sign in sign up
<<
Home , Lyase, Protein family, Pyrenoid, RuBisCO

MIAMI UNIVERSITY The Graduate School

Certificate for Approving the Dissertation

We hereby approve the Dissertation

of Jenna Dolhi

Candidate for the Degree:

Doctor of Philosophy

Dr. Rachael Morgan-Kiss, Director

Dr. Annette Bollmann, Reader

Dr. Gary Janssen, Reader

Dr. D.J. Ferguson

Dr. Melany Fisk, Graduate School Representative ABSTRACT

ENVIRONMENTAL IMPACTS ON RUBISCO: FROM GREEN ALGAL LABORATORY ISOLATES TO ANTARCTIC LAKE COMMUNITIES

by Jenna M. Dolhi

Ribulose-1,5-bisphosphate carboxylase (RubisCO) is found in a variety of autotrophic ranging from green , , and chemoautotrophic . As this has the potential to catalyze () or oxygenation () reactions, it is regulated in response to environmental variables at the levels of transcription, translation, and post-translation by the enzyme, RubisCO activase. A combination of laboratory experiments on green algal isolates and field experiments were utilized to gain insight on carbon fixation in permanently ice-covered Antarctic lakes. RubisCO was investigated as a potential target for cold adaptation of carbon fixation in the psychrophilic green alga, raudensis UWO241 (UWO241), isolated from Lake Bonney, Antarctica. RubisCO activity, stability, and whole cell carbon fixation were measured for the psychrophile and compared to a closely related mesophilic alga, C. raudensis SAG49.72 (SAG49.72). The effect of environmental factors including light and temperature on UWO241 and SAG49.72 RubisCO activation state, an indirect measurement of RubisCO activase activity, and abundance was investigated using a modified RubisCO carboxylase assay and immunoblotting, respectively. Lastly, maximum potential RubisCO carboxylase activity was determined using a modified activity assay in multiple ice covered Antarctic lakes including Lake Bonney. This data was complemented with lake depth profiles of enzyme abundance determined by quantitative real- time PCR and RubisCO-harboring organism diversity. While purified RubisCO of the psychrophilic green alga did not function optimally at low temperature, whole cell carbon fixation was greater under such conditions, suggesting that the overall process of carbon fixation is modified to function in UWO241. Increased RubisCO abundance at low temperature may contribute to this phenomenon. Low light levels may be important in regulation of RubisCO via RubisCO activase and should be further investigated. Based on community level RubsiCO activity and enzyme abundance, light and RubisCO harboring organisms including eukaryotic algae and cyanobacteria were positively correlated, but this was variable between lakes. Dark carbon fixation was potentially important in lakes west lobe Bonney and Fryxell and this community was negatively correlated with light. Results of targeted physiology and community level experiments led to development of a carbon fixation model for Lake Bonney.

ENVIRONMENTAL IMPACTS ON RUBISCO: FROM GREEN ALGAL LABORATORY ISOLATES TO ANTARCTIC LAKE COMMUNITIES A DISSERTATION

Submitted to the faculty of Miami University

in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

Department of

by

Jenna M. Dolhi

Miami University

Oxford, OH

2014

Advisor: Dr. Rachael Morgan-Kiss

Reader: Dr. Annette Bollmann

Reader: Dr. Gary Janssen

TABLE OF CONTENTS Page Chapter I: Introduction 1 1.1. Natural habitat and identity of UWO241 2 1.2. Photosynthetic electron transport in UWO241 3 1.2.1. Photosynthetic electron transport 3 1.2.2. Adaptations and acclimations of photosynthetic electron transport to irradiance 4 1.2.3. Adaptations of energetics to low temperature 6 1.3. Photosynthetic carbon fixation reactions in UWO241 7 1.3.1. Photosynthetic carbon fixation reactions 7 1.3.2. Adaptations and acclimations of to low temperature 7 1.3.3. RubisCO: the rate limiting enzyme of the CBB cycle 8 1.3.4. Adaptation of RubiscO to atmospheric CO2 and O2 concentrations 9 1.3.5. Regulation of carbon fixation: RubisCO 10 1.4. Photosynthetic microbial diversity and the impacts of 12 in the dry valley lakes 1.5. Dissertation objectives 13

Chapter II: Functional characterization of RubisCO from psychrophilic and 20 mesophilic green algal isolates 2.1. INTRODUCTION 21 2.2. METHODS 24 2.2.1. Growth conditions 24 2.2.2. rbcL and rbcS sequencing 24 2.2.3. RubisCO enzyme extraction and purification 25 2.2.4. RubisCO carboxylase activity assays 26 2.2.5. RubisCO carboxylase activity assay of crude lysates 27 2.2.6. Carbon fixation by whole cells 27 2.2.7. Chlorophyll fluorescence measurements 28 2.3. RESULTS 29 2.3.1. RubisCO subunit sequences 29 2.3.2. RubisCO purification 29 2.3.3. Temperature response of partially purified RubisCO carboxylase activity 29 2.3.4. Temperature response of RubisCO carboxylase activity of crude lysates 30 2.3.5. Temperature response of whole cell carbon fixation and 31 2.4. DISCUSSION 32

Chapter III: The effect of environmental factors (growth irradiance and temperature) 45 on modulation of RubisCO activity and abundance in psychrophilic and mesophilic green algal isolates 3.1. INTRODUCTION 46 3.2. METHODS 48 3.2.1. Strains and growth conditions 48 3.2.2. Chlorophyll fluorescence measurements 49 3.2.3. Lysate extraction and determination 49 3.2.4. RubisCO carbamylation assay 50

ii

3.2.5. SDS-PAGE and Western blotting 50 3.2.6. RubisCO activase gene sequencing 51 3.3. RESULTS 52 3.3.1. Effect of growth irradiance and temperature on growth rates 52 3.3.2. Effect of growth irradiance and temperature on photochemical function 53 3.3.3. Effect of growth irradiance and temperature on RubisCO activity and 54 carbamylation state 3.3.4. Effect of growth irradiance and temperature on RubisCO abundance 55 3.3.5. Determination of the RubisCO activase sequence for C. raudensis UWO241 55 3.4. DISCUSSION 56

Chapter IV: Diversity and distribution of carbon fixation genes in ice-covered lakes 67 of the McMurdo Dry Valleys, Antarctica 4.1. INTRODUCTION 68 4.2. METHODS 71 4.2.1. Site description 71 4.2.2. Sample collection 72 4.2.3. Limnological parameters 72 4.2.4. DNA extraction 73 4.2.5. Quantitative PCR 73 4.2.6. PCR, cloning, and sequencing 74 4.2.7. Phylogenetic analysis 74 4.2.8. Statistical analysis 75 4.3. RESULTS 75 4.3.1. Lake chemistry and 75 4.3.2. Spatial distribution of rDNA genes 76 4.3.3. Spatial distribution of major autotrophic genes 76 4.3.4. Functional gene diversity 77 4.3.5. Structure of autotrophic communities in relation to abiotic variables 79 4.4. DISCUSSION 79 4.5. ACKNOWLEDGEMENTS 84

Chapter V: Functional characterization of autotrophic and -specific 104 heterotrophic activity in permanently ice-covered lakes of the McMurdo Dry Valleys, Antarctica 5.1. INTRODUCTION 105 5.2. METHODS 108 5.2.1. Field sampling 108 5.2.2. Limnological parameters 108 5.2.3. Lysate extraction and RubisCO carboxylase activity assay 109 5.2.4. Lysate extraction and βGAM activity assay 109 5.2.5. Protein concentration determination 110 5.2.6. Bacterial enumeration 110 5.2.7. Statistical analyses 110 5.3. RESULTS 110 5.3.1. Limnological parameters 110

iii

5.3.2. MDV lake community RubisCO carboxylase activity 111 5.3.3. MDV lake community βGAM activity 112 5.3.4. Correlations of autotrophic and heterotrophic enzyme activity with 112 physicochemical lake parameters 5.4. DISCUSSION 113

Chapter VI: Conclusions 125

References 135

iv

LIST OF TABLES

Page 1. Impact of low temperatures on photostasis in Chlorella vulgaris and C. 15 raudensis UWO241 2. Anion exchange column chromatography purification of RubisCO from C. 36 raudensis UWO241 and SAG49.72 3. General physical and chemical characteristics for study sites 85 4. Primer set sequences with annealing temperatures for qPCR and/or clone library 86 construction of functional carbon fixation genes 5. Checklist for MiQE guidelines 87 6. Autotrophic gene clone libraries 89 7. Pearson’s correlation coefficients for the relationship between lake biological 90 and environmental parameters and abundance of functional genes in MDV lakes 8. Pearson correlation coefficient values for average RubisCO and βGAM activity 118 with lake physical, chemical, and biological parameters for all lakes combined 9. Pearson correlation coefficient values for average RubisCO and βGAM activity 119 with lake physical, chemical, and biological parameters for East and West lobe Bonney, Fryxell, and Vanda 10. UWO241 cDNA sequence library contigs coding for homologs of β- 133 carboxylation and TCA cycle enzymes

v

LIST OF FIGURES

Page 1. General physicochemical characteristics of Lake Bonney in Taylor Valley, 16 Antarctica 2. Photosynthetic electron transport chain in and plants 17 3. Calvin Benson Bassham cycle showing carboxylation and oxygenation reactions 18 catalyzed by the RubisCO enzyme 4. Predicted climate effects in Antarctic lakes 19 5. Neighbor-joining phylogenetic trees based on translated DNA sequences of C. 37 raudensis UWO241 and SAG49.72 RbcL and RbcS 6. SDS-PAGE of purified protein fractions from exponentially growing C. 38 raudensis UWO241 and SAG49.72 7. Activity of purified RubisCO from the psychrophilic and mesophilic C. raudensis 39 measured at various assay temperatures 8. Thermolability of purified RubisCO from C. raudensis UWO241 and SAG49.72 40 incubated at 50 °C 9. Thermolability of purified RubisCO from C. raudensis UWO241 and SAG49.72 41 incubated at various temperatures 10. RubisCO activity in crude lysate of psychrophilic (Chlamydomonas sp. CCMP 42 681, Chlamydomonas sp. ARC, Chlorella sp. BI, C. raudensis UWO241) and mesophilic green algae (C. raudensis SAG49.72) 11. Inorganic carbon uptake at various temperatures and 50 µmol photons m-2 s-1 for 43 C. raudensis UWO241 and C. raudensis SAG49.72 12. Effect of temperature on electron transport efficiency and energy distribution in 44 C. raudensis UWO241 and SAG49.72 13. Effects of irradiance and temperature on growth rates of psychrophilic C. 60 raudensis UWO241 and mesophilic C. raudensis SAG49.72 15. Effect of growth irradiance on photochemical function 61 16. Effect of growth temperature on photochemical function 62 17. Effect of growth irradiance on initially activated , maximally activated RubisCO 63 activity, and carbamylation state of C. raudensis UWO241 and SAG49.72 18. Effect of growth temperature on initially activated, maximally activated RubisCO 64 activity, and carbamylation state of C. raudensis UWO241 and SAG49.72 19. Effect of growth irradiance and growth temperature on RbcL (the large subunit of 65 RubisCO) abundance determined by densitometry 20. Neighbor-joining phylogenetic tree based on translated DNA sequence of C. 66 raudensis UWO241 RubisCO activase 21. Depth profiles of physical and chemical characteristics of east lobe Bonney, west 91 lobe Bonney, Fryxell, and Vanda 22. Depth profiles for phytoplankton diversity and biomass 92 23. Depth profiles of 16S and 18S rRNA genes in east lobe Bonney, west lobe 93 Bonney, Fryxell, and Vanda 24. East lobe and west lobe Bonney lake depth profiles of autotrophic functional gene 94 abundance 25. Lake Fryxell depth profile of autotrophic functional gene abundance 95

vi

26. Lake Vanda depth profile of autotrophic functional gene abundance 96 27. Maximum likelihood phylogenetic tree of representative form IA/B rbcL, 97 chlorophyte subgroup 28. Maximum likelihood phylogenetic tree of representative form IA/B rbcL, 98 cyanobacteria subgroup 29. Maximum likelihood phylogenetic tree of representative form IA/B rbcL, 99 subgroup 30. Tamura 3-parameter phylogenetic tree of representative form ID rbcL 100 31. Maximum likelihood phylogenetic tree of representative form II rbcL 101 32. Maximum likelihood phylogenetic tree of representative nifJ 102 33. Depth profiles of physical and chemical characteristics of East lobe Bonney, 120 West lobe Bonney, Frxyell, and Vanda 34. Preliminary of RubisCO carboxylase activity assay from filtered lake water 121 or protist enriched cultures 35. Depth profiles of autotrophic RubisCO activity and heterotrophic βGAM activity 122 measured in crude lysates extracted from filtered East lobe and West lobe Bonney 36. Depth profiles of Lake Fryxell autotrophic RubisCO activity and heterotrophic 123 βGAM activity measured in crude lysates extracted from filtered lake water 37. Depth profiles of Lake Vanda autotrophic RubisCO activity and heterotrophic 124 βGAM activity measured in crude lysates extracted from filtered lake water 38. Diagram of Lake Bonney spatial trends of form IA/B RubisCO harboring 134 organisms and those capable of dark carbon fixation (chemolithoautotrophs)

vii

ACKNOWLEDGEMENTS

The Microbiology PhD program has taken me places where I never imagined I would go. There are so many individuals who have helped to get me through the program. First, I want to acknowledge my parents for their unwavering support (on many levels). You made me believe that I could be whatever I wanted and gave me encouragement when I needed it most. You helped celebrate my successes, listened through my failures, and provided advice along the way. Thank you for everything. And thank you to my little brother who inspires me.

