Ocean acidification affects iron speciation during a coastal
seawater mesocosm experiment. Biogeosciences 7, 1065–1073 (2010).
12. Fishwick, M. P. et al. The impact of changing surface ocean conditions on the
dissolution of aerosol iron. Global Biogeochem. Cycles 1–16 (2014).
doi:10.1002/(ISSN)1944-9224
13. Shi, D., Xu, Y., hopkinson, B. M. & morel, F. M. M. Effect of Ocean
Acidification on Iron Availability to Marine Phytoplankton. Science 327, 676–
26
679 (2010).
14. Morris, J. J., Lenski, R. E. & Zinser, E. R. The Black Queen Hypothesis:
Evolution of Dependencies through Adaptive Gene Loss. mBio 3, e00036–12–
e00036–12 (2012).
15. Hutchins, D. A., Fu, F.-X., Webb, E. A., Walworth, N. & Tagliabue, A. Taxon-
specific response of marine nitrogen fixers to elevated carbon dioxide
concentrations. Nat. Geosci. 6, 1–6 (2013).
16. Behrenfeld, M. J., Siegel, D. A. & O'Malley, R. T. Global ocean phytoplankton
and productivity. State of the Climate in (2007).
17. Moore, C. M. et al. Processes and patterns of oceanic nutrient limitation. Nat.
Geosci. 6, 701–710 (2013).
18. Sunda, W. G. & Huntsman, S. A. Iron uptake and growth limitation in oceanic
and coastal phytoplankton. Marine Chemistry 50, 189–206 (1995).
19. Garcia, N. S., Fu, F., Sedwick, P. N. & Hutchins, D. A. Iron deficiency increases
growth and nitrogen-fixation rates of phosphorus-deficient marine cyanobacteria.
ISME J. 9, 238–245 (2014).
20. Arrigo, K. R. Marine microorganisms and global nutrient cycles. Nature 437,
349–355 (2005).
21. Cavender-Bares, K. K., Mann, E. L., Chisholm, S. W., Ondrusek, M. E. &
Bidigare, R. R. Differential response of equatorial Pacific phytoplankton to iron
fertilization. Limnol. Oceanogr. 44, 237–246 (1999).
22. Geider, R. & La Roche, J. Redfield revisited: variability of C:N:P in marine
microalgae and its biochemical basis. European Journal of Phycology 37, 1–17
(2011).
23. Redfield, A. C. The biological control of chemical factors in the environment.
American scientist (1958). doi:10.2307/27827150
24. Quigg, A. et al. The evolutionary inheritance of elemental stoichiometry in
marine phytoplankton. Nature 425, 291–294 (2003).
25. Van Mooy, B. A. S. et al. Phytoplankton in the ocean use non-phosphorus lipids
in response to phosphorus scarcity. Nature 458, 69–72 (2009).
26. Tuit, C., Waterbury, J. & Ravizza, G. Diel variation of molybdenum and iron in
marine diazotrophic cyanobacteria. Limnol. Oceanogr. 49, 978–990 (2004).
27. Moore, C. M. et al. Relative influence of nitrogen and phosphorous availability
on phytoplankton physiology and productivity in the oligotrophic sub-tropical
North Atlantic Ocean. Limnol. Oceanogr. 53, 291–305 (2008).
28. de Baar, H. J. W. von Liebig's law of the minimum and plankton ecology (1899–
1991). Progress in Oceanography 33, 347–386 (1994).
29. Saito, M. A., Goepfert, T. J. & Ritt, J. T. Some thoughts on the concept of
colimitation: Three definitions and the importance of bioavailability. Limnol.
Oceanogr. 53, 276–290 (2008).
30. Falkowski, P. G. Evolution of the nitrogen cycle and its influence on the
biological sequestration of CO 2 in the ocean. Nature 387, 272–275 (1997).
31. Tyrrell, T. The relative influences of nitrogen and phosphorus on oceanic
primary production. Nature 400, 525–531 (1999).
32. Mills, M. M., Ridame, C., Davey, M., LaRoche, J. & Gelder, R. J. Iron and
phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic.
27
Nature 429, 292–294 (2004).
33. Saito, M. A. & Goepfert, T. J. Zinc–cobalt colimitation of Phaeocystis antarctica.
Limnol. Oceanogr. 53, 1–11 (2008).
34. Saito, M. A. Multiple nutrient stresses at intersecting Pacific Ocean biomes
detected by protein biomarkers. Science 345, 1170–1173 (2014).
35. Chappell, P. D., Moffett, J. W., Hynes, A. M. & Webb, E. A. Molecular evidence
of iron limitation and availability in the global diazotroph Trichodesmium. ISME
J. 6, 1728–1739 (2012).
36. Boyd, P. W. et al. Mesoscale Iron Enrichment Experiments 1993-2005:
Synthesis and Future Directions. Science 315, 612–617 (2007).
37. Deutsch, C., Sarmiento, J. L., Sigman, D. M., Gruber, N. & Dunne, J. P. Spatial
coupling of nitrogen inputs and losses in the ocean. Nature 445, 163–167 (2007).
38. Moutin, T. et al. Phosphate availability and the ultimate control of new nitrogen
input by nitrogen fixation in the tropical Pacific Ocean. Biogeosciences
Discussions 4, 2407–2440 (2007).
39. Thingstad, T. F. et al. Nature of Phosphorus Limitation in the Ultraoligotrophic
Eastern Mediterranean. Science 309, 1068–1071 (2005).
40. Tanaka, T. et al. Lack of P-limitation of phytoplankton and heterotrophic
prokaryotes in surface waters of three anticyclonic eddies in the stratified
Mediterranean Sea. Biogeosciences 8, 525–538 (2011).
41. Klausmeier, C. A., Litchman, E., Daufresne, T. & Levin, S. A. Phytoplankton
stoichiometry. Ecol Res 23, 479–485 (2008).
42. Chappell, P. D. & Webb, E. A. A molecular assessment of the iron stress
response in the two phylogenetic clades of Trichodesmium. Environ. Microbiol.
12, 13–27 (2010).
43. Fu, F. X., Tatters, A. O. & Hutchins, D. A. Global change and the future of
harmful algal blooms in the ocean. Mar. Ecol. Prog. Ser. 470, 207–233 (2012).
44. Boyd, P. W. et al. Marine Phytoplankton Temperature versus Growth Responses
from Polar to Tropical Waters – Outcome of a Scientific Community-Wide
Study. PLoS ONE 8, e63091–17 (2013).
45. Tatters, A. O. et al. Short- and long-term conditioning of a temperate marine
diatom community to acidification and warming. Philosophical Transactions of
the Royal Society B: Biological Sciences 368, 20120437–20120437 (2013).
46. Sigman, D. M., Hain, M. P. & Haug, G. H. The polar ocean and glacial cycles in
atmospheric CO2 concentration. Nature 466, 47–55 (2010).
47. Martínez-Garcia, A. et al. Southern Ocean dust-climate coupling over the past
four million years. Nature 476, 312–315 (2011).
48. Ren, H. et al. Foraminiferal Isotope Evidence of Reduced Nitrogen Fixation in
the Ice Age Atlantic Ocean. Science 323, 244–248 (2009).
49. Sarmiento, J. L., Gruber, N., Brzezinski, M. A. & Dunne, J. P. High-latitude
controls of thermocline nutrients and low latitude biological productivity. Nature
427, 56–60 (2004).
50. Steinacher, M. et al. Projected 21st century decrease in marine productivity: a
multi-model analysis. Biogeosciences 7, 979–1005 (2010).
51. Boyd, P. W. & Brown, C. J. Modes of interactions between environmental
drivers and marine biota. Front. Mar. Sci. 2, 716–8 (2015).
28
52. Sarmiento, J. L. et al. Response of ocean ecosystems to climate warming. Global
Biogeochem. Cycles 18, n/a–n/a (2004).
53. Stramma, L., Johnson, G. C., Sprintall, J. & Mohrholz, V. Expanding Oxygen-
Minimum Zones in the Tropical Oceans. Science 320, 655–658 (2008).
54. Mahowald, N. M. et al. Observed 20th century desert dust variability: impact on
climate and biogeochemistry. Atmospheric Chemistry and Physics 10, 10875–
10893 (2010).
55. Seitzinger, S. P. et al. Global river nutrient export: A scenario analysis of past
and future trends. Global Biogeochem. Cycles 24, n/a–n/a (2010).
56. Poore, A. G. B. et al. Direct and indirect effects of ocean acidification and
warming on a marine plant-herbivore interaction. Oecologia 173, 1113–1124
(2013).
57. Ries, J. Acid ocean cover up. Nature Clim Change (2011).
58. Aronson, R. B., Thatje, S., Clarke, A. & Peck, L. S. Climate change and
invasibility of the Antarctic benthos. Annual Review of … (2007).
doi:10.2307/30033855
59. SEWELL, M. A. & HOFMANN, G. E. Antarctic echinoids and climate change:
a major impact on the brooding forms. Global Change Biol 17, 734–744 (2011).
60. Hutchins, D. A. et al. Irreversibly increased nitrogen fixation in Trichodesmium
experimentally adapted to elevated carbon dioxide. Nat. Commun. 6, (2015).
61. Schlichting, C. D. Hidden Reaction Norms, Cryptic Genetic Variation, and
Evolvability. Annals of the New York Academy of Sciences 1133, 187–203
(2008).
62. Merilä, J. & Hendry, A. P. Climate change, adaptation, and phenotypic plasticity:
the problem and the evidence. Evol Appl 7, 1–14 (2014).
63. Schaum, E., Rost, B., Millar, A. J. & Collins, S. Variation in plastic responses of
a globally distributed picoplankton species to ocean acidification. Nat. Clim.
Chang. 3, 298–302 (2012).
64. Schaum, C. E. & Collins, S. Plasticity predicts evolution in a marine alga. Proc.
R. Soc. B 281, 20141486 (2014).
65. Schaum, C.-E., Rost, B. O. R. & Collins, S. E. A. Environmental stability affects
phenotypic evolution in a globally distributed marine picoplankton. ISME J. 10,
75–84 (2016).
66. Boyd, P. W., Strzepek, R., Fu, F. & Hutchins, D. A. Environmental control of
open-ocean phytoplankton groups: Now and in the future. Limnol. Oceanogr. 55,
1353–1376 (2010).
67. Fu, F.-X., Warner, M. E., Zhang, Y., Feng, Y. & Hutchins, D. A. EFFECTS OF
INCREASED TEMPERATURE AND CO 2ON PHOTOSYNTHESIS,
GROWTH, AND ELEMENTAL RATIOS IN MARINE
SYNECHOCOCCUSAND PROCHLOROCOCCUS(CYANOBACTERIA). J.
Phycol. 43, 485–496 (2007).
68. Traving, S. J., Clokie, M. R. J. & Middelboe, M. Increased acidification has a
profound effect on the interactions between the cyanobacterium Synechococcus
sp. WH7803 and its viruses. FEMS Microbiology Ecology 87, 133–141 (2014).
69. Lomas, M. W., Hopkinson, B. M. & Losh, J. L. Effect of ocean acidification on
cyanobacteria in the subtropical North Atlantic. Aquat. Microb. Ecol. 66, 211–
29
222 (2012).
70. Rusch, D. B., Martiny, A. C. & Dupont, C. L. Characterization of
Prochlorococcus clades from iron-depleted oceanic regions. in (2010).
71. Sohm, J. A. et al. Co-occurring Synechococcus ecotypes occupy four major
oceanic regimes defined by temperature, macronutrients and iron. ISME J. 1–13
(2015). doi:10.1038/ismej.2015.115
72. Paulino, A. I., Egge, J. K. & Larsen, A. Effects of increased atmospheric
CO₂ on small and intermediate sized osmotrophs during a nutrient
induced phytoplankton bloom. Biogeosciences 5, 739–748 (2008).
73. Sobrino, C., Ward, M. L. & Neale, P. J. Acclimation to elevated carbon dioxide
and ultraviolet radiation in the diatom Thalassiosira pseudonana: Effects on
growth, photosynthesis, and spectral sensitivity of …. Limnol. Oceanogr. (2008).
74. Li, Y., Xu, J. & Gao, K. Light-Modulated Responses of Growth and
Photosynthetic Performance to Ocean Acidification in the Model Diatom
Phaeodactylum tricornutum. PLoS ONE 9, e96173 (2014).
75. Hoogstraten, A., Timmermans, K. R. & de Baar, H. J. W. MORPHOLOGICAL
AND PHYSIOLOGICAL EFFECTS IN PROBOSCIA ALATA
(BACILLARIOPHYCEAE) GROWN UNDER DIFFERENT LIGHT AND CO2
CONDITIONS OF THE MODERN SOUTHERN OCEAN1. J. Phycol. 48, 559–
568 (2012).
76. Ratti, S., Giordano, M. & Morse, D. CO2-CONCENTRATING MECHANISMS
OF THE POTENTIALLY TOXIC DINOFLAGELLATE PROTOCERATIUM
RETICULATUM (DINOPHYCEAE, GONYAULACALES)1. J. Phycol. 43,
693–701 (2007).
77. Tatters, A. O., Fu, F.-X. & Hutchins, D. A. High CO 2 and Silicate Limitation
Synergistically Increase the Toxicity of Pseudo-nitzschia fraudulenta. PLoS ONE
7, e32116 (2012).
78. Sun, J., Hutchins, D. A., Feng, Y. & Seubert, E. L. Effects of changing pCO2
and phosphate availability on domoic acid production and physiology of the
marine harmful bloom diatom Pseudo-nitzschia multiseries. Limnology and …
(2011).
79. Fu, F. X., Place, A. R. & Garcia, N. S. CO2 and phosphate availability control
the toxicity of the harmful bloom dinoflagellate Karlodinium veneficum. Aquatic
Microbial … (2010).
80. Tatters, A. O., Flewelling, L. J., Fu, F., Granholm, A. A. & Hutchins, D. A. High
CO2 promotes the production of paralytic shellfish poisoning toxins by
Alexandrium catenella from Southern California waters. Harmful Algae 30, 37–
43 (2013).
81. Bach, L. T. et al. Dissecting the impact of CO2 and pH on the mechanisms of
photosynthesis and calcification in the coccolithophore Emiliania huxleyi. New
Phytol. 199, 121–134 (2013).
82. Iglesias-Rodriguez, M. D. et al. Phytoplankton Calcification in a High-CO2
World. Science 320, 336–340 (2008).
83. Beaufort, L. et al. Sensitivity of coccolithophores to carbonate chemistry and
ocean acidification. Nature 476, 80–83 (2011).
84. Zondervan, I., Zeebe, R. E., Rost, B. & Riebesell, U. Decreasing marine biogenic
30
calcification: A negative feedback on rising atmospheric pCO2. Global
Biogeochem. Cycles 15, 507–516 (2001).
85. Lefebvre, S. C. et al. Nitrogen source and pCO2 synergistically affect carbon
allocation, growth and morphology of the coccolithophore Emiliania huxleyi:
potential implications of ocean acidification for the carbon cycle. Global Change
Biol 18, 493–503 (2012).
86. Feng, Y. et al. Interactive effects of increased pCO2, temperature and irradiance
on the marine coccolithophore Emiliania huxleyi (Prymnesiophyceae). European
Journal of Phycology 43, 87–98 (2008).
87. Riebesell, U. Effects of CO2 Enrichment on Marine Phytoplankton. Journal of
Oceanography 60, 719–729 (2004).
88. Tortell, P. D., DiTullio, G. R. & Sigman, D. M. CO~ 2 effects on taxonomic
composition and nutrient utilization in an Equatorial Pacific phytoplankton
assemblage. Marine Ecology-Progress … (2002).
89. Tortell, P. D. et al. CO2 sensitivity of Southern Ocean phytoplankton.
Geophysical Research Letters 35, L04605 (2008).
90. Hare, C. E., Leblanc, K. & DiTullio, G. R. Consequences of increased
temperature and CO~ 2 for phytoplankton community structure in the Bering Sea.
MARINE ECOLOGY- … (2007).
91. Riebesell, U. et al. Technical Note: A mobile sea-going mesocosm system – new
opportunities for ocean change research. Biogeosciences 10, 1835–1847 (2013).
92. Hoppe, C. J. M. et al. Iron Limitation Modulates Ocean Acidification Effects on
Southern Ocean Phytoplankton Communities. PLoS ONE 8, e79890–10 (2013).
93. Böhme, H. Regulation of nitrogen fixation in heterocyst-forming cyanobacteria.
Trends in Plant Science 3, 346–351 (1998).
94. Finzi-Hart, J. A. et al. Fixation and fate of C and N in the cyanobacterium
Trichodesmium using nanometer-scale secondary ion mass spectrometry. Proc.
Natl. Acad. Sci. U.S.A. 106, 6345–6350 (2009).
95. Ohki, K. Intercellular localization of nitrogenase in a non-heterocystous
cyanobacterium (cyanophyte), Trichodesmium sp. NIBB1067. Journal of
Oceanography 64, 211–216 (2008).
96. Fredriksson, C. & Bergman, B. Ultrastructural characterisation of cells
specialised for nitrogen fixation in a non-heterocystous
cyanobacterium,Trichodesmium spp. Protoplasma 197, 76–85 (1997).
97. Lin, S., Henze, S., Lundgren, P., Bergman, B. & Carpenter, E. Whole-cell
immunolocalization of nitrogenase in marine diazotrophic cyanobacteria,
trichodesmium spp. Appl. Environ. Microbiol. 64, 3052–3058 (1998).
98. El-Shehawy, R., Lugomela, C., Ernst, A. & Bergman, B. Diurnal expression of
hetR and diazocyte development in the filamentous non-heterocystous
cyanobacterium Trichodesmium erythraeum. Microbiology 149, 1139–1146
(2003).
99. Barcelos e Ramos, J., Biswas, H., Schulz, K. G., LaRoche, J. & Riebesell, U.
Effect of rising atmospheric carbon dioxide on the marine nitrogen fixer
Trichodesmium. Global Biogeochem. Cycles 21, GB2028 (2007).
100. Levitan, O. et al. Elevated CO
2
enhances nitrogen fixation and growth in the
marine cyanobacterium Trichodesmium. Global Change Biol 13, 531–538 (2007).
31
101. Levitan, O. et al. Combined effects of CO2 and light on the N2-fixing
cyanobacterium Trichodesmium IMS101: a mechanistic view. Plant Physiol. 154,
346–356 (2010).
102. Levitan, O., Sudhaus, S., LaRoche, J. & Berman-Frank, I. The Influence of
pCO2 and Temperature on Gene Expression of Carbon and Nitrogen Pathways in
Trichodesmium IMS101. PLoS ONE 5, e15104–11 (2010).
103. Garcia, N. S. et al. Interactive effects of irradiance and CO
2
on CO
2
fixation and
N
2
fixation in the diazotroph Trichodesmium erythraeum (cyanobacteria). J.
Phycol. 47, 1292–1303 (2011).
104. Spungin, D., Berman-Frank, I. & Levitan, O. Trichodesmium's strategies
to alleviate phosphorus limitation in the future acidified oceans. Environ.
Microbiol. 16, 1935–1947 (2014).
105. Shi, D., Kranz, S. A., Kim, J.-M. & morel, F. M. M. Ocean acidification slows
nitrogen fixation and growth in the dominant diazotroph Trichodesmium under
low-iron conditions. Proc. Natl. Acad. Sci. U.S.A. 109, E3094–100 (2012).
106. Moisander, P. H. et al. Unicellular Cyanobacterial Distributions Broaden the
Oceanic N
2
Fixation Domain. Science 327, 1512–1514 (2010).
107. Fu, F.-X. et al. Interactions between changing pCO
2
, N
2
fixation, and Fe
limitation in the marine unicellular cyanobacterium Crocosphaera. Limnol.
Oceanogr. 53, 2472–2484 (2008).
108. Garcia, N. S. et al. Combined effects of CO2 and light on large and small isolates
of the unicellular N2-fixing cyanobacterium Crocosphaera watsonii from the
western tropical Atlantic Ocean. European Journal of Phycology 48, 128–139
(2013).
109. Garcia, N. S., Fu, F. & Hutchins, D. A. Colimitation of the unicellular
photosynthetic diazotroph Crocosphaera watsonii by phosphorus, light and
carbon dioxide. Limnol. Oceanogr. (2013).
110. Eichner, M., Rost, B. & Kranz, S. A. Diversity of ocean acidification effects on
marine N2 fixers. Journal of Experimental Marine Biology and Ecology 457,
199–207 (2014).
111. Wannicke, N. et al. Response of Nodularia spumigena to
pCO₂ – Part 1: Growth, production and nitrogen cycling.
Biogeosciences 9, 2973–2988 (2012).
112. Luo, Y. W. et al. Database of diazotrophs in global ocean: abundance, biomass
and nitrogen fixation rates. Earth System Science Data 4, 47–73 (2012).
113. Zehr, J. P. et al. Globally Distributed Uncultivated Oceanic N2-Fixing
Cyanobacteria Lack Oxygenic Photosystem II. Science 322, 1110–1112 (2008).
114. Law, C. S. et al. No stimulation of nitrogen fixation by non-filamentous
diazotrophs under elevated CO2 in the South Pacific. Global Change Biol 18,
3004–3014 (2012).
115. Lima-Mendez, G. et al. Determinants of community structure in the global
plankton interactome. Science 348, 1262073–1262073 (2015).
116. Bergman, B., Sandh, G., Lin, S., Larsson, J. & Carpenter, E. J. Trichodesmium–
a widespread marine cyanobacterium with unusual nitrogen fixation properties.
FEMS Microbiol Rev 37, 286–302 (2013).
117. Elena, S. F. & Lenski, R. E. Microbial genetics: Evolution experiments with
32
microorganisms: the dynamics and genetic bases of adaptation. Nat Rev Genet 4,
457–469 (2003).
118. Lohbeck, K. T., Riebesell, U. & Reusch, T. B. H. Adaptive evolution of a key
phytoplankton species to ocean acidification. Nat. Geosci. 5, 346–351 (2012).
119. Jin, P., Gao, K. & Beardall, J. EVOLUTIONARY RESPONSES OF A
COCCOLITHOPHORID GEPHYROCAPSA OCEANICA TO OCEAN
ACIDIFICATION. Evolution 67, 1869–1878 (2013).
120. Benner, I. et al. Emiliania huxleyi increases calcification but not expression of
calcification-related genes in long-term exposure to elevated temperature and
pCO2. Philosophical Transactions of the Royal Society B: Biological Sciences
368, 20130049–20130049 (2013).
121. Lohbeck, K. T., Riebesell, U., Collins, S. & Reusch, T. B. H. FUNCTIONAL
GENETIC DIVERGENCE IN HIGH CO2 ADAPTED EMILIANIA HUXLEYI
POPULATIONS. Evolution 67, 1892–1900 (2013).
122. Reusch, T. B. H. & Boyd, P. W. EXPERIMENTAL EVOLUTION MEETS
MARINE PHYTOPLANKTON. Evolution 67, 1849–1859 (2013).
123. Collins, S. & Bell, G. Phenotypic consequences of 1,000 generations of selection
at elevated CO2 in a green alga. Nature 431, 566–569 (2004).
124. Low-Décarie, E., Jewell, M. D., Fussmann, G. F. & Bell, G. Long-term culture at
elevated atmospheric CO2 fails to evoke specific adaptation in seven freshwater
phytoplankton species. Proc. R. Soc. B 280, 20122598–20122598 (2013).
125. Low-Décarie, E., Fussmann, G. F. & Bell, G. The effect of elevated CO2 on
growth and competition in experimental phytoplankton communities. Global
Change Biol 17, 2525–2535 (2011).
126. Collins, S. Competition limits adaptation and productivity in a photosynthetic
alga at elevated CO2. Proc. R. Soc. B 278, rspb20101173–255 (2010).
127. Tatters, A. O. et al. SHORT- VERSUS LONG-TERM RESPONSES TO
CHANGING CO2 IN A COASTAL DINOFLAGELLATE BLOOM:
IMPLICATIONS FOR INTERSPECIFIC COMPETITIVE INTERACTIONS
AND COMMUNITY STRUCTURE. Evolution 67, 1879–1891 (2013).
128. Sunday, J. M. et al. Evolution in an acidifying ocean. Trends in Ecology &
Evolution 29, 117–125 (2014).

33

Chapter 2
Trichodesmium genome maintains abundant, widespread noncoding DNA in situ,
despite oligotrophic lifestyle

Trichodesmium genome maintains abundant,
widespread noncoding DNA in situ, despite
oligotrophic lifestyle
Nathan Walworth
a
, Ulrike Pfreundt
b
, William C. Nelson
c
, Tracy Mincer
d
, John F. Heidelberg
a
, Feixue Fu
a
,
John B. Waterbury
e
, Tijana Glavina del Rio
f
, Lynne Goodwin
f
, Nikos C. Kyrpides
f
, Miriam L. Land
g
, Tanja Woyke
f
,
David A. Hutchins
a
, Wolfgang R. Hess
b
, and Eric A. Webb
a,1
a
Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089;
b
Genetics and Experiment Bioinformatics, University of
Freiburg, 79098 Freiburg, Germany;
c
Fundamental and Computational Sciences, Pacific Northwest National Laboratory, Richland, WA 99352;
d
Marine
Chemistry and Geochemistry Department and
e
Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543;
f
Joint Genome
Institute, Walnut Creek, CA 94598; and
g
Oak Ridge National Laboratory, Oak Ridge, TN 37831
Edited by Edward F. DeLong, University of Hawaii, Manoa, Honolulu, HI, and approved February 10, 2015 (received for review November 26, 2014)
Understanding the evolution of the free-living, cyanobacterial,
diazotroph Trichodesmiumisofgreatimportancebecauseofits
criticalroleinoceanicbiogeochemistryandprimaryproduction.Un-
like the other>150 available genomes of free-living cyanobacteria,
only 63.8% of the Trichodesmium erythraeum (strain IMS101) ge-
nomeispredictedtoencodeprotein,whichis20–25%lessthanthe
averageforothercyanobacteriaandnonpathogenic,free-livingbac-
teria. We use distinctive isolates and metagenomic data to show
that low coding density observed in IMS101 is a common feature
of the Trichodesmium genus, both in culture and in situ. Transcrip-
tome analysis indicates that 86% ofthe noncoding space is expressed,
although the function of these transcripts is unclear. The density of
noncoding,possibleregulatoryelementspredictedinTrichodesmium,
when normalized per intergenic kilobase, was comparable and two-
fold higher than that found in the gene-dense genomes of the
sympatriccyanobacterialgeneraSynechococcusandProchlorococcus,
respectively. Conserved Trichodesmium noncoding RNA secondary
structures were predicted between most culture and metagenomic
sequences, lending support to the structural conservation. Conser-
vation of these intergenic regions in spatiotemporally separated
Trichodesmiumpopulations suggests possible genus-wide selection
for their maintenance. These large intergenic spacers may have de-
veloped during intervals of strong genetic drift caused by periodic
bloomsofasubsetofgenotypes,whichmayhavereducedeffective
population size. Our data suggest that transposition of selfish
DNA, low effective population size, and high-fidelity replication
allowed the unusual “inflation” of noncoding sequence observed
in Trichodesmium despite its oligotrophic lifestyle.
marinemicrobiology
|
oligotrophic
|
evolution genomics
|
nitrogenfixation
T
he low availability of N (and fixed carbon) in the midlatitude
upper oceans provides an important niche for autotrophic
organisms that can fix atmospheric nitrogen, which can exert
control over global primary production (1–3). Nitrogen fixation
is a prokaryotic process with a high-energy demand, and oceanic
cyanobacteria are known to be significant sources of this “new”
nitrogen (nitrogen that is fixed from the atmosphere or NO
3
advected from depth) (4, 5). Molecular field data have shown
that a handful of cyanobacterial diazotrophs responsible for ol-
igotrophic nitrogen fixation can reach relatively high cell num-
bers (6–12) and be of significant biogeochemical importance
(13–16). These include photosynthetic free-living forms, such as
thefilamentous Trichodesmiumandthe unicellular Crocosphaera
and Cyanothece,andphotosyntheticandnonphotosynthetic
symbiotic forms, such as heterocystous Richelia and Candidatus
Atelocyanobacterium thalassa (3, 5, 17).
Trichodesmium cells can grow eitheras trichomes (i.e., filaments)
or aggregates and form three types of classically described colonies,
including radial puffs, vertically aligned fusiform tufts, and bowties
(e.g., refs. 18–20). Trichodesmium spp. form blooms throughout the
nitrogen-limited Atlantic and Pacific Oceans (21), as well as the
ArabianandRedSeas(19,22),andintheNorthPacificSubtropical
Gyre, they dominate a recurrent annual phytoplankton bloom (23).
The up-to-multimillimeter-sized Trichodesmium colony environ-
ment can be an oasis of fixed N and C in the oligotrophic oceans
(24, 25) and has been observed to contain a varied assemblage of
organisms,rangingfromprokaryotesandunicellulareukaryotesto
juvenile copepods and decapods (26–31) with metabolisms in-
cludingheterotrophsandanoxygenicandoxygenicphototrophs,as
well as mixotrophic eukaryotes. Thus, in contrast to most other
unicellularoligotrophs,Trichodesmiumcaneitherliveinacolonial
habitat dominated by extensive physical interactions with both
sister cells and other taxa and/or as free trichomes that can con-
stitute a significant fraction of the Trichodesmium water column
biomass (32). These varying ecological lifestyles can partition
Trichodesmium into different subpopulations with dynamic states
including multiple morphologies, genotypes, and varied physical
interactions. Furthermore, a specific population may interchange
between trichome- and colony-dominated biomass.
Significance
Thefree-livingcyanobacteriumTrichodesmiumisamajorsource
ofnewnitrogenandfixedcarbontothetropicalandsubtropical
oceans, but despite its importance, we know little about the
molecular mechanisms it uses to succeed in its oligotrophic
habitat.Hereweshowthatitsgene-sparsegenomeislittered
with large, conserved, expressed intergenic spaces, which is
atypicalformostknownfree-livingprokaryotes.Paradoxically,
although its genome is enriched in predicted transposases and
repeatsequences,itexhibitsconservedintragenussyntenyand
similar intergenic architecture relative to its sympatric, gene-
dense relatives Prochlorococcus and Synechococcus. This ob-
servationdemonstratesasuccessfulalternativetothegenomic
streamlining strategy observedin other free-livingoligotrophs
such as Prochlorococcus or Pelagibacter.
Author contributions: N.W., T.M., D.A.H., W.R.H., and E.A.W. designed research; N.W.,
U.P., W.C.N., T.M., J.F.H., F.F., J.B.W., T.G.d.R., L.G., N.C.K., M.L.L., T.W., and E.A.W. per-
formed research; N.W., U.P., W.C.N., T.M., J.F.H., W.R.H., and E.A.W. analyzed data; and
N.W.,U.P.,W.C.N.,T.M.,J.F.H.,F.F.,J.B.W.,T.G.d.R.,L.G.,N.C.K.,M.L.L.,T.W.,D.A.H.,W.R.H.,
and E.A.W. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
Data deposition: The Fastq files have been deposited into the NCBI Sequence Read Archive
(SRA), www.ncbi.nlm.nih.gov/sra (accession nos. SAMN02199363 and SAMN02199364). The
genomes of IMS101, 2175, and H94 have been deposited into the NCBI WGS database,
www.ncbi.nlm.nih.gov (accessionnos.SAMN02598485,SAMN03421191,andSAMN03421272.
1
To whom correspondence should be addressed. Email: eawebb@usc.edu.
Thisarticlecontainssupportinginformationonlineatwww.pnas.org/lookup/suppl/doi:10.
1073/pnas.1422332112/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1422332112 PNAS Early Edition | 1of6
ENVIRONMENTAL
SCIENCES
EVOLUTION
Trichodesmium erythraeum IMS 101 (hereafter IMS101) was
isolated in 1991 by Prufert-Bebout et al. from the Gulf Stream
off the coast of North Carolina (33), and since then a handful
of other strains have been isolated by Waterbury and others
(20, 34, 35); however, virtually nothing is known about the sim-
ilarity of its unusual genome architecture to the genome struc-
ture of natural populations. Additionally, efforts to cryopreserve
Trichodesmium have been unsuccessful, which presents a challenge
due to the requirement of constant culturing and maintenance.
In recent years, numerous studies on unicellular, oligotrophic
microbes have observed relatively high coding percentages and
genomic streamlining in sympatric, nondiazotrophic cyanobacteria
(e.g., refs. 36–41) as well as other dominant, marine heterotrophic
bacteria (e.g., refs. 42–44). In these aforementioned studies,
streamlining has been used to describe both gene loss and general
genome reduction primarily as a result of selection for greater
metabolic efficiency in nutrient-poor regimes rather than mainly
derivingfrompopulation-levelprocesses.Accordingly,population-
level processes suggest that the reduced impact of genetic drift as
a result of elevated effective population sizes exhibited by free-
living microbial taxa is sufficient to allow for selection against ac-
cumulationofexcessDNAwithinapopulation(45,46).Thislatter
population-level stance argues that microbial genome reduction
has been primarily a result of weakened nonadaptive forcesrather
than any one strong selective force increasing metabolic efficiency.
Hence, fitting with a small predicted effective population size
as well as the absence of nitrogen limitation, many cyanobacterial
diazotrophs (e.g., Crocosphaera, Cyanothece, Cylindrospermopsis,
etc.)donothavethe conventionally defined streamlinedgenomes
seeninothermarinetaxa(40,42,45–47).SimilartoIMS101,these
genomes are enriched in predicted insertion sequences, repeats,
andregulatoryproteins(41,48,49),yetdespitethisfactstillretain
much larger coding percentages (≥80%) than Trichodesmium.
Until now, the low gene density and large intergenic spacers have
only been observed in the genome of IMS101 that has been
maintained in culture for approximately two decades. Here, we
explorethedegreeofgenomearchitectureconservationbetween
spatiotemporally segregated Trichodesmium isolates as well as
natural populations.
A defining feature of Trichodesmium ecology involves popu-
lations with dynamic, multicellular morphotypes (e.g., single tri-
chomes or different colony types), where one of these forms may
dominatethepopulation(includingblooms)atdifferenttimes(e.g.,
refs. 19, 31, and 32). Therefore, we hypothesize that high trans-
poson load coupled to periodic nonadaptive, bloom-driven reduc-
tions in effective population size from a subset of morphological
genotypes harboring different epibiotic consortia may have sub-
stantially contributed to production and maintenance of large
intergenic regions, proliferation of repetitive DNA, and subse-
quent selection for noncoding regulatory regions. This trajectory
may have ultimately allowed Trichodesmium to evolve one of
most intergenic-rich genomes of any free-living prokaryote.
Results and Discussion
IMS101 Genome Does Not Exhibit Streamlining. Genomic streamlining
has been associated with vitamin and amino acid auxotrophies,
simplifiedcarbonandnitrogenmetabolism,alowabundanceof
pseudogenes, limited motility, and little to no selfish DNA (e.g.,
insertion sequences, transposons, etc.) (36–38, 42, 50–52). Com-
pared with four sympatric reference marine picoplankton genomes
(Synechococcus WH8102 and CC9311 and Prochlorococcus SS120
and MED4), the IMS101 genome displays characteristics that are
inconsistent with streamlining. IMS101 has a large suite of regula-
tory proteins (COG K, transcriptional regulators subgroup; 141 vs.
55averageofthefourgenomes),motility-relatedproteins(COGN;
56 vs. 6.5 average), increased number of transport-related proteins
(COG P; 125 vs. 75.25 average), signal transduction proteins
(COG T; 146 vs. 41.75), a large number of transposase sequences
(165 vs. zero), and numerous pseudogenes (625 vs. 3). Normalizing
the number of these genes in IMS101 either by genome size
(IMS101 is 3.7 times greater than the picoplankton average) or
gene count (IMS101 is 2.2 times greater than the picoplankton
average) shows that regulators and transporters roughly scaled
with the increases in size, while transposases, motility genes, signal
transducers, and pseudogenes (Figs. S1 and S2) were enriched
inIMS101.Thesedatasuggestthatunliketherecentlydescribed
photofermentative, cyanobacterial symbiont, Candidatus Atelocya-
nobacterium thalassa (17, 38), and marine Synechococcus and
Prochlorococcus,themetabolismpredictedfor Trichodesmiumisnot
minimized.
Basedon Integrated Microbial Genomes(IMG)COG analysis,
Crocosphaera (strains WH8501, WH0003, and WH0401, average
valuesreportedbelow)andTrichodesmiumretainsimilarnumbers
ofsignaltransductionproteins(131vs.146)andtransporter-related
proteins(125vs.110),whereasCrocosphaerapossessesslightlymore
transcriptional regulator proteins (29 vs. 13) and motility proteins
(33 vs. 29) (Dataset S1). In terms of gene content, Crocosphaera
does not necessarily exhibit a streamlined profile but its coding
percentage is still that of an average free-living prokaryote
(∼75–80%), whereas Trichodesmium also retains similar pro-
tein content to Crocosphaera but has a substantially reduced
coding percentage.
Accordingly, the IMS101 genome encodes 5,076 proteins (per
the IMG annotation https://img.jgi.doe.gov/), yielding a coding
percentage of∼60%, whereas its sympatric, picoplanktonic cousins
Prochlorococcus,marine Synechococcus, Crocosphaera watsonii,and
Cyanothece (41, 48, 53, 54), and all of the 45 currently sequenced
membersoftheOscillatoriales(thecyanobacterialorderwithwhich
IMS101isphylogeneticallyplaced)havecodingpercentages>75%.
The Oscillatoriales demonstrates variation both in gene count and
genome size (average gene number = 5,663.6 ± 1,150.1; average
genome size= 6,346,139± 1,207,411), and although the number of
genesinIMS101iswithinthisrange,thecodingpercentageandthe
genome size are at opposite ends of the spectrum, respectively
(Dataset S1). Furthermore, principal component analysis (PCA)
of cyanobacterial genome features segregates Trichodesmium
from the rest of the taxa based on its substantially lower coding
percentage relative to other characteristics (Fig. S1 and Dataset
S1). Similarly, a PCA including only cyanobacteria that possess
annotated transposases also shows Trichodesmium segregating
away from other genomes opposite the axis of coding percent-
age, whereas C. watsonii WH 8501 segregates away from others
basedonthe1,000+annotatedtransposasesinitsgenomeunlike
other sequenced Crocosphaera genomes (41) (Fig. S1B). These
data imply that, although IMS101 has a “normal” number of
genes for a filamentous, diazotrophic cyanobacterium, its non-
coding space is unique.
Long Intergenic Regions Are Conserved in Trichodesmium. IMS101
was in culture for >10 y (33) before genome sequencing was
initiated in 2003. To determine whether the unusual genomic
characteristics observed in IMS101 were in common in the
T. erythraeumspecies,wegenerateddraftgenomesequencefrom
a more recently isolated strain of T. erythraeum [strain 21–75
(2175) isolated from the Tropical Atlantic in 6/2006] that was
only in culture for∼1 y before sequencing (Datasets S2 and S3).
Both T. erythraeum strains have relatively large genomes (7–7.78
Mbps),lowGCcontent(∼33–34%),andareducedprotein-coding
percentage (∼61%). We also obtained a partial genome sequence
from T. thiebautii H94, a Hawaiian isolate from 2004 that is rep-
resentative of the other major Trichodesmium clade currently in
culture (20), and although this 2009 sequencing run returned low
coverage of the H94 genome, the contigs that assembled showed
a similarly low coding percentage of ∼61% (Dataset S4). Thus,
from this limited genomic analysis of three cultured isolates, it
appears that the low coding percentage and large genome ob-
served in IMS101 are commonplace in the genus.
High levels of synteny were also observed between the two
T. erythraeum isolates, even through the long, noncoding inter-
genic regions. According to a MAUVE alignment, the IMS101
and 2175 genomes contain 28 colinear blocks ranging in size
from 1,700 bp up to∼2.5 Mbp (Fig. 1A).
2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1422332112 Walworth et al.
Additionally,although2175isadraftgenome,wewereableto
recover almost all (98%) IMS101 protein-encoding and inter-
genic regions (96%) using BLASTn (Dataset S5). Homologous
intergenic regions between IMS101 and 2175 averaged 510 bp
in length with a median of 380 bp. In contrast, the intergenic
regions unique to each strain were considerably shorter, with
an average length of 53 bp and a median of 21 bp. These results
suggest that the unusually long intergenic regions may confer
somesortofselectableadvantagethatensurestheirconservation
or, conversely, a robust maintenance mechanism that does not
act upon the shorter regions.
To place these noncoding characteristics in the context of
otherbacteriaacrossabroadrangeofrepresentativephylogenetic
origins (55), we performed PCA on multiple genomic features
(Dataset S1), including the relative size and distribution of inter-
genic regions. Fig. 1B shows both Trichodesmium genomes seg-
regating along the axis associated with the median intergenic
spacer length, suggesting that maintenance or accumulation of
noncodingDNAisdistributedthroughoutthegenomeratherthan
confined to a few intergenic regions. This evidence contrasts with
previous observations of predominant regulatory protein accu-
mulation in free-living bacteria rather than noncoding DNA (47,
55, 56), anditfurther emphasizesthe unique size and distribution
of Trichodesmium intergenic regions relative to other genome fea-
tures in bacteria.
Environmental Populations of Trichodesmium Have Low Gene Density.
To extend our analysis of Trichodesmium genome structure to en-
vironmental populations, a colony-enriched metagenome was gen-
erated from Trichodesmium colonies in the North Atlantic
Subtropical Gyre. Fragment recruitment of metagenomic reads
against IMS101 shows nearly complete coverage in mixed natural
Trichodesmium populations (Fig. 1C, Upper). Although many of
the assembled metagenomic contigs (n= 460,494) were relatively
small (N50= 1,217), a considerable amount of larger contigs was
generated (n= 1,032; N50= 4,335; max length= 12,688) as well.
When these larger contigs are mapped to the IMS101 scaffold
using nucmer, the subsequent alignment plot strongly suggests
genome synteny between IMS101 and natural populations in situ
(Fig. 1C, Lower). Furthermore, the metagenomic dataset con-
tained∼94% of the IMS101 and 2175 intergenic sequences. Un-
detected intergenic sequences for both IMS101 and 2175 again
hadsmallaverageintergeniclengthsof85and57bpandmedians
of 27.5 and 21 bp, respectively. The near-complete in situ de-
tection of each genome’s longer intergenic sequences in contrast
to the short averages and medians of the undetected intergenic
sequences further suggests that in natural populations, longer
intergenicregionsmaybeselectivelymaintained,lendingevidence
totheirpotentialphysiologicalimportancetotheinsituecologyof
Trichodesmium spp.
Intergenic Regions and Repetitive Elements.The intergenic regions
of the IMS101 genome contain numerous DNA repeats, ranging
from very small noncoding elements [e.g., highly interspersed
palindromicsequences(35,57)orotherrepeatingsequences(58,
59)] to larger, gene-encoding insertion sequences (e.g., ref. 60).
Because many of the repeating elements can overlap and/or be
nested inside of each other, we assessed the contribution of
intergenic repeats to the total intergenic space of 2,801,094 bp
(SI Materials and Methods). We counted an intergenic sequence
as a repeat if it occurred two or more times in the genome.
Hence, when comparing IMS101 intergenic regions against the
IMS101 scaffold by using BLASTn and summing the length of
each repeat sequence hit, it yielded ∼4.1 Mb of nested over-
lapping repeat sequence. However, when these nested repeats
were consolidated into discrete nonoverlapping repeat sequen-
ces, only one-third of IMS101 noncoding DNA (804,807 bp)
consisted of repetitive elements, or ∼10% of the total genome
(SIMaterialsandMethods).Similarly,approximatelyone-thirdof
2175 noncoding DNA (799,980 of 2,708,763 bp of total inter-
genic space) consisted of nonoverlapping repetitive elements
(11.5% of the 2175 genome). Some of these repetitive elements
are likely to be mobile, because 58% of the 625 IMS101 pseu-
dogenes are interrupted by a repeat sequence. Regardless of the
function of these repeats, their conservation across time and
spacefrombottleneckedisolatestonaturalcommunitiesstrongly
suggests that selection for maintenance of these elements exists
within this genus.
The distribution of repeats around a genome can give insights
intothemechanismbywhichtheypropagate[i.e.,DNAreplication
slippage for tandem and transposition/recombination for distrib-
uted repeats, respectively (58, 61, 62)]. Here, we compared the
distribution of putative transposase genes and insertion sequences
in the IMS101 genome to the top 10 intergenic regions containing
the most abundant repeats identified in our pipeline (Table S1).
Although the predicted insertion sequences are distributed around
the genome, areas of increased density were apparent at∼3, 8, 9,
and 12 o’clock on the genome (Fig. 2). Focusing on the∼7.49- to
7.54-Mbp regional cluster, we usedBLASTntoidentifylocations
containing numerous overlapping sequence elements, including
insertion sequences, predicted transposase-related genes, and re-
petitive elements (Fig. 2, Inset), suggesting that this region may be
arecombination “hot spot,” with both DNA polymerase slippage
and transposition causing genetic elements to be stacked on top
of each other. Further, we identified sequences from this re-
gion in a publicly available metatranscriptome containing
Trichodesmium colonies that is geospatially distinct from our
metagenomesamples(SouthPacificvs.NorthAtlantic,respectively)
(27). These results either suggest that this genomic region is gen-
erally conserved and active in the genus, oratleast thatthesingle
copy elements comprising the region in IMS101 are active and
conserved at high identity in other members of the genus, even if
the arrangement observed in region 7.49–7.54 is not.
Inagene-centricstudycomparingCrocosphaeragenomes(41),
itwasobservedthatmoststrainsdidnotcontainhighlyrepetitive
ORFs, with the exception of Crocosphaera watsonii WH 8501,
-2 0 2 4 6
-3 -2 -1 0 1 2
Genome Size
Percent coding
Total Gene Number
Average
Gene Length
Mean Intergenic
Spacer Length
Median Intergenic
Spacer Length
90%
80%
70%
100%
IMS101
2175
Nucleotide Identity
IMS101 Chromosome
IMS101
2175
Metagenome
PC1 (62%)
PC2 (19%)
A
B
C
Fig. 1. Differences between Trichodesmium genome content and other
bacteria and conservation of Trichodesmium genome both in culture and
in situ. (A) MAUVE alignment showing high synteny between the fin-
ished IMS101 genome and the 2175 draft genome. (B) PCA on multiple
genomicfeaturessegregatesIMS101and2175alongthemedian intergenic
spacer length axis relative to small (<2Mb,green),medium(2–4Mb,
red), and large (>4 Mb, black) bacterial genomes (55), indicating that
noncoding sequence accumulation is distributed throughout the ge-
nome. (C)Upper is afragmentrecruitmentmapofenrichedTrichodesmium
metagenomicreadsfromtheNorthAtlanticSubtropicalGyremappedtothe
IMS101 genome. Lower is a mummer plot generated from a nucmer align-
ment between the IMS101 chromosome and the large metagenomic contigs
(N50 = 4,335; max = 12,688), suggesting genome synteny of the IMS101
genome in natural populations (red diagonal) along with repeats (dots) in
different positions around the genome.
Walworth et al. PNAS Early Edition | 3of6
ENVIRONMENTAL
SCIENCES
EVOLUTION
and only WH 8501 had substantially more annotated trans-
posases than the rest of the Crocosphaera genomes. This evi-
dence isalsoseeninthePCAanalysisincludingtransposaseand
paralog numberspergenome(Fig.S1B),inwhichno Crocosphaera
genomessegregatealongthe“Paralogs”axis.C.watsoniiWH8501
segregates away from other genomes along the “Transposase”
axis,whereastheotherCrocosphaeragenomesremainmoretightly
clustered within the plot. Hence, although WH 8501 possesses
agreatlyenhancedtransposonloadrelativeto Trichodesmium,
it did not develop the large noncoding regions observed in
Trichodesmium.
Intergenic Regions and ncRNA Elements. In the absence of coding
genes, it is possible that noncoding structural RNAs, regulatory
RNAs, or ribozymes within the intergenic regions are the select-
able traits driving conservation of these regions, which have been
recognized asimportantcomponents in cyanobacterial expression
networks(63).SixknownstructuralRNAelementswereidentified
(using RFAM) within IMS101 intergenic regions (Table S2). We
also used a pipeline that has previously identified cyanobacterial
noncoding RNAs (ncRNAs) (63, 64) to look for Trichodesmium
structuralconservationbetweenintergenicregionsamongIMS101
and2175andtheirinsitucounterpartswithinmetagenomicreads
(SI Materials and Methods). For an in situ sympatric comparison,
thesameprocedure was donewitheither Prochlorococcus(n= 5)
orSynechococcus(n=6)genomes,alongwithalloftheassembled
Sargasso Sea sequences from the Global Ocean Sampling dataset
(65).Althoughitwouldhavebeeninformativetorunthispipeline
with Crocosphaera genomes, we feel our results would not have
been comparable for several reasons (SI Text). However, of the
4,033predictedncRNAsinTrichodesmium(seebelow),only0.6%
were detected in a Crocosphaera metatranscriptome (66), as well
asCrocosphaeraandCyanothecegenomesusingBLASTn(Dataset
S6). This comparison suggests that much of the development and
expansion of noncoding sequences in Trichodesmium may be
uniquely specific to the evolution of the genus rather than shared
with sympatric diazotrophs that also do not display streamlining.
The pipeline predicted 365 putative noncoding regulatory
elements in Synechococcus environmental sequences and 190 in
Prochlorococcus (Fig. 3A). This trend is consistent with previous
studies showing widespread Prochlorococcus streamlining relative
to Synechococcus evidenced in both publicly available genomes
(67–69)andenvironmentalsingleamplified genomes(40).From
the Trichodesmium cultures and metagenome, the pipeline pre-
dicted 4,033 nonredundant, noncoding regulatory elements (Fig.
3A), of which 3,027 (75%) were expressed in the IMS101 tran-
scriptome (Dataset S6) (see below). When putative regulatory
elements were normalized per intergenic kilobase (Kb), Tricho-
desmium and Synechococcus yielded∼1.5 per intergenic Kb (Fig.
3A, line), whereas Prochlorococcus yielded<1 regulatory element
perintergenicKb.Althoughthesepredictionsalsoincludepossible
regulatory untranslated regions (UTRs), riboswitches, and termi-
natorsequences,theresultsshowevidenceforgenerallyconserved
intergenic composition among all three sympatric cyanobacteria,
with Trichodesmium and Synechococcus possibly possessing slightly
more regulatory elements per intergenic Kb.
Expression of Transposases and Intergenic Sequence. To determine
global, annotated transposase, and intergenic expression pat-
terns, transcriptomes of biological duplicate IMS101 cultures
growing semicontinuously in Aquil medium were sampled near
the middle of the photoperiod and sequenced by using the
IlluminaHi-Seqplatform.Thisanalysisshowedthat∼86%ofthe
intergenic regions in Trichodesmium, ∼91% of the IMG anno-
tated transposases, and 75% of the 4,033 predicted regulatory
elements from the above pipeline were expressed (Datasets S6
and S7). In a laboratory-based study using a different approach,
directed toward the identification of transcriptional start sites, it
was determined that at least 18.2% of the intergenic space was
transcribed as either ncRNAs or 5′ UTRs of protein-coding
mRNAs (70). This analysis further revealed that, of all bacteria
examined to date, T. erythraeum has the highest percentage of
transcriptional start sites from which ncRNAs originate. The
combination of such widespread noncoding and transposase ex-
pression as well as conservation of these sequences across isolates
0.5
4.5
1.5
Insertion Sequence
0
500
1000
Length of Blast Hit (bp)
75
80
85
90
95
100
Percent ID
Type
IS
Repeats
7.49 7.5 7.51 7.53 7.54
GC Skew (-)
GC Skew (+)
IMS101 Repeats
Metagenomic
Metatranscriptomic
Top 10 Most Abundant
IMS101 Chromosome IMS101 Chromosome
Composition ~7.49-7.54 Mb Composition ~7.49-7.54 Mb
IMS101
2175
6.77 6.77 6.78 6.78 6.80 6.80
7.5 (Mb)
6.5
5.5
2.5
3.5
Fig. 2. Genomic map of the 10 intergenic regions possessing the most re-
peated sequences in the IMS101 genome. These IMS101 intergenic regions
were mapped to the Trichodesmium metagenome and publicly available
metatranscriptomes. For the circular map, going from out to in are insertion
sequences (IS) (labeled red), the top 10 most abundant intergenic repeats,
metatranscriptomic reads, metagenomic reads, IMS101 repeats, and GC skew
withinsertionsequencelocationsoverlaidontoitshowninred.Thescatterplot
showstheISandrepeatcompositionofarepeathotspotbetween∼7.49and
7.54 Mb. Inset is the IMS101 genome alignment of this hotspot with 2175
showing nonhomologous (shaded boxes) regions between the two chromo-
somes aligning over the segment containing numerous repeats and IS sites.
Fig. 3. Predicted noncoding structurally conserved elements via comparative
genomics pipeline. The bar graph shows the increased amounts of in situ pu-
tative conserved elements in Trichodesmium relative to sympatric Synecho-
coccus and Prochlorococcus. The blue line indicates the relative amount of
predicted elements per intergenic Kb of sequence among the three cyano-
bacteria. Although many more structurally conserved elements arepredicted
in Trichodesmium relative to Prochlorococcus and Synechococcus, Trichodes-
mium and Synechococcus retain similar frequencies of structurally conserved
elements relative to intergenic kilobase (KB), whereas Prochlorococcus retains
approximately half.
4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1422332112 Walworth et al.
and natural populations suggest that possible widespread RNA-
based regulation along with active transposition may be common-
place in the genus.
Although the combination of structural predictions and in-
tergenic RNA sequencing lends strong support to active, widely
distributed ncRNAs in Trichodesmium, it is still difficult to de-
termine whether expressed portions of intergenic regions are
discrete ncRNAs or part of expressed UTRs in mRNA tran-
scripts, or both. To corroborate both our sequencing data and
informatic predictions, 18 noncoding elements with consistently
strong Illumina expression profiles (Dataset S6) between the
biological replicates were chosen for Northern blot analysis, and
allyieldedpositivehybridizations(Fig.S3andTablesS3andS4).
To determine the degree of secondary structural conservation
betweenculture-derivedncRNAsequencesandtheircounterparts
in the metagenome and metatranscriptome, we used RNAfold to
compare computationally predicted structures (71–74). RNAfold
predicted very similar core secondary structures between the cul-
ture and environmental sequences for most ncRNAs among the
topmatches(Fig.S4),withseveralvariationsduetoshorter/larger
loops and hairpins. These conserved features may be selectable
traits that drive the conservation of the long intergenic regions in
globally distributed Trichodesmium populations.
Population Level Processes vs. Natural Selection.Lynch and Conery
proposethatmicrobialgenomesarestreamlinedprimarilybecause
their effective population sizes are generally large enough to
preventsignificantcolonizationofmobileelementsandnoncoding
sequences, whereas effective population sizes in multicellular
eukaryotes are low enough to allow a permissive environment for
the expansion of noncoding DNA (51). The abundant noncoding
sequences in Trichodesmium relative to most other free-living
bacteria and marine oligotrophs, along with its general genome
architecture, may be due to a combination of small effective
population size derived from differing morphological genetic
subpopulationswithvaryingassociatedepibionts(e.g.,ref.28),as
well as potential rampant active mobile elements via transposase
activity. This feature is noteworthy because, unlike Trichodesmium,
other bloom-forming cyanobacteriawithmanyrepetitivesequences
such as Microcystis have an ∼80% coding average, which suggests
thatotherprominentforcesareinfluencing Trichodesmiumgenome
evolution in combination with reductions in effective population
size (75). Hence, the absolute causes of the large, intergenic-rich
genomes observed relative to other free-living prokaryotes(76) and
marine oligotrophs within the same habitat (40) remain obscure.
Because it is thought that many bacteria are deletion-biased
(47, 77), stable maintenance of these elements from laboratory
isolates to the natural samples suggest that they may be required
insomefashionforgrowthbothincultureandinsitu.Ithasbeen
shown in numerous systems that repeating elements (repeats
and/or IS; Dataset S8) can be mediators of genomic plasticity
(61, 62, 78); however, the direct impacts of these repeats are not
always so clear. For example, high IS density in the genome of
Lactobacillus acidophilus has been described (79), and despite the
propensity of these elements to inactivate genes and facilitate
recombination of genomic structure (61), the genome of this
isolate still displays high levels of synteny with other sequenced
Lactobacilli. Because it has also been shown that partial IS
sequences can inhibit transposition (78, 80), it is possible that
these repeats/pseudogenes have not been deleted because they
are controlling transposition in the transposase-heavy IMS101.
Others have hypothesized that the conserved repeat structures
observed in some bacteria could function as recombination-
dependent “promoter banks” for adaptation to new conditions,
thereby allowing relatively quick “rewiring” of metabolism in sub-
populations (59, 62, 81).
Summary
This study highlights a previously unidentified, environmentally
conserved genomic architecture of a successful oligotrophic,
free-living cyanobacterial diazotroph that is biogeochemically
important across global oceanic regimes (3, 5, 24). Free-living,
cyanobacterial diazotrophs such as Crocosphaera and Tricho-
desmium contain a wealth of transposases, chemotaxis, signal
transduction, and pseudogenes that directly contradict the ge-
nome streamlining observed in other oligotrophic prokaryotic
genomes. Hence, because of these gene content commonalities
among some cyanobacterial diazotrophs, it is very peculiar that
such low coding percentage and gene density has persisted ge-
nus-wide in Trichodesmium populations, both in culture and in
situ. One possible explanation is that the intergenic regions ex-
perience gradual inflation during certain evolutionary intervals
characterized by bloom-driven selective sweeps. Additionally,
a central difference in Trichodesmium spp. oligotrophic ecology
includes periodic aggregate formation, with possibly varying
physical interaction with epibiotic prokaryotes and eukaryotes.
Although specific causal factors contributing to the unusual
IMS101 genome still remain unclear, these data do confirm the
environmental relevance of the Trichodesmium genome architec-
ture, as well as a nonstreamlined, alternative route to a free-living
oligotrophic lifestyle.
Materials and Methods
See SI Materials and Methods for logistical protocols. In brief, DNA from
batch Trichodesmium cultures was isolated, frozen at−20 °C, and processed
for sequencing at the Joint Genome Institute (JGI) (IMS101) or at the Uni-
versity of Southern California (USC) (2175 and H94), and annotation was
performed using the JGI genome annotation pipeline. Trichodesmium meta-
genomesamples were collectedunder nonbloomconditionsin October 2010
on the R/V Oceanus cruise number OC469-1 near the Bermuda Atlantic Time
Series(BATS)station (28°37.474N, 66°0.606W).Colonies and trichomeswere
gently picked, and colony DNA was extracted immediately, stored at−20 °C,
and shipped to the JGI for pyrosequencing. IMS101 genes and intergenic
regions were downloaded from https://img.jgi.doe.gov and were used for
fragmentrecruitmentplots,comparativegenomics,andprincipalcomponent
analyses. RNA was isolated from flash-frozen biological duplicates of IMS101
cultures growing semi-continuously and sequenced at the USC Epigenome
Center. Northern blots and structural sequence predictions were conducted
as previously described (see SI Materials and Methods).
ACKNOWLEDGMENTS. WethankFrankLarimer,JillSohm,SuzanneEdmands,
MichaelLee,ChristopherDupont, andAndrewAllenforinsightfuldiscussions.
TheworkconductedbytheU.S.DepartmentofEnergyJointGenomeInstitute
is supported by the Office of Science of the Department of Energy under
ContractsDE-AC02-05CH11231andDE-AC03-76SF00098.Otherportionsofthis
workweresupportedbyNationalScienceFoundationGrantOCE-1260490and
the University of Southern California.
1. Falkowski PG, Barber RT, Smetacek V (1998) Biogeochemical controls and feedbacks
on ocean primary production. Science 281(5374):200–207.
2. Moore CM, et al. (2013) Processes and patterns of oceanic nutrient limitation. Nat
Geosci 6(9):701–710.
3. Sohm JA, Webb EA, Capone DG (2011) Emerging patterns of marine nitrogen fixa-
tion. Nat Rev Microbiol 9(7):499–508.
4. Sohm JA, et al. (2011) Nitrogen fixation in the South Atlantic Gyre and the Benguela
Upwelling System. Geophys Res Lett 38(16):L16608.
5. ZehrJP(2011)Nitrogenfixationbymarinecyanobacteria.TrendsMicrobiol19(4):162–173.
6. Foster RA, et al. (2007) Influence of the Amazon Riverplume on distributions of free-
living and symbiotic cyanobacteria in the western tropical north Atlantic Ocean.
Limnol Oceanogr 52(2):517–532.
7. Foster RA, Subramaniam A, Zehr JP (2009) Distribution and activity of diazotrophs in
the Eastern Equatorial Atlantic. Environ Microbiol 11(4):741–750.
8. Foster RA, Zehr JP (2006) Characterization of diatom-cyanobacteria symbioses on the
basis of nifH, hetR and 16S rRNA sequences. Environ Microbiol 8(11):1913–1925.
9. Moisander PH, et al. (2010) Unicellular cyanobacterial distributions broaden the
oceanic N2 fixation domain. Science 327(5972):1512–1514.
10. Church MJ, et al. (2009) Physical forcing of nitrogen fixation and diazotroph com-
munity structure in the North Pacific subtropical gyre. Global Biogeochem Cycles
23(2):GB2020.
11. Church M, Björkman K, Karl D, Saito M, Zehr J (2008) Regional distributions of
nitrogen-fixing bacteria in the Pacific Ocean. Limnol Oceanogr 53(1):63–77.
12. Langlois RJ, Hümmer D, LaRoche J (2008) Abundances and distributions of the
dominant nifH phylotypes in the Northern Atlantic Ocean. Appl Environ Microbiol
74(6):1922–1931.
13. Montoya JP, et al. (2004) High rates of N2 fixation by unicellular diazotrophs in the
oligotrophic Pacific Ocean. Nature 430(7003):1027–1032.
Walworth et al. PNAS Early Edition | 5of6
ENVIRONMENTAL
SCIENCES
EVOLUTION
14. KarlDM,ChurchMJ,DoreJE,LetelierRM,MahaffeyC(2012)Predictableandefficient
carbon sequestration in the North Pacific Ocean supported by symbiotic nitrogen
fixation. Proc Natl Acad Sci USA 109(6):1842–1849.
15. Subramaniam A, et al. (2008) Amazon River enhances diazotrophy and carbon se-
questration in the tropical North Atlantic Ocean. Proc Natl Acad Sci USA 105(30):
10460–10465.
16. GroßkopfT,etal.(2012)Doublingofmarinedinitrogen-fixationratesbasedondirect
measurements. Nature 488(7411):361–364.
17. ThompsonAW,etal.(2012)Unicellularcyanobacteriumsymbioticwithasingle-celled
eukaryotic alga. Science 337(6101):1546–1550.
18. Webb EA, Jakuba RW, Moffett JW, Dyhrman ST (2007) Molecular assessment of
phosphorus and iron physiology in Trichodesmium populations from the western
Central and western South Atlantic. Limnol Oceanogr 52:2221–2232.
19. Post AF, et al. (2002) Spatial and temporal distribution of Trichodesmium spp. in the
stratified Gulf of Aqaba, Red Sea. Mar Ecol Prog Ser 239:241–250.
20. Hynes AM, Webb EA, Doney SC, Waterbury JB (2012) Comparison of cultured Tri-
chodesmium (Cyanophyceae) with species characterized from the field.JPhycol 48:
196–210.
21. Letelier R, Karl D (1996) Role of Trichodesmium spp in the productivity of the sub-
tropical North Pacific Ocean. Mar Ecol Prog Ser 133:263–273.
22. Capone DG, et al. (1998) An extensive bloom of the N2-fixing cyanobacterium Tri-
chodesmium erythraeum in the central Arabian Sea. Mar Ecol Prog Ser 172:281–292.
23. DoreJE,LetelierRM,ChurchMJ,LukasR(2008)Summerphytoplanktonbloomsinthe
oligotrophic North Pacific Subtropical Gyre: Historical perspective and recent ob-
servations. Prog Oceanogr 76(6):2–38.
24. Capone D, Zehr J, Paerl H, Bergman B, Carpenter E (1997) Trichodesmium,aglobally
significant marine cyanobacterium. Science 276(5316):1221.
25. Mulholland M (2007) The fate of nitrogen fixed by diazotrophs in the ocean. Bio-
geosciences 4:37–51.
26. Sheridan C, Steinberg D, Kling G (2002) The microbial and metazoan community as-
sociated with colonies of Trichodesmium spp: A quantitative survey. J Plankton Res 24:
913–922.
27. Hewson I, et al. (2009) Microbial community gene expression within colonies of the
diazotroph,Trichodesmium, from theSouthwestPacificOcean.ISMEJ3(11):1286–1300.
28. Hmelo LR, Van Mooy B, Mincer TJ (2012) Characterization of bacterial epibionts on
the cyanobacterium Trichodesmium. Aquat Microb Ecol 67:1–14.
29. Wyman M, Hodgson S, Bird C (2013) Denitrifying alphaproteobacteria from the
ArabianSeathatexpressnosZ,thegeneencodingnitrousoxidereductase,inoxicand
suboxic waters. Appl Environ Microbiol 79(8):2670–2681.
30. PaerlHW,BeboutBM,PrufertLE(1989)Bacterialassociationswithmarineoscillatoria
Sp.(Trichodesmiumsp.)populations:Ecophysiologicalimplications.JPhycol25:773–784.
31. Hynes AM, Chappell PD, Dyhrman ST, Doney SC, Webb EA (2009) Cross-basin com-
parison of phosphorus stressandnitrogenfixationinTrichodesmium.LimnolOceanogr
54:1438–1448.
32. Orcutt KM, et al. (2001) A seasonal study of the significance of N2 fixation by Tri-
chodesmiumspp.attheBermudaAtlanticTime-seriesStudy(BATS)site.DeepSeaRes
Part II Top Stud Oceanogr 48:1583–1608.
33. Prufert-Bebout L, Paerl HW, Lassen C (1993) Growth, nitrogen fixation, and spectral
attenuation in cultivated trichodesmium species. Appl Environ Microbiol 59(5):
1367–1375.
34. Bell P, et al. (2005) Laboratory culture studies of Trichodesmium isolated from the
great Barrier Reef Lagoon, Australia. Hydrobiologia 532:9–21.
35. Orcutt KM, et al. (2002) Characterization of Trichodesmium spp. by genetic techni-
ques. Appl Environ Microbiol 68(5):2236–2245.
36. Palenik B, et al. (2003) The genome of a motile marine Synechococcus. Nature
424(6952):1037–1042.
37. Rocap G, et al. (2003) Genome divergence in two Prochlorococcus ecotypes reflects
oceanic niche differentiation. Nature 424(6952):1042–1047.
38. Tripp HJ, et al. (2010) Metabolic streamlining in an open-ocean nitrogen-fixing cya-
nobacterium. Nature 464(7285):90–94.
39. Scanlan DJ, et al. (2009) Ecological genomics of marine picocyanobacteria. Microbiol
Mol Biol Rev 73(2):249–299.
40. Swan BK, et al. (2013) Prevalent genome streamlining and latitudinal divergence of
planktonic bacteria in the surface ocean. Proc Natl Acad Sci USA 110(28):11463–11468.
41. Bench SR, et al. (2013) Whole genome comparison of six Crocosphaera watsonii
strains with differing phenotypes.JPhycol 49:786–801.
42. Giovannoni SJ, et al. (2005) Genome streamlining in a cosmopolitan oceanic bacte-
rium. Science 309(5738):1242–1245.
43. Dupont CL, et al. (2012) Genomic insights to SAR86, an abundant and uncultivated
marine bacterial lineage. ISME J 6(6):1186–1199.
44. Lauro FM, et al. (2009) The genomic basis of trophic strategy in marine bacteria.Proc
Natl Acad Sci USA 106(37):15527–15533.
45. Lynch M (2007) The frailty of adaptive hypotheses for the origins of organismal
complexity. Proc Natl Acad Sci USA 104(Suppl 1):8597–8604.
46. Lynch M (2006) Streamlining and simplification of microbial genome architecture.
Annu Rev Microbiol 60:327–349.
47. Mira A, Ochman H, Moran NA (2001) Deletional bias and the evolution of bacterial
genomes. Trends Genet 17(10):589–596.
48. Welsh EA, et al. (2008) The genome of Cyanothece 51142, a unicellular diazotrophic
cyanobacterium important in the marine nitrogen cycle. Proc Natl Acad Sci USA
105(39):15094–15099.
49. Sinha R, et al. (2014) Comparative genomics of Cylindrospermopsis raciborskii strains
with differential toxicities. BMC Genomics 15:83.
50. Tripp HJ, et al. (2008) SAR11 marine bacteria require exogenous reduced sulphur for
growth. Nature 452(7188):741–744.
51. Lynch M, Conery JS (2003) The origins of genome complexity. Science 302(5649):
1401–1404.
52. Morris JJ, Lenski RE, Zinser ER (2012) The Black Queen Hypothesis: Evolution of de-
pendencies through adaptive gene loss. MBio 3(2):e00036–12.
53. Bench SR, Ilikchyan IN, Tripp HJ, Zehr JP (2011) Two strains of Crocosphaera watsonii
with highly conserved genomes are distinguished by strain-specific features. Front
Microbiol 2:261.
54. Bandyopadhyay A, et al. (2011) Novel metabolic attributes of the genus cyanothece,
comprising a group of unicellular nitrogen-fixing Cyanothece. MBio 2(5):e00214–11.
55. KuoC-H,MoranNA,OchmanH(2009)Theconsequencesofgeneticdriftforbacterial
genome complexity. Genome Res 19(8):1450–1454.
56. Konstantinidis KT, et al. (2009) Comparative systems biology across an evolutionary
gradient within the Shewanella genus. Proc Natl Acad Sci USA 106(37):15909–15914.
57. Smith JK, Parry JD, Day JG, Smith RJ (1998) A PCR technique based on the Hip1 in-
terspersed repetitive sequence distinguishes cyanobacterial species and strains. Mi-
crobiology 144(Pt 10):2791–2801.
58. TreangenTJ,AbrahamA-L,TouchonM,RochaEPC(2009)Genesis,effectsandfatesof
repeats in prokaryotic genomes. FEMS Microbiol Rev 33(3):539–571.
59. Matus-Garcia M, Nijveen H, van Passel MWJ (2012) Promoter propagation in pro-
karyotes. Nucleic Acids Res 40(20):10032–10040.
60. Lin S, et al. (2011) Genome-wide comparison of cyanobacterial transposable ele-
ments, potential genetic diversity indicators. Gene 473(2):139–149.
61. Siguier P, Gourbeyre E, Chandler M (2014) Bacterial insertion sequences: Their ge-
nomic impact and diversity. FEMS Microbiol Rev 38(5):865–891.
62. Zhou K, Aertsen A, Michiels CW (2014) The role of variable DNA tandem repeats in
bacterial adaptation. FEMS Microbiol Rev 38(1):119–141.
63. Voss B, GeorgJ, Schön V,Ude S, HessWR (2009)Biocomputationalprediction of non-
coding RNAs in model cyanobacteria. BMC Genomics 10:123.
64. GiergaG,VossB,HessWR(2012)Non-codingRNAsinmarineSynechococcusandtheir
regulation under environmentally relevant stress conditions. ISME J 6(8):1544–1557.
65. Venter JC, et al. (2004) Environmental genome shotgun sequencing of the Sargasso
Sea. Science 304(5667):66–74.
66. Hewson I, et al. (2009) In situ transcriptomic analysis of the globally important key-
stone N2-fixing taxon Crocosphaera watsonii. ISME J 3(5):618–631.
67. Kettler GC, et al. (2007) Patterns and implications of gene gain and loss in the evo-
lution of Prochlorococcus. PLoS Genet 3(12):e231.
68. Steglich C, et al. (2008) The challenge of regulation in a minimal photoautotroph:
Non-coding RNAs in Prochlorococcus. PLoS Genet 4(8):e1000173.
69. García-Fernández JM, de Marsac NT, Diez J (2004) Streamlined regulation and gene
loss as adaptive mechanisms in Prochlorococcus for optimized nitrogen utilization in
oligotrophic environments. Microbiol Mol Biol Rev 68(4):630–638.
70. Pfreundt U, Kopf M, Belkin N, Berman-Frank I, Hess WR (2014) The primary tran-
scriptomeofthemarinediazotrophTrichodesmiumerythraeumIMS101.SciRep4:6187.
71. Lorenz R, et al. (2011) ViennaRNA Package 2.0. Algorithms Mol Biol 6:26.
72. Hofacker IL, Fekete M, Stadler PF (2002) Secondary structure prediction for aligned
RNA sequences.JMolBiol 319(5):1059–1066.
73. Will S, Reiche K, Hofacker IL, Stadler PF, Backofen R (2007) Inferring noncoding RNA
familiesandclassesbymeansofgenome-scalestructure-basedclustering.PLOSComput
Biol 3(4):e65.
74. Bernhart SH, Hofacker IL, Will S, Gruber AR, Stadler PF (2008) RNAalifold: Improved
consensus structure prediction for RNA alignments. BMC Bioinformatics 9:474.
75. Kaneko T, et al. (2007) Complete genomic structure of the bloom-forming toxic cy-
anobacterium Microcystis aeruginosa NIES-843. DNA Res 14(6):247–256.
76. OchmanH,DavalosLM(2006)Thenatureanddynamicsofbacterialgenomes.Science
311(5768):1730–1733.
77. Nilsson AI, et al. (2005) Bacterial genome size reduction by experimental evolution.
Proc Natl Acad Sci USA 102(34):12112–12116.
78. Delihas N (2011) Impact of small repeat sequences on bacterial genome evolution.
Genome Biol Evol 3:959–973.
79. Callanan M, et al. (2008) Genome sequence of Lactobacillus helveticus, an organism
distinguished by selective gene loss and insertion sequence element expansion.
JBacteriol 190(2):727–735.
80. Gueguen E, Rousseau P, Duval-Valentin G, Chandler M (2006) Truncated forms of
IS911 transposase downregulate transposition. Mol Microbiol 62(4):1102–1116.
81. Gemayel R, Vinces MD, Legendre M, Verstrepen KJ (2010) Variable tandem repeats
accelerate evolution of codingandregulatorysequences. Annu Rev Genet 44:
445–477.
6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1422332112 Walworth et al.
40
Chapter 3
The molecular transition underlying plasticity-mediated CO
2
adaptation in the
globally distributed marine cyanobacterial nitrogen-fixer Trichodesmium