I am grateful to my mentor, and Team Protist leader, Dr. Rachael Morgan-Kiss, who has been generous, patient, and supportive through the years. You’ve taught me much of what I know about science and it’s been a pleasure to learn and work in your lab. Under your guidance, I have had amazing opportunities to do field research, attend conferences, and build a publication record, which have made me a well-rounded scientist. I couldn’t have done it without you. Thank you!

I must also thank my labmates, Amber, Wei, Nick, and Sarah for their help and comic relief in the lab. You have been great critics, sounding boards, hiking buddies, and fellow-adventurers. Your smarts, wit, and, positive attitudes have made bad days good and good days great. To the rest of my fellow graduate students who have become dear friends, you have made this more than just a “work place”. Your camaraderie and support have been more than I could have asked for. You know this journey well and sharing the graduate school experience with you has been invaluable.

Thank you to my friends from home, Shippensburg University, and beyond for standing by my side for all of these years-you are my people. Despite all of the miles between us your friendship has kept me grounded.

To Andrew, thank you for keeping me sane over the past year. You have helped me to focus, yet stay in touch with the real world. You have tolerated the brunt of my frustrations and have been a source of sense when all I could see was nonsense. You’re my rock.

The Microbiology Department as a whole has been an amazing place to learn to be a scientist. I appreciate the opportunities to teach, attend meetings, receive achievement awards, and network

viii with other scientists. I am grateful to my committee members Dr. Annette Bollmann, Dr. DJ Ferguson, Dr. Melany Fisk, and Dr. Gary Janssen for their support and tough love when I needed it. I would like to acknowledge the late Dr. John Hawes for his support and willingness to serve on my committee. I am also indebted to Barb Stahl and Darlene Davidson who always have the answers.

I would also like to thank and acknowledge the Center for and Functional Genomics at Miami University including Dr. Andor Kiss and Xiaoyun Deng for the experimental support.

The past six years have been painful and fun, challenging and rewarding, nerve-wracking and eye-opening. I am grateful for all of those moments and to you for being a part of them.

ix

CHAPTER 1

Introduction

Jenna Dolhi, Rachael Morgan-Kiss

Most of this chapter appeared as:

[1] Dolhi J.M., Maxwell D.P., Morgan-Kiss R.M. (2013) Review: The Antarctic Chlamydomonas raudensis: an emerging model for cold adaptation of photosynthesis. Extremophiles 17:711-722. [2] Morgan-Kiss R.M., Dolhi J. Microorganisms and Plants: a Photosynthetic Perspective. In: Storey K., Tanino K. [Eds.] Nature at Risk: Temperature in a Changing Climate. CABI 2011; pp. 24-44.

Author contributions: JD and RMK wrote and edited the manuscripts. DM contributed editorial suggestions to ref [1]

1

CHAPTER 1 INTRODUCTION Microorganisms dominate low temperature ecosystems at the level of biodiversity and abundance (Morgan-Kiss et al. 2006). Research on cold-adapted microorganisms has resulted in a considerable breadth of knowledge on elucidating the adaptations to low temperature in laboratory isolates and natural communities (D'amico et al. 2006; Margesin and Miteva 2011; Morgan-Kiss et al. 2006). A phylogenetically diverse group of microbial (, unicellular algae, fungi, chytrids, yeasts) have been detected in a variety of cold habitats (Alexander et al. 2009; Bielewicz et al. 2011; Jungblut et al. 2012; Lopez-Garcia et al. 2001; Lovejoy et al. 2006). Yet the study of psychrophilic microbial eukaryotes lags far behind their bacterial counterparts. After more than two decades of research on the Antarctic green alga Chlamydomonas raudensis UWO241 (UWO241), it is one of the most thoroughly characterized psychrophilic microbial eukaryotes and a model for adaptation of photosynthetic processes to permanent cold. Originally isolated from a permanently ice-covered lake in the dry valleys of Antarctica (Neale and Priscu 1995), this alga is currently maintained in a number of laboratories as well as the National Center for Marine Algae and Microbiota culture collection (https://ncma.bigelow.org/; strain CCMP 1619). UWO241 is not capable of photoautotrophic growth at temperatures above 16 °C and is therefore a true photopsychrophile (i.e., photosynthetic organism adapted to low temperatures) (Morgan et al. 1998). While UWO241 has been characterized thoroughly at the level of its photochemical apparatus and photosynthetic membranes (Morgan-Kiss 2002b), only one other enzymology study exists for this alga (Inman 2013). A major focus of this project is the investigation of cold adaptation of the key carbon fixation enzyme, -1,5-bisphosphate carboxylase oxygenase (RubisCO) in UWO241. Directed studies on RubisCO function in laboratory cultures of UWO241 were also complemented with carbon fixation potential and diversity of autotrophic communities residing in this organism’s natural habitat.

Natural habitat and identity of UWO241

UWO241 was isolated from the water column of an ice-covered lake (Lake Bonney) located in the McMurdo Dry Valleys (MDV), Antarctica (Fig. 1.1A). UWO241 was originally

2 classified by morphology as Chlamydomonas subcaudata (Neale and Priscu, 1995). However, more recently it was discovered that the Internal Transcribed Spacer (ITS) sequences of UWO241 are identical to those of the type strain, C. raudensis SAG49.72 (SAG49.72), while the ITS sequences of both C. raudensis strains were distantly related to C. subcaudata and the model green alga C. reinhardtii (Pocock et al. 2004). Despite identical ITS sequences, C. raudensis SAG49.72 was isolated from a pond in the Czech Republic and has been shown to be mesophilic as it has an optimum temperature for growth of 29 oC (Szyszka et al. 2007). The sister strains of C. raudensis represent a unique comparative system for understanding environmental adaptation in permanent low temperature habitats, whereby differences that may have arisen due to the of psychrophily versus speciation can be distinguished. Lake Bonney is separated into two basins named East and West Lobe Bonney (ELB and WLB, respectively). Ice covers minimize wind-driven mixing as well as allochthonous inputs, resulting in permanently stratified water chemistry and isolated microbial communities. The vast majority of organic carbon is provided by photosynthetic microbial eukaryotes (Bielewicz et al. 2011; Kong et al. 2012). UWO241 was isolated from the deep photic zone in ELB at a depth of 17 m. The unusual light environment at this depth is extreme shade below the light compensation point for photosynthesis (~5 µmol photons m-2 s-1 during mid-day in the summer; Fig. 1.1B), a light spectral distribution that is heavily biased to short wavelengths (450 – 550 nm), and extreme seasonality in intensity (24 hr summer daylight /24 hr winter darkness) (Lizotte and Priscu 1992; Lizotte et al. 1996). There is minimal exchange of gases between the water column and the atmosphere, with dissolved levels being very high and peaks at >200% saturation at the chemocline where primary productivity is at maximum levels (~ 15 m depth; Fig. 1.1B). Under- ice surface waters of Lake Bonney are ultra-oligotrophic and phosphorus-limited (Fig. 1.1C), while deeper layers of the lake are hypersaline (up to 10 times seawater) (Fritsen et al. 1988; Priscu 1995). Thus, UWO241 is a polyextremophile, possessing adaptations and acclimations to survive a range of extreme limits on growth.

1.2. Photosynthetic electron transport in UWO241

1.2.1. Photosynthetic electron transport. UWO241 is a green alga and thus the organization of its photosynthetic apparatus is very similar to plants. The transformation of light energy into chemical energy occurs through photosynthetic electron transport, which is comprised of a group

3 of supramolecular pigment-binding protein complexes and related associated with the membrane of the (Fig. 1.2). Photosynthetic electron transport is driven by absorption of light energy by the pigment/ that constitute the light harvesting complexes (LHCs) of photosystem II (PSII) and (PSI). Electron transport between PSII and

PSI is facilitated by a pool of plastoquinone and the (Fig. 1.2). NADPH is the product of electron transport; ATP is generated by chloroplastic ATP , which uses the proton gradient that is established across the thylakoid membrane during electron transport. As the demand for ATP and NADPH are not always constant, the photosynthetic apparatus can switch rapidly from linear electron transport that generates both molecules, to cyclic electron flow which generates additional ATP (Fig. 1.2; dotted line). The balance between light absorption and its utilization is a combination of acclimatory responses termed photostasis (Hüner et al. 2003). Maintaining photostasis is essential to provide maximum amounts of usable energy to the cell, while preventing the absorption of excess light, which may be damaging. Low temperature presents unique challenges to maintaining photostasis because the primary photochemical events of photon capture and charge separation are largely temperature independent, but the rate of enzyme-catalyzed reactions of the and other metabolic processes decrease as temperature decreases (Ensminger et al. 2006). UWO241 exhibits distinct acclimation strategies in comparison with mesophilic algae. In cultures grown at low temperature, an 8-fold increase in irradiance resulted in the predicted increase in PSII excitation pressure (represented by state) in both species, yet the value was always significantly lower in UWO241 than cultures of Chlorella vulgaris (Table 1.1). As well, only C. vulgaris responded to the increase in irradiance by downsizing light harvesting capacity and increasing xanthophyll cycle pigment content (Maxwell et al. 1994). That UWO241 failed to show these responses can be explained by the fact that, unlike C. vulgaris, UWO241 was able to increase its growth rate in response to the higher irradiance which is consistent with higher rates of carbon fixation potential (Morgan-Kiss et al. 2006). The increase in growth rate would require increased photosynthetic energy utilization, which has previously not been found in green algae acclimated to low temperature and provided the impetus for studying RubisCO in UWO241.

1.2.2. Adaptations and acclimations of photosynthetic electron transport to irradiance. Photoautotrophs inhabiting low light environments tend to have larger LHC's compared to

4 organisms found in high light habitats (Falkowski and Owens 1980). LHC size can be inferred by the ratio of chlorophyll (chl) a to chl b. Chl b is found exclusively in LHCs, so a lower chl a/b ratio reflects increased LHC size. UWO241 has been found to have a very low chl a/b of ~1.5 (Morgan et al. 1998; Pocock 2004), compared to a ratio > 3 for most other species of green algae. This was accompanied by a relatively high oligomeric:monomeric LHC ratio in UWO241 (Morgan et al. 1998). Unlike the vast majority of other laboratory-studied algal species, LHC size in UWO241 is generally invariant regardless of growth regime (Morgan-Kiss 2002a). These data indicate that UWO241 is adapted for growth under low light conditions and has limited ability to acclimate to changes in light intensity by adjusting photosystem antenna size. Photosynthetic organisms sense and respond to changes in the wavelength distribution of incoming light (i.e., light quality). Such adjustments are required because while PSII and PSI have distinctly different absorption characteristics (PSII preferentially blue-green light and PSI red light), light absorption by both photosystems needs to be balanced to maintain linear electron transport (Wilson et al. 2006). State transitions is a short term mechanism used by plants and algae to redistribute absorbed energy between PSII and PSI (Wollman 2001). The ability to perform state transitions requires reversible phosphorylation of LHCII proteins and rearrangements within the thylakoid membranes (Bennett et al. 1980). Data from chlorophyll fluorescence spectroscopy indicates that UWO241 does not appear to perform state transitions, and biochemical data using antibodies against phosphorylated proteins provides evidence that LHCII proteins are not phosphorylated in UWO241 (Morgan-Kiss et al. 2002a; Szyszka et al. 2007; Takizawa et al. 2009). However, unidentified protein(s) within the PSI core complex of UWO241 are reversibly phosphorylated in response to high salinity, and may play a role in controlling rates of PSI-driven cyclic electron transport and ATP synthesis (Morgan et al. 1998; Szyszka et al. 2007). Last, UWO241 cDNA sequence library identified a homologue of the LHCII stt7 (Morgan-Kiss, Kiss & Raymond, in prep.) which is essential for LHCII phosphorylation in C. reinhardtii (Depège et al. 2003), indicating that the lack of a state transition response in the psychrophile is not due to the absence of the kinase. The limited ability of UWO241 to adjust its PSI/PSII stoichiometry is reflected, in part, by its surprising inability to grow under red light (Morgan-Kiss et al. 2005). This lack of plasticity has been explained by its constitutive down-regulation of PSI that includes the presence of a very small amount of LHCI; however, surprisingly, the transcriptome of UWO241

5 possesses multiple homologues of lhcA genes which encode for LHCI proteins (Morgan-Kiss, Kiss & Raymond, in prep.). That UWO241 has lost the ability to acclimate to red light is likely the result of long-term adaptation to a native habitat that is dominated by blue-green wavelengths as long-wavelength red light is attenuated by the water column. It has been postulated that the unusual light quality response observed under lab-controlled conditions mimics adaptation in natural UWO241 communities to the Antarctic winter (Morgan-Kiss et al. 2006). This model is based on the premise that the photochemical apparatus of this photopsychrophile remains intact, but is shutdown during the Antarctic winter. This strategy is similar to that used by overwintering evergreens which modulate photochemical efficiency on a seasonal level, exhibiting a prolonged state of lowered energy conversion efficiency for the entire winter season and quickly converting to efficient energy capture during the short growing season (Demmig- Adams et al. 2012). The winter adaptation model was recently tested in a series of experiments that monitored responses of UWO241 cultures transplanted back to the organism’s original habitat (i.e., 17 m sampling depth in ELB) in a novel algal dialysis frame. In response to the loss of light availability during the seasonal transition between Antarctic summer and winter, transplanted cultures downregulated expression of genes essential for carbon fixation and photochemistry (rbcL and psbA, respectively), but maintained essential photochemical proteins during the polar night transition (Morgan-Kiss et al. in revision).

1.2.3. Adaptations of energetics to low temperature. A major trend that has been observed across many psychrophiles (including UWO241) is high intracellular adenosine 5'- triphosphate (ATP) concentrations (Napolitano and Shain 2004, 2005). In conjunction with high cellular levels of ATP, UWO241 exhibits higher levels of two major subunits of the chloroplastic ATP synthase (Morgan 1998) and up to 2-fold greater cyclic electron transport rates compared with the mesophile, C. reinhardtii (Morgan 2002b). These increases may offset the reduction in diffusion rates as temperature declines and allow for the maintenance of sufficiently high concentrations of ATP required for the activity of enzymes that catalyze endergonic (biosynthetic) reactions. Higher ATP levels may also be needed in natural communities of UWO241 residing in the hypersaline waters of the deep photic zone to actively pump Na+ across the , which is a common adaptive mechanism in salt-tolerant algae such as the model halophile, (Liska et al. 2004). In addition to higher ATP levels, a recent survey of the KEGG database shows that the genomes of many psychrophilic organisms are enriched in

6 adenosine 5’-monophosphate (AMP) synthetic pathways, while mesophilic organisms tend to possess more AMP degradative enzymes (Parry and Shain 2011).

1.3. Photosynthetic carbon fixation reactions in UWO241

1.3.1. Photosynthetic carbon fixation reactions. ATP and NADPH generated in the photochemical reactions of photosynthetic electron transport are utilized in the enzyme catalyzed carbon fixation reactions of the Calvin-Benson-Bassham (CBB) cycle, whereby inorganic carbon is reduced to sugars that become incorporated into cell biomass. The CBB cycle is responsible for the majority of carbon in the and exists in plants, green algae, cyanobacteria, and α, β, and γ (Bar-Even 2012). Carbon fixation by the CBB cycle in eukaryotes occurs in the chloroplast . Like all cold adapted organisms, UWO241 must cope with the exponential loss of reaction rates at low temperature.

1.3.2. Adaptations and acclimations of enzymes to low temperature. Many fully characterized cold-adapted enzymes exhibit high catalytic rates (kcat) at low temperatures as a result of decreased activation enthalpy (ΔH). The relationship between these parameters according to transition state theory is: -ΔG#/RT kcat = (kBT/h)e 23 -1 -34 where kB is the Boltzman constant (1.38 x 10 J K ), h is the Plank constant (6.63 x 10 J s), ΔG# is the activation free energy barrier between and transition state, or the difference between ΔH# and change in activation (ΔS#) at absolute temperature (T), and R is the universal gas constant (8.314 J mol-1 K-1) (Siddiqui and Cavicchioli 2006). Decreased ΔH# is achieved by structural modifications (Siddiqui and Cavicchioli 2006) which can also result in increased structural flexibility, reduced thermostability and inactivation at moderate temperatures in comparison to homologous mesophilic counterparts (D'amico et al. 2006; Doyle et al. 2011; Siglioccolo et al. 2010). Several studies have reported polar algal species exhibiting cold active enzymes. The Antarctic chlorophyte Koliella antarctica exhibits maximum nitrate

(NR) activity in crude lysate at lower temperatures (Tmax = 15 °C), as well as an increase in thermal lability of the enzyme compared to the mesophile Chlorella sorokiniana (Rigano et al.

2006). A low NR Tmax was also observed in the Antarctic Chloromonas sp. ANT1 (Loppes et al. 1996) and in sea-ice cultures dominated by the , Nitzschia stellate Mangin and Amphiprora kufferathii Mangin (Priscu et al. 1989). 6-

7

(G6PDH) catalyzes the first reaction in the oxidative pentose phosphate pathway, and has been implicated in cold hardening in plants and animals (Bredemeijer and Esselink 1995; Joanisse and Storey 1994). G6PDH from K. antarctica exhibits catalytic activity at low temperatures (10 oC) (Ferrara et al. 2013), and two different isoforms of G6PDH are thought to play a role in freezing tolerance in Chlorella vulgaris C-27 (Honjoh et al. 2003). Increased abundance of critical enzymes is an acclimation strategy psychrophiles employ to overcome low turnover rate of specific enzymes in crucial metabolic pathways, including the CBB cycle (Siddiqui and Cavicchioli 2006, Raven and Geider 2003, Hüner et al. 1998, Hikosaka et al. 2006). This has been observed in the green alga C. vulgaris which contains 2-3 fold more of the CBB cycle enzyme, RubisCO, when grown at 5 °C than at 27 °C (Savitch et al. 1996). Similarly, two psychrophilic Chloromonas sp. (ANT1 and ANT3) grown at 5 °C have greater RubisCO abundance than several mesophilic Chloromonas and Chlamydomonas species grown at 25 °C (Devos et al. 1998). An increased abundance of photosynthetic enzymes (RubisCO and other CBB cycle enzymes) has been observed in cold acclimation of plants including spinach, Arabidopsis, rye and wheat (Holaday et al. 1992, Strand et al. 1999, Hüner et al. 1998, Hikosaka et al. 2006).

1.3.3. RubisCO: the rate limiting enzyme of the CBB cycle. The rate limiting step of the CBB cycle is catalyzed by the enzyme RubisCO which binds a 5-carbon sugar, ribulose-1,5- bisphosphate (RuBP) and either CO2 (carboxylation) or O2 (oxygenation or photorespiration) and forms a C6 or C5 intermediate, respectively (Spreitzer and Salvucci 2002, Peterhansel and Maurino 2011). Protonation and hydration results in the products of carboxylation: two molecules of 3-phosphoglycerate (3PGA), or oxygenation: one of 3PGA and one molecule of 2-phosphoglycolate (2PG) (Spreitzer and Salvucci 2002). 3PGA is further reduced to glyceraldehyde 3-phosphate (G3P) which can be utilized in sucrose or starch synthesis (biomass) or recycled to generate RuBP. In contrast, 2PG ultimately gets recycled back to the

CBB cycle through conversion to serine in the while CO2 and NH3 are released (Fig. 1.3). This counter-productive reaction contributes to RubisCO’s role as the rate-limiting step in the CBB cycle (Peterhansel and Maurino 2011); although photorespiration is also hypothesized to be a mechanism of dissipating energy under some stressful conditions, including nutrient deficiency and excessive irradiance (Osmond 1981). Last, RubisCO has an extremely slow turnover rate and these inefficiencies necessitate that autotrophic organisms synthesize

8 large amounts of the enzyme (Spreitzer and Salvucci 2002) which is localized to a region of the stroma called the in eukaryotic algae (Raven and Beardall 2003). RubisCO exists in various forms: I, II, III and IV or RubisCO-like-proteins (RLPs), based on differences in the primary sequence of the large RubisCO subunit (Tabita et al. 2008). Form I RubisCO, which is found in almost all photosynthetic microorganisms and plants, is composed of eight large subunits (LSU; RbcL; 55 kDa) and eight small subunits (SSU; RbcS; 15 kDa). rbcL is encoded by the chloroplast genome while rbcS is nuclear encoded and the holoenzyme is assembled in the chloroplast (Spreitzer and Salvucci 2002). The large subunits of form I RubisCO are arranged in dimers which are capped by small subunits. Each LSU contains two active sites (Tabita et al. 2008) which must be cleared of inhibitory sugar to 2+ spontaneously bind non-substrate CO2 to Lys-201 (carbamylation) as well as Mg in order to be catalytically competent (Spreitzer and Salvucci 2002). RubisCO includes RuBP binding the carbamylated , forming 2,3-enediolate RuBP (-RuBP) (Spreitzer and Salvucci 2002). Following RuBP binding, loop 6 (between β sheet 6 and α helix 6 of the large subunit α/β barrel carboxy-terminus) closes to cover the active site, forming a closed state (Andersson 2008, Spreitzer and Salvucci 2002). Cleavage of the transition state intermediate C-C bond triggers opening of the closed active site for product release (Spreitzer and Salvucci 2002, Duff et al. 2000).

1.3.4. Adaptation of RubisCO to atmospheric CO2 and O2 concentrations. Substrate (CO2 or O2) binding is dictated by the enzyme’s specificity factor, the preference for CO2 versus O2 which is represented by the equation:

Ω = VcKo/VoKc where Vc and Vo are the maximal velocities (Vmax) of carboxylation and oxygenation and Ko and

Kc are the Michaelis-Menten constants (Km is the substrate concentration at which half of Vmax is achieved) for O2 and CO2 (Spreitzer and Salvucci 2002, Andersson 2008). Specificity varies amongst groups of organisms, with non-green eukaryotic algae having the highest. Green algae and land plants have intermediate specificities (Spreitzer and Salvucci 2002). The structural mechanism of carbamylation was recently determined and specificity modeled in the red alga, . This organism has a high specificity factor which is dependent on a strong electric field gradient at the active site, temperature during carbamylation (protein mobility), and directionality of ligands (Stec 2012). Organisms with low specificity can increase

9 carboxylation efficiency by increasing the CO2 concentration in the area of the active site (Andersson 2008).

Algae adapt to increasing atmospheric O2 by acquiring carbon concentrating mechanisms

(CCM), energy dependent reactions that increase the concentration of CO2 in the proximity of - RubisCO (Giordano et al. 2005). In C. reinhardtii, HCO3 and CO2 are actively transported across the plasma membrane and the chloroplast envelope. Carbonic anhydrases which convert - HCO3 to CO2 have been identified in the chloroplast stroma and thylakoid lumen (i.e., close proximity to RubisCO) (Karlsson et al. 1998, Mitra et al. 2004) as well as in the periplasmic space of C. reinhardtii (Fujiwara et al. 1990, Giordano et al. 2005). As all pyrenoid containing algae have been found to have CCMs, it is assumed that UWO241 also has a CCM (Giordano et al. 2005, Raven and Beardall 2003). Furthermore, the transcriptome of UWO241 contains homologues of carbonic anhydrases and homologues of two types of bicarbonate transport systems (Morgan-Kiss, Kiss & Raymond, in prep). Further investigation is required to determine if UWO241 possesses a CCM and if the CCM is functional in situ, as environmental conditions including CO2 concentration, pH, temperature, and salinity can affect CCM activity. Additionally, irradiance can affect CCM activity as carbon transport requires ATP generated by cyclic PSI electron transport (Giordano et al. 2005, Palmqvist et al. 1990, Spalding et al. 1984).