Nathan G. Walworth, Michael D. Lee, Fei-Xue Fu, David A. Hutchins, and Eric A. Webb

Note: All supplementary information for this chapter can be found at:
https://www.dropbox.com/sh/1ema43bdthxezza/AAD7pY58mqAHW3w4_SHbGcGVa?d
l=0

Author contributions
Experiments were conceived and carried out by D.A. Hutchins, F.-X. Fu, N.G. Walworth,
and E.A. Webb with experimental and analytical contributions from M.D. Lee. N.G.
Walworth wrote the manuscript, with assistance from M.D. Lee, D.A. Hutchins, E.A.
Webb and F.-X. Fu.

Acknowledgments
Grant support was provided by U.S. National Science Foundation OCE 1143760
to D.A. Hutchins, E.A. Webb and F.-X. Fu.

41
Abstract
Globally rising temperatures and increasing carbon dioxide (CO
2
) emissions have
prompted assessments of both short- and long-term responses of biogeochemically-
critical carbon- and nitrogen-fixing (i.e., diazotrophic) marine microbes. Most studies
have focused on plastic (short-term) phenotypic responses in the absence of genetic
change, and a few have investigated adaptive (long-term) phenotypic responses.
However, no studies to date have investigated the molecular progression underlying the
transition from plasticity to adaptation under elevated CO
2
for a marine diazotroph. To
address this, we cultured the globally important, cyanobacterial nitrogen-fixer
Trichodesmium at both low and high CO
2
for 4.5 years, followed by reciprocal CO
2

transplantation to generate transcriptomes for both long-term treatments and reciprocal
transplants. By leveraging paralleling phenotypic and transcriptional profiles between
treatments, we identified expression changes and pathway enrichments in both immediate
and long-term responses including increases in photosystem II-related genes and
decreases in regulatory elements involved in iron and redox sensing. These changes
occurring on both short and long timescales implicate their involvement in instigating the
plastic phenotype as well as maintaining the adaptive one. Certain metabolic shifts were
only observed in the adaptive response including specific RNA Polymerase (RNAP)
sigma factors, indicating these expression changes to primarily contribute to adaptation.
This is the only known system to date where fitness actually increased in all high-CO
2
adapted cell lines when returned to the ancestral environment, a response possibly
mediated by differential regulation of RNAP sigma factors and transposable elements.
These molecular insights are not only critical for identifying pathways under selection as
drivers in plasticity and adaptation, but also for interpreting in situ genomic surveys as
proxies for global change-driven evolutionary potential in natural populations.