1.3.5. Regulation of carbon fixation: RubisCO. Early reports identified RubisCO, fructose-1,6- bisphosphatase (FBPase), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and (PRKase) as having regulatory roles in the CBB cycle (Kelly et al. 1976). More recently, antisense plant experiments in which CBB cycle enzymes were individually targeted and knocked-down allowed for the effect of an individual enzyme on carbon fixation cycle flux to be determined. These experiments have challenged the knowledge of enzyme regulation of photosynthetic carbon fixation. The dominate regulatory enzymes were determined to be sedoheptulose-1,7-bisphosphatase (SBPase) and RubisCO. Two additional CBB cycle enzymes, and aldolase, which catalyze freely reversible reactions were also shown to affect carbon flux. GAPDH, FBPase, and PRKase were not shown to have a regulatory role on carbon fixation in plants (Raines 2003). RubisCO is highly regulated including the requirement of spontaneous carbamylation which is dependent on RubisCO activase (Rca) for removal of inhibitory sugar phosphates which can tightly bind RubisCO active sites resulting in the closed state (i.e., catalytically incompetent)

10

(Portis 2003). Inhibitory sugar phosphates, including 2-carboxy-D-arabinitol 1-phosphate (CA1P), xylulose-1,5-bisphosphate (XuBP), pentadiulose-1,5-bisphosphate (PDBP), and RuBP are naturally occurring and resemble transition state intermediates of catalysis. Inhibitors can bind carbamylated (CA1P) or uncarbamylated (CA1P, XuBP, PDBP, RuBP) active sites, thereby precluding carbamylation and/or substrate binding (Parry et al. 2008). The structure of tobacco activase was recently determined and functions as a hexamer. Release of inhibitors is suggested to occur by RubisCO activase encircling RubisCO and partial threading of RubisCO segments with activase pore loops that are required for ATPase activity and subsequent removal of RubisCO inhibitors, but further mechanistic studies are required (Stotz 2011). RubisCO activase is nuclear encoded and exists in two isoforms (41-43 and 45-46 kDa) due to alternative splicing in spinach (Werneke et al. 1989), Arabidopsis (Werneke et al. 1989), barley (Rundle and Zielinkskis 1991) and rice (To et al. 1999), however the algae, C. reinhardtii (Roesler and Ogren 1990) and Chlorococcum littorale (Beuf et al. 1999), contain only the smaller isoform (Portis 2003). The larger form of RubisCO activase is very sensitive to chloroplast stroma ADP/ATP ratios and this sensitivity is modulated by oxidation/reduction of its carboxy-terminal domain (sensitivity increased when oxidized) (Zhang and Portis 1999, Portis 2003). For organisms that lack the large activase isoform, the mechanism of regulation by light is unclear. An investigation of tobacco which lacks the large activase isoform suggests mechanisms such as ΔpH, degree of reduction of the acceptor side of PSI, or an unknown mechanism involving the system (Ruuska et al. 2000). RubisCO can be further regulated by the reversible binding of effectors to the active site which affects carbamylation. Positive effectors increase carbamylation under sub-saturating 2+ conditions of CO2 and Mg by increasing the stability of the activated enzyme and include 6- phosphogluconate, NADPH, and inorganic orthophosphate (Pi). Oppositely, negative effectors that bind uncarbamylated active sites include ribose-5-phosphate and fructose 6-phosphate and inhibit carboxylation depending on the concentration of effector and RuBP (Badger and Lorimer 1981, Parry et al. 2008). In addition to limitation in RubisCO activity, photosynthesis can be limited by light harvesting and electron transport to regenerate RuBP, or regeneration of Pi for photophosphorylation via sucrose and starch synthesis (Sage et al. 2008, Sharkey 1985). The

11

CBB cycle plays a central role in regulation of photosynthesis and is impacted by environmental parameters such as light and temperature.

1.4. Photosynthetic microbial eukaryote diversity and the impacts of climate change in the dry valley lakes

UWO241 is a member of a diverse community of microbial eukaryotes residing within Lake Bonney. A recent paper reported on the phylogenetic diversity and distribution of microbial eukaryotes residing in both basins of Lake Bonney using 18S rRNA sequencing libraries coupled with real time quantitative PCR (qPCR) (Bielewicz et al. 2011). Both lobes are dominated by photosynthetic including a cryptophyte species related to Geminigera cryophila, a (Isochrysis sp.) and a stramenopile (Nannochloropsis sp.), which represent the major primary producers in this closed-basin aquatic ecosystem. Analyses of environmental gene expression using the major gene of RubisCO (rbcL) as an indicator of carbon fixation support the finding that these organisms are the major primary producers (Kong et al. 2012). The deepest photic zone (18 m) harbors a variety of chlorophyte species including UWO241 (Bielewicz et al. 2011). This alga has been reported in a number of Antarctic ponds and lakes (Koob and Leister 1972), while other Chlamydomonas sp. were found to exist in a eutrophic Antarctic lake which was high in inorganic phosphorus and nitrogen (Mataloni et al. 1998). The capacity of UWO241 to adapt its photochemistry and energetics to low light and temperature, in addition to recent findings of physiological plasticity in response to high temperature (Possmayer et al. 2011) and salinity (Pocock et al. 2011) likely contribute to this organism’s widespread distribution in Antarctic aquatic systems. As the regulation of photosynthesis is complex and related to a number of environmental factors, it is likely that lake photosynthetic communities will be affected by climate change. Perturbation of the primary producer communities will directly affect functioning of the MDV lake food chain as a whole. It is therefore critical to understand the impact of climate change on the photoautotrophic communities in these low temperature ecosystems. Climate change predictions include warming of the Antarctic continent in the next 50-100 years (Chapman and Walsh 2007, Walsh 2009), which will impact the MDV ecosystems by increased hydrologic activity through increased occurrence in pulse flood events and glacial melting.

12

Episodic events causing short duration variations in major abiotic drivers are now recognized to have profound effects on aquatic ecosystem functioning, and are expected to intensify at the level of frequency and magnitude (Alley et al. 2003, Meehl et al. 2007, Jentsch et al. 2007). Current climate change models predict warmer, wetter summers in the dry valleys that will result in higher stream flow, larger perimeter moats and thinner ice covers resulting in increased nutrients and dissolved inorganic carbon into the water column (Fig. 1.4). Under current climatic conditions, interactions between the streams and water column are usually restricted to the upper 1-5 m of the water column (Fig. 1.4 a); however, as the moat size increases we predict that the mixing layer will extend deeper into the lake, causing a displacement of the permanent chemocline (Fig. 1.4 b). The long-term impact of increased flood events on the dry valley lake primary producers is currently unknown, but at a minimum, thinner ice covers in combination with alleviation of nutrient deprivation could stimulate photosynthesis and may favor “aquatic weeds” such as Chlamydomonas sp. that exhibit acclimatory plasticity over a range of environmental stresses.

1.5. Dissertation objectives

The overall goal of this project is to characterize the effects of environmental factors (specifically, temperature and light) on RubisCO in photopsychrophiles of the unique low temperature MDV lake ecosystems using an integrated approach of directed laboratory investigation on algal isolates with field studies on natural lake samples. This combined approach will provide a novel view of how major environmental drivers influence inorganic carbon fixation in low temperature ecosystems as well as contribute to a more thorough understanding of how aquatic ecosystems will respond to environmental changes. To this end, the following goals related to algal isolate physiology and lake communities will be addressed:

Algal isolate physiology: (1) Determine the sequences of RubisCO large and small subunits and regulatory enzyme, RubisCO activase, of psychrophilic and mesophilic green algae, UWO241 and SAG49.72, respectively. (2) Compare thermal properties of purified UWO241 and SAG49.72 RubisCO. (3) Compare thermal properties of RubisCO in UWO241 and several other psychrophilic green algal species.

13

(4) Analyze thermal properties of whole cell carbon fixation between psychrophilic and mesophilic algae. (5) Compare the effects of growth irradiance and temperature on RubisCO activity and abundance for UWO241 vs. SAG49.72. MDV lake communities: (6) Characterize the diversity and abundance of autotrophic communities residing in several dry valley lakes. (7) Optimize the RubisCO activity assay for measuring carbon fixation potential in natural aquatic samples. (8) Identify major environmental drivers for carbon fixation potential in natural autotrophic communities residing in several dry valley lakes.

14

Table 1.1. Impact of low temperatures on photostasis in Chlorella vulgaris and C. raudensis UWO241. Low light (LL): 20 µmol photons m-2 s-1, high light (HL): 150 µmol photons m-2 s-1. (modified from Morgan-Kiss and Dolhi 2011).

Chlorella Chlamydomonas Organism vulgaris raudensis UW0241

Irradiance level LL HL LL HL

Redox state (1 - qP) 0.20 0.60 0.10 0.35

-1 Growth rate (day ) 0.24 0.24 0.24 0.48

15

Figure 1.1. Site of isolation of C. raudensis UWO241, Lake Bonney in Taylor Valley, Antarctica (A). General physicochemical characteristics of the east lobe of Lake Bonney. C. raudensis was isolated from a lake sample collected at 17 m. Hatched square shows permanent ice-cover (B). PAR, photosynthetically available radiation; DIN, dissolved inorganic nitrogen. Nutrient data kindly provided by McMurdo Long Term Ecological Research Program (www.mcmlter.org).

16

Figure 1.2. Photosynthetic electron transport chain in green algae and plants. Light energy absorption occurs via light harvesting complexes/antenna and is transferred to chlorophyll reaction centers in PSII and PSI. Oxidized PSII reaction centers drives electron separation from a water molecule by the oxygen evolving complex. Electrons pass through PSII to the mobile electron carrier plastoquinone which delivers electrons to Cyt b6f. The lumenal protein plastocyanin carries electrons to oxidized PSI reaction centers where NADP+ is reduced to NADPH via Ferredoxin-NADP+ reductase. Lumenal protons are utilized by chloroplastic ATP synthase to generate ATP. The dotted line shows the alternative cyclic electron transport pathway. The redox state of the PQ pool is an early sensor of imbalances between turnover rates of PSII and PSI (grey box). LHC; light harvesting complex, OEC; oxygen evolution complex, PS; photosystem, PQ; plastoquinone, Cyt; cytochrome, PC; plastocyanin, ATPase; ATP synthase.

17

Figure 1.3. Simplified CBB cycle showing carboxylation (black) and oxygenation (gray) reactions catalyzed by the RubisCO enzyme and downstream fates of the products generated by each reaction. Products of carboxylation (3PGA; 3-phosphoglycerate) are incorporated into cell biomass via sugars or contribute to regeneration of substrate (RuBP; ribulose 1,5-bisphosphate). Products of oxygenation include 3PGA and 2-phosphoglycerate (2PG). 2PG is recycled back to the CBB cycle via photorespiration. This process is counter-productive in that CO2 is generated. 1: RubisCO, 2: 2PG , 3: glycolate , 4: glutamine-glyoxylate aminotransferase, 5: decarboxylase, 6: serine-glyoxylate aminotransferase, 7: hydroxypyruvate reductase. Adapted from Peterhansel and Maurino 2011.

18

Figure 1.4. Predicted climate effects in Antarctic lakes. Dry valley lakes are permanently ice- capped for most of the year, with the exception of narrow open water moats which form around the perimeter of the lake during the short austral summer (mid-November to late-February). The moat allows input from glacial-fed streams into the lake (dotted lines) and limited mixing occurs between the moat and layers of the water column directly under the ice. Interactions between the water column and the atmosphere are minimal (double arrows) (A). As the dry valleys slowly warm, it is predicted that seasonal “pulse” events will lead to warmer, wetter summers, resulting in more stream input, larger moats and reduced ice covers (B). Images show examples of a typical summer moat (A) and an unusual large moat on the east lobe of Lake Bonney during the summer of 2010 (B). Larger moats are anticipated to be a more common occurrence as the dry valleys slowly warm.

19

CHAPTER 2

Functional characterization of RubisCO from psychrophilic and mesophilic green algal isolates

Jenna Dolhi, Rachael Morgan-Kiss

Author contributions: JD developed methods, performed data collection and analyses, and wrote the manuscript

20

CHAPTER 2

2.1. INTRODUCTION

Microalgae play essential roles in aquatic ecosystems as primary producers by fixing inorganic carbon through the process of photosynthesis. Photosynthesis is the absorption of light energy via pigment-protein complexes (light harvesting complexes; LHC) and conversion of this energy to chemical energy and reductant (ATP and NADPH, respectively) during photosynthetic electron transport. ATP and NADPH are utilized in the enzyme-driven process of carbon fixation, the Calvin-Benson-Bassham (CBB) cycle. The balance of photochemical processes with carbon processes (termed “photostasis”) poses a challenge for low temperature adapted photosynthetic organisms primarily due to reduced enzyme reaction rates. This balance of source energy with sink energy is represented by the equation: -1 σPSII x Ek = τ where σPSII is the effective absorption cross-section of PSII, Ek is the irradiance at which the maximum quantum photosynthetic yield balances photosynthetic capacity and τ-1 is the rate of consumption of photosynthetic electrons. Not only can an imbalance in photostasis occur at low temperatures, but high light conditions can cause an imbalance due to excess energy input during light absorption and photosynthetic electron transport. Therefore, both of these conditions can result in excessive PSII excitation energy, or high excitation pressure (HEP) which can lead to photoinhibition. The mechanisms that photoautotrophs use to decrease HEP and attain photostasis vary between species (Hüner et al. 1998, Ensminger et al. 2006). Strategies to dissipate HEP and maintain photostasis include modifications to energy production processes such as a decrease in the functional absorption cross-section of PSII. This can be achieved through a decrease in the size or abundance of light harvesting antenna via state transitions, reduction in transcription and translation of Lhcb which encode LHCII polypeptides, or alteration of light harvesting pigments to energy quenching pigments (i.e., xanthophyll cycle). The xanthophyll cycle and PSII reaction center quenching are mechanisms of non-photochemical quenching (NPQ) or thermal dissipation of energy, and are important in avoidance of photoinhibition. Last, photosystem stoichiometry may be modulated to control rates of electron flow through the reaction centers by activation/repression of genes encoding PSI and PSII reaction center proteins. Photosynthetic organisms must be able to sense an energy imbalance in

21 order to alter gene expression to attain photostasis (Ensminger et al. 2006). The plastoquinone

(PQ) pool is an important redox sensor and occurs upstream of cytochrome b6f, the rate limiting step in the photosynthetic electron transport chain (Hüner et al. 2012). The redox state of the PQ pool is indicative of excitation pressure and is measured using a pulse amplitude-modulated (PAM) fluorometer; a reduced PQ pool indicates HEP (Ensminger et al. 2006). Other potentially important redox sensors include PSII, thylakoid proton motive force, ferredoxin, thioredoxin and on the acceptor-side of PSI. Last, reactive oxygen species (ROS) may play a role in sensing excitation pressure. A combination of these sensors is likely to contribute to signaling and modulation of gene transcription and expression in the cell chloroplast and nucleus (Hüner et al. 2012, 2013). Alternative to modifying energy source processes, an imbalance in energy poise can be dealt with by up-regulation of energy sink processes such as C, N, and S assimilation, and ultimately through cellular growth and maintenance (Ensminger et al. 2006). Cold adapted enzyme activity and/or enhanced enzyme synthesis are required for efficient metabolic rates at low temperature. Cold adapted enzymes have an increased kcat (turnover rate) at low temperatures which can be achieved by a decrease in ΔH (enthalpy; i.e., lower occurrence of enthalpy related structural interactions that must be broken during transition state formation) (Siddiqui et al. 2013, Siddiqui and Cavicchioli 2006, Papaleo et al. 2011, D'amico et al. 2006). Such structural modifications can result in increased flexibility of cold adapted enzymes which is correlated to decreased stability. Thermolability of cold adapted enzymes relative to their homologues from mesophilic organisms is a hallmark characteristic of such enzymes. Cold adapted proteins can harbor any combination of structural modifications to gain flexibility, including: increased surface hydrophobicity and hydrophilicity, decreased core hydrophobicity, decreased arginine/ ratio, increased glycine residues, weaker interdomain and intersubunit interactions, more and longer loops with fewer prolines, increased prolines in α-helices, decreased metal binding sites, decreased disulfide bridges and electrostatic interactions. An increase in kcat can also be achieved through larger and more easily accessible active sites for substrate binding (Siddiqui et al. 2013, Siddiqui and Cavicchioli 2006). When grown under low temperature-induced HEP conditions (5 °C/150 µmol photons m- 2 s-1), the mesophilic green alga, Chlorella vulgaris, appears to be limited at the ability to modulate rates of carbon fixation to ameliorate HEP and maintain photostasis (Hüner et al. 1998,

22

Morgan-Kiss et al. 2006, Maxwell et al. 1994, 1995). Therefore, at low temperature-induced HEP, C. vulgaris must rely on maintaining energy balance by downregulating light absorption through the replacement of light harvesting pigments with energy-dissipating carotenoids as well as reducing the size of the major light harvesting antenna. In stark contrast with C. vulgaris, low temperature grown cultures of the psychrophile, C. raudensis UWO241, do not significantly adjust pigmentation or light harvesting capacity, but instead exhibit an increase in growth rate in response to an increase in irradiance levels (Morgan-Kiss et al. 2006). These early studies provided the first evidence that UWO241 possesses alternative mechanisms for maintaining photostasis under low temperature-induced HEP compared with mesophilic algae (Morgan-Kiss et al. 2006). One potential mechanism to adjust to low temperatures is an ability to upregulate fixation of inorganic carbon, which would be similar to acclimation strategies in cold tolerant crop plants such as winter rye and wheat (Savitch et al. 2002, Dahal et al. 2012). A likely bottleneck in carbon fixation at low temperatures is the enzyme ribulose-1, 5-bisphosphate carboxylase oxygenase (RubisCO). Therefore, this current study investigates whether UWO241 exhibits altered carbon fixation ability at low temperature and focuses on the rate-limiting enzyme of the CBB cycle, RubisCO, including experiments focused on activity at a range of temperatures, thermolability of the enzyme, and the protein sequences of the large and small subunits (encoded by rbcL and rbcS). RubisCO is the most abundant enzyme on earth and can be found in organisms from all domains of life (Tabita et al. 2008). It catalyzes a carboxylation reaction combining the substrate ribulose-1, 5-bisphosphate (RuBP) with inorganic CO2 to generate 3-phosphoglycerate (PGA) which is further reduced to sugars and incorporated into biomass. This enzyme also catalyzes a counter-productive oxygenation reaction which results in CO2 generation in a process called photorespiration. Regulation of RubisCO activity is complex: prior to catalyzing these reactions,

RubisCO active sites must be cleared of inhibitory sugar molecules and bind non-substrate CO2 2+ 2+ as well as Mg . Active sites having CO2 and Mg bound are referred to as carbamylated and are catalytically competent. The uncoupling of inhibitory sugar molecules is carried out by the enzyme, RubisCO activase (Rca) (Spreitzer and Salvucci 2002). In this study, RubisCO was investigated as a potential cold adapted enzyme in UWO241 which would support the finding of enhanced growth rate in a low temperature and high light environment compared to C. vulgaris. This enzyme was targeted due to its important role in the CBB cycle and high abundance.

23

RubisCO activity and carbon fixation potential were predicted to be active at lower temperatures for UWO241 compared with that of the closely related mesophile, C. raudensis SAG49.72. Additionally, UWO241 RubisCO was predicted to have increased thermolability than that of SAG49.72. In order to extend our findings to other psychrophilic organisms, enzyme activity in crude lysate was also investigated in other algal isolates that are from different low temperature ecosystems, including Chlamydomonas sp. ARC, Chlamydomonas sp. CCMP 681, and Chlorella sp. BI.

2.2. MATERIALS AND METHODS

2.2.1. Growth conditions. Axenic cultures of Chlamydomonas raudensis UWO241 and C. raudensis SAG49.72 were grown in 250 mL Pyrex tubes submerged in temperature regulated aquaria (Morgan et al. 1998). UWO241 was grown in Bold’s Basal Medium (BBM) supplemented with 0.7 M NaCl and SAG49.72 was grown in standard BBM (Nichols and Bold 1965). UWO241 was grown at 8 °C and 20 µmol photons m-2 s-1, which is comparable with the natural light and temperature conditions in its natural environment (Neale and Priscu 1995). The mesophilic strain was grown at 29 °C and moderate growth irradiance, 110 µmol photons m-2 s-1. Growth irradiance was generated by fluorescent tubes (Sylvania CW40) and measured with a quantum sensor attached to a Li- 250A radiometer (Li-Cor, Lincoln, NE). Growth kinetics were measured as the change in optical density at 750 nm. All experiments were performed on samples derived from exponentially growing cultures.

2.2.2. rbcL and rbcS sequencing. Primers were designed in Primer3 (Rozen et al. 2000) against the large and small RubisCO subunits (rbcL and rbcS, respectively). rbcL primers were designed using UWO241 (Gudynaite-Savitch et al. 2006) as template, whereas rbcS primers were designed against Chlamydomonas reinhardtii cDNA. PCR amplifications were performed on a Mastercycler ep (Eppendorf AG, Hamburg, Germany) thermocycler. A fragment of SAG49.72 rbcL (1,334 bp) was PCR amplified with forward primer, rbcL UWO F (5’ ATG GTC CCT CAA ACA 3’), and reverse primer, rbcL UWO R (5’ CGG TGG TGG TAC TTT AGG 3’). The majority of UWO241 rbcL (1, 368 bp) was PCR amplified using the same forward primer and rbcL SAG R (5’ ATT ACA TCT CCA CCT TCA CGA 3’) as the reverse primer. Taq DNA was used as supplied in the GoTaq Green Master Mix (Promega Corporation, Madison, WI) to amplify rbcL in a three-step PCR program (one cycle at

24

95 °C for five minutes, 30 cycles of denaturing: 95 °C for one minute, annealing: 44 °C for one minute, and elongation: 72 °C for one minute, and a final cycle at 72 °C for five minutes). The PCR products were purified using a Wizard SV gel and PCR clean-up system (Promega Corporation, Madison, WI). The resultant PCR product was ligated into a pGEM vector (pGEM- T Easy vector systems, Promega) and transformed into chemically competent E. coli DH5α. The cells were plated on Luria Burtani plates supplemented with Ampicillin (100 µg/mL) for selection of resistant transformants. Recombinant E. coli colonies were identified by blue/white screening and confirmed by PCR using pUC/M13 sequencing forward primer (5’ GTT TTC CCA GTC ACG AC 3’) and reverse primer (5’ CAG GAA ACA GCT ATG AC 3’). Colony PCR products were run on 1.2% agarose gels and DNA bands matching the size of the target gene were cut out and gel purified. The resulting DNA concentration was measured on a Nanodrop1000 (Thermo Scientific, West Palm Beach, FL) and sequenced in both directions by Sanger sequencing using the pUC/M13 sequencing primers. Sequences were translated and alignments containing closely related Rbcl sequences chosen from results of NCBI BLAST tblastn were generated in CLC Main Workbench 6. Last, neighbor-joining trees were generated in MEGA 4.0. C. raudensis SAG49.72 rbcL was submitted to NCBI GenBank (Accession no. JF439459). A fragment of UWO241 rbcS (376 bp) was PCR amplified from cDNA using forward primer UWO-rbcS-F1 (5’ ATG ATG GTC TGG ACC CCG GTC 3’) and reverse primer UWO- rbcS-R1 (5’ GCC CTG GAC CAG GAA GCC CAT 3’). A fragment of SAG49.72 rbcS (256 bp) was PCR amplified from cDNA using the forward primer UWO-rbcS-F3 (5’ TGC CTG GAG TTC GCT GAG GCC 3’) and the UWO-rbcS-R1 primer. rbcS genes were amplified using Taq DNA polymerase as above and a three-step PCR program (one cycle at 95 °C for five minutes, 30 cycles of denaturing: 95 °C for one minute, annealing: 59 °C for one minute, and elongation: 72 °C for one minute, and a final cycle at 72 °C for five minutes). PCR clean up, cloning and sequencing of rbcS genes were performed as described above.

2.2.3. RubisCO enzyme extraction and purification. UWO241 and SAG49.72 were grown in 3 L glass jars in a temperature regulated aquaria at 8 °C/20 µmol photons m-2 s-1 and 29 °C/110 µmol photons m-2 s-1, respectively. Mid-log phase cultures were pelleted by centrifugation at 3,000 x g for 10 min at 4 °C. Cell pellets were flash frozen in liquid nitrogen and stored at -80 °C for no longer than two months before use.

25

Cell pellets were thawed on ice and resuspended in 10 mL extraction buffer (50 mM bicine, 15 mM MgCl2, 1 mM EDTA, 10 mM NaHCO3, 10 mM DTT, 3.3 mM amino-caproic acid, and 0.7 mM benzamidine, pH 8.0; the last four chemicals were added fresh) per gram of pellet (Taylor et al. 2001). The resuspended cells were passed through a chilled French press two times at 10,000 lb/in2 to lyse the cells. Lysis of cells was checked microscopically. Cell lysate was centrifuged at 23,700 x g for 30 minutes at 4 °C. The supernatant was removed and put into a clean centrifuge tube and the centrifugation was repeated a second time to remove all insoluble cell material. Soluble cell lysates were transferred to a new tube and glycerol was added to a final concentration of 12.5%. The lysates were flash frozen in liquid nitrogen and stored at -80 °C for one week before RubisCO was partially purified by anion exchange column (AEC) chromatography at 4 °C. Crude soluble lysate obtained from both algal species was loaded onto a Macro-Prep High Q AEC (Bio-Rad, Hercules, CA) following equilibration of the column with extraction buffer. RubisCO was eluted from the column with an 80% salt gradient provided by the extraction buffer plus 1 mM NaCl at a flow rate of 5 mL min-1 (Taylor et al. 2001). Fractions containing protein were identified by absorbance at 280 nm. RubisCO carboxylase activity was measured as described in the next section to identify which fractions contained RubisCO. Protein concentration was measured in the RubisCO containing lysate by Bradford assay (Bradford 1976) using BSA as a standard. The lysate was electrophoretically separated by SDS-PAGE on a 15% (w/v) polyacrylamide resolving gel and 8% (w/v) polyacrylamide stacking gel using a Mini-Protean II apparatus (Bio-Rad, Hercules, CA) to observe the degree of purification. Gels were stained with 0.05% Commassie Brilliant Blue R. Glycerol (12.5%) was added to RubisCO containing lysate and was flash frozen in liquid nitrogen and stored at -80 °C.