42
Warming temperatures and increasing anthropogenic carbon dioxide (CO
2
)
emissions have galvanized recent research attempts assessing evolutionary potential as a
result of global change factors in numerous biological systems. Studies assessing
responses of both carbon- (primary producers) and nitrogen-fixing (diazotrophs)
organisms to ocean acidification have been of particular interest due to their bottom-up
control of global biogeochemical cycles and food webs
1-3
. However, attributing observed
phenotypic changes to specific environmental perturbations in situ remains an ongoing
challenge, particularly when delineating between phenotypic plasticity and adaptive
evolution
2
. Phenotypic plasticity occurs when individuals in a population of a given
genotype amend their phenotype as part of a rapid response to environmental change,
while adaptive evolution occurs when the underlying genetic (allelic) composition of a
population changes the phenotype as a result of natural selection
3
. It is also worth noting
that population level phenotypic changes may also ultimately result from environmental
stress
2
.
Additionally, it has been shown that a range of phenotypic plasticity can exist
within a single species
4,5
and that phenotypic plasticity itself can evolve and aid in
adaptation
3,6,7
. As such, plasticity can potentially affect evolution in opposing ways. It
may either facilitate adaptation by having natural selection act on a beneficial plastic
phenotype (defined hereafter as “plasticity-mediated” adaptation), or it can shield certain
genotypes from natural selection if optimal phenotypes may be produced by plasticity
alone
3
. Hence, these phenomena necessitate investigations into the effects of plasticity on
populations-level adaptation during periods of environmental pressure.
Laboratory-based experimental evolution studies enable analysis of organismal
and population responses to defined experimental conditions as they transition from
plastic to adaptive
7
. These insights better inform environmental phenotypic observations
and offer more constrained time scales of plasticity versus adaptation. However, aside
from being typically restricted to rapidly dividing microorganisms, the main experimental
challenge resides in extrapolating laboratory evolutionary potential to predicting adaptive
capacities in natural populations. Thus, comprehensively interpreting in situ genetic and
phenotypic datasets remains challenging due to limited knowledge of fundamental
biology, gene flow, population sizes, mutation, and recombination rates
3
.
43
One promising approach is to couple molecular techniques with experimental
evolution in order to elucidate the coordination of underlying molecular changes as they
influence both the phenotypic plasticity phenotype and/or evolutionary genotype
8
. For
example, one recent study examining the effect of high CO
2
on gene expression changes
in the eukaryotic calcifying alga Emiliana huxleyi found that opposing plastic and
adaptive phenotypes were also reflected by their corresponding gene expression
changes
9
. In a preceding study, Lohbeck et al. confirmed adaptation through observing
significant growth rate increases in high CO
2
-selected lines relative to those of low-CO
2

selected lines under elevated CO
2
conditions
10
. Reduced growth and calcification in the
plastic response correlated with reductions in expression of genes involved in pH
regulation, photosynthesis, carbon transport, and calcification while partly restored
growth and calcification observed in the adaptive response were associated with the
significant recovery of the expression of these genes. Hence, this experiment elucidated
an opposing phenotypic relationship between plasticity and adaptation mirrored by
underlying gene expression changes in E. huxleyi. However, no studies to date have
characterized the molecular progression underlying this transition for marine microbial
nitrogen-fixing organisms.
Due to the biogeochemical importance of the globally-distributed diazotrophic
cyanobacteria Trichodesmium spp., many field and laboratory studies have set the stage
for evolutionary investigations by characterizing its physiology in terms of factors like
iron (Fe), phosphorus (P), and CO
2
11,12-15
. For example, multiple short-term (weeks to
months) laboratory
13,16-20
and field
1,21
experiments have demonstrated increases in both
growth and N
2
fixation in Trichodesmium spp. in response to elevated CO
2
, revealing a
plastic high-CO
2
phenotype under specific nutrient conditions. Furthermore, Hutchins et
al.
22
observed divergent taxon-specific responses to increased CO
2
among
biogeographically-distinct diazotrophic cyanobacteria including Trichodesmium; a result
which has also been observed in eukaryotic phytoplankton
5,23,24
. Hence, genus-level
conservation of physiological variability across broad eukaryotic and prokaryotic genera
highlights the potential importance of maintaining evolutionary plasticity in certain
environmental regimes and suggests differential taxonomic CO
2
selection over geological
time, which could have ultimately influenced biogeographic distributions. Additionally,
44
Walworth et al.
25
recently validated the conservation of genome architecture and coding
potential of the cyanobacterial diazotroph Trichodesmium erythraeum IMS101 (hereafter
IMS101) with different Trichodesmium isolates as well as with natural Trichodesmium
populations sampled decades apart. These data help to environmentally contextualize the
molecular changes described here, which further aid in substantiating the adaptation of
IMS101 in a laboratory setting with the genetic potential contained in natural
populations.
Leveraging these
phenotypic and genetic
data, we investigated the
global transcriptional
underpinnings of long-
term CO
2
selection of a
single IMS101 starting
population as its
phenotype transitioned
from plastic to adaptive.
One cell line was divided
into two CO
2
treatments of
6 biological replicates each
and experimentally
adapted at both present
day (380 micro-
atmospheres(μatm)) and projected year 2100 (750 μatm) concentrations for ~4.5 years
(~570-850 generations depending on CO
2
treatment) with growth rate as a proxy for
reproductive fitness
8,26
. At the onset of this incubation, cell lines placed in 750 μatm CO
2

(750 ancestral) rapidly increased both growth and N
2
fixation rates whereas cell lines in
380 μatm CO
2
(380 ancestral) sustained lower physiological rates (Fig. 1)
26
. This
immediate fitness increase in response to high CO
2
is consistent with the classically
observed plastic response of IMS101 (high-CO
2
phenotype) as previously shown (see
above). After 4.5 years of low and high CO
2
selection, no further changes in growth or N
2

Figure 1 Growth (bottom) and N
2
fixation (top) rates of the
ancestral CO
2
-selected cultures both prior to and after 4.5 years
of selection. Stars denote statistical significance and error bars are
standard errors.
Growth d
-1
380
ancestral
750
ancestral
380
selected
750
selected
380-
selected
to 750
switch
750-
selected
to 380
switch
0.3
0.32
0.34
0.36
0.38
>4.5 years
0
5
10
15
20
25
N
2
fixation rate
(pmol N ng Chl a
-1
hr
-1
)
Plastic + Adaptive
Adaptive
45
fixation were observed for either the 380- or 750-selected lines relative to their
corresponding 380- and 750-ancestral time points, respectively (Fig. 1). All 6 replicates
in the 750-selected cell lines still maintained significantly higher growth and N
2
fixation
rates relative to the 380-selected cell lines but showed no further fitness increase
following the initial plastic growth rate response despite ~850 subsequent generations of
selection at high CO
2
(Fig. 1, yellow bars). Once subcultures of the 380-selected cell
lines were placed in high (750 μatm) CO
2
for 2 weeks following the 4.5-year incubation
at low (380 μatm) CO
2
(380s-to-750), both growth (fitness) and N
2
fixation rapidly
increased similar to the 750-ancestral response and consistent with the aforementioned
experiments (Fig. 1, green bars). Surprisingly however, when subcultures of the 750-
selected cell lines were reciprocally transplanted back to the ancestral CO
2
condition
(750s-to-380), a 44% fitness increase was observed relative to both the 380-selected and
ancestral cell lines (Fig. 1; blue bars, bottom panel). As such, the adaptive response
following long-term high-CO
2
selection was characterized not by steady fitness increases
in the selection environment, but instead by a large fitness increase in the ancestral
environment, which is an extremely rare phenomenon in microbial evolution literature
3,8
.
Hence, since both the plastic and adaptive responses of IMS101 to increased CO
2

exhibited the same high-CO
2
phenotype, the transition from plasticity to adaption appears
to have remained phenotypically neutral in the selection environment, yet resulted in an
altered phenotype in the ancestral environment.
In order to elucidate the molecular progression of this phenotypically cryptic
transition, we sequenced biological triplicate transcriptomes of both long-term CO
2

treatments after 4.5 years of selection (380-selected and 750-selected), as well as both
reciprocal transfers (380s-to-750 and 750s-to-380) after 2 weeks in the reciprocal CO
2

concentration. One of the most striking insights separating the short- and long-term
responses relative to the 380-selected phenotype (i.e. low-CO
2
phenotype) was the
differential regulation of RNA polymerase sigma factors, which have been previously
shown to induce broad shifts in metabolic pathways in response to carbon and nitrogen
fluctuations in other microbial systems
27,28
. Additionally, differential expression of
transposition was also detected in both short- and long-term CO
2
responses. Hence, our
data suggest that CO
2
may impact regulation of transposition and sigma factor expression
46
in plasticity-mediated adaptation potentially leading to broad, downstream changes in
metabolic pathways.
Results and Discussion
The 380s-to-750 expression profiles characterized the true plastic response while
those of the 750-selected condition characterized the adaptive one relative to 380 μatm
CO
2
. The unexpected high-CO
2
phenotype (increased growth and N
2
fixation rates
relative to the 380-selected) exhibited by the 750s-to-380 enabled us to leverage its
transcriptome in several ways. First, since the 750-selected, the 380s-to-750, and the
750s-to-380 all harbored the high-CO
2
phenotype (Fig. 1), genes sharing parallel
expression profiles among all 3 treatments represent those that both rapidly responded to
increased CO
2
as part of the plastic response and subsequently maintained these profiles
as part of the adaptive response. Second, concurrent analysis of parallel gene expression
changes shared between the 750-selected and 750s-to-380 treatments represent
transcriptional changes driving the maintenance of the high-CO
2
phenotype even after
transfer back to ancestral CO
2
levels. Since the plastic high-CO
2
phenotype (380s-to-750)
was sustained in long-term adaptation (750-selected), it is important to note that the
maintenance of this high-CO
2
phenotype in the 380 µatm CO
2
ancestral condition (750s-
to-380) may either be a type of plastic response by the 750-selected cell lines relative to
750 μatm CO
2
selection
2
or a general inability to respond to low CO
2
reflected in a
potential loss of the low-CO
2
phenotype. Regardless, the adaptive changes acquired by
long-term high-CO
2
exposure in the 750-selected cell lines should also be the ones
influencing the plastic 750s-to-380 phenotype that maintained increased fitness and N
2

fixation. Hence, the 750s-to-380 treatment will be referred to hereafter as part of the
adaptive response relative to 380 μatm CO
2
. Therefore, gene expression profiles shared
between the 750-selected and 750s-to-380 conditions that parallel this phenotypic
maintenance indicate their contribution to sustaining the high-CO
2
phenotype following
long-term, high-CO
2
adaptation.
By analyzing genes exhibiting parallel expression changes among different
treatments, we were able to identify processes involved in both the short- (380s-to-750;
plastic) and long-term responses (750-selected and 750s-to-380; adaptive) relative to the
380-selected treatment, potentially implicating them as being important for instigating
47
and sustaining the high-CO
2
phenotype (Supp. File 1). Genes exhibiting significant
decreases in expression (downregulation) in the high-CO
2
phenotype (Fig. 2a “Plastic +
Adaptive”) relative to low-CO
2
phenotype levels were enriched in gene ontology (GO)
metabolisms involving broad metabolic processes—particularly of note sigma factor
activity and carbon transport (Fig. 2a, hypergeometric test with Benjamin-Hochberg
correction FDR <= 0.1
29
). Differential expression of specific transposition types (Fig. 3,
see below) was also correlated to the high-CO
2
phenotype suggesting them to be potential
targets of CO
2
selection
or the result of
prolonged CO
2

exposure. Similarly,
genes with increased
expression
(upregulation) shared
across all 3 treatments
exhibiting the high-CO
2

phenotype were detected
in widespread pathways
(Fig. 4, see below).
In order to test
for the probability of
different experimental
conditions sharing a
given amount of genes
with parallel expression changes, pairwise hypergeometric tests were conducted for both
up- and downregulated gene sets. These results revealed very low probabilities that the
number of genes exhibiting parallel expression profiles were shared by chance alone
among downregulated gene sets (Fig. 2a) between 380s-to-750 and 750s-to-380 (p < 10
-
13
), 750-selected and 380s-to-750 (p < 10
-29
), and 750-selected to 750s-to-380 (p < 10
-42
).
Similarly, low probabilities were also observed for upregulated gene sets (Fig. 4) between
the 380s-to-750 and 750s-to-380 (p < 10
-50
), the 750-selected and 380s-to-750 (p < 10
-48
),
38
3
308
45 2
14
6
750 380s-to-750
750s-to-380
Downregulated
Carbon Metabolism Transcription/Translation
pyruvate metabolic process sigma factor activity
phosphoribosylformylglycinamidine
synthase
Energy/General Metabolism
lipoate biosynthesis
polysaccharide transport
a
b
Adaptive
Plastic + Adaptive
DNA Binding Transcription Factor
regulation of transcription
Cell signaling/communication
phosphoric diester hydrolase
380-selected
380s-to-750
750-selected
750s-to-380
0
500
1000
1500
2000
2500
0
100
200
300
normalized counts
sigC sigF
furB
(Tery_1958)
(Tery_3404)
furA
furA
Figure 2 Shown are downregulated GO-enriched pathways relative to
380-selected and transcriptional profiles of sigma factors, sigC and
sigF as well as Fur proteins. (a) shows downregulated GO-enriched
pathways for the all high-CO
2
phenotype treatments. (b) shows differential
expression of transcriptional regulators with stars representing statistical
significance relative to the 380-selected and error bars being standard
errors.
48
and the 750-selected to 750s-to-380 (p < 10
-79
). Due to the low probabilities of sharing
these many genes or more between treatments by chance alone, it is likely that some of
the downstream effects of these shared expression changes are associated with the plastic
and/or adaptive responses to high CO
2
.
Sigma succession underlying plasticity-mediated adaptation
Our data shows that increased CO
2
correlates with lower expression of RNA
polymerase sigma factors sigC (Tery_1956; group 2) and sigF (Tery_3916; group 3)
(Fig. 2b)
30
. Differential regulation of sigma factors, “sigma switching”, aids in both
stress responses and adaptation via transcriptional initiation of gene sets specific to
particular environmental or internal cellular changes
31
. For example, sigC transcripts
have been shown to increase under short-term nitrogen limitation in diazotrophic
cyanobacteria
27
, which is consistent with the simultaneous decrease in sigC (ortholog to
Anabaena sigC, reciprocal best blast hit, evalue < 1e-10) expression and increase in
nitrogen fixation in all high-CO
2
phenotypes (Fig. 2b, red bars).
Furthermore, the iron-induced regulatory protein, Fur, has been previously shown
to bind to the promoter region of sigC in cyanobacteria, implicating sigC to have a
connective role in both nitrogen and iron homeostasis, and potentially oxidative stress as
well
32
. Accordingly, furA (Tery_1958) and furB (Tery_1953) genes (orthologs to
Anabaena, reciprocal best blast hit, evalue < 1e-10) exhibited parallel decreases in
expression with sigC under prolonged high CO
2
(Fig. 2b). In contrast, a fur paralog
(Tery_3404) showed no changes in expression after prolonged exposure to high CO
2
in
replete iron. These transcriptional reductions of fur homologs in high CO
2
may influence
tetrapyrrole production (Fig. 4, see below) as observed in the cyanobacterial diazotroph
Anabaena in order to energetically aid in growth increases
33
. Furthermore, furA
(Tery_1958) and furB slightly increased expression in the 750s-to-380 treatment as
compared to the 750-selected treatment, suggesting a reciprocal response to low CO
2

exposure contrasting with their decreased expression seen in the 750-selected under high
CO
2
. Taken together, these data provide some evidence of the co-regulation of sigC and
specific fur genes as part of both short-term plastic and long-term adaptive responses.
Conversely, sigF (ortholog to Synechocystis sigF, reciprocal best blast hit, evalue
< 1e-10) transcription was only significantly decreased in 750-selected and 750s-to-380
49
treatments, suggesting significant downregulation of sigF to be primarily involved in
adaptation rather than initiation of the high-CO
2
phenotype (Fig. 2b, purple bars). sigF is
involved in a variety of cellular processes and has been shown to target other
transcriptional regulators such as rsfA in the gram-positive Bacillus subtilis
34
as well as a
phytochrome-like histidine kinase (HK) in the cyanobacterium Synechocystis PCC6803
30
.
Interestingly, an IMS101 hypothetical protein (Tery_2530), containing an rsfA domain
(BLASTx, default settings), as well as a PAS/PAC signal transduction histidine kinase
(Tery_4221) containing several overlapping portions of conserved domains including
bacteriophytochrome (COG4251), phosphate regulon sensor kinase (PhoR;
TIGR02966)
35
, and NtrY (COG5000) also exhibited parallel downregulation with sigF
(Supp. File 1). Intriguingly, NtrY modulates nifA expression that specifically controls
expression of N
2
-fixing nif genes in the symbiotic diazotroph Azorhizobium caulinodans
ORS571
36
. However, no IMS101 nifA homologs to that of ORS571 were detected. PAS-
containing histidine kinases have also been shown to bind to a wide array of cofactors
and are important signaling modules that monitor changes in light, redox potential, small
ligands, and cellular energy
37,38
. Hence, Tery_4221 may regulate several different
metabolic functions aiding in increased growth and N
2
fixation. Regardless, its
significantly decreased expression in conjunction with sigF following long-term high
CO
2
exposure implicates a role in influencing the high-CO
2
phenotype.
Taken together these data suggest that adaptive increased growth and N
2
fixation
is mediated through plastic mechanisms and aided by downregulation of key upstream
sigma factors. Reduced sigC transcription is associated with the initiation of the high-
CO
2
phenotype, while decreased expression of both sigC and sigF ultimately contribute
to its adaptive maintenance.
Transposition regulation in plasticity and adaptation
In addition to sigma switching, shifts in regulation of transposable elements (TEs)
have been previously shown to be involved in both environmental plasticity and
adaptation by causing subsequent downstream changes in broad cellular processes
39-41
.
However, reliably quantifying expression of repetitive DNA sequences (repeats) such as
TEs (e.g. Insertion Sequences (IS)) remains a significant challenge for next-generation
sequencing methods utilizing short read technology (e.g. Illumina sequencing typically
50
50-150 base pairs) due to difficultly mapping repetitive sequences to a single genomic
location (e.g. multi-reads)
42
. Multi-read mapping can result in erroneous or biased read
counts per repetitive element (e.g. transposase), which inevitably skews downstream
quantification. Several statistical methods have been developed in attempts to more
accurately assign multi-reads to specific genomic locations but nearly all of them have
been developed for novel isoform detection in eukaryotic genomes (reviewed in
42
).
Hence, in order to analyze expression of paralogous TEs associated with the high-CO
2

phenotype in the TE-heavy IMS101 genome
25
, we developed a method to quantify
relative transcription of TE clusters irrespective of genomic position. Briefly, TE
sequences were obtained from Walworth et al.
25
and clustered at 70% identity using
USEARCH
43
in which representative sequences (centroids) for each of the 69 clusters
detected were identified (Supplementary File 2). Next, BLASTn
44
was used to search for
all paralogous sequences for each centroid sequence (cluster) within the IMS101 genome
(Methods) followed by mapping of RNA-Seq reads
45
to all sequences within every
cluster. Then, read counts for all sequences within a cluster were summed to produce
aggregated read counts per cluster followed by normalization and differential expression
analysis using edgeR
46
. Finally, any cluster differentially expressed in at least one high
CO
2
-phenotype treatment (380s-to-750, 750-selected, or 750s-to-380) relative to the 380-
selected reference was selected for further analysis, and log2 fold changes were
calculated for each high-CO
2
phenotype treatment of each cluster relative to the 380-
selected condition.
In order to identify TE-cluster expression correlated to the high-CO
2
phenotype,
hierarchical clustering with multiscale bootstrap resampling (replicates=1000) of log2
fold changes for all high-CO
2
phenotype treatments was conducted
47
resulting in two
well-defined groups of TE clusters whose average high-CO
2
phenotype expression was
either increased or decreased relative to the 380-selected reference (Fig. 3a). Due to the
difficulty in quantifying location-specific TE expression (see above) in addition to the
lack of knowledge of TE regulation in IMS101, this method was developed to
conservatively correlate TE types (clusters) to increased growth and N
2
fixation. In order
to test for the strength of positive or negative associations of specific TE clusters to the
high-CO
2
phenotype, Welch’s t-tests were conducted between the 380-selected biological
51
replicates (n=3) and all high-CO
2
phenotype replicates (n=9) for each TE cluster
exhibiting differential expression in at least one high-CO
2
phenotype treatment (Fig. 3a,
n=16 clusters). These analyses resulted in 75% (12/16) of the clusters exhibiting
significantly different mean expression values (p < 0.05) between the low- (i.e. 380-
selected) and high-CO
2
phenotypes (Fig. 3a,c; bolded/asterisk TE’s). These results
indicate that expression between the high-CO
2
phenotype treatments relative to the 380-
selected phenotype for these TE clusters was generally consistent in sign as a result of
high CO
2
exposure, implicating the simultaneous enrichment of specific types of TE
elements along with the concurrent repression of others to influence increased growth and
N
2
fixation.
52
Additionally, 77% (53/69) of total TE
clusters (n=69; Supplementary File 2) showed no differences in expression indicating
maintained cluster expression irrespective of phenotype, which suggests widespread TE
activity devoid of selection (Fig. 3c, light grey links). Taken together, these patterns
corroborate the maintenance of transposition as a result of neutral processes
41
and/or
weak selection
48
where stable coexistence occurs between TEs and the host genome.
Hence, as external selective pressure increases (e.g. increasing CO
2
), the ability of weak
selection to maintain global TE activity becomes limited, which could then allow for
Figure 3 Shown are (a) hierarchical
clustering of log2 fold changes of TE
centroids differentially expressed in at
least one high-CO
2
phenotype
treatment, (b) the distribution of TE
copies within different genomic
elements, and (c) the distribution of
TE centroids and their corresponding
genome copies. (a) shows hierarchical
clustering of log2 fold changes of
differentially expressed TE centroids in
at least one high-CO
2
phenotype
treatment (see methods) resulting in two
well-defined clusters with one
representing TE centroids with increased
average log2 fold changes in the high-
CO
2
phenotype (blue) treatments
relative to the 380-selected and vice
versa for the “decreased” cluster (red).
Bolded/asterisked labels indicate
differentially expressed TE clusters in
all high-CO
2
phenotype treatments
versus the 380-selected treatment. (b)
shows a pie chart of the distribution of
all detected TE copies in the genome. (c)
shows a genome plot of TE centroids
and their corresponding copies as well as
all detected TE copies in the genome.
Going from out to in: Track 1 shows TE
copies in the genome on the forward
strand and is colored according to
genomic element. Track 2 is the same
but on the minus strand. Track 3
contains the names and symbols for
differentially expressed TE centroids in
(a) relative to the 380-selected treatment.
Track 4 shows the distribution of the
corresponding copies of differentially
expressed TE centroids in Track 3 via
blue and red colored links. Underlying
grey links represent paralagous copies
from TE clusters showing no change in
expression.
0 kb
400 kb
800 kb
1200 kb
1600 kb
2000 kb
2400 kb
2800 kb
3200 kb
3600 kb
4000 kb
4400 kb
4800 kb
5200 kb
5600 kb
6000 kb
6400 kb
6800 kb
7200 kb
TE_58*
TE_59*
TE_56*
TE_26
TE_22
TE_20*
TE_68*
TE_40*
TE_66*
TE_12*
TE_11
TE_17*
TE_30*
TE_36
TE_37*
TE_6*
Increased
Decreased
Gene/Hypothetical/Pseudogene
Intergenic
b
Intergenic
Pseudogene
Gene
Hypothetical
a
TE_36
TE_17*
TE_26
TE_22
TE_11
TE_56*
TE_6*
TE_37*
TE_12*
TE_20*
TE_68*
TE_58*
TE_66*
TE_40*
TE_30*
TE_59*
0.0 0.5 1.0 1.5 2.0
Height
100
91 94
92 100
96 43
98
96
93
99
93
c
53
specific types of TE enrichment
41
. Furthermore, ~75% of all detected TE paralogs reside
within either genic or pseudo-genic bodies (Fig. 3b) suggesting pseudogenization as a
mechanism for generating degenerate TE genome copies. Accordingly, numerous partial
TE genome sequences are marks of neutral maintenance based on TE deletion bias
41
.
Upon plotting the locations of all differentially expressed centroid sequences along with
their corresponding paralogs (Fig. 3c), some TE clusters contained numerous copies with
widespread distributions across the genome (e.g. TE_30, TE_56) while others contained
only 1 or 2 copies (e.g. TE_12, TE_40). The mechanisms involved in the differing
degrees of TE cluster proliferation remain unknown, but no apparent trends were
observed between increased/decreased TE-cluster expression and strand, genic/intergenic
body, cluster copy number, or location. Regardless, it has been shown that repetitive
elements such as TEs can selectively mediate genome plasticity while partial TEs can
also act to inhibit transposition
49
, which may be partly why IMS101 has retained its
numerous repetitive elements and TE pseudogenes. These potential TE-controlling
mechanisms may also contribute to the high genome conservation observed between
IMS101, other isolates, and natural populations
25
. In summary, the fact that most TE
clusters show no changes in expression between the low- and high-CO
2
phenotypes
suggests that these are being maintained under neutral processes and/or weak selection
while the few that did exhibit differences between phenotypes are potential candidates
under selection. Hence, these data implicate differential regulation of certain
transposition types to be involved in or caused by both plastic and adaptive high-CO
2

phenotypes.
Functional GO-enriched transcription in plasticity and adaptation
The plastic response residing in the 380s-to-750 expression profile exhibited
significant GO enrichment of tetrapyrrole biosynthesis driven by the upregulation of
Tery_3684 (Anabaena sp. wa102 homolog, hemB, AA650_24065) consistent with prior
observations of hemB induction via decreases in Fur transcription in the cyanobacterial
nitrogen-fixer, Anabaena
33
. González et al.
33
also showed dual roles of Fur regulation in
some Heme proteins involved in tetrapyrrole biosynthesis including transcriptional
repression (e.g. hemB, hemC, and ho1) and activation (e.g. hemK, hemH). However, other
hem genes showed either broad variability or no changes in expression and seemed to be
54
regulated independently of Fur leading the authors to suggest each hem gene to be under
different regulatory mechanisms. Similar to Anabaena, we detected the upregulation of
hemB while other detected hems either showed variable or no changes in expression.
Additionally, the significant fraction of shared upregulated genes between the 380s-to-
750 and the 750s-to-380 (hypergeometric test, p < 10
-50
) were enriched in GTP and
cytochrome b6f complex activity, which transfers electrons from PSII to PSI. Enrichment
of PSII light-harvesting and electron transport metabolisms was also detected consistent
with previously observed decreases in PSI:PSII ratios under short-term exposure to high
CO
2
16
(Fig. 4). Taken together, shifts in transcription of genes involved in electron flow
deriving from PSII seem to be part of a rapid plastic response to general fluctuations in
CO
2
regardless of sign.
Interestingly, all high-CO
2
phenotype treatments demonstrated slightly different
enrichments of photosynthesis GO sub-pathways but all shared upregulation of proteins
involving electron flow and light reactions suggesting these transcriptional changes to be
important in the plasticity-mediated adaptation of the high CO
2
phenotype. Hutchins and
Levitan et al. observed no significant changes in either photosynthetic rates or
photochemical activity of PSII, respectively, between short-term, low- and high-CO
2

treatments in IMS101. This led them to suggest that increased growth and N
2
fixation is
energized from decreased energetic demands in other cellular processes (e.g. alleviation
of carbon limitation) rather than increased photosynthetic electron flow
13,16
. However,
Levitan et al. also observed lower PSI:PSII ratios under high CO
2
indicating decreased
investment in PSI biosynthesis generally consistent with our expression results (see
above), leading them to hypothesize that a reduction in iron-heavy PSI would free up
available Fe for nitrogenase. Our results generally support these observations through
decreases in carbon transport and increases in PSII-associated gene expression. However,
more studies investigating differential gene expression involved in photosystem electron
flow coupled to simultaneous measurement of photosynthetic rates in high CO
2
regimes
are necessary to determine the specific roles of the photosystems associated with
increased growth and N
2
fixation.
55
GO-enriched
groups in both
plastic and adaptive
responses (Fig. 4,
middle section
“Plastic +
Adaptive”) included
enhanced energy
production (Fig. 4,
blue symbols),
carbon fixation
(orange symbols;
consistent with
13
),
nitrogen storage
(orange symbols),
and carbon storage
(magenta symbols). As such, these enriched metabolisms highlight pathways potentially
influenced by CO
2
concentrations on short timescales, which may have been
subsequently fixed upon prolonged CO
2
exposure in a stable, nutrient replete
environment.