2.2.4. RubisCO carboxylase activity assays. Optimal temperature for maximum potential RubisCO carboxylase activity was determined by measuring carboxylase activity at a range of temperatures (1-60 °C). Partially purified RubisCO (5-10 µg protein) was aliquoted into assay tubes on ice and brought up to 125 µL with assay buffer (50 mM Bicine-NaOH, pH 8.0). Assay tubes were equilibrated to 25 °C for three minutes in a heat block then maximally activated with 14 the same buffer spiked with NaH CO3 (ViTrax, Placentia, CA) (final NaHCO3 concentration in -1 reaction: 20 mM at specific activity: 0.1 µCi mmol ) and MgCl2 (final concentration 10 mM) for

26

5 minutes at 25 °C. After activation, RubisCO activity was initiated by addition of substrate, 0.8 mM RuBP (Sigma-Aldrich, St. Louis, MO). The reaction proceeded at the desired assay temperature for five minutes. The reaction was stopped by addition of 100% propionic acid (Fisher, Pittsburgh, PA) and centrifuged at 2,000 x g for 1.5 hours to exhaust unincorporated 14C. The remaining volume was added to 3 mL Bio-Safe II scintillation counting cocktail (Research Products International Corporation, Mount Prospect, IL) (R. Tabita and B. Witte, personal communication). A multipurpose scintillation counter LS6500 (Bekcman Coulter, Miami, FL) was used to determine the cpm of acid stable reaction end products. Thermolability of RubisCO carboxylase activity was determined using a variation of the assay above which measures residual activity after high temperature incubation. Following activation, the reaction was incubated at various temperatures for 10 minutes before RubisCO activity was initiated and measured at 25 °C. Alternatively, thermolability was tested by incubating the reaction at 50 °C for varying amounts of time. The reaction was initiated with RuBP and incubated for five minutes at 25 °C. Acid stable end products were measured as above.

2.2.5. RubisCO carboxylase activity assay of crude lysates. Three additional psychrophilic green algae, Chlamydomonas sp. ARC, Chlamydomonas sp. CCMP 681, and Chlorella sp. BI, were grown under the same light and temperature conditions as UWO241 (8 °C and 20 µmol photons m-2 s-1). SAG49.72 was grown under the same light and temperature conditions as above. Cultures (10 mL) were harvested during exponential growth and pelleted by centrifugation at 3,000 x g for 10 minutes at 4 °C. The pellets were resuspended in 1 mL extraction buffer. Cells were disrupted using a Minibead beater (BioSpec Products, Bartlesville, OK) three times for 30 seconds on speed setting 48. To prevent heating of the sample, beadbeating was alternated with 1 minute ice incubations. The supernatant was transferred to a 1.5 mL centrifuge tube and centrifuged at 15,000 x g for 2 minutes at 4 °C to remove insoluble material. The crude soluble lysate (5-10 µg protein) was used to measure RubisCO carboxylase activity. The optimal temperature carboxylase assay was performed as described above and the reaction temperatures included 25, 40 and 55 °C.

2.2.6. Carbon fixation by whole cells. To determine if other processes involved in gross carbon fixation were temperature dependent, whole cell carbon fixation was measured in psychrophilic

27 and mesophilic algae. Exponential growth phase cultures were collected and diluted with fresh media to approximately 4x105 cells mL-1. Diluted cultures (4.5 mL) were pre-incubated in clear tubes with clear caps (VWR, Radnor, PA) for 15 minutes at the respective assay temperature in a temperature gradient block with an irradiance of 50 µmol photons m-2 s-1. Unlabeled stock 14 NaHCO3 (50 mM) in 50 mM bicine buffer (pH 8.0) was spiked with NaH CO3 (ViTrax, Placentia, CA, specific activity: 0.5 µCi mmol-1) and 0.5 mL was added to each culture (Morris and Glover 1974). Foil-covered tubes were run alongside samples for negative controls. The carbon fixation assay was carried out for 2 hr after which cultures were vacuum filtered (7 in Hg) onto 25 mm GF/F filters. Filters were placed in scintillation vials and 0.5 mL of 3N HCl was added. Filters were dried on a hot plate at 60 °C for 5 hrs. Scintillation counting cocktail was added and acid stable end products were measured as above. Radioactive 14C counts in the negative control were subtracted from cpm of the samples.

2.2.7. Chlorophyll fluorescence measurements. Chlorophyll a (chl a) fluorescence parameters are tools for estimating the in vivo activity of the photochemical apparatus and photosynthetic electron transport. Photochemical performance was monitored alongside the whole cell carbon fixation measurements using a pulse amplitude modulated chlorophyll fluorescence detection system, Dual-PAM-100 (Walz, Effeltrich, Germany). A 2 mL sample of exponentially growing culture was dark adapted and incubated with 4 mM NaHCO3 immediately prior to chl a fluorescence measurements. Fluorescence parameters generated from PAM measurements include FV/FM (or FM-FO/FM), or theoretical maximum of functional PSII reaction centers, where

FO is the measure of chl a fluorescence of oxidized PSII reaction centers (dark adapted) and FM is the maximal chl a fluorescence resulting after a saturating light pulse (equivalent to the culture growth irradiance) is applied to the dark adapted cells (Morgan-Kiss et al. 2005). Additional parameters generated from chl a fluorescence included ɸPSII; quantum yield of PSII electron transport, qL; PSII reduction state, qN regulated non-photochemical dissipation, and ɸNO non- regulated energy dissipation (Szyszka et al. 2007). ɸPSII reflects more closely the true functional proportion of PSII reaction centers than does FV/FM. The quenching parameters qL, qN and ɸNO (photochemical and non-photochemical quenching, respectively) provide estimates of the proportion of energy which is used for photochemical reactions vs. lost as heat dissipation. qL provides an estimate of the PQ pool redox state; qL >0.5 indicates an oxidized PQ pool, whereas qL <0.5 indicates a reduced PQ pool. An over-reduced PQ pool is potentially dangerous as it may

28 result in photoinhibition. To aide in protection against this situation, non-photochemical quenching or energy dissipation processes occur. These include quenching by xanthophyll cycle pigments (measured by qN) and photorespiration and reaction center quenching (measured by ɸNO).

2.3. RESULTS

2.3.1. RubisCO subunit sequences. As cold adapted enzymes are known to harbor unique sequence and structural modifications compared to mesophilic homologues, the large (rbcL) and small (rbcS) subunit nucleotide sequences were determined and translated to protein sequences for investigation of sequence differences. Partial protein sequences were generated for UWO (456 residues, 96% complete) and SAG (445 residues, 94% complete) RbcL, as well as for UWO (163 residues, 87% complete) and SAG (143 residues, 76% complete) RbcS. The RbcL neighbor joining phylogenetic tree showed that RbcL from UWO241 was most closely related to that of Chlamydomonas parkeae (Accession no. BAE92572), while SAG49.72 and Chlamydomonas moewusii (Accession no. ABL61336), were closely related as indicated by a high bootstrap percentage. However, the neighbor joining phylogenetic tree based on RbcS showed that UWO241 was more closely related to other psychrophilic algae: Chloromonas sp. ANT3 and Chlamydomonas sp. HS-5, while RbcS from SAG49.72 was more closely related to the mesophilic alga, C. moewusii (Fig. 2.1).

2.3.2. RubisCO purification. RubisCO was partially purified from UWO241 and SAG49.72 using AEC chromatography to use in measurements of carboxylase activity. RubisCO was purified 5.03-fold and 2.63-fold for UWO241 and SAG49.72, respectively (Table 2.1). The protein fractions were analyzed for purity by SDS-PAGE. Fractions which were enriched in RbcL (55 kDa) and RbcS (14 kDa) were used in carboxylase activity assays (fractions 35-38 for UWO241 and 37-40 for SAG49.72) (Fig. 2.2).

2.3.3. Temperature response of partially purified RubisCO carboxylase activity. RubisCO carboxylase activity of the partially purified RubisCO fractions was measured over a range of incubation temperatures using a radioactive (14C) assay. In general, RubisCO activity for both organisms was comparable with previously reported values from eukaryotic algae (Read and Tabita 1994, Devos et al. 1998). RubisCO activities of UWO241 and SAG49.72 increased with

29 increasing temperature and were similar at all temperatures measured (0-60 °C). The temperature optimum for RubisCO activity ranged between 50-60 °C and was similar for UWO241 and SAG49.72 (Fig 2.3), suggesting that UWO241 does not possess a cold adapted RubisCO. Thermolability in response to incubation at high temperatures is a typical characteristic of cold adapted enzymes (D’amico et al. 2006, Siddiqui and Cavicchioli 2006). We tested RubisCO enzyme thermolability at the level of incubation time (Fig. 2.4) and temperature (Fig. 2.5) for partially purified enzymes in both the mesophilic and psychrophilic algae using variations of the 14C radioactivity assay which measured thermolability via residual enzyme activity. Residual RubisCO activity was measured at 25 °C following incubation at high temperatures. Both algal RubisCOs showed a decrease in activity with increasing exposure time at 50 °C (Fig. 2.4). The average decrease in activity from 10 to 30 mins was 3-fold greater for UWO241 vs. SAG49.72 (- 0.033 and -0.011 relative % RubisCO activity min-1, respectively); however, the decrease from 30 to 60 mins was similar between the two organisms. In contrast, over the time course of 60 to 120 mins, the average decrease in activity was 1.4-fold greater for SAG49.72 than UWO241 (- 0.015 and -0.021 relative % RubisCO activity min-1, respectively). Residual RubisCO activity also decreased in partially purified RubisCO fractions when exposed to increasing temperatures (25-60 °C) for 10 minutes. The greatest loss of enzyme activity in both algal species occurred following incubation at 60 °C relative to RubisCO incubated at 25 °C. RubisCOs from UWO241 and SAG49.72 were similar in their thermolabilities following incubation at the range of temperatures tested: at 50 °C the loss in activities were 29% and 25% for UWO241 and SAG49.72, respectively (Fig. 2.5).

2.3.4. Temperature response of RubisCO carboxylase activity of crude lysates. To determine whether the trends in temperature-dependence of RubisCO activity observed in UWO241 were common across other psychrophilic algal species, carboxylase activity in crude lysates was measured in the following psychrophilic algae: Chlamydomonas sp. ARC, Chlamydomonas sp. CCMP 681, and Chlorella sp. BI. RubisCO activity was measured in crude soluble lysates at three incubation temperatures (25, 40 and 55 °C) to determine the optimal temperature for carboxylation activity. Maximal RubisCO activity in crude lysates extracted from all psychrophilic algae exhibited an increase in response to a rise in incubation temperature which was comparable with the response observed in the mesophilic alga SAG49.72 (Fig. 2.6). This data supports that RubisCO activity is not cold adapted in all of the psychrophilic algae

30 examined. Interestingly, crude lysates of UWO241 exhibited significantly lower activity at higher incubation temperatures (Fig. 2.6), compared with that of the partially purified enzyme (Fig. 2.3), suggesting the presence of thermally sensitive component(s) in the former which is lost in the latter enzyme preparations.

2.3.5. Temperature response of whole cell carbon fixation and photosynthesis. Our results regarding RubisCO activity from partially purified enzyme fractions versus that of the crude lysates provided some evidence that UWO241 does not possess a cold active RubisCO when it is purified away from other cellular factors. These results were somewhat surprising, considering past reports that UWO241 exhibits the unique ability to modulate its growth rates at low temperature (Hüner et al. 1998, Morgan-Kiss et al. 2006, Maxwell et al. 1994). Moreover, there are clearly some differences in the thermal response of the enzyme in purified vs. crude enzyme preparations (Fig. 2.3 and 2.6). Therefore, we also investigated whether UWO241 exhibits temperature dependent differences in whole cell carbon fixation into biomass. Incorporation of inorganic carbon over a short incubation period (2 hours) likely represents gross photosynthesis (Dring and Jewson 1982, Thomas et al. 1992). In contrast with the RubisCO enzyme data, our results based on whole cell fixation were markedly different between cells of the psychrophilic UWO241 and the mesophilic SAG49.72. First, the psychrophile showed carbon incorporation rates that were significantly higher compared to the mesophile at all incubation temperatures below 25 °C (Fig. 2.7). Carbon incorporation rates were 2.6-fold greater at 1.5-6 °C than 25 °C, indicating that UWO241 exhibited a shift to lower temperatures for optimum inorganic carbon incorporation. UWO241 and SAG49.72 exhibited o comparable rates of CO2-incorporation at a measuring temperature of 25 C, while the psychrophile exhibited CO2-incorporation rates that were near dark fixation controls at the higher incubation temperature of 32 °C (Fig. 2.7). Since carbon fixation rates are dependent on the supply of photochemically-derived energy products, the sensitivity of the photochemical apparatus was also monitored under the same incubation conditions. Thermal sensitivity of four fluorescence parameters were monitored: qL, which estimates the proportion of oxidized PSII reaction centers, as well as a series of three parameters (PSII, quantum yield of PSII photochemical quenching; qN, quantum yield of nonphotochemical energy quenching processes; NO, quantum yield of nonregulated energy quenching processes) which monitor the

31 partitioning of energy associated with PSII. The mesophile, SAG49.72, exhibited minimal changes in all of the measured photochemical parameters regardless of the incubation temperature. On the other hand, UWO241 exhibited thermal sensitivity at the level of both redox poise and energy partitioning (Fig. 2.8A, B). Higher incubation temperatures induced a reduction in the parameter qL (photochemical quenching) by nearly 50% which indicated that mobile electron pools were relatively reduced under higher incubation temperatures (Fig. 2.8A). Over- reduction of the PQ pool in UWO241 samples was accompanied by a reduction in the quantum yield of PSII (ɸPSII) coupled with an increase in energy dissipation mechanisms (ɸNO, ɸqN;

Figure 2.8B). Specifically, at moderate temperatures (26 °C) ɸqN accounted for the majority of the loss in ɸPSII, while at higher temperatures constitutive energy dissipation (i.e., ɸNO) was a major contributor to the reduction in PSII quantum yield.

2.4. DISCUSSION

Photosynthetic organisms that have evolved under permanent low temperature environments possess a suite of adaptive strategies to maintain photostasis and avoid photooxidative damage. In the model Antarctic alga, C. raudensis UWO241, earlier studies showed that this psychrophile exhibits novel mechanisms, including reorganization of its photochemical apparatus as well as highly unsaturated photosynthetic membranes (Morgan-Kiss et al. 1998, 2002b). In addition, its ability to maintain high growth rates at low temperatures indicated that downstream metabolism, including pathways involved in fixation of inorganic carbon, should also be targets for cold adaptation. While the poor catalytic properties of the main enzyme of the CBB pathway made it an attractive target, this comparative study of closely related organisms, UWO241 vs. SAG49.72, involving characterization of partially purified as well as crude lysates, indicates that the carboxylase activity of the psychrophile RubisCO enzyme is not significantly altered to function better at low temperatures. Enzyme activity from crude lysates across three other psychrophilic algae (Chlamydomonas sp. CCMP681, Chlamydomonas sp.ARC, Chlorella sp. BI) supported this result that RubisCO is not a major target for cold adaptation in psychrophilic algae. In contrast with the results based on enzyme activity, UWO241 and SAG49.72 RubisCOs showed differences in subunit sequences between the two organisms. Phylogenetic analysis showed that UWO241 RbcS had greater sequence identity to cold adapted algae than

32 mesophilic algae (Fig. 2.1). However, these differences do not correlate with a cold functioning RubisCO enzyme, as no cold activity was measured in UWO241 RubisCO carboxylase assays. Similar results were found in an investigation of psychrophilic Chloromonas sp. ANT1 and ANT3; RubisCO activity of the psychrophilic algae was comparable to that of the mesophile, C. reinhardtii, despite differences in RubisCO sequences between ANT3 and the mesophile (Devos et al. 1998). Investigation of RubisCO sequence differences between UWO241 and SAG49.72 in relation to holoenzyme structure could suggest whether these differences have any effect on function, as certain regions have been shown to be more likely to impact function (activity and or specificity) (Spreitzer and Salvucci 2002). Another sign that a process has evolved to function more efficiently at lower temperatures is evidence of altered thermolability relative to the same process in a mesophilic species. For example, UWO241 exhibits a high degree of polyunsaturated fatty acids which contribute to extreme thermosensitivity of the photochemical apparatus as well as its ability to evolve oxygen (Morgan-Kiss et al. 2002b). These early studies support the hypothesis that low temperature adaptation of energy generation is a key evolutionary strategy in this psychrophilic alga. While photochemical reactions appear to be highly thermally sensitive, UWO241 RubisCO carboxylase activity exhibited minimal thermal sensitivity in assays with partially purified enzyme preparations (Fig. 2.4, 2.5). However, it should be noted that there was a difference between the thermal responses of carboxylase activity in the psychrophile, UWO241, between crude lysates vs. partially purified samples (Figs. 2.3, 2.6). The correlation of RubisCO activity at high temperatures with thermal stability was also demonstrated in a comparative study of an Antarctic hairgrass and a desert shrub, creosote bush (Salvucci and Crafts-Brandner 2004b). Despite the vast differences in environmental temperatures to which these organisms are adapted, carboxylase activity of purified RubisCO from both organisms exhibited a positive relationship between enzyme activity and incubation temperature (up to the maximum temperature tested, 50 °C). Additionally, RubisCOs from both the Antarctic and temperate plants were thermally inactivated at 40 °C at a comparable rate. However, the temperature for optimal 14 photosynthesis (i.e., uptake of CO2) was 10 °C higher in creosote bush compared with that of the Antarctic hairgrass, indicating that gross photosynthesis in the Antarctic plant was low temperature-shifted relative to the temperate plant. Further investigation of this phenomenon

33 showed that the differences in the temperature optimum were attributable to cold adaptation of the RubisCO regulatory enzyme, RubisCO activase (Salvucci and Crafts-Brandner 2004b). Given the differences between the thermal properties in the crude vs. the partially purified fractions, it is likely that additional factor(s) (e.g., RubisCO activase) may play a role in environmental adaptation of RubisCO in UWO241. A homolog of rca was recovered from a transcriptome of UWO241, indicating that the psychrophile actively expresses RubisCO activase (Morgan-Kiss, Kiss & Raymond, in prep). However, multiple attempts to heterologously express the enzyme in the current study were unsuccessful. As an additional approach to further investigate whether carbon fixation could be cold active in UWO241, gross cellular 14 incorporation of CO2 was investigated. In marked contrast with the thermal trends observed in the lysates, gross photosynthesis rates in UWO241 exhibited both cold activity as well as thermal sensitivity (Fig. 2.7). This could account for the enhanced ability of UWO241 (relative to mesophilic algae) to maintain high growth rates under a range of growth irradiances at low temperature (Morgan-Kiss et al. 2006). This may involve a cold active/thermally sensitive RubisCO activase, alternatively other enzymes in the CBB cycle could be cold active or present in greater abundance. C. vulgaris grown under HEP conditions exhibited restricted carbon metabolism which was attributed to changes in enzyme abundance rather than kinetic rates for the enzymes tested (RubisCO, sucrose phosphate synthase, and fructose-1,6-bisphosphatase), providing evidence that enzyme abundance may also play a role in modulation of carbon fixation (Savitch et al. 1996). The impact of RubisCO enzyme abundance on environmental adaptation is explored in Chapter 3. The results from the gross photosynthesis experiment indicate that: (i) UWO241 possesses significantly higher rates of CO2 incorporation relative to SAG49.72 at low temperatures, and (ii) CO2 incorporation in UWO241 is thermally sensitive to even moderate incubation temperatures (Fig. 2.7). The process of carbon fixation via the CBB cycle is an energy-consuming pathway, which requires a supply of ATP and NADPH for the regeneration of the RubisCO substrate, RuBP. The vast majority of this energy is supplied by light driven photochemical reactions. Past evidence showed that UWO241 possesses a thermally labile photochemical apparatus, with thermal inactivation processes occurring at temperatures above 40 oC (Morgan-Kiss et al. 2002b). Here, we observed a significant loss in photochemical function in the psychrophile at higher incubation temperatures, while the mesophile was unaffected (Fig.

34

2.8). Therefore, the apparent thermal sensitivity of gross photosynthesis in the psychrophile is likely to be at least partially explained by a highly thermally-sensitive photochemical apparatus in UWO241. This study contributes to the efforts to understand potential adaptations of RubisCO and carbon fixation to low temperature by sequencing RubisCO subunits and characterizing the enzyme activity from a unique green alga. We determined that the RubisCO carboxylase activity of the psychrophilic alga, C. raudensis UWO241, does not exhibit cold activity or enhanced thermolability. However, it is likely that UWO241 possesses other adaptations that indirectly affect the activity of RubisCO, including cold adaptation of the energy generating processes, such as enhanced LHC (Morgan et al. 1998), high intracellular ATP (Napolitano and Shain 2004, 2005), and increased abundance of two ATP synthase subunits (Morgan et al. 1998), which are responsible for providing the majority of the energy and reducing equivalents for the CBB cycle. In addition, alterations in the abundance of RubisCO and/or changes in the thermal properties of RubisCO activase may also contribute to the enhanced ability of the psychrophile to sustain high rates of CO2 fixation and growth rates at low temperatures.

35

Table 2.1. Anion exchange column (AEC) chromatography purification of RubisCO from crude lysates (CL) of C. raudensis UWO241 and SAG49.72.

Step Protein Activity (µmol Specfic activity (µmol Yield Fold -1 -1 -1 -1 (mg mL ) CO2 hr ) CO2 hr mg protein ) (%) purification UWO241 4.77 51.48 2.57 100.00 - CL UWO241 0.46 129.74 12.92 9.64 5.03 AEC SAG49.72 7.29 170.14 8.64 100.00 - CL SAG49.72 1.40 227.10 22.73 19.20 2.63 AEC

36

A

B

Figure 2.1. Neighbor-joining phylogenetic trees based on protein sequences of C. raudensis UWO241 and SAG49.72 RbcL (A) and RbcS (B) and nearest neighbors according to results of NCBI BLAST tblastn. Values at node represent bootstrap support (%) observed in 500 replicates. The scale bars represent a difference of 0.005 or 0.05 substitutions per site.

37

A

55

15 10

B

55

15

10

Figure 2.2. SDS-PAGE of anion exchange column (AEC) purified protein fractions from exponentially growing C. raudensis UWO241 (A) and SAG49.72 (B) cultures. A. Lanes from left to right: protein ladder, fraction 6, 33, 34, 35, 36, 37, 38. B. Lanes from left to right: protein ladder, fraction 37, 38, 39, 40, and 41. All lanes contain 4 µg of protein. Standard proteins labeled to show molecular weight of RbcL (55 kDa) and RbcS (14 kDa).

38

)

-1

hr -1 35 SAG49.72 UWO241 30

mg mg protein 25 2

20 mol mol CO

 15

10

5

0 0 10 20 30 40 50 60 RubisCOactivity ( Temperature (C)

Figure 2.3. Activity of AEC purified RubisCO (5-10 µg protein) from the psychrophilic green alga, C. raudensis UWO241, and mesophilic green alga, C. raudensis SAG49.72, measured at various assay temperatures. Error bars represent standard deviation of 3-4 replicates.

39

110 SAG49.72 100 UWO241

90

80

70

60

50

40

30 RubisCOactivity (relative %) 20 0 20 40 60 80 100 120 Time (mins)

Figure 2.4. Thermolability of AEC purified RubisCO (5-10 µg protein) from C. raudensis UWO241 and SAG49.72 following incubation at 50 °C for various times. Residual activity determined by normalizing to specific activity after a 10 min incubation at 50 °C. Error bars represent standard deviation of 3-4 replicates.

40

110 SAG49.72 100 UWO241

90

80

70

60

50

40

30 RubisCOactivity (relative %) 20 20 25 30 35 40 45 50 55 60 65 Temperature (C)

Figure 2.5. Thermolability of AEC purified RubisCO (5-10 µg protein) from C. raudensis UWO241 and SAG49.72 following incubation at various temperatures for 10 min. Residual activity determined by normalizing to specific activity following a 10 min incubation at 25 °C. Error bars represent standard deviation of 3 replicates.

41

) -1

hr 25 -1 -1

C. 681 20 C. ARC C. BI

SAG49.72

mg protein mg 2

15 UWO241 mol CO mol

 10

5

0

25 40 55 RubisCOactivity ( Temperature (C)

Figure 2.6. RubisCO activity in crude lysate (5-10 µg protein) of psychrophilic (Chlamydomonas sp. CCMP 681, Chlamydomonas sp. ARC, Chlorella sp. BI, C. raudensis UWO241) and mesophilic green algae (C. raudensis SAG49.72) measured at various temperatures. C. sp. BI and C. raudensis UWO241 are an average of 2 biological replicates; error bars represent standard deviation of 3-4 replicates.

42

) 0.25

-1 hr

-1 SAG49.72

0.20 UWO241 cell 2 2 * * 0.15 *

0.10

0.05 *

assimilation (pg CO 2

CO 0.00 1.5 6 11 25 32 Temperature (C)

Figure 2.7. Inorganic carbon assimilation during a 2 hr incubation at various temperatures and 50 µmol photons m-2 s-1 for psychrophilic (C. raudensis UWO241) and mesophilic (C. raudensis SAG49.72) green algae. Asterisks show statistically significant differences (p<0.05) as determined by a paired student’s t-test. Error bars represent standard error of 3-4 replicates.

43

PSII NO qN

1.0 A 1.0 B

0.8 0.8

0.6 0.6 qL 0.4 0.4

0.2 0.2

0.0 0.0

C 6 16 26 32 Energy partitioning(rel. proportion) C 6 16 26 32

1.0 C 1.0 D

0.8 0.8

0.6 0.6 qL 0.4 0.4

0.2 0.2

0.0 0.0

C 9 26 32 Energy partitioning(rel. proportion) C 9 26 32 Temperature (C) Temperature (C)

Figure 2.8. Effect of temperature on electron transport efficiency and energy distribution in

UWO241 (A and B) and SAG49.72 (C and D). PSII reduction state represented by qL (A and C). Energy partitioning at PSII upon 1 hr incubation at each temperature and in a control culture (C) without incubation (B and D). Error bars show standard error of 3 and 4 replicates for UWO241 and SAG49.72, respectively. ɸPSII; quantum yield of PSII electron transport, ɸqN yield of regulated non-photochemical dissipation, and ɸNO yield of non-regulated non-photochemical dissipation.