Teasing apart the molecular succession underlying plasticity-mediated adaptation
Although the physiological transition from plasticity to adaptation was
phenotypically neutral in the selection environment, the shared phenotype between the
reciprocal transfers and the 750-selected adaptive condition enabled identification of
expression changes that were initially involved in short-term increased growth and
sustained in long-term adaptive maintenance. However, other portions of the metabolic
pool exhibited clear expression differences solely in the 750-selected and 750s-to-380
treatments relative to the 380-selected, thereby suggesting these gene expression changes
to be specific to the maintenance of the high-CO
2
phenotype even in the ancestral
Figure 4 Shown are upregulated GO-enriched pathways for the all high CO
2

phenotype treatments relative to the 380-selected replicates.
114 12
303
92 19
44
8
750 380s-to-750
750s-to-380
Upregulated
photosynthesis
ATP binding
Photosynthesis (P) Cell signaling/communication
Transcription/Translation
Energy/General Metabolism
Nitrogen/Carbon Metabolism
Lipid Synthesis
photosystem II (PS II)
oxidoreductase activity
aspartate kinase acitivty
homoserine dehydrogenase
PSII electron transport
Chla binding
protein phosphorylation
cell communication
GTP activity
Chla biosynthesis
aconitate hydratase
acetyl-CoA biosynthesis
P light reactions
Phosphoglycerate kinase
calcium binding
magnesium chelatase binding
coenzyme binding
alanine transport
lipid biosynthesis
polysaccharide biosynthesis
amino acid biosynthesis
macromolecule metabolism
urea metabolism
tRNA activity
MET6 activity
nitrogen compound metabolism
glutamate biosynthesis
ribosome
fatty acid biosynthsis
carbon fixation
protein binding
protein kinase activity
proteolysis
transcription/RNA polymerase
D-alanine-D-alanine ligase
thylakoid membrane
chorismate synthase
phycobilisome
sulfate transport
glycine dehydrogenase
histidine biosynthesis
methionine biosynthesis
Adaptive Plastic
Plastic + Adaptive
P electron transporter
coproporphyrinogen oxidase activity
tetrapyrrole biosynthesis
56
environment. Several other lines of evidence also corroborate these molecular and
physiological parallels.
First, out of all differentially expressed genes, it is unlikely that the amount of
genes exhibiting parallel expression between the 750-selected and 750-to-380s conditions
were shared by chance alone for both the downregulated (hypergeometric test, p < 10
-42
)
and upregulated (p < 10
-79
; Figs. 2 and 4, “Adaptive” sections) fractions, implicating their
expression to be non-randomly associated to these phenotypes. Second, the strong
statistical support for the differential downregulation of specific sigma factors (and other
paralleling gene expression) in the plastic versus adaptive responses suggests differing
roles in short- and long-term CO
2
phenotypes, respectively. Specifically, differential
regulation of sigC seemed to help mediate the transition from plasticity to adaptation
while differences in sigF expression may be primarily involved in the adaptive conditions
(Fig. 2).
Additionally, the plastic response of Trichodesmium to high CO
2
in stable light
and replete nutrients may initially shield it from adaptation on short timescales since an
optimum phenotype is achieved by plasticity alone. However, upon prolonged selective
CO
2
pressure, initial short-term tradeoffs in energy reallocation towards increased growth
and N
2
fixation may become fixed if held under constant conditions, which in this case
seems to have led to a loss of the low-CO
2
phenotype (Fig. 1, blue bars). Paralleling this
physiological trend are both downregulated pathways like transcriptional regulation (e.g.
sigF) and carbon transport, as well as numerous upregulated pathways (Fig. 4).
In summary, the adaptation of IMS101 to high CO
2
in stable light and replete
nutrients is mediated through an initial plastic response reflected in parallel changes in
both phenotype and gene expression. Our data suggest upstream regulatory elements (e.g.
sigma factors) and differential regulation of transposition clusters to influence both short-
and long-term CO
2
responses. The maintenance of the adaptive phenotype in the
ancestral condition may either be a plastic response of the 750s-to-380 treatment relative
to high CO
2
utilizing the genetic architecture acquired by long-term high CO
2
exposure
(750-selected) or an inability to respond due to the loss of the low-CO
2
phenotype.
Additionally, increased transcription of photosystem electron flow and its mechanical
components (e.g. histidine enrichment
50,51
) in concert with the differential expression of
57
potential iron and redox sensing regulation possibly suggests constant light and replete
iron to be synergistically acting with enhanced CO
2
to initiate and maintain increased
growth and N
2
fixation. Indeed the short-term achievement of the plastic high-CO
2

phenotype has been demonstrated in natural populations when conditions were
appropriate
1,21
, but our observed form of laboratory adaptation, defined by the apparent
loss of the low-CO
2
phenotype, will likely depend on both genotype
22
and the consistent
availability of in situ compensatory environmental factors (e.g. replete phosphorus and
iron) in order to maintain increased growth and N
2
fixation.
The plastic high-CO
2
phenotype seemed to have been fixed upon adaptation in
IMS101. Hence, upon prolonged selective pressure, optimal plastic phenotypes that may
have initially shielded adaptive genotypes can be acted upon by natural selection to
facilitate adaptation. Varying physiological results have been observed in other algal
systems in which the plastic response is either maintained during evolution as in this
study or ultimately reversed by adaptation
52,53
. However, this is the only known system to
date wherein fitness in the ancestral environment (low CO
2
) actually increased in all high
CO
2
-selected lines, demonstrating the apparent loss of the ancestral phenotype. Hence,
sequencing both the low and high CO
2
-selected cell lines along with their reciprocal
transfers ultimately allowed for the appropriate delineation of transcriptional changes
specific to the plastic response relative to 380 μatm CO
2
and its contribution to adaptive
phenotypic maintenance. These genetic insights are not only critical for identifying
metabolic pathways under intense selection as evolutionary drivers in adaptation, but for
also interpreting in situ genetic surveys as proxies for global change-driven evolutionary
potential in natural populations. Furthermore, these types of data also provide
environmentally relevant genetic context to physiological adaptation. Future efforts
examining both genetic and epigenetic effects on adaptation should provide insight into
potential mechanisms driving ultimate differences in expression levels between
experimental conditions
8,39,54
. In summary, this study both supports past observed
phenomena and contributes novel evolutionary observations underlined by molecular
drivers specific to controlled environmental conditions in the globally-distributed and
biogeochemically-important Trichodesmium.

58
Methods
Physiology

Growth and N
2
fixation data were obtained from Hutchins et al.
26
.

RNA isolation for Illumina Sequencing
Treatments were analyzed in biological triplicate with both sampling and RNA
isolation being conducted as previously described
25
.

Differential expression analysis
Raw fastq files were quality trimmed and filtered as previously described
25
and
mapped onto IMS101, IMG-called genes (https://img.jgi.doe.gov/) using Bowtie2
v2.2.6
45
with default settings followed by differential expression analysis using edgeR
46
.
Genes containing more than 1 read per million reads in at least 3 samples were retained
followed by library normalization using the trimmed mean of M-values (TMM) method.
Common dispersion was estimated by fitting a generalized linear model (GLM) using the
estimateGLMCommonDisp() function, and differentially expressed genes (except
transposases) were determined by fitting the negative binomial GLM using glmFit()
followed by likelihood ratio tests with glmLRT(). Genes with a Benjamin and
Hochberg’s false discovery rate (FDR) < 0.05 were deemed differentially expressed.
Venn diagrams were produced using differentially expressed gene lists per
treatment
55
to determine both shared and exclusive genes between treatments. The
“phyper” function in “R” (R Core Team 2014) was used in order to determine the
probability of sharing n or more genes (where n is a number within the shared portions of
the Venn diagrams; Figs. 2 and 4) by chance between two treatments.

Gene Ontology (GO) enrichment analysis
Gene Ontology (GO) annotations for Trichodesmium were downloaded from the
Genome2D web server (http://pepper.molgenrug.nl/index.php/bacterial-genomes). Next,
the “phyper” function in “R” (R Core Team 2014) was used to test for significant
59
enrichment of GO categories among the treatments and p-values were corrected with the
Benjamini and Hochberg method
56
using the “p.adjust” function (p <= 0.1)
29
.

Transposable element (TE) expression
TE sequences were downloaded from Walworth et al.
25
and clustered at
70% identity using USEARCH
43
, which yielded representative centroid sequences for
each of the resulting 69 clusters. Next, BLASTn
44
was used to search for all paralogous
sequences for each centroid (cluster) within the IMS101 genome with e-value <= 1e-5
and a minimum length threshold of >= 70% of the original centroid sequence length.
Quality trimmed RNA-Seq reads were then mapped (see above) to all paralogous
sequences within every cluster, and read counts for all sequences within a cluster were
summed to produce aggregated read counts per cluster followed by aforementioned
normalization and differential expression (see above). Finally, any cluster differentially
expressed in at least one high CO
2
-phenotype treatment (380s-to-750, 750-selected, or
750s-to-380) relative to the 380-selected reference was selected for downstream analysis,
and log2 fold changes were calculated for each cluster of every high-CO
2
phenotype
treatment relative to the 380-selected condition.
Hierarchical clustering with multiscale bootstrap resampling
(replicates=1000) of log2 fold changes for all high-CO
2
phenotype treatments was
conducted using “pvclust”
47
resulting in two well-defined groups of TE clusters whose
average high-CO
2
phenotype expression was either increased or decreased relative to the
380-selected reference (Fig. 3a). Welch’s T-tests assuming heteroscedasticity were
conducted in Microsoft Excel.

60
References
1. Hutchins, D. A., Mulholland, M. R. & Fu, F.-X. Nutrient cycles and Marine
Microbes in a CO
2
-Enriched Ocean. Oceanography 22, 128–145 (2009).
2. Merilä, J. & Hendry, A. P. Climate change, adaptation, and phenotypic plasticity:
the problem and the evidence. Evol Appl 7, 1–14 (2014).
3. Collins, S., Rost, B. & Rynearson, T. A. Evolutionary potential of marine
phytoplankton under ocean acidification. Evol Appl 7, 140–155 (2013).
4. Boyd, P. W. et al. Marine Phytoplankton Temperature versus Growth Responses
from Polar to Tropical Waters – Outcome of a Scientific Community-Wide Study.
PLoS ONE 8, e63091–17 (2013).
5. Schaum, E., Rost, B., Millar, A. J. & Collins, S. Variation in plastic responses of a
globally distributed picoplankton species to ocean acidification. Nature Climate
change 3, 298–302 (2012).
6. Draghi, J. A. & Whitlock, M. C. Phenotypic plasticity facilitates mutational
variance, genetic variance, and evolvability along the major axis of environmental
variation. Evolution 66, 2891–2902 (2012).
7. Schaum, C.-E., Rost, B. O. R. & Collins, S. E. A. Environmental stability affects
phenotypic evolution in a globally distributed marine picoplankton. ISME J. 10,
75–84 (2015).
8. Elena, S. F. & Lenski, R. E. Microbial genetics: Evolution experiments with
microorganisms: the dynamics and genetic bases of adaptation. Nat Rev Genet 4,
457–469 (2003).
9. Lohbeck, K. T., Riebesell, U. & Reusch, T. B. H. Gene expression changes in the
coccolithophore Emiliania huxleyi after 500 generations of selection to ocean
acidification. Proceedings of the Royal Society B: Biological Sciences 281,
20140003–20140003 (2014).
10. Lohbeck, K. T., Riebesell, U. & Reusch, T. B. H. Adaptive evolution of a key
phytoplankton species to ocean acidification. Nat. Geosci. 5, 346–351 (2012).
11. Sañudo-Wilhelmy, S. A. et al. Phosphorus limitation of nitrogen fixation by
Trichodesmium in the central Atlantic Ocean. Nature 411, 66–69 (2001).
12. Mills, M. M., Ridame, C., Davey, M., LaRoche, J. & Gelder, R. J. Iron and
phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic. Nature
429, 292–294 (2004).
13. Hutchins, D. A. et al. CO
2
control of Trichodesmium N
2
fixation, photosynthesis,
growth rates, and elemental ratios. Limnol. Oceanogr. 52, 1293–1304 (2007).
14. Sohm, J. A., Webb, E. A. & Capone, D. G. Emerging patterns of marine nitrogen
fixation. Nat Rev Micro 9, 499–508 (2011).
15. Chappell, P. D., Moffett, J. W., Hynes, A. M. & Webb, E. A. Molecular evidence
of iron limitation and availability in the global diazotroph Trichodesmium. ISME J.
6, 1728–1739 (2012).
16. Levitan, O. et al. Elevated CO 2enhances nitrogen fixation and growth in the
marine cyanobacterium Trichodesmium. Global Change Biol 13, 531–538 (2007).
17. Ramos, J. B. E. & Biswas, H. Effect of rising atmospheric carbon dioxide on the
marine nitrogen fixer Trichodesmium. Glob. Biogeochem. Cycles 21, GB2028
(2007).
61
18. Kranz, S. A. et al. Combined effects of CO
2
and light on the N
2
-fixing
cyanobacterium Trichodesmium IMS101: physiological responses. Plant
Physiology 154, 334–345 (2010).
19. Levitan, O. et al. Regulation of nitrogen metabolism in the marine diazotroph
Trichodesmium IMS101 under varying temperatures and atmospheric CO
2

concentrations. Environmental Microbiology 12, 1899–1912 (2010).
20. Garcia, N. S. et al. Interactive effects of irradiance and CO
2
on CO
2
fixation and
N
2
fixation in the diazotroph Trichodesmium erythraeum(Cyanobacteria). Journal
of Phycology 47, 1292–1303 (2011).
21. Shetye, S., Sudhakar, M. & Jena, B. Occurrence of Nitrogen Fixing
Cyanobacterium Trichodesmium under Elevated pCO
2
Conditions in the Western
Bay of Bengal. International Journal of Oceanography (2013).
http://dx.doi.org/10.1155/2013/350465
22. Hutchins, D. A., Fu, F.-X., Webb, E. A., Walworth, N. & Tagliabue, A. Taxon-
specific response of marine nitrogen fixers to elevated carbon dioxide
concentrations. Nat. Geosci. 6, 1–6 (2013).
23. Langer, G., Nehrke, G., Probert, I., Ly, J. & Ziveri, P. Strain-specific responses of
Emiliania huxleyi to changing seawater carbonate chemistry. Biogeosciences 6,
2637-2646 (2009).
24. Kremp, A., Godhe, A., Egardt, J. & Dupont, S. Intraspecific variability in the
response of bloom-forming marine microalgae to changed climate conditions.
Ecology Evol. 2, 1195-1207 (2012).
25. Walworth, N. et al. Trichodesmium genome maintains abundant, widespread
noncoding DNA in situ, despite oligotrophic lifestyle. Proc. Natl. Acad. Sci. U.S.A.
112, 4251–4256 (2015).
26. Hutchins, D. A. et al. Irreversibly increased nitrogen fixation in Trichodesmium
experimentally adapted to elevated carbon dioxide. Nat. Commun. 6, (2015).
27. Brahamsha, B. & Haselkorn, R. Identification of multiple RNA polymerase sigma
factor homologs in the cyanobacterium Anabaena sp. strain PCC 7120: cloning,
expression, and inactivation of the sigB and sigC genes. J. Bacteriol. 174, 7273–
7282 (1992).
28. Caslake, L. F., Gruber, T. M. & Bryant, D. A. Expression of two alternative sigma
factors of Synechococcus sp. strain PCC 7002 is modulated by carbon and nitrogen
stress. Microbiology 143, 3807–3818 (1997).
29. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and
dispersion for RNA-seq data with DESeq2. Genome Biol. (2014). doi:
10.1186/s13059-014-0550-8
30. Imamura, S. & Asayama, M. Sigma factors for cyanobacterial transcription. Gene
Regulation and Systems Biology 3, 65-87 (2009).
31. Camsund, D. & Lindblad, P. Engineered Transcriptional Systems for
Cyanobacterial Biotechnology. Front. Bioeng. Biotechnol. 2, 177–10 (2014).
32. López-Gomollón, S. et al. Cross-talk between iron and nitrogen regulatory
networks in Anabaena (Nostoc) sp. PCC 7120: identification of overlapping genes
in FurA and NtcA regulons. Journal of Molecular Biology 374, 267–281 (2007).
33. González, A., Bes, M. T., Valladares, A., Peleato, M. L. & Fillat, M. F. FurA is the
master regulator of iron homeostasis and modulates the expression of tetrapyrrole
62
biosynthesis genes in Anabaena sp. PCC 7120. Environmental Microbiology 14,
3175–3187 (2012).
34. Wu, L. J. & Errington, J. Identification and Characterization of a New Prespore-
Specific Regulatory Gene. J. Bacteriol. 182, 418–424 (1999).
35. Su, Z., Olman, V. & Xu, Y. Computational prediction of Pho regulons in
cyanobacteria. BMC Genomics 8, 156–12 (2007).
36. Pawlowski, K., Klosse, U. & De Bruijn, F. J. Characterization of a novel
Azorhizobium caulinodans ORS571 two-component regulatory system,
NtrY/NtrX, involved in nitrogen fixation and metabolism. Molecular and General
Genetics MGG 231, 124–138 (1991).
37. Narikawa, R., Okamoto, S., Ikeuchi, M. & Ohmori, M. Molecular evolution of
PAS domain-containing proteins of filamentous cyanobacteria through domain
shuffling and domain duplication. DNA Res. 11, 69–81 (2004).
38. Ashby, M. K. & Houmard, J. Cyanobacterial two-component proteins: structure,
diversity, distribution, and evolution. Microbiol. Mol. Biol. Rev. 70, 472–509
(2006).
39. Casacuberta, E. & González, J. The impact of transposable elements in
environmental adaptation. Mol Ecol 22, 1503–1517 (2013).
40. Raeside, C. et al. Large Chromosomal Rearrangements during a Long-Term
Evolution Experiment with Escherichia coli. mBio 5, e01377–14–e01377–14
(2014).
41. Iranzo, J., Gómez, M. J., López de Saro, F. J. & Manrubia, S. Large-Scale
Genomic Analysis Suggests a Neutral Punctuated Dynamics of Transposable
Elements in Bacterial Genomes. PLoS Comput Biol 10, e1003680–11 (2014).
42. Treangen, T. J. & Salzberg, S. L. Repetitive DNA and next-generation sequencing:
computational challenges and solutions. Nat Rev Genet 13, 36-46 (2011).
doi:10.1038/nrg3117
43. Edgar, R. C. Search and clustering orders of magnitude faster than BLAST.
Bioinformatics 26, 2460–2461 (2010).
44. Altschul, S., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic Local
Alignment Search Tool. Journal of Molecular Biology 215, 403–410 (1990).
45. Ben Langmead & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat
Meth 9, 357–359 (2012).
46. Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package
for differential expression analysis of digital gene expression data. Bioinformatics
26, 139–140 (2010).
47. Suzuki, R. & Shimodaira, H. Pvclust: an R package for assessing the uncertainty in
hierarchical clustering. Bioinformatics 22, 1540–1542 (2006).
48. Navarro-Quezada, A. & Schoen, D. J. Sequence evolution and copy number of
Ty1-copia retrotransposons in diverse plant genomes. Proc Natl Acad Sci USA 99,
268–273 (2002).
49. Delihas, N. Impact of Small Repeat Sequences on Bacterial Genome Evolution.
Genome Biology and Evolution 3, 959–973 (2011).
50. Tang, X.-S. et al. Identification of histidine at the catalytic site of the
photosynthetic oxygen-evolving complex. Proc Natl Acad Sci USA 91(2), 704–708
(1997).
63
51. Onidas, D., Stachnik, J. M., Brucker, S., Krätzig, S. & Gerwert, K. Histidine is
involved in coupling proton uptake to electron transfer in photosynthetic proteins.
European Journal of Cell Biology 89, 983–989 (2010).
52. Schaum, C. E. & Collins, S. Plasticity predicts evolution in a marine alga.
Proceedings of the Royal Society B: Biological Sciences 281, 20141486–20141486
(2014).
53. Reusch, T. B. H. Climate change in the oceans: evolutionary versus phenotypically
plastic responses of marine animals and plants. Evol Appl 7, 104–122 (2013).
54. Kronholm, I. & Collins, S. Epigenetic mutations can both help and hinder adaptive
evolution. Mol Ecol (2015). doi:10.1111/mec.13296
55. Heberle, H., Meirelles, G. V., da Silva, F. R., Telles, G. P. & Minghim, R.
InteractiVenn: a web-based tool for the analysis of sets through Venn diagrams.
BMC Bioinformatics 16, 1 (2015).
56. Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and
powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol 57,
289–300 (1995). doi:10.2307/2346101

64
Chapter 4
Mechanisms of increased Trichodesmium fitness under iron and phosphorus co-
limitation in the present and future ocean
Nathan G. Walworth
1
, Fei-Xue Fu
1
, Eric A. Webb
1
, Mak A. Saito
2
, Dawn Moran
2
,
Matthew R. Mcllvin
2
, Michael D. Lee
1
, and David A. Hutchins
1
*

Publication Note
This chapter was formatted for and submitted to Nature Communications. All
supplementary information for this chapter can be found at:
https://www.dropbox.com/sh/7ey2oo3v423f3e5/AACWzzOXVFlDhnbTiqcKFfHUa?dl=
0

Author contributions
Experiments were conceived and carried out by D.A. Hutchins, F.-X. Fu, N.G. Walworth,
E.A. Webb, and M.A. Saito, with experimental and analytical contributions from D.
Moran, M. R. McIlvin, and M.D. Lee. N.G. Walworth wrote the manuscript, with
assistance from D.A. Hutchins, E.A. Webb, M.A. Saito, M.D. Lee, and F.-X. Fu.

Acknowledgments
Grant support was provided by U.S. National Science Foundation OCE 1260490 to D.A.
Hutchins, E.A. Webb and F.-X. Fu, and OCE OA 1220484 and G.B. Moore Foundation
3782 and 3934 to M.A. Saito.

1
Marine and Environmental Biology, Department of Biological Sciences, University of
Southern California, Los Angeles, California, 90089, USA.
2
Marine Chemistry and
Geochemistry Department, Woods Hole Oceanographic Institution, Woods Hole, MA,
USA.
65
Abstract
Nitrogen fixation by cyanobacteria supplies critical bioavailable nitrogen to marine
ecosystems worldwide, but field and lab data have demonstrated it to be limited by
iron, phosphorus, and/or CO
2
. To address unknown future interactions among these
factors, we grew the nitrogen-fixing cyanobacterium Trichodesmium for 1 year
under Fe/P co-limitation following 7-years of both low and high CO
2
selection. Fe/P
co-limited cell lines demonstrated a complex cellular response including increased
growth rates, broad proteome restructuring, and cell size reductions relative to
steady state growth limited by either Fe or P alone. Fe/P co-limitation increased
abundance of a protein containing a conserved domain previously implicated in cell
size regulation, suggesting a similar role in Trichodesmium. Increased CO
2
further
induced nutrient-limited proteome shifts in widespread core metabolisms. N
2
-fixing
microbes will thus be significantly impacted by interactions between elevated CO
2

and nutrient limitation, with broad implications for global biogeochemical cycles in
the future ocean.

66
Biological atmospheric nitrogen (N
2
) fixation by cyanobacteria including
filamentous Trichodesmium spp. is a globally significant biogeochemical process, as it
contributes a major fraction of the new N supporting food webs in ocean basin-scale
ecosystems
1-4
. Although prevailing N limitation of the central gyre ecosystems provides
an important ecological niche for diazotrophic cyanobacteria, field data have
demonstrated Fe or P limitation of N
2
fixation in both the Atlantic and Pacific Oceans
1,5-9
.
Traditionally, biomass limitation by the single nutrient in shortest supply (i.e. Liebig
limitation
10
) has been invoked as the controlling mechanism for marine primary
production and carbon sequestration
11,12
. Depletion of this primary limiting nutrient can
then lead to a secondary limitation by the next most limiting nutrient.
More recently, several studies have demonstrated nutrient co-limitation, whereby two
nutrients can limit growth simultaneously rather than sequentially
1,13-16
. These
observations suggest that marine microbes persistently experience periods of selective
pressure under widespread nutrient co-limitation
17
, which may have favored the evolution
of specific metabolic responses to co-limiting conditions. For instance, diazotrophic
cyanobacteria simultaneously limited by Fe and P (Fe/P co-limitation) grow and fix N
2

faster than when limited by either nutrient alone, suggesting that they may possess
adaptations specific to co-limited oligotrophic environments
18
.
Despite the apparent importance of co-limitation in marine systems, we know
little about the molecular mechanisms employed by co-limited microbes and how they
may respond to a changing ocean environment
9,13,18
. For instance, increasing
anthropogenic carbon dioxide (CO
2
) is decreasing seawater concentrations of both
hydroxide and carbonate ions (OH
-
and CO
3
2-
), thereby reducing ocean pH
19
. Hence,
long-term ocean acidification is likely to have major consequences for key nutrient
biogeochemical processes, including N
2
fixation
20
. Past work has observed divergent
responses of Trichodesmium spp. isolates to CO
2
enrichment suggesting that temporal
CO
2
fluctuations throughout Earth’s history, perhaps combined with regional
physicochemical forcings, could have resulted in differential ecotypic selection. This
niche specialization relative to CO
2
may have in turn influenced current relative
abundances and biogeographic distributions of diazotrophic cyanobacteria
21
. However,
virtually nothing is known about how adaptation of Trichodesmium to changing CO
2
will
67
in turn interact with the pervasive, long-term Fe and P co-limitation implied by in situ
observations
1
.
To begin to address these issues, we examine the cellular responses of
Trichodesmium erythraeum strain IMS101 (hereafter IMS101) to Fe and/or P (co)-
limitation using a global proteomics approach in the context of long-term adaptation to
both current CO
2
concentrations, and projected future ocean acidification conditions
22
.
Prior to the nutrient limitation experiments, one cell line was split into two CO
2
treatments of 6 biological replicates each, including a low “present day” control treatment
(380 micro-atmospheres (µatm)) and a high “year 2100” treatment (750 µatm). Both sets
of cell lines were then experimentally selected at their respective CO
2
concentrations for
~7 years. Intriguingly, at the 4.5 year mark (of the 7 years), constitutive increases in
growth and N
2
fixation rates in all high CO
2
-selected cell lines were observed, even after
they were switched back to low CO
2
(380 µatm) for 2 years
23
. These findings help to
reveal the potential responses of a key nitrogen biogeochemical cycle process to the
evolutionary consequences of natural selection by future ocean acidification. Next, we
subjected both 380- and 750-selected cell lines to long-term (~1 year) Fe/P co-limitation
selection, followed by either Fe or P additions to Fe/P co-limitation-selected subcultures
to generate steady state Fe- and P-single limitation treatments.
Our results demonstrate a complex response of cellular metabolism specific to
Fe/P co-limitation, which includes increased growth rates, broad proteome restructuring,
and cell size reductions relative to growth limited by a single nutrient. This global
cellular response may have resulted from long-term selection by widespread Fe/P co-
limitation, whereby cell size reductions help to relieve both diffusion and ligand-
exchange kinetic limitation, thereby facilitating increased growth
24
. Reduced elemental
quotas of smaller cells may also allow cells to maintain more rapid division rates when
multiple resources are limiting
23
. Furthermore, elevated CO
2
interacting with Fe/P co-
limitation induces additional proteome shifts relative to present day CO
2
, characterized
by increased abundances of proteins involved in broad cellular metabolic functions.
Together with increased growth rates, this restructuring reveals a unique co-limited
phenotype under “balancing” co-limitation, in which simultaneous limitation by two
nutrients may be more advantageous than “imbalanced” nutrient supply consisting of
68
severe limitation by one resource and an excess of the other. This response fundamentally
alters traditional interpretations of interactive nutrient limitations
10
and their
consequences for key global biogeochemical processes in both the present and future
ocean.
Results and Discussion
We generated Fe/P co-limited treatments from the 380- and 750-selected IMS101
cell lines, using semi-continuous culturing methods in biological triplicate for >1 year at
each CO
2
level (Methods). Following this extended co-limitation growth period, either Fe
or P was added to subsamples of the Fe/P co-limited cell lines, which were then allowed
to acclimate for ~2 months prior to sampling in order to create steady-state, triplicate Fe-
and P-single limitation treatments at both CO
2
levels. This experimental design enables
us to examine the effects of both short- and long-term nutrient limitation by Fe and/or P
on protein biochemistry of Trichodesmium following ~1000 (380-selected) to ~1500
(750-selected) generations of selection by CO
2
. It has been observed that the mid-latitude
oligotrophic oceans may be persistently Fe/P co-limited, a situation which is episodically
relieved by pulses of either Fe or P from sources like atmospheric dust or advection
1,9
.
Accordingly, our experimental design was intended to mimic these processes using
diazotrophs adapted to both present and future CO
2
concentrations. Hence, we used these
two CO
2
concentrations to generate a simulated chronological progression examining
interactions between Fe and/or P (co)-limitation scenarios and changing CO
2
levels in
Trichodesmium, from today up to the year 2100.
69
Fe/P co-limited growth rates of IMS101
adapted to 380 and 750 µatm CO
2
were lower than
those of Fe- and P-replete cells, but were
significantly increased relative to both Fe- and P-
single limitation treatments (by 100-110% and 22-
43%, respectively, p < 0.05)(Fig. 1a, top). This
growth response is consistent with previous results
at current CO
2
levels
18
. Additionally, in the Fe/P
co-limited cell lines, particulate organic carbon
(POC) per unit of trichome length was 19-31%
(380-selected) and 15-29% (750-selected) lower
than in Fe-limited, P-limited, and replete
treatments, also similar to Garcia et al.
18
(Fig. 1a,
bottom). When compared with either single
limitation treatment, this distinctive co-limitation
phenotype suggests a large but unexplained
advantage under a two nutrient “balancing
limitation” regime, which is associated with
reductions in cell size and volume. This
demonstration of a substantial reproductive fitness
advantage under Fe/P co-limiting conditions
contrasts to the lower growth rate and biomass
production typically seen under single nutrient
limitations, and which have been the focus of most
prior work on diazotroph physiology.
Interaction of elevated CO
2
and nutrient limitation on proteome variation
Proteome analysis detected 1836 proteins using a <1% false discovery rate (0.3%
actual; minimum of two peptides per protein; Methods), resulting in 307,509 spectra
containing identified peptides from 24 discrete samples. Mapping to the ~5076 proteins
predicted in the IMS101 genome
25
yielded 37% coverage of the potential proteome
(1836/5076). Relative protein abundances were measured by normalized spectral counts
Figure 1. Cell sizes, growth rates and
global proteome redundancy analysis
(RDA) of all eight factorial Fe, P, and CO
2

interaction treatments (a) Cell-specific
growth rates are significantly increased under
Fe/P co-limitation, relative to either Fe- or P-
limitation. Concurrently, cell size (assessed
using the proxy carbon content per filament
length (μg C/μm)) was significantly reduced
in Fe/P co-limited cell lines relative to all
other treatments. Error bars are standard
errors (b) RDA analysis separates biological
replicates into their respective nutrient
regimes. Low iron- treatments (Fe and Fe/P)
segregate by their differing responses to CO
2

concentration. Ellipses are 95% confidence
limits.

-10 -5 0 5 10
-10 -5 0 5 10
RDA1 (65%)
RDA2 (28%)
0 1
-P
-Fe
380 μatm CO
2
750 μatm CO
2
Fe/P co-limited
Fe/P replete
P-limited
Fe-limited
CO
2
b
0.05
0.15
0.25
0.35
0
0.01
0.02
380 750 380 750 380 750 380 750
a
Growth Rate d
-1
μg C/μm
filament
70
relative to total spectra collected across all 24 samples, with the normalization reflecting
a very minor change.
In order to test for nutrient-specific proteome variation, nonmetric
multidimensional scaling (NMDS) (Supplementary Fig. 1) and redundancy analysis
(RDA) (Fig. 1b) were applied to the normalized protein abundances revealing consistent,
nutrient-limited abundance patterns across replicate cultures for Fe/P replete, Fe-limited,
P-limited, Fe/P co-limited, and/or high and low CO
2
interactive profiles. Permutational
Multivariate Analysis of Variance (MANOVA) was used to test for statistically
significant correlations between limiting nutrient concentrations and proteome
variation
26
, revealing limitation by Fe, P, and Fe/P to all have individually significant
effects on proteome variation (p < 0.05). Elevated CO
2
alone had no significant effect on
the proteome (p > 0.05) despite inducing significant growth rate increases in replete cell
lines as previously observed
23
(Fig. 1a). However, the interactive effects of CO
2
with
each nutrient scenario (Fe, P, or Fe/P) were all significant (p < 0.05), suggesting that
increased CO
2
concentrations will significantly interact with nutrient limited proteomes
in the future ocean. Hence, the proteomes examined in this study strongly grouped by
treatment and were significantly correlated to limiting nutrient concentrations, thus
reflecting specific nutrient-limited or co-limited metabolisms. In particular, the
distinctive segregation of Fe/P co-limited proteomes away from other single limitation
treatments (Fig. 1b and Supplementary Fig. 1) suggests a broad biochemical response
underlying the concurrent increase in growth and decrease in cell size.
71
In order to analyze nutrient-specific protein responses, we tested for pairwise
changes in protein abundances in each
nutrient limitation scenario (P-limited,
Fe-limited, and Fe/P co-limited) relative
to either the replete 380 µatm CO
2
-
selected (r380) or the replete 750 µatm
CO
2
-selected

(r750) controls using the
Power Law Global Error Model
27
(Fig.
2). In this simulated chronological CO
2

timeline, comparison of the r380 versus
each nutrient-limited treatment in the
380 µatm CO
2
-selected cultures (n380s)
reflects proteome changes due to
nutrient (co)-limitation in the absence of
elevated CO
2
interactions, as in a present
day scenario. Similarly, comparison of
the r380 versus each nutrient treatment
in the 750 µatm CO
2
-selected cell lines
(n and r750s) illuminates how nutrient
(co)-limited proteomes may interact with
future higher CO
2
conditions

(circa year
2100)
22
. Finally, analyses of the nutrient
replete and nutrient limited high CO
2
-
selected cell lines (r750 versus the
n750s) reflect proteome nutrient (co)-
limitation responses between cell lines
already acclimated to high CO
2
as in a
future ocean. In addition, we compared
Fe- and P-limited proteomes to either the
380 or 750 µatm Fe/P co-limitation
(380-Fe/P or 750-Fe/P, respectively) treatments, in order to capture steady state co-
1
0
0
1
0
0
3
0 0
1
0
3
15
22
0
1 1
9
26
0
0
0
0
0
1 1
13
42
1
14
0
3 2
4
4
0
4
1
3 9
2
21
1
8
0
35 1
14
0
3
0
1
17 20
2
2
0
21
r380 vs n380
-Fe
-Fe-P
-P
(n380)
replete
(r380)
b
a
P-limited
Fe-limited
Fe/P co-limited
r380 vs n750
-Fe
-Fe-P
CO
-P
(r380)
r750 vs n750
-Fe
-Fe-P
-P
(r750)
CO replete
(750 μatm)
2
replete
replete
2
(n750)
(n750)
750 μatm
380 μatm
750 μatm
750 μatm
380 μatm
380 μatm
= Increased
White
font
= Decrease
Black
font
Figure 2 (a) Experimental treatment comparative
matrices and (b) differential abundance of shared and
treatment-specific proteins. (a) Depiction of the method
of differential abundance analysis (see Methods) with
smaller (top) hexagons representing the replete reference
treatments (r380 and r750) being compared versus the
nutrient limitation treatments (n380s and n750s, arrows).
(b) Venn diagrams denote the relationships between the
differentially abundant proteins from the analysis in (a).
White font corresponds to proteins with significantly
increased abundances relative to replete conditions, and
black font denotes significant decreases.

72
limited proteome changes in response to single nutrient inputs (Fe or P; see below). By
assessing global proteome changes in this manner, we are able to track the effect of CO
2

on the proteomes in each of the three nutrient limitation scenarios (Fe, P, and Fe/P)
through simulated time, while also elucidating the molecular mechanisms underlying cell
size decreases and increased growth of the co-limited (Fe/P) phenotype relative to each of
the two single limitation treatments (P-limited or Fe-limited).
In the “r380 vs n380” scenario (Fig. 2a,b, top), Fe/P co-limitation (380-Fe/P)
retained the largest fraction of significantly increased protein abundances (~93% of total
increased protein abundances) while P-limitation saw the largest amount of decreased
abundances (~61%). However, the interaction of elevated CO
2
and nutrient limitation
(e.g. r380 vs n750s; Fig. 2a,b, middle) drastically increased differentially abundant
proteins in iron limitation (750-Fe), suggesting increased CO
2
to intensify iron limitation
as previously noted
23
(also see below). Additionally, numerous proteins (see below) that
only increased in Fe/P co-limitation under 380 µatm CO
2
(380-Fe/P) also increased in
750-Fe. Accordingly, in the “r750 vs n750” high-CO
2
scenario (Fig. 2a,b, bottom), Fe-
limitation dominated the increased protein fraction (~63% of total increased protein
abundances) while Fe- and P-limitation made up similar fractions of the decreased
protein abundance pool (~53% and 56%, respectively). Taken together, increasing CO
2
in
Fe-limitation induces widespread changes in protein abundances while in P-limitation
(750-P), decreases in protein abundances were primarily observed for this subset of the
proteome.
Cell size and increased growth
The ability to sustain Fe/P co-limited, increased growth relative to single
limitation appears to be facilitated through general reductions in cell size, thereby
alleviating high cell elemental quota requirements and physical nutrient acquisition
limitations on large cells imposed by ligand-exchange kinetics and diffusion. Since
uptake rate per unit volume will vary inversely with cell diameter
24
, an increased surface
area to volume quotient should help to relieve these limitations via increased transporter
density per unit area, which can be assessed when transporter protein abundances are
normalized to a proxy for cell size (µg C/µm filament length). Table 1 shows the
average percent changes in abundance for detected Fe and P acquisition proteins between
73
the r380 to 380-Fe/P, the r380 to 750-Fe/P, and the r750 to 750-Fe/P before (circle) and
after (asterisk) normalizing to µg C/µm filament length. Protein names in italics indicate
protein abundances significantly affected by CO
2
, that is from 380- to 750-Fe/P co-
limitation conditions (Supplementary Fig. 2). Due to cell size reductions seen under Fe/P
co-limitation, transporter protein abundances increase per unit surface area relative to
replete and single-limited cell sizes, which could explain the significant increases in co-
limited but not single-limited growth rates. Hence, decreasing cell size under co-
limitation enables cells to either increase or maintain uptake rates by either conserving or
only
marginally
increasing
particular
transporter
abundances
(e.g. IdiA,
PstS) per unit
area, thus
facilitating
energy and material reallocation away from protein synthesis and towards other
processes.
Biochemical evidence for cell size reduction specific to Fe/P co-limitation derives
from a hypothetical protein (Tery_1090), which was found to include an EzrA
(pfam06160) domain (web-based BLASTx)
28
. An EzrA-containing protein is required
for regulating cell size in Staphylococcus aureus, as average cell diameter significantly
increases after deletion of the EzrA protein
29
. This result is consistent with the lower
cellular mass (Fig. 1a) and concurrent significant increase in Tery_1090 abundance
exclusively under Fe/P co-limitation (Fig. 3a). Homologs (≥ 95% of the protein length
with evalue < 1e-10; Methods) for Tery_1090 were detected in a handful of colony-
forming bacteria based on a search of the current NCBI database (NCBI search using
BLASTx; Methods), and of these, maximum likelihood phylogenetic analysis places
Tery_1090 homologs among the small fraction of the colony-forming cyanobacterial
340 515 302 430
853
1036
PhnD
PstB
IdiA
388 603 137 223
121
158
Sphx 193 306 84 142
104
138
PstS 21 70 25 68
29
50
PhnM 2372 3507 776 1113
568
759
2 36 16 53
0.68
15
r380 to 380-Fe/P r380 to 750-Fe/P r750 to 750-Fe/P
* * *
Table 1 The absolute average percent changes in Fe and P acquisition protein
abundance going from the replete 380 to 380-Fe/P co-limited, the replete 380 to 750-
Fe/P co-limited, and replete 750 to 750-Fe/P co-limited before (circle) and after
(asterisk) normalizing to cell size. Italicized protein names indicate significant changes
in protein abundance under increased CO
2
in Fe/P co-limitation. Percent changes per unit
area increase for all proteins, indicating that cells may sustain or increase uptake rates
under co-limitation by conserving absolute transporter protein abundances as cell size
decreases.

74
diazotrophs (Fig. 3b). However, the EzrA domain itself is found in proteins distributed
across a broad phylogenetic range of both unicellular and colony-forming bacteria whose
global sequence homology
to Tery_1090 (e.g.
Staphylococcus EzrA
protein) fell well below our
homolog threshold,
suggesting adjacent
sequences to have
considerably diverged over
time while the domain itself
remained functionally
conserved. The exact
mechanism involved in the
proliferation of the
conserved EzrA domain
remains to be determined,
but nonetheless, the use of
different and/or divergent
cell size protein machinery
with conserved domains
suggests that there is strong
selective pressure to
maintain cell size reduction
capabilities in different
habitats among a variety of
distantly related bacteria.
These data are consistent
with prior observations of
cell size reductions with
various nutrient-limiting
Figure 3 Normalized spectral counts of the EzrA-containing protein
putatively involved in cell size regulation and a maximum likelihood
phylogeny of detected homologs of the EzrA-containing IMS101
protein. (a) The EzrA-containing protein significantly increases in
abundance exclusively under Fe/P co-limitation. Squares indicate
significance relative to the replete 380 and triangles relative to both the
replete 750 and 380. Error bars are standard errors. (b) Maximum
likelihood phylogeny of detected protein homologs in NCBI (Methods) to
the IMS101 EzrA-containing protein, with the Staphylococcus EzrA
protein previously shown to regulate cell size as the outgroup. Bootstrap
values ≥50 are noted. All detected homologs reside in colony-forming
bacteria, potentially indicating this protein to be lifestyle-specific.

0.3
Leptospira inadai serovar Lyme str. 10
Streptomyces peucetius
Burkholderia mimosarum
Streptomyces flavidovirens
Neurospora crassa OR74A
Conexibacter woesei DSM 14684
Lewinella cohaerens
Streptomyces sp. 769
Herpetosiphon aurantiacus
Microcystis aeruginosa SPC777
Chondromyces apiculatus DSM 436
Microcystis aeruginosa PCC 9808
Microcystis aeruginosa NIES-44
Streptomyces aureocirculatus
Cystobacter fuscus
Microcystis aeruginosa NIES-843
Staphylococcus aureus
Microcystis aeruginosa
Microcystis aeruginosa PCC 9807
Microcystis aeruginosa PCC 9432 a
Streptomyces sp. NRRL B-3229
Bradyrhizobium sp. ORS 285
Microcystis aeruginosa
Solirubrobacter sp. URHD0082
Microcystis aeruginosa PCC 9806
Microcystis aeruginosa PCC 9432
Trichodesmium erythraeum IMS101
Streptomyces monomycini
Saccharothrix sp. ST-888
Microcystis aeruginosa PCC 7806
Cylindrospermum stagnale PCC 7417
Stigmatella aurantiaca DW4/3-1
Kitasatospora griseola
Microcystis aeruginosa NIES-2549
Leptospira broomii serovar Hurstbridge str. 5399
Leptospira fainei serovar Hurstbridge str. BUT 6
Herpetosiphon aurantiacus DSM 785
Fischerella sp. PCC 9339
Pseudovibrio sp. JE062
Chryseobacterium sp. YR477
Microcystis aeruginosa PCC 7941
Microscilla marina ATCC 23134
Streptomyces scabrisporus
100
100
100
55
100
54
100
52
82
100
56
96
100
100
100
74
56
77
Cyanobacteria
Fungi
Actinobacteria
Chloroflexi
Bacteroidetes
Spirochaetes
Firmicutes
Proteobacteria
r380 r380 and r750
b
a
0
5
10
15
20
25
30
35
normalized spectral counts
EzrA domain
380-Fe
750-Fe
r380
r750
380-P
750-P
380 Fe/P
750 Fe/P
75
treatments in other microbial systems
30
.
Additionally, abundance profiles of other IMS101 orthologs to cell size/division
proteins are generally consistent with prior observations in other bacterial systems
30
. For
instance, cell size increases in Escherichia coli (E. coli) when the rod shape-determining
gene, mreB, is inhibited
31
. Accordingly, an MreB protein homolog (Tery_1150)
significantly increased in abundance in our IMS101 Fe/P co-limited cell lines, consistent
with decreased cell size. While the MreB protein and another cell division regulator,
MinD, showed increased expression under Fe/P co-limitation, their expression also
significantly increased under Fe-single limitation, indicating strong control by iron
limitation despite there being no observed change in cell size under this single limitation
scenario (Supplementary Fig. 3, Supplementary File 1). So, although IMS101 does not
significantly reduce cell size under iron limitation alone, limiting Fe concentrations still
impart some control over certain cell size/division machinery.
Sunda et al.
24
observed decreases in both growth rates and cell size with
decreasing iron concentrations across a range of eukaryotic phytoplankton. In contrast,
we saw reductions in growth, but not cell size, under Fe-single limitation in
Trichodesmium. Cell size only decreased once IMS101 was co-limited by both iron and
phosphorus, as in the unicellular diazotrophic cyanobacterium Crocosphaera
18
, while
growth rate simultaneously increased relative to Fe-limited growth. Both this unicellular
group and colony-forming N
2
-fixing cyanobacteria (Trichodesmium) share some cell
size/division homologs, including MinD which we observed was more abundant in
IMS101 under both Fe/P co-limitation and Fe single limitation (above), despite the lack
of cell size changes under Fe limitation alone. Hence, either unknown mechanisms in
diazotrophic cyanobacteria maintain cell size in the face of decreased growth under iron
limitation, and/or nutrient-controlled mechanisms governing cell size reductions are only
triggered under co-limiting conditions leading to increased growth. It remains to be seen
whether this coordination is specific to Fe/P co-limitation, or if other forms of co-
limitation induce a similar response.
Nutrient (co)-limited proteome profiles under increasing CO
2

Both the single- (Fe or P) and co-limited (Fe/P) cell lines shared analogous
abundance profiles of several well-characterized iron or phosphorus nutrient stress
76
proteins relative to the replete treatments, respectively (see below; Supplementary Fig. 2;
Supplementary File 1). Intriguingly, once either iron or phosphorus was added to co-
limited cell cultures to achieve new single limitation steady states, 71-86% and 97-100%
of the differentially abundant Fe- and P-limited proteins exhibited significantly reduced
abundances relative to their corresponding Fe/P co-limited abundances, respectively
(Supplementary File 2). These substantial fractions of reduced nutrient stress protein
abundances following additions of either Fe or P to co-limited cells are accompanied by
concurrent decreases in growth and increases in cell size in both Fe- and P-limited cells
(Fig. 1a). Hence, the drastic decrease in protein abundances in cells that transitioned from
Fe/P co-limited to single-limited steady states (Fe or P) via single-nutrient injections may
either be a product of reduced growth, and/or a reallocation of energy towards cell size
increases at the cost of reduced growth (Fig. 1). In order to tease apart the respective
influences of P- and Fe-limitation on both present day and future Fe/P co-limited protein
biochemistry, we examined differences in protein composition between single and co-
limited scenarios following selection by increasing CO
2.

P-limitation versus Fe/P co-limitation under increasing CO
2

Phosphorus acquisition proteins (see below; Supplementary Fig. 2) were
significantly more abundant in both the P-limited and Fe/P co-limited treatments relative
to the replete cell lines, signaling P-limitation under both conditions. However, Fe/P co-
limited cells significantly increased a large protein complement specific to co-limitation
(Figure 2b, see below). This difference between P-limitation and Fe/P co-limitation
steady states further highlights a broad, coordinated transition between single and co-
limited states, with P-limitation still persisting in both conditions.
There were also large differences between the 380-Fe/P and 750-Fe/P co-limited
proteomes. For instance, adaptation to the interaction of elevated CO
2
with Fe/P co-
limitation (750-Fe/P) induced significant reductions in the abundance of particular P
acquisition proteins (Supplementary Fig. 2), relative to 380-Fe/P. Log
2
fold changes
ranged from -1.5 to -3.3 going from 380- to 750-Fe/P co-limitation for proteins involved
in phosphonate acquisition (PhnD, PhnL, PhnK, PhnM), inorganic phosphate (P
i
) binding
(SphX), and one protein of unknown function (Tery_3845) containing the phosphorus
response regulator SphR motif, which is an ortholog to PhoB in Eschrerichia coli
32

77
(Welch’s t-test, p < 0.05; Supplementary Fig. 2). In contrast, the high affinity inorganic
phosphate (P
i
) uptake subunits of the Pst transporter complex, PstB (ATP binding) and
PstS (P
i
binding), remained unchanged in 750-Fe/P, as did the exopolyphosphatase
enzyme, SurE (Welch’s t-test, p > 0.05; Supplementary Fig. 2). The sphX gene encoding
the SphX subunit is an additional P
i
binding subunit of the Pst transporter complex only
found in a handful of cyanobacteria
32
, and is located upstream of the Pst transporter
complex operon in IMS101, suggesting its regulation to be independent of the other Pst
subunits. Specific reasons for both the reduction of the additional SphX subunit but not
other Pst subunits as well as other subunits of the phosphonate transporter complex in
750-Fe/P need further investigation, but this divergence in P-stress protein abundance
suggests that increased CO
2
may have varying effects on P-acquisition complexes under
Fe/P co-limitation, thus potentially affecting uptake efficacy of different forms of P.
Fe-limitation versus Fe/P co-limitation under increasing CO
2
Similar to published studies
33,34
, several iron stress proteins (iron-starvation-
induced-protein-A (IsiA), IsiB, iron-deficiency-induced-protein-A (IdiA)) were enriched
in both Fe-limited and Fe/P co-limited cell lines relative to the replete treatments, thereby
signaling general iron limitation (Supplementary Fig. 2; Supplementary File 1). The
smaller amount of differentially abundant proteins observed in Fe-single limitation under
present day CO
2
(380-Fe) relative to a previous Fe-limitation study conducted at present
day CO
2
34
is likely a product of the different methods employed to generate Fe-limitation,
with iron stress proteins still signaling iron limitation in both studies. In particular, Fe-
single limitation in the present study was generated via P additions to cultures already
acclimated to Fe and P co-limiting conditions for 1 year, which may impart
fundamentally different physiological pressure relative to abruptly removing iron from
replete cultures as was done in Snow et al.
34
and most other previous lab-based
experiments
33
(see Supplemental for more discussion). Once increased CO
2
interacted
with iron limitation (750-Fe), similar proteomic trends to Snow et al.
34
involving major
energy, carbon, and nitrogen pathways were observed including photosystem segregation
(Fig. 4a, see below) and reduced FBP (fructose-1,6-bisphosphate aldolase) and
nitrogenase (NifH) abundances (Supplementary Fig. 4). Additionally, IsiA, IsiB, and
IdiA significantly increased in abundance in both 750-Fe and 750-Fe/P proteomes
78
relative to the corresponding low CO
2
treatments (Welch’s t-test, p < 0.05). Taken
together, these proteome shifts suggest that elevated CO
2
intensified iron limitation
23
in
Trichodesmium N
2
-fixing metabolism, irrespective of P concentration.
Iron limitation in high CO
2
-selected cell lines (750-Fe) caused significant
decreases in abundance of all detected photosystem I (PSI) proteins, as seen in previous
present day CO
2
Fe-limitation studies
34,35
, while all detected photosystem II (PSII)
proteins either increased or maintained abundance (Fig. 4a). However, under 750-Fe/P
co-limitation, all PSI proteins exhibited abundances statistically indistinguishable from
both the replete 380 and 750 treatments, thereby indicating PSI recovery under long-term
co-limitation. Additionally, several significantly increased PSII proteins in 750-Fe
c a
b
-4
-3
-2
-1
0
1
2
3
NdhB (Tery_2529)
PsaL (Tery_1204)

PsbC (Tery_2484)
PSII antennae (Tery_2485)
PSII antennae (Tery_2483)
phycobilisome (Tery_0984)
PSII
PSI
750-Fe/P 750-Fe
Log
2
fold changes in
protein abundance
r380
r380 and r750
Increased
Decreased
750-Fe
750-Fe/P
380-Fe
380-Fe/P
r750
380-P
r380
750-P
0.00 0.05 0.10 0.15 0.20 0.25
Height
750-Fe
380-Fe
750-P
380-P
r380
r750
750 Fe/P
380 Fe/P
-2
-1
0
1
2
Figure 4 (a) Bar graph showing Log2 fold changes of photosystem proteins in the 750-Fe and 750-Fe/P
treatments, respectively, and dendrograms showing hierarchical clustering of Bray-Curtis
dissimilarities of (b) photosystem proteins and (c) proteins exclusively responding to Fe/P co-limitation
(Fe/P co-limited protein complement). Heatmap scale bars represent individual protein abundances
standardized across treatments. The symbols in (a) denote statistical significance of differentially
abundant proteins relative to either the replete 380 (square) or both the replete 380 and 750 (hexagon).
The striped fill denotes statistically significant increases in abundance relative to the replete
condition(s), and significant decreases for the black fills. In (a), PSI proteins are significantly reduced in
high CO
2
/low iron (750-Fe), but partially recover under Fe/P co-limitation (750-Fe/P), to the extent that the
750-Fe/P PSI protein abundances are not significantly different from the replete conditions. Significantly
increased PSII proteins in 750-Fe also reduced in abundance under 750-Fe/P. In (b), hierarchical clustering of
PS protein abundances segregate iron treatments (Fe and Fe/P) away from the others suggesting iron to be a
strong environmental driver of PS segregation even with varying P. In (c), hierarchical clustering of the Fe/P
protein complement groups iron limitation treatments together suggesting these proteins (the Fe/P protein
complement) to be strongly influenced by limiting iron limitation in Fe/P co-limitation.

79
reduced their abundances. Hence, the interaction of Fe-limitation with high CO
2
-selected
cell lines at steady state significantly segregates PSI and PSII protein abundance, but this
phenomenon is largely remedied under a Fe/P co-limiting regime. Hierarchical clustering
of all detected photosystem components segregated iron limited treatments (Fe and Fe/P)
away from both replete and P-limited conditions, implicating iron limitation as a primary
driver in PSI and PSII segregation even as P availability varies (Fig. 4b).
Hierarchical clustering of protein abundances exhibiting significant changes
solely under co-limitation (the Fe/P protein complement; Supplementary File 3) groups
Fe-single and Fe/P co-limitation treatments together, suggesting the Fe/P protein
complement to also be strongly influenced by Fe (Fig. 4c). Additionally, numerous
proteins that changed abundances exclusively under 380-Fe/P co-limitation also
increased in abundance under 750-Fe (Fig. 2; Supplementary File 3). Accordingly, these
trends in conjunction with the greater number of broad-function proteins shared between
Fe-single and Fe/P co-limitation (Fig. 2, Supplementary File 1 and 3) suggest the
interaction of high CO
2
adaptation and iron limitation to be a primary driver of cell size
reduction and increased growth characterizing the Fe/P phenotype.
Exclusive Fe/P co-limitation response
Of the differentially abundant proteins in 380-Fe/P co-limitation relative to the
r380, 65% (n=46) were unresponsive in either Fe- or P-limitation alone (Fig. 2b, top),
evidence that this subset of proteins responds exclusively to co-limitation (Supplementary
File 3). Proteins with increased abundances constituted the majority of the Fe/P protein
complement under all three comparative CO
2
scenarios (86%, 88%, and 90%,
respectively) with most showing no changes in abundance in single limitation treatments
(Fig. 2b). Although inherent growth rate-dependent differences between the nutrient
limitations may contribute to these proteome shifts, the notably large amount of
differentially abundant proteins (n=46) unique to 380-Fe/P co-limitation relative to the
corresponding 380-single limitation treatments (Fig. 2b, top) suggests the possibility of
an evolutionarily conserved, coordinated biochemical response controlled by unknown
regulatory systems underlying balanced limitation. This complex, distinct co-limitation
response may have evolved due to intense selection by global Fe/P co-limitation
regimes
1,14
.
80
Interestingly, although no growth rate differences were observed between the low
and high CO
2
treatments in iron-limited (Fe and Fe/P) scenarios (Fig. 1a), large
proteomic differences were observed between the 380-Fe and 750-Fe as well as between
the 380-Fe/P and 750-Fe/P (Fig. 1b; Fig. 2b; Supplementary Fig. 1; Supplementary File 1
and 3). These differences suggest that these Fe-limited, CO
2
induced proteome shifts are
not growth rate-driven, but instead are likely to be CO
2
-specific. In fact, they may
represent the cellular compensatory mechanism(s) that allow Trichodesmium to maintain
similar growth rates under Fe-limited conditions, regardless of changing CO
2
levels.
Taken together, these proteome shifts provide mechanistic insights into the departure of
the co-limiting from the single limiting response as outlined in the Liebig model
10
. They
also reveal changes to proteome architecture mediated by the interactions of CO
2
and iron
limitation that are independent of growth rate.
COG (Cluster of Orthologous Genes) categories were assigned to the distinctive
Fe/P co-limited protein complement (~72% of the proteins; Supplementary File 3;
Methods). Although full pathway characterization was not possible, possibly due to
potential limitations in protein detection, mapping these COG-assigned proteins to
KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways
36
revealed them to reside
in widespread cellular metabolisms (Supplementary File 3). These included membrane
stability and biogenesis, carbon catabolism and storage, cofactor biosynthesis, carbon
fixation, photosynthesis, and various precursor metabolisms, all of which together
indicate a broad coordinated shift in numerous general cellular processes consistent with
cell size decreases and growth rate increases.
More specifically, the enrichment of proteins mapping to metabolisms involved in
cofactor and precursor biosynthesis in concert with the increased abundances of both P-
and Fe-limitation stress proteins suggests cellular reallocation under co-limitation to the
biosynthesis of versatile precursor biomolecules
81
residing at various
metabolic junctures,
possibly allowing for
greater metabolic
flexibility (Fig. 5). Cells
experiencing (co)-
limitation typically
respond rapidly to
nutrient additions, which
involves global changes
to cellular metabolism
reflected in cell size and
growth rate changes
37
.
Hence, biosynthesis of
general precursor
molecules that can
potentially be used by
multiple pathways when
nutrient fluxes are persistently variable may enable greater cellular plasticity and energy
usage efficiency.
For example, increased protein abundance of isopentenyl pyrophosphate
isomerase (IDI; Tery_1589) and squalene synthase (SQS; Tery_2043) suggests increased
isoprenoid biosynthesis, which serve as critical components in various biochemical
functions including quinones in electron transport chains, membrane components,
photosynthetic pigments, and others
38
. These isoprenoid enzyme increases are consistent
with the concurrent increase in the protochlorophyllide reductase subunit, ChlL
(Tery_1532), and the protoporphyrinogen oxidase, HemY (Tery_2218), where ChlL is
involved in precursor production for chlorophyll (Chla) biosynthesis
39
and HemY in
precursor production for both Chla and hemes important for electron transport
40
.
Furthermore, increased abundance of enzymes such as OTC (ornithine
carbamoyltransferase; Tery_1323) involved in arginine/cyanophycin biosynthesis and
Isoprenoid
Biosynthesis
Phosphonate
Carbon catabolism
and storage
Cell membrane
biogenesis
Amino acid and
nucleotide sugar
metabolism
Nicotinate and
nicotinamide metabolism
Cell size metabolism
Photosynthesis and
carbon fixation
Fe(III)
Phosphate
PstB, PstS,
SphX
PhnD, PhnG, PhnK,
PhnL, PhnM
IdiA
Figure 5 COG-assigned proteins exclusively increased in Fe/P co-
limitation mapped to KEGG pathways. “Cell size metabolism” does
not exist in KEGG pathways, but was added to acknowledge the cell size
protein (Fig. 3a) exhibiting increased abundance in conjunction with these
broad metabolic pathways under Fe/P co-limitation. These proteins
residing in widespread precursor and carbon/energy metabolisms
highlight a broad biochemical response to co-limiting conditions.
Detected P and Fe transporters are shown indicating both Fe- and P-
limitation, and italicized names are components affected by increased
CO
2
(see main text).

82
hence nitrogen storage
41
(Supplementary Fig. 3), Glga (glycogen synthase; Tery_2147)
involved in carbon storage, ManC and RfaE (lipopolysaccharide (LPS) biosynthesis
enzymes; Tery_1856, Tery_3495) involved in membrane stability, and NadE (NAD
synthetase; Tery_1984) involved in cofactors for photosynthesis and respiration all
corroborate the re-apportionment of cellular energy towards synthesizing flexible
precursors and intermediates involved in a variety of pathways tied to core carbon
metabolism (Supplementary File 3).
Increased abundance of many precursor pathway proteins in Fe/P co-limited cells
relative to Fe-limited and P-limited steady state treatments could be related to the higher
growth rates observed in the former condition (Fig. 1a). However, the replete treatments
had significantly higher growth rates than Fe/P co-limited cultures (Fig. 1a), which were
not reflected by higher levels of precursor pathway proteins, suggesting that this cannot
be explained as a simple growth rate-driven phenomenon. Further investigations (e.g.
metabolite analyses) are necessary to validate the increased production of these
intermediates under Fe/P co-limitation, but the increased abundance of these precursor
biosynthesis proteins involved in various pathways looks to be a direct product of a
metabolic shift under balancing Fe/P co-limitation not seen under single limitations.
The cellular regulation controlling these broad, coordinated proteomic shifts in
widespread metabolic pathways under co-limitation is potentially controlled by upstream
regulatory mechanisms at a whole systems level. For example, the switching of different
RNA polymerase sigma factors (i.e. “sigma switching”) has been shown to aid in both
stress and adaptive responses via transcriptional initiation of gene sets that are specific to
particular environmental or internal cellular changes
42
. Predicted IMS101 sigma factors
in the genome were either undetectable, not expressed, or below our analysis threshold,
thus preventing confident analysis of their differential abundances. However, future
efforts can include more targeted studies looking at these proteins under nutrient
limitation. Additionally, other mechanisms such as DNA modifications
43
(e.g.
epigenetics) and transposition
44
have also been shown to aid in stress and adaptation,
which in turn affect downstream transcription and translation. Hence, widespread
changes to the proteome may be a product of coordinated changes from a smaller number
of upstream regulatory systems, each controlling numerous biochemical pathways
83
contingent upon environmental stimuli. The results described here offer insight into
downstream biochemical pathways affected by both independent and interactive nutrients
and CO
2
, and thus provide a foundation for future investigations of the regulatory
mechanisms governing these broad biochemical changes.
The biochemical/physiological coordination described here offers the first
molecular and mechanistic insight into the underlying cellular mechanisms governing the
“balancing limitation” phenotype selected for by simultaneous iron and phosphorus co-
limitation, and characterized by increased growth rates and reductions in cellular size and
volume. Looking ahead, the interaction of future increasing CO
2
with single and multiple
nutrient limitation scenarios induces fundamental metabolic shifts away from those seen
under present day CO
2
, including many responses that are clearly not simply growth rate
driven. Surprisingly, numerous proteins (e.g. photosynthetic proteins) that are
differentially abundant under either Fe- or P-limitation are maintained at similar
abundances to replete conditions under Fe/P co-limitation at high CO
2
. Concurrently,
Fe/P co-limitation also induces a significant increase of a wholly different protein
complement involved in broad cellular functions including a variety of precursor
metabolisms. The exact nature of this coordinated molecular and physiological response
under Fe/P co-limitation needs further study, but the increased abundance of the Ezra-
containing protein along with concurrent cell size reductions provides mechanistic insight
into achievement of increased fitness (growth) relative to single limitation scenarios.
Widespread co-limited oceanic conditions may have selected for a master
regulatory pathway specifically evolved to sense co-limiting regimes, which may then
affect downstream pathways to produce the co-limited phenotype. Alternatively,
sovereign Fe and P regulatory mechanisms may initially respond independently to their
respective limiting nutrients, followed by subsequent cellular coordination to produce the
observed Fe/P co-limitation proteome profiles and phenotype. Regardless, it remains to
be determined how widespread this response is among broad ranges of both prokaryotic
and eukaryotic photoautotrophs, and what the ultimate consequences will be for the
controls on key global biogeochemical processes in the present day and future ocean.
84
Methods
Culturing Methods
Stock cultures of Trichodesmium strain IMS101 were maintained in modified
Aquil medium without added combined nitrogen. They were grown at 26 °C under a
light-dark cycle of 12:12 high: dark (LD) and maintained under a light intensity of
120µmol photons m
-2
s
-1
incident irradiance. Experimental cultures were maintained in
0.2um-filtered, microwave-sterilized artificial seawater. Artificial seawater and Aquil
nutrient stocks (except for the trace metal stock) were passed through a chelex-100
column to remove contaminating Fe before medium preparation.
Semi-continuous culturing methods using optically thin cultures were conducted
to avoid nutrient limitation and targeted CO
2
level perturbations
45-47
. All experimental
conditions used three biological replicate bottles, and each replicate was diluted
individually according to growth rates calculated daily for that bottle using in vivo
chlorophyll fluorescence measurements with a Turner 10 AU fluorometer
21,47
. For all
experiments, final sampling occurred once steady-state growth (no significant difference
in growth rates) was reached for at least 10 generations, and reported growth rates were
calculated based on microscopic cell counts (see below).
To examine interactive effects of iron (Fe) and phosphorus (P) limitation on
growth of the cyanobacteria, T. erythraeum were grown in four treatments as follows: 1)
Nutrient replete, 10 uM PO
4
3-
and 250 nM Fe; 2) P-limited, 0.25 uM PO
4
3-
and 250 nM
Fe; 3) Fe-limited, 10nM Fe and 10 uM PO
4
3-
; and Fe/P co-limited, 0.25 uM PO
4
3-
and
10nM Fe. EDTA concentrations were 25 µM irrespective of Fe conditions. Each of these
four experimental nutrient treatments were generated using adapted cell lines that had
been previously selected under 380 or 750 ppm CO
2
for ~7 years
23
, for a total of eight
treatments. The replete and Fe/P co-limited cultures were grown in steady state semi-
continuous cultures for ~ 12 months at each CO
2
level. Following this long-term
incubation, either Fe or P concentration was increased in subcultures of the Fe/P co-
limited cell lines, thus creating cultures limited by either P or Fe alone, respectively.
These two sets of single nutrient limited cultures were then grown at each pCO
2
level for
~ 2 months prior to being sampled together with the replete and Fe/P co-limited cultures.
Seawater medium was bubbled with 0.2um-filtered prepared air/CO
2
mixtures
85
(Praxair) to maintain stable targeted CO
2
concentration treatments of 380 and 750 ppm.
In-line high-efficiency particulate air (HEPA) filters were employed to avoid Fe
contamination from particles in the gas tanks or lines, and pH was monitored daily with
DIC being measured at the final sampling. Once steady state growth was achieved,
Trichodesmum filament abundance and lengths were measured in a 1ml-phytoplankton
counting chamber using epifluorescence microscopy and significant differences between
treatments were calculated using two-way Anova along with the Tukey test. For
proteome analysis, cultures were swiftly and gently filtered onto 5 um polycarbonate
filters (Whatman) during the middle of the photoperiod, immediately flash frozen, and
stored in liquid nitrogen until protein extraction.
Carbonate buffer system analysis
Dissolved inorganic carbon (DIC) and pH were measured according to standard
protocols
48
. DIC samples were collected from the experimental cultures and immediately
poisoned with 200µL of a saturated HgCl
2
solution in 25mL combusted borosilicate glass
bottles. Samples were stored at room temperature until analysis on a UIC CO
2
coulometer
conducted in triplicate, as in Fu et al.
45
. Certified reference materials were used to
calibrate total DIC measurements obtained from A. Dickson (UCSD). In order to
confirm carbonate system equilibration, pH measurements were used for real-time
monitoring and made using an Orion model 8102 combination electrode.
Proteome analysis
Protein extraction, label-free mass spectrometry, and spectral count normalization
for global proteome analysis were performed as previously described
23
.
Multivariate Analysis
Filtering of the proteome was performed before multivariate analysis to eliminate
consistently low normalized spectral counts across all treatments. In many cases where a
protein contained zero counts in at least half of the treatments, the normalized spectral
counts in the remaining treatments also proved to be uninformatively low (< 3), which
could skew ordination methods. However, certain proteins were only detected when a
specific nutrient was limiting (e.g. Tery_2498, Tery_0463) in which case these proteins
had zero spectral counts in half of the treatments as above but now consistently higher
86
(>5) counts in treatments where that nutrient was limiting (e.g. iron). Thus, in order to
remove proteins with little to no expression across all treatments while simultaneously
retaining proteins that showed substantial expression only in treatments where a
particular nutrient was limiting, for each protein we summed the spectral counts across all
replicates of all treatments (24 libraries) and divided by half (12) of the number of
libraries. If this quotient was greater or equal to 5 normalized spectral counts, the protein
was retained in the ordination analysis.
Nonmetric multidimensional scaling (NMDS) on normalized spectral counts of
the filtered proteome was conducted in “R” (R Core Team 2014) using the “metaMDS”
function from the vegan package
49
with default settings (Bray-Curtis dissimilarities
computed using Wisconsin double standardization) except for “autotransform=FALSE”.
Variance partitioning was performed using the “varpart” function in vegan, and
redundancy analysis (RDA) with permutation tests was used to test the significance of the
partitions via the “rda” and “anova” functions in vegan, respectively. The input data
were the normalized spectral counts of the filtered proteome and a table containing the
nutrient concentrations per treatment.
For the Permutational Multivariate Analysis of Variance (MANOVA), Bray-
Curtis dissimilarities were calculated from the normalized spectral counts using the
“vegdist” function and subsequently input into the “adonis” function in vegan with 2000
permutations.
Exploratory based pairwise differential expression tests were administered using a
power law global error model provided in the PLGEM package with default settings
27
on
normalized spectral counts between the nutrient replete treatments and nutrient limited
treatments at each respective CO
2
. Following each pairwise test, each statistically
significant protein within the differentially expressed pool was manually checked.
In order to analyze well-characterized nutrient stress proteins specific to certain
nutrient condition, Welch’s one-tailed t-tests assuming heteroscedasticity (unequal
variances)
50
were conducted in Microsoft Excel and are noted throughout the manuscript.
T-tests were also conducted to test for significant differences between mean values of
specific proteins that changed going from single to co-limitation and are noted throughout
the text.
87
Metabolic Map
All proteins displaying differential expression exclusive to Fe/P co-limitation in
each of the three CO
2
scenarios (see main text) were matched to their corresponding
COG (Cluster of Orthologous Genes) ids as per the IMG (Integrated Microbial Genomes)
annotation (https://img.jgi.doe.gov). These COG ids were used as input into the iPath2.0
tool
36
to generate the metabolic map and beautified in Adobe Illustrator.
Hierarchical clustering of protein expression
For all heatmaps, Bray-Curtis dissimilarities were calculated on normalized
spectral counts using the “vegdist” function in the vegan package, and hierarchical
clustering was performed with the Heatplus package
51
using the average (UPGMA)
clustering method.
Maximum Likelihood Phylogeny
The Tery_1090 protein containing the Ezra domain (see main text) was searched
against the NCBI non-redundant (nr) database using the web-based BLASTx tool
28
, and
hits with evalue < 1e-10 and high scoring pairs covering >95% of the original query
length were kept in order to ensure robust homolog identification and multiple sequence
alignments. The Staphylococcus aureus
29
Ezra protein was used as the outgroup.
MAFFT was used for amino acid multiple sequence alignment
52
with default settings,
and MEGA v6.06
53
was used to construct maximum likelihood phylogeny with 100
bootstrap replicates and the following eight parameters:
(1) Substitution Type = Amino Acid (2) Substitution Model = WAG with Freqs. (+F)
Model (3) Rates Among Sites = Gamma Distributed (G) (4) No of Discrete Gamma
Categories = 5 (5) Gaps/Missing Data Treatment = Complete Deletion (6) ML Heuristic
Method = Nearest-Neighbor-Interchange (NNI) (7) Initial Tree for ML = NJ/BioNJ (8)
Branch Swap Filter = Very Strong

88
References
1. Sohm, J. A., Webb, E. A. & Capone, D. G. Emerging patterns of marine nitrogen
fixation. Nat Rev Micro 9, 499–508 (2011).
2. Moisander, P. H. et al. Unicellular Cyanobacterial Distributions Broaden the Oceanic
N
2
Fixation Domain. Science 327, 1512–1514 (2010).
3. Capone, D. G., Zehr, J. P., Paerl, H. W., Bergman, B. & Carpenter, E. J.
Trichodesmium, a Globally Significant Marine Cyanobacterium. Science 276, 1221–1229
(1997).
4. Mulholland, M. R. The fate of nitrogen fixed by diazotrophs in the ocean.
Biogeosciences 4, 37–51 (2007).
5. Rueter, J. G., Hutchins, D. A., Smith, R. W. & Unsworth, N. L. in Marine Pelagic
Cyanobacteria: Trichodesmium and other Diazotrophs 362, 289–306 (Springer
Netherlands, 1992).
6. Sañudo-Wilhelmy, S. A. et al. Phosphorus limitation of nitrogen fixation by
Trichodesmium in the central Atlantic Ocean. Nature 411, 66–69 (2001).
7. Hynes, A. M., Chappell, P. D., Dyhrman, S. T., Doney, S. C. & Webb, E. A. Cross-
basin comparison of phosphorus stress and nitrogen fixation in Trichodesmium. Limnol.
Oceanogr. 54, 1438–1448 (2009).
8. Chappell, P. D., Moffett, J. W., Hynes, A. M. & Webb, E. A. Molecular evidence of
iron limitation and availability in the global diazotroph Trichodesmium. ISME J. 6, 1728–
1739 (2012).
9. Moore, C. M. et al. Processes and patterns of oceanic nutrient limitation. Nat. Geosci.
6, 701–710 (2013).
10. de Baar, H. J. W. von Liebig's law of the minimum and plankton ecology (1899–
1991). Progress in Oceanography 33, 347–386 (1994).
11. Falkowski, P. G. Evolution of the nitrogen cycle and its influence on the biological
sequestration of CO
2
in the ocean. Nature 387, 272–275 (1997).
12. Tyrrell, T. The relative influences of nitrogen and phosphorus on oceanic primary
production. Nature 400, 525–53 (1999).
13. Saito, M. A., Goepfert, T. J. & Ritt, J. T. Some thoughts on the concept of
colimitation: Three definitions and the importance of bioavailability. Limnol. Oceanogr.
89
53, 276–290 (2008).
14. Mills, M. M., Ridame, C., Davey, M., LaRoche, J. & Gelder, R. J. Iron and
phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic. Nature 429,
292–294 (2004).
15. Saito, M. A. & Goepfert, T. J. Zinc–cobalt colimitation of Phaeocystis antarctica.
Limnol. Oceanogr. 53, 1–11 (2008).
16. Saito, M. A. et al. Multiple nutrient stresses at intersecting Pacific Ocean biomes
detected by protein biomarkers. Science 345, 1173–1177 (2014).
17. Harpole, W. S. et al. Nutrient co-limitation of primary producer communities. Ecol.
Lett. 14, 852–862 (2011).
18. Garcia, N. S., Fu, F., Sedwick, P. N. & Hutchins, D. A. Iron deficiency increases
growth and nitrogen-fixation rates of phosphorus-deficient marine cyanobacteria. ISME
J. 9, 238–245 (2014).
19. Caldeira, K. & Wickett, M. E. Oceanography: Anthropogenic carbon and ocean pH.
Nature 425, 365–365 (2003).
20. Hutchins, D. A., Mulholland, M. R. & Fu, F.-X. Nutrient cycles and Marine Microbes
in a CO
2
-Enriched Ocean. Oceanography 22, 128–145 (2009).
21. Hutchins, D. A., Fu, F.-X., Webb, E. A., Walworth, N. & Tagliabue, A. Taxon-
specific response of marine nitrogen fixers to elevated carbon dioxide concentrations.
Nat. Geosci. 6, 1–6 (2013).
22. IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change (eds Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J.
Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley). (Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA, 2013) 1535 pp,
doi:10.1017/CBO9781107415324.
23. Hutchins, D.A. et al. Irreversibly increased nitrogen fixation in Trichodesmium
experimentally adapted to elevated CO2. Nat. Commun. 6:8155
doi: 10.1038/ncomms9155 (2015).
24. Sunda, W. G. & Huntsman, S. A. Iron uptake and growth limitation in oceanic and
coastal phytoplankton. Marine Chemistry 50, 189–206 (1995).
90
25. Walworth, N. et al. Trichodesmium genome maintains abundant, widespread
noncoding DNA in situ, despite oligotrophic lifestyle. Proc. Natl. Acad. Sci. U.S.A. 112,
4251–4256 (2015).
26. Sohm, J. A. et al. Co-occurring Synechococcus ecotypes occupy four major oceanic
regimes defined by temperature, macronutrients and iron. ISME J. 1–13 (2015).
doi:10.1038/ismej.2015.115
27. Pavelka, N. et al. A power law global error model for the identification of
differentially expressed genes in microarray data. BMC Bioinformatics 5, 203 (2004).
28. Altschul, S., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic Local
Alignment Search Tool. Journal of Molecular Biology 215, 403–410 (1990).
29. Jorge, A. M., Hoiczyk, E., Gomes, J. P. & Pinho, M. G. EzrA contributes to the
regulation of cell size in Staphylococcus aureus. PLoS ONE 6, e27542 (2011).
30. Chien, A.-C., Hill, N. S. & Levin, P. A. Cell Size Control in Bacteria. Current
Biology 22, R340–R349 (2012).
31. Tropini, C. et al. Principles of bacterial cell-size determination revealed by cell-wall
synthesis perturbations. Cell Rep 9, 1520–1527 (2014).
32. Su, Z., Olman, V. & Xu, Y. Computational prediction of Pho regulons in
cyanobacteria. BMC Genomics 8, 156–12 (2007).
33. Chappell, P. D. & Webb, E. A. A molecular assessment of the iron stress response in
the two phylogenetic clades of Trichodesmium. Environ. Microbiol. 12, 13–27 (2010).
34. Snow, J. T. et al. Quantifying Integrated Proteomic Responses to Iron Stress in the
Globally Important Marine Diazotroph Trichodesmium. PLoS ONE 10, e0142626 (2015).
35. Raven, J. A. Iron acquisition and allocation in stramenopile algae. Journal of
Experimental Botany 64, 2119–2127 (2013).
36. Yamada, T., Letunic, I., Okuda, S., Kanehisa, M. & Bork, P. iPath2.0: interactive
pathway explorer. Nucleic Acids Research 39, W412–5 (2011).
37. Arrigo, K. R. Marine microorganisms and global nutrient cycles. Nature 437, 349–
355 (2005).
38. Lange, M. B., Rujan, T., Martin, W. & Croteau, R. Isoprenoid biosynthesis: The
evolution of two ancient and distinct pathways across genomes. Proc Natl Acad Sci USA
97, 13172–13177 (2000).
91
39. Fujita, Y., Tsujimoto, R. & Aoki, R. Evolutionary Aspects and Regulation of
Tetrapyrrole Biosynthesis in Cyanobacteria under Aerobic and Anaerobic Environments.
Life 5, 1172–1203 (2015).
40. Kobayashi, K., Masuda, T., Tajima, N., Wada, H. & Sato, N. Molecular Phylogeny
and Intricate Evolutionary History of the Three Isofunctional Enzymes Involved in the
Oxidation of Protoporphyrinogen IX. Genome Biology and Evolution 6, 2141–2155
(2014).
41. Stephan, D. P., Ruppel, H. G. & Pistorius, E. K. Interrelation between Cyanophycin
Synthesis, L-Arginine Catabolism and Photosynthesis in the Cyanobacterium
Synechocystis Sp. Strain PCC 6803. Zeitschrift für Naturforschung C 55, (2000).
42. Camsund, D. & Lindblad, P. Engineered Transcriptional Systems for Cyanobacterial
Biotechnology. Front. Bioeng. Biotechnol. 2, 177–10 (2014).
43. Kronholm, I. & Collins, S. Epigenetic mutations can both help and hinder adaptive
evolution. Mol. Ecol. (2015). doi:10.1111/mec.13296
44. Casacuberta, E. & González, J. The impact of transposable elements in environmental
adaptation. Mol. Ecol. 22, 1503–1517 (2013).
45. Fu, F.-X., Warner, M. E., Zhang, Y., Feng, Y. & Hutchins, D. A. Effects of Increased
Temperature and CO
2
on Photosynthesis, Growth, and Elemental Ratios in marine
Synechococcus and Prochlorococcus(Cyanobacteria). Journal of Phycology 43, 485–496
(2007).
46. Hutchins, D. A. et al. CO
2
control of Trichodesmium N
2
fixation, photosynthesis,
growth rates, and elemental ratios. Limnol. Oceanogr. 52, 1293–1304 (2007).
47. Fu, F.-X. et al. Interactions between changing pCO
2
, N
2
fixation, and Fe limitation in
the marine unicellular cyanobacterium Crocosphaera. Limnol. Oceanogr. 53, 2472–2484
(2008).
48. Cooley, S. R. & Yager, P. L. Physical and biological contributions to the western
tropical North Atlantic Ocean carbon sink formed by the Amazon River plume. Journal
of Geophysical Research 111, (2006).
49. Okasanen, J., Blanchet, F. G., Kindet, R. & Legendre, P. vegan: Community Ecology
Package, R package version 2.0-4. (2012). doi:10.1111/nph.12935/full
50. Zhang, B. et al. Detecting differential and correlated protein expression in label-free
92
shotgun proteomics. J. Proteome Res. 5, 2909–2918 (2006).
51. Ploner, A. Heatplus: heatmaps with row and/or column covariates and colored
clusters. (R package version, 2012). doi:10.1002/pmic.201400392/full
52. Katoh, K. & Standley, D. M. MAFFT multiple sequence alignment software version
7: improvements in performance and usability. Molecular Biology and Evolution 30,
772–780 (2013).
53. Tamura, K., Dudley, J., Nei, M. & Kumar, S. MEGA4: Molecular Evolutionary
Genetics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24,
1596–1599 (2007).

93

Conclusions

94
This dissertation aimed to address outstanding gaps in knowledge concerning the
molecular evolution of N
2
-fixing metabolism in Trichodesmium erythraeum IMS101
under multiple global change scenarios, and how future food webs and biogeochemical
cycling may be affected. Through next-generation sequencing and other “-omics”
technologies (e.g. proteomics), this work relates laboratory evolutionary potential with
genetic potential contained in natural populations, which will be important in assessing in
situ evolutionary trajectories available to organisms in the future oceans. The results
presented here lie at the interface of several disciplines including microbial physiology,
evolution, biogeochemistry, and oceanography and thus highlight the necessity to analyze
global change impacts through interdisciplinary approaches. These data can also lay the
groundwork for future modeling efforts aiming to account for the effects of evolutionary
processes in future ocean biogeochemistry, as recently noted by Follows et al
1
.
Relating the unusual genome architecture of IMS101 to genetic potential in situ
Due to the substantial fraction of new nitrogen supplied by Trichodesmium spp. to
the subtropical and tropical oceans, it is important to understand the ecological and
evolutionary forces underlying its unusual genome architecture in order to inform
predictions of its responses to the future changing oceans. By using a combination of
spatiotemporally distinct Trichodesmium isolate genomes, metagenomes, and
transcriptomes, we showed that the unusual low-coding density (~60%) observed in the
intensively studied Trichodesmium erythraeum IMS101 is a genus-wide feature both in
culture and in situ, and that its pseudogenes are enriched in degraded transposable
elements as opposed to functional genes. This genomic strategy is in striking contrast to
the commonly observed genome streamlining and/or high (>75%) coding density
exhibited by nearly all other cohabitating, oligotrophic prokaryotes. Prokaryotic genomes
in low nutrient oceanic regimes exhibit selective gene loss (i.e. streamlining)
2
, which has
also been associated with N-limitation
3
. Like Trichodesmium, other free-living,
unicellular marine N
2
-fixers (e.g. Crocosphaera and Cyanothece) that are also less
susceptible to N-limitation do not exhibit streamlining, as evidenced by their repetitive
and/or transposon-heavy genomes. In contrast to Trichodesmium, though, their coding
densities (>75%) still mirror those of other free-living prokaryotes. Hence, the selective
forces driving the low-coding density evolutionary trajectory exhibited by ecologically
95
successful Trichodesmium populations relative to other diazotrophs remain obscure, but
its lifestyle offers hypothetical clues into its genome formation.
The bio-ecology of Trichodesmium substantially departs from other open ocean,
free-living prokaryotes in that multiple cells are physically attached in the form of free-
floating single trichomes, or as aggregates of trichomes with dynamic morphologies.
These Trichodesmium colonies contain varied assemblages of microbial consortia,
ranging from unicellular prokaryotes and eukaryotes to juvenile copepods and decapods
containing both photoautotrophic and heterotrophic metabolisms. Hence, Trichodesmium
cells are enveloped in interactive microhabitats filled with both sister cells and other
microbial communities. This appears to be quite unlike the commonly observed open-
ocean unicellular lifestyle. Additionally, genetically distinct subpopulations of
Trichodesmium harboring varying morphotypes and microbial consortia periodically
form massive blooms, thereby potentially reducing effective population sizes of most
other genetic subpopulations and thus possibly allowing for the non-adaptive
proliferation of noncoding DNA. Our detection of structurally conserved, expressed
noncoding sequences in the majority of intergenic space further suggests that much of
Trichodesmium noncoding DNA contains regulatory elements that may have been
selectively maintained over time. Thus, the combination of the Trichodesmium-specific
complex microhabitat in addition to its bloom-forming ecology may have allowed for the
unusual inflation of its intergenic space.
These findings are important on two major fronts. First, this low coding density
exhibited by Trichodesmium compared to other ecologically successful, oligotrophic N
2
-
fixing microbes would not have been predicted based on evolutionary theory underlying
widespread observations of microbial deletion bias
4
and/or selective genome
streamlining
2,5
. Thus, these findings elucidate a rare evolutionary trajectory employed in
situ by a free-living diazotroph in a low nutrient regime, which highlights an unusually
successful, low-coding genome architecture that may affect the evolutionary responses of
Trichodesmium to interactive global change variables. Therefore, these results add
important insight into alternative evolutionary strategies in low nutrient environments.
Second, by showing the environmental conservation of low coding density in
global Trichodesmium populations, experimental evolution studies examining genome
96
evolution in IMS101 can relate potential adaptive strategies observed in the lab to the
genetic potential observed in natural populations. Since extrapolating observed laboratory
evolution to future environmental scenarios remains challenging, these data help to
provide environmental context for future lab-based experiments examining
Trichodesmium evolution in response to global change.
Relating CO
2
-driven plasticity to adaptation in Trichodesmium
In order to investigate both the short- and long-term effects of elevated CO
2
on
the transcriptional and physiological dynamics of N
2
-fixing metabolism, we grew
IMS101 cell lines for ~4.5 years at both low (380 μatm) and high (750 μatm) CO
2
,
followed by reciprocal transplantation of each of the CO
2
-selected cell lines to the
respective opposing CO
2
concentrations. Cell lines in 380 μatm CO
2
rapidly increased
their growth and N
2
fixation rates once placed into 750 μatm CO
2
(380s-to-750),
demonstrating a classic plastic response (high-CO
2
phenotype) consistent with numerous
previous studies (see Ch. 3). Following 4.5 years of both low and high CO
2
selection,
neither the low (380-selected) nor high CO
2
-selected (750-selected) cell lines exhibited
any further fitness (growth) changes apart from the initial rapid increase in growth and N
2

fixation observed in cell lines placed in high CO
2
at the onset of the experiment. Hence,
both the plastic and adaptive responses exhibited an analogous high-CO
2
phenotype
indicating phenotypic plasticity to be maintained in adaptation in IMS101. This
observation is further corroborated by the maintenance of high growth and N
2
fixation by
the 750-selected cell lines relative to 380-selected lines, even after they were placed back
into low CO
2
(750s-to-380). The fitness increase in the 750-selected cell lines in the
ancestral (380 μatm) relative to the selection (750 μatm) CO
2
concentration is
unprecedented in microbial evolution literature
6
and suggests a potential loss of plasticity
to low CO
2
in the 750-selected cell lines. Hence, shared gene expression changes
common to the high-CO
2
phenotype treatments (380s-to-750, 750-selected, 750s-to-380)
represent transcriptional responses underlying the plastic response that were subsequently
maintained in the adaptive one. Similarly, expression changes occurring in either the
plastic or adaptive response suggests transcriptional dynamics specific to each,
respectively.
97
Differential regulation of upstream, regulatory elements was enriched in both the
plastic and adaptive responses including RNA polymerase sigma factors and
transposition clusters, suggesting these transcriptional regulators may impart downstream
effects in broad metabolic pathways over both short- and long-term time scales.
Concurrent transcriptional responses in downstream gene ontology (GO) pathways
involved enrichment of increased photosynthetic electron flow and the PSII:PSI ratio,
with simultaneous decreases in carbon transport, Fur regulation, and long wavelength-
light harvesting. Additionally, differential expression of particular genes (e.g. sigF) was
only significant in the adaptive response (750-selected and 750s-to-380) suggesting their
full repression to be time-dependent and primarily involved in sustained adaptation rather
than plastic instigation of the high-CO
2
phenotype. Thus, elevated CO
2
combined with
constant light and replete nutrients increase growth and N
2
fixation of IMS101 on both
short and long timescales. Transcriptional dynamics possibly mediated by differential
sigma and transposase regulation suggest this high-CO
2
phenotype to be achieved in part
through enhanced photosynthetic electron flow at the expense of decreased carbon
transport and light and iron/oxidative stress regulation.
These findings provide the first analyses of underlying molecular dynamics
relating CO
2
-driven phenotypic plasticity to adaptation in marine microbes, and hold
important implications concerning the effects of plasticity on adaptation in IMS101. The
role of plasticity in evolution has been recently highlighted as an important determinant
in adaptation where the plastic phenotype can either be maintained as an adaptive
phenotype as described here, or can reverse in sign and/or magnitude
7,8
. Hence, these data
are not only important to fundamental evolutionary theory but also to predictions
concerning the adaptation of microbial populations to elevated CO
2
and its effects on
future biogeochemistry. Different microbial species within the same genus exhibit
varying plastic phenotypic responses to elevated CO
2
, suggesting differing degrees of
adaptive potential and therefore competitive fitness in the future acidified oceans
9-11
.
Thus, it is paramount to study the role of plasticity in evolution for different,
biogeochemically-important marine microbes as it may enable better interpretation of
short-term responses as indicators of long-term adaptation. Additionally, recent modeling
efforts have begun to incorporate both plastic and adaptive empirical physiologies into
98
global change-driven ocean circulation models
12
, which may produce misleading results
if plastic phenotypes are not representative of adaptive ones.
Another complicating factor to understanding CO
2
-driven plasticity and
adaptation in the future oceans is the interaction of rising CO
2
and other variables like
nutrient limitation. The study of both short- and long-term dynamics with CO
2
as the sole
variable is an appropriate starting point to investigate the isolated effects of rising CO
2
on
microbial metabolism, and sets the stage to study the interactive effects of CO
2
and
relevant nutrient limitation. However, microbes such as Trichodesmium must adapt to
rising CO
2
in the context of a persistently nutrient-limited environment- perhaps even
more so in the future more stratified ocean. Accordingly, we next conducted both short-
and long-term nutrient limitation experiments on CO
2
-adapted IMS101 cell lines to
examine different global change scenarios simultaneously impacting IMS101
metabolism.
The interactive effects of CO
2
and nutrient limitation
To address the fundamental knowledge gap concerning the interaction of CO
2
-
driven adaptation and pervasive nutrient (co)-limitation, we generated iron and
phosphorus (Fe/P) co-limited treatments in biological triplicate from the aforementioned
380- and 750-selected cell lines (380-Fe/P and 750-Fe/P, respectively) using semi-
continuous culturing for >1 year. Following this long-term Fe/P co-limited incubation,
either Fe or P were added back to subsamples of the 380- and 750-Fe/P co-limited
treatments, respectively, and allowed to acclimate for ~2 months to create steady-state
Fe- and P-single limitation treatments. Since persistent Fe/P co-limitation has been
observed in the mid-latitude, oligotrophic oceans with episodic pulses of either Fe and/or
P deriving from atmospheric dust or advection, this experimental design was intended to
mimic these physicochemical processes in both the present (380 μatm) and future (750
μatm) ocean.
Significantly increased growth rates were observed in high CO
2
under replete and
P-single limitation scenarios, respectively, consistent with prior studies. However, no
growth differences were observed between CO
2
levels in iron limited (Fe or Fe/P)
treatments. Hence, higher fitness under increased CO
2
concentrations occurs under
99
certain nutrient regimes (e.g. replete and P-limitation) while Fe-limitation seems to
neutralize this effect.
Intriguingly, Fe/P co-limited growth rates were reduced relative to replete cell
lines but significantly increased relative to both Fe- and P-single limitation treatments.
Hence, the Fe/P co-limited cell lines exhibited a complex, distinctive cellular response
characterized by increased growth rates, broad proteome restructuring, and cell size
reductions relative to steady state growth limited by either Fe or P alone. This Fe/P co-
limited response demonstrates a substantial fitness advantage under a two-nutrient,
“balancing” co-limitation regime relative to that of a single-nutrient, imbalanced one as
in Fe- or P-single limitation. The increased growth under co-limitation is mediated by cell
size reductions, thereby increasing cellular surface area to volume and alleviating high
cell elemental quotas and physical limitations imposed by ligand-exchange kinetics and
diffusion. Decreased cell size allows for increased nutrient transporter density on the cell
surface, which is evidenced when normalizing relative transporter abundances to co-
limited cell size relative to replete/single-limited size. This molecular/physiological
coordination may enable cells to maintain or increase uptake rates per unit area with
minimal energy allocation towards more biosynthesis of new transporter proteins. We
also identified a hypothetical protein containing an Ezra domain that significantly
increased abundance exclusively under Fe/P co-limitation, implicating its role in cell size
regulation. Accordingly, an Ezra-containing protein has been shown to regulate cell size
in Staphylococcus as cell size increased upon deletion of the gene, which is consistent
with our observed increase of the Ezra-containing protein and the simultaneous decrease
in cell size under Fe/P co-limitation.
Phosphorus acquisition proteins were significantly increased in both P-single and
Fe/P co-limitation relative to replete cultures, signaling P-limitation in both scenarios.
Intriguingly, increased CO
2
significantly reduced abundances of certain P-stress proteins
under co-limitation (e.g. 380-Fe/P versus 750-Fe/P), including those involved in
phosphonate acquisition and inorganic P binding while other P-acquisition protein
abundances remained unchanged (e.g. exopolyphosphatase). Hence, elevated CO
2
may
impart varying effects on P-acquisition protein complexes in future Fe/P co-limited
100
regimes, thus potentially affecting overall acquisition efficacy of this key nutrient in
IMS101.
Iron stress proteins were significantly increased in both Fe-single and Fe/P co-
limitation. Elevated CO
2
seemed to intensify Fe-limitation (750-Fe), as evidenced by
further increases in Fe-stress proteins as well as the significant enrichment of the
PSII:PSI ratio. Intriguingly, co-limiting 750-Fe with P-limitation (750 Fe/P) yielded a
reduced PSII:PSI ratio relative to 750-Fe, as all PSI proteins exhibited abundances
statistically indistinguishable from those in replete nutrients. Hierarchical clustering of all
photosystem proteins and the Fe/P protein complement (proteins only changing
abundance under co-limitation), respectively, grouped iron limitation (Fe and Fe/P)
treatments together, indicating Fe to heavily influence both photosystem segregation and
the Fe/P protein complement even as P concentrations vary.
Finally, the Fe/P protein complement was enriched in COG metabolic categories
involving cofactor and precursor pathways, suggesting enhanced biosynthesis of versatile
precursor biomolecules residing at various metabolic junctures. This apparently
increased biochemical flexibility could possibly allow for greater cellular plasticity under
co-limitation. The abundances of these proteins were significantly higher than those in
replete cell lines, despite the higher growth rates observed in replete versus co-limited
treatments. This indicates that increases in these precursor enzymes are not solely a
growth rate-driven phenomenon going from single- to co-limitation.
Taken together, these data offer the first coordinated molecular and physiological
insight into the interaction of CO
2
-adaptation with the distinctive Fe/P co-limited
phenotype as they relate to both replete and single nutrient limitations. The varying
metabolic responses observed under a matrix of CO
2
and nutrient limitation scenarios
highlight the complexity in predicting future evolutionary trajectories in the face of
interacting global change variables. Additionally, these cellular responses may also be
affected by interactions with cohabitating microbial taxa, necessitating approaches
looking at both individual and community-level adaptation
13
. More broadly speaking, this
multivariate approach opens a new door in the emerging multi-stressor, global change
field in microbial physiology and evolution by highlighting previously unidentified
evolutionary strategies during adaptation to co-limiting nutrient regimes.
101

Overall research significance and future directions
This work aimed to explore and contextualize the CO
2
-driven adaptation of
Trichodesmium erythraeum IMS101 in terms of genetic potential contained within
natural populations, as well as with realistic nutrient limitation scenarios commonly
observed in oligotrophic regimes. Coincidentally, the research progression described here
mirrors a broader movement in global change biological research as lessons from genetic
surveys and single-variable experiments influence a new wave of more realistic
multivariate, experimental designs incorporating physicochemical predictions in the
future oceans. Whenever possible in studying the adaptive potential of underlying
molecular/biochemical mechanisms to global change, it is invaluable to first understand
both the short- and long-term individual effects of global change variables (e.g. CO
2
,
temperature, pH, etc.) on cellular metabolism. When next transitioning to multi-stressor
experiments, interactive impacts to metabolic pathways can be appropriately teased apart
given the knowledge of their individual effects.
Additionally, since short-term dynamics (plasticity) heavily influence long-term
adaptation (see above), it is crucial that future research emphasize the investigation of
plasticity impacts on evolutionary dynamics for diverse marine taxa. Since it has been
shown that plastic responses can reverse in sign upon adaptation, I would argue that
investigating biogeochemically-important microbes that do not initially show a short-
term, plastic response to global change variables could potentially yield an opposing
adaptive one (i.e. reverse in sign) upon long-term exposure, especially under multi-
stressor regimes. Examining both short- and long-term dynamics of ecosystem-critical
organisms to global change by definition requires long-term experiments, which can be
challenging due to varying difficulty in organism-specific culturing, cryopreservation,
genetic transformation, etc. Nonetheless, the variability in global change responses by
widespread microbial taxa necessitates characterization of evolutionary dynamics in
representative organismal systems.
The multifaceted approaches employed in this work offer a framework for future
global change studies enabled by the recent wealth of genetic data characterizing both the
global oceans and organismal metabolic responses to global change. The advances in
102
next-generation sequencing offer unprecedented insight into metabolic potential
contained within organisms as they face a rapidly changing ocean. These genetic maps
of “global” metabolism have also enabled new hypotheses concerning metabolic
components responding to environmental stimuli that were once “invisible” due to
limitations in technology. However, carefully designed experiments measuring a battery
of phenotypic parameters will always yield the most comprehensive insights, irrespective
of how much sequencing is involved.
Indeed, interpretation of next-generation “-omics” results presented here would
not have been possible without thoughtful experimental design and interdisciplinary,
collaborative efforts. For example, interpreting the molecular underpinnings relating CO
2

plasticity to adaptation (Ch. 3) or to nutrient (co)-limitation dynamics (Ch. 4) would have
been less comprehensive without meticulous physiological and biogeochemical
characterization within carefully designed short- and long-term, interrelated treatments.
Of course, limitations to interdisciplinary work including lack of personnel and/or
funding will always persist, but given the increasing ease in worldwide connectivity,
seeking out collaboration in conducting global change research should be a top priority to
researchers as numerous, interacting abiotic and biotic phenomena will require
interdisciplinary interpretation.
The wealth of moving parts residing under global change is especially evident in
the challenges facing ocean circulation and biogeochemical modeling, which is beginning
to utilize the rapidly increasing global change empirical data in broad organismal taxa.
Due to the inherent interdisciplinary nature required in building and modifying such
models, experimentally based global change researchers should also inform their designs
based on its eventual integration into future modeling efforts. Collaboration between
these two fields has yielded important insights. However, much of the empirical research
incorporated into global change-driven ocean circulation models have derived from short-
term experiments
12
, which may or may not be good predictors of long-term trends.
Hence, modelers and evolutionary biologists using combinations of plastic and adaptive
data need to be cognizant of these phenomena when assessing global change.
Finally, more broadly speaking, the results described here have widespread
implications for the fields of applied science and biotechnology. The burgeoning field of
103
synthetic biology and its application to marine photoautotrophs holds promising
prospects for biosynthetic production of desirable compounds
14
, which inherently
requires knowledge of both systems biology and evolutionary processes. Hence,
biotechnological research draws heavily from broad philosophical disciplines in order to
design product-driven algal systems including microbial physiology, evolution, genetics,
and biochemistry. The results described here thus contain desirable knowledge for
application as they describe both short- and long-term dynamics under a range of
environmental conditions, and present novel evolutionary phenomena that broaden the
evolutionary scope of cyanobacteria.
The imminent effects of anthropogenic-driven global change will cause
fundamental local and global alterations to ecosystems and the economies relying on their
proper functioning. Perturbations from coastal waste, overfishing, ocean acidification and
general sea temperature warming are all occurring both simultaneously and interactively,
thereby obscuring the mechanistic nature of fundamental impacts to marine and terrestrial
biomes. Hence, these accelerating changes are creating new sub-disciplines within the
different fields of environmental science, as “global change” variables are rapidly
becoming the “new” normal. This shifting landscape has quickened the pace to try and
understand both new and long-studied systems under the context of global change, in
addition to researching solutions to try and counteract the deleterious processes
disrupting the global biosphere (e.g. from marine reserves to renewable energy). Only
through cross-disciplinary collaboration among scientific disciplines as well as a variety
of other industries including business, policy, education, and entertainment will we begin
to build a more conscious public that can hopefully work together to collectively preserve
the only planet we live on. Science can only do so much to “right our wrongs” and only
through a shift in global consciousness of the general public will we see a sustainable,
environmental movement for generations to come.

104
References

1. Follows, M. J. & Dutkiewicz, S. Modeling Diverse Communities of Marine
Microbes. Annu. Rev. Marine. Sci. 3, 427–451 (2011).
2. Swan, B. K. et al. Prevalent genome streamlining and latitudinal divergence of
planktonic bacteria in the surface ocean. Proc Natl Acad Sci USA 110, 11463–
11468 (2013).
3. Grzymski, J. J. & Dussaq, A. M. The significance of nitrogen cost minimization in
proteomes of marine microorganisms. ISME J. 6, 71–80 (2011).
4. Mira, A., Ochman, H. & Moran, N. A. Deletional bias and the evolution of
bacterial genomes. Trends in Genetics 17, 589–596 (2001).
5. Giovannoni, S. J., Thrash, J. C. & Ben Temperton. Implications of streamlining
theory for microbial ecology. ISME J. 8, 1553–1565 (2014).
6. Hutchins, D. A. et al. Irreversibly increased nitrogen fixation in Trichodesmium
experimentally adapted to elevated carbon dioxide. Nat. Commun. 6, (2015).
7. Collins, S., Rost, B. & Rynearson, T. A. Evolutionary potential of marine
phytoplankton under ocean acidification. Evol Appl 7, 140–155 (2013).
8. Schaum, C. E. & Collins, S. Plasticity predicts evolution in a marine alga.
Proceedings of the Royal Society B: Biological Sciences 281, 20141486–20141486
(2014).
9. Schaum, E., Rost, B., Millar, A. J. & Collins, S. Variation in plastic responses of a
globally distributed picoplankton species to ocean acidification. Nature Climate
change 3, 298–302 (2012).
10. Boyd, P. W. et al. Marine Phytoplankton Temperature versus Growth Responses
from Polar to Tropical Waters – Outcome of a Scientific Community-Wide Study.
PLoS ONE 8, e63091–17 (2013).
11. Hutchins, D. A., Fu, F.-X., Webb, E. A., Walworth, N. & Tagliabue, A. Taxon-
specific response of marine nitrogen fixers to elevated carbon dioxide
concentrations. Nat. Geosci. 6, 1–6 (2013).
12. Dutkiewicz, S. et al. Impact of ocean acidification on the structure of future
phytoplankton communities. Nature Climate change 5, 1002–1006 (2015).
13. Tatters, A. O. et al. Short- and long-term conditioning of a temperate marine
diatom community to acidification and warming. Philosophical Transactions of
the Royal Society B: Biological Sciences 368, 20120437–20120437 (2013).
14. Wang, B., Wang, J., Zhang, W. & Meldrum, D. R. Application of synthetic
biology in cyanobacteria and algae. Front. Microbiol. 3, (2012). doi:
10.3389/fmicb.2012.00344

Abstract (if available)

Abstract The colony-forming photoautrophic nitrogen fixer, Trichodesmium, is among the most important contributors of newly fixed nitrogen (N₂) to the global oceans. Hence, its evolutionary fate in the face of multiple, interacting global change factors will have large impacts on both marine food webs and global biogeochemistry. In order to comprehensively link global change-driven evolutionary mechanisms with future biogeochemical impacts, Trichodesmium erythraeum IMS101 (IMS101) was adapted to low and high CO₂ for ~550-900 generations under multiple nutrient limitation scenarios and subjected to a battery of physiological, biogeochemical, and molecular analyses. ❧ Analysis of genomes from multiple Trichodesmium isolates as well as metagenomes and metatranscriptomes from natural populations revealed genus-wide conserved, genome architecture unusually littered with numerous large intergenic spacers, repetitive elements, and transposases. These findings not only defy streamlining observations generally seen in free-living prokaryotes in the oligotrophic oceans but also importantly confirm the genetic potential of the lab-based evolution of IMS101 with evolutionary potential contained within natural populations. After 4.5 years of adaptation, all 6 replicates in the high-CO₂ cell lines maintained significantly higher growth and N₂ fixation rates relative to the low-CO2 cell lines under either replete nutrients or phosphorus limitation. Surprisingly, a 44% fitness increase was observed relative to the low-CO₂ cell lines when high-CO₂ cell lines were reciprocally transplanted back to the ancestral CO₂ condition. This observation is extremely rare in microbial evolution literature and has broad implications for nitrogen and carbon cycling in the future oceans. Transcriptome analysis revealed CO₂-specific transcriptional regulation of key upstream processes including transposition and sigma factors, which exert control over vast gene regulatory networks, implicating these regulatory mechanisms to be important in both short- and long-term responses to CO₂ increases. Low- and high CO₂-adapted cell lines were also subjected to iron- (Fe), phosphorus- (P), and Fe/P co-limitation in which Fe/P co-limited cell lines exhibited higher growth and N₂ fixation rates relative to either Fe- or P-limitation alone accompanied by reductions in cell size. This unexpected increase suggests Trichodesmium to be independently adapted to Fe/P co-limited regimes, which redefines classical single-nutrient Liebig limitation typically invoked to describe what is most limiting in different oceans. Global proteome restructuring in the Fe/P co-limited cell lines included alterations to both core and precursor metabolic pathways, increased abundance of proteins implicated in cell size regulation, and an increase in abundance of a protein complement specific to Fe/P co-limitation. Additionally, high CO₂ induced fundamental proteomic shifts from low to high CO₂ co-limitation, which revealed specific metabolic pathways that may come under CO₂ selective pressure in the future ocean. Accordingly, this thesis uncovered multiple levels of transcriptional and proteomic responses underlying the adaptation of Trichodesmium to high CO₂ under multiple nutrient limitation scenarios, which will further serve as invaluable data for future modeling efforts assessing the impacts of interactive global change factors on the global biosphere.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Effects of global change on the physiology and biogeochemistry of the N₂-fixing cyanobacteria Trichodesmium erythraeum and Crocosphaera watsonii

PDF

Thermal acclimation and adaptation of key phytoplankton groups and interactions with other global change variables

PDF

Iron-dependent response mechanisms of the nitrogen-fixing cyanobacterium Crocosphaera to climate change

PDF

B-vitamins and trace metals in the Pacific Ocean: ambient distribution and biological impacts

Asset Metadata

Creator Walworth, Nathan G. (author)

Core Title The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Biology

Degree Conferral Date 2016-05

Publication Date 03/16/2016

Defense Date 03/10/2016

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

Tag biological oceanography,cyanobacteria,evolution,genomics,marine microbiology,microbiology CO₂,nitrogen fixation,OAI-PMH Harvest,ocean acidification,phytoplankton,proteomics,transcriptomics,Trichodesmium

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Webb, Eric A. (committee chair), Ehrenreich, Ian M. (committee member), El-Naggar, Moh Y. (committee member), Hutchins, David A. (committee member), Nealson, Kenneth H. (committee member)

Creator Email nathan.walworth@gmail.com,nwalwort@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-222711

Unique identifier UC11279416

Identifier etd-WalworthNa-4208.pdf (filename),usctheses-c40-222711 (legacy record id)

Legacy Identifier etd-WalworthNa-4208-0.pdf

Dmrecord 222711

Document Type Dissertation

Format application/pdf (imt)

Rights Walworth, Nathan G.

Type texts

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

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

Repository Name University of Southern California Digital Library

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