44

CHAPTER 3

The effect of environmental factors (growth irradiance and temperature) on modulation of RubisCO carbamylation state and abundance in psychrophilic and mesophilic green algal isolates

Jenna Dolhi, Melissa Morris, Wei Li, Rachael Morgan-Kiss

Author contributions: JD developed methods, performed data collection and analyses, and wrote the manuscript; MM helped with chlorophyll fluorescence data collection; WL helped with growth rate data collection

45

CHAPTER 3 3.1. INTRODUCTION

RubisCO is a highly regulated enzyme with the potential to catalyze two competing reactions: photorespiration (oxygenation) and carbon fixation (carboxylation) (Jordan and Ogren 1984, Parry et al. 2008). Regulation of RubisCO activity and/or abundance occurs in response to environmental parameters including irradiance, temperature, and CO2 concentration (Sage et al. 2008). RubisCO activity is regulated by the enzyme RubisCO activase (Rca) which modulates inhibitor binding to RubisCO active sites and carbamylation, or the binding of non-substrate CO2 to Lys-201 and stabilization by Mg2+ in active sites (Galmes et al. 2013). RubisCO activase is in the AAA+ ( associated with diverse cellular activities) which act in macromolecular complexes that perform chaperone-like functions (Spreitzer and Salvucci 2002, Portis 2003). In an ATP dependent manner, activase removes inhibitory sugar phosphates from RubisCO active sites, allowing carbamylation (Spreitzer and Salvucci 2002) which is a requirement for RubisCO to catalyze an oxygenase or carboxylase reaction (Portis 2003). Carbamylation state, or activation state, represents the proportion of RubisCO active sites that are catalytically competent and this varies with environmental conditions. Carbamylation state is determined by calculating the ratio of initial activity of RubisCO and maximal activity, whereby extracted lysate is incubated with saturating 2+ concentrations of CO2 and Mg (Sharkey et al. 1986). RubisCO activity can be modulated by a family of inhibitors that are produced in plant and algal cells under different conditions. Sugar phosphate inhibitors bind tightly in RubisCO active sites, precluding carbamylation and therefore catalysis. Inhibitors are naturally occurring and resemble RubisCO reaction transition state intermediates (Parry et al. 2008). Some inhibitors (xylulose-1,5-bisphosphate; XuBP and pentadiulose-1,5-bisphosphate; PDBP) are a result of catalytic misfire during the oxygenase or carboxylase reactions and bind uncarbamylated active sites (Kane et al. 1998, Edmondson et al. 1990, Parry et al. 2008). Alternatively, the inhibitor, 2- carboxyarabinitol 1-phosphate (CA1P) binds carbamylated active sites, is synthesized in greater abundance in dark and low light conditions (Seemann et al. 1985, Seemann et al. 1990, Parry 2008), and is a mechanism for regulating RubisCO activity during the day/night cycle in plants. After CA1P is removed by activase, it is dephosphorylated by CA1P phosphatase rendering it inactive. However, under conditions of low irradiance, the compound can be re-phosphorylated

46 and therefore gain back its ability to bind RubisCO (Parry et al. 2008). RubisCO substrate, ribulose-1,5-bisphosphate (RuBP), can also act as a tight binding inhibitor when it binds uncarbamylated RubisCO (Jordan and Chollet 1983, Parry et al. 2008). This has been observed in low light conditions where stromal Mg2+ concentrations are limiting (Brooks and Portis 1988, Portis 1981). The degree to which RubisCO concentration, carbamylation state, and inhibitor binding are modulated in RubisCO regulation is variable across different photosynthetic organisms (Galmes et al. 2013, Kobza and Seemann 1988, Seemann et al. 1990). In response to increasing irradiance, RubisCO was regulated via carbamylation state and inhibitor binding for Beta vulgaris (sugar beet) and Spinacea oleracea (spinach) (Kobza and Seemann 1988). However, regulation under dark/low light conditions was mainly under the control of binding of the inhibitor, CA1P, in Phaseolus vulgaris (bean) (Seemann et al. 1985, Kobza and Seemann 1988, Seemann et al. 1990, Sharkey et al. 1986). Inhibitor binding by PGA and RuBP to decarbamylated RubisCO active sites was found to occur in the green alga, Scenedesmus ecornis, as an artifact during harvesting of cells under dark conditions (i.e., centrifugation) (Mouget et al. 1993). Temperature can also impact activity and/or abundance of RubisCO. In general, moderate to severe heat stress results in a loss of RubisCO activity, likely as a result of increased inhibitor (i.e., XuBP and PDBP) production as misfire reactions increase in frequency (Galmes et al. 2013, Schrader et al. 2006, Salvucci and Crafts-Brandner 2004a). The carbamylation state of RubisCO from cotton was shown to decrease in vitro and in vivo at 35 and 42 °C (Salvucci and Crafts-Brandner 2004a). This deactivation was attributed to heat sensitivity and loss of RubisCO activase activity under conditions of increased inhibitor production, thus preventing removal of inhibitors from RubisCO active sites (Salvucci and Crafts-Brandner 2004a). Heat sensitivity of activase has been shown in multiple plants including spinach, wheat, cotton, tobacco, and Antarctic hairgrass (Deschampsia antarctica) (Robinson and Portis 1989, Feller et al. 1998, Salvucci et al. 2001, Salvucci and Crafts-Brandner 2004a, b). Alternatively, at low temperatures, increased RubisCO concentration has been observed in order to maintain activity under conditions of low turnover rate (Cavanaugh and Kubien 2013, Galmes et al. 2013). RubisCO content was increased in low temperature grown crops, such as: cucumber, winter rye, tomato, wheat, spinach, and broad bean (Yamori et al. 2009, Dahal et al. 2012), and green algae

47 including: Chlorella vulgaris, Chloromonas sp. ANT1 and ANT3 (Savitch et al. 1996, Devos et al. 1998). Temperature also affects relative rates of photorespiration and carboxylation; generally the former is increased over the latter as temperature increases (Sage et al. 2008, Thomas et al. 1992, Jordan and Ogren 1984, Atkin and Tjoelker 2003). The response of plant RubisCOs to environmental conditions including irradiance, temperature, and CO2 concentration has been extensively studied. However, few studies on the response of algal RubisCO to such conditions exist (Mouget et al. 1993, MacIntyre et al. 1997). As green algae are closely related to plants, it is predicted that their RubisCOs may be regulated in comparable ways as that of plants; including increased inhibitor binding and decreased carbamylation state in response to low irradiance and high temperature, and increased RubisCO abundance at low temperature. In this study we analyzed the response of RubisCO from a unique green alga, Chlamydomonas raudensis UWO241, to a range of growth irradiance and temperature conditions. This alga was isolated from Lake Bonney, a permanently ice-covered lake in the Taylor Valley of the McMurdo Dry Valleys, Antarctica. The lake is characterized by low temperature (0-6 °C) and low light (~5-40 µmol photons m-2 s-1), which are both stressful conditions for photosynthetic organisms (Priscu et al. 1999, Dolhi et al. 2013). Low temperature and light-induced stress occurs as a result of the need for algae to maintain a balance in light energy absorbed by photochemical components and energy expended by enzyme driven metabolic reactions. In order to determine if adaptation to such conditions has affected RubisCO regulation in response to environmental parameters, we compared RubisCO carbamylation state and abundance of C. raudensis UWO241 to that of a closely related mesophilic alga, C. raudensis SAG49.72.

3.2. MATERIALS AND METHODS

3.2.1. Strains and growth conditions. Axenic cultures of C. raudensis UWO241 and C. raudensis SAG49.72 were grown in temperature regulated aquaria as previously described (Morgan et al. 1998). UWO241 was grown in Bold’s Basal Medium (BBM) supplemented with 0.7 M NaCl and SAG49.72 was grown in standard BBM (Nichols and Bold 1965). For growth temperature experiments, cultures of UWO 241 were grown at 2, 8, and 12 °C and light conditions comparable with its natural environment (20 µmol photons m-2 s-1) whereas the mesophilic strain

48 was grown at 12, 20, and 29 °C and moderate growth irradiance (100 µmol photons m-2 s-1). Irradiance was generated by fluorescent tubes (Sylvania CW40) and measured with a quantum sensor attached to a Li- 250A radiometer (Li-Cor, Lincoln, NE). Growth irradiance experiments were carried out at 8, 14, 25, 45, 65, and 110 µmol photons m-2 s-1 for both UWO241 and SAG49.72 at growth temperatures consistent with their natural environments, 8 and 29 °C, respectively. Growth kinetics were monitored as the change in optical density at 750 nm. Experiments were performed on exponentially growing cultures.

3.2.2. Chlorophyll fluorescence measurements. Chlorophyll a (chl a) fluorescence parameters are tools for estimating the in vivo activity of the photochemical apparatus and photosynthetic electron transport. As photochemically-derived energy products are required for the regeneration of RuBP, it is important to consider how growth conditions may modulate photochemical performance and thus functioning of the Calvin cycle. Measurements were generated in vivo for all algal cultures using a pulse amplitude modulated chlorophyll fluorescence detection system, Dual-PAM-100 (Walz, Effeltrich, Germany). The measuring temperature was matched to the growth temperature by use of a water-jacketed cuvette accessory. A 2 mL sample was axenically collected from an exponentially growing culture and was dark adapted for 10 mins and incubated with 4 mM NaHCO3 immediately prior to chl a fluorescence measurements. Fluorescence parameters generated from PAM measurements include FV/FM (or FM-FO/FM), or maximum photochemical efficiency, where FO is the measure of chl a fluorescence of oxidized PSII reaction centers (dark adapted) and FM is the maximal chl a fluorescence resulting after a saturating light pulse is applied to the dark adapted cells. Additional parameters generated from chl a fluorescence include ɸPSII; quantum yield of PSII electron transport, qL; photochemical quenching, and qN; non-photochemical quenching (Morgan-Kiss et al. 2005).

3.2.3. Lysate extraction and protein determination. Exponentially growing cultures were sampled and cell lysate was extracted for the RubisCO carbamylation assay. Samples (10 mL) were pelleted by centrifugation at 3,000 x g for 10 min at 4 °C. Pellets were resuspended in 1.5 mL of carbamylation buffer (100 mM Bicine, 15 mM MgCl2, 5 mM DTT, 3.3 mM amino- caproic acid, 0.7 mM benzamidine, pH 7.8) and disrupted by bead beating (BioSpec Products, Bartlesville, OK) for four cycles of 30 s with 1 min ice incubations between cycles. Cell lysates were centrifuged at 15,000 x g for 2 min at 4 °C and the supernatant was used for RubisCO

49 carbamylation assays. Protein concentration of the lysates was determined according to the Bradford method (1976) following the Bio-Rad Protein Assay microassay protocol with BSA as a standard.

3.2.4. RubisCO carbamylation assay. RubisCO carbamylation state was measured in the soluble cell lysates within 30 min of extraction. The cell lysate was aliquoted between two tubes. One tube was used to measure maximal RubisCO activity: this tube was supplemented with 10 mM

NaHCO3 to ensure CO2 saturation to achieve RubisCO carbamylation. The second tube was used to measure initial RubisCO activity: no external CO2 source was added. This aliquot was representative of the proportion of activated RubisCO active sites in the culture at the time of sampling. Soluble protein (5-20 µg) was used to determine RubisCO activity in initially and maximally activated lysates using the standard 14C RubisCO assay method at 25 °C (R. Tabita and B. Witte, personal communication). The lysate to be maximally activated was incubated for 14 5 min with an additional 50 mM NaHCO3 spiked with NaH CO3 (ViTrax, Placentia, CA) at 25 14 °C before initiation of RubisCO activity (specific activity of NaH CO3 in final reaction: 0.06 -1 14 µCi mmol ). The initially activated lysate was not pre-incubated with NaH CO3, as the reaction 14 was initiated immediately by addition of 0.8 mM RuBP following the addition of NaH CO3 (specific activity: 0.1 µCi mmol-1). The RubisCO reaction was stopped after 5 min by adding concentrated propionic acid. Acid stable reaction products were measured by scintillation 14 counting following centrifugation for 1.25 hr to aerosolize unfixed NaH CO3. The proportion of initial activity to maximal RubisCO activity yields the carbamylation state of the sample.

3.2.5. SDS-PAGE and Western Blotting. Algal cultures (45-50 mL) were pelleted by centrifugation at 3,000 x g for 10 min at 4 °C. The pellets were flash frozen in liquid nitrogen and stored at -80 °C until lysate extraction. The pellet was resuspended in 1.5 mL LDS extraction buffer (140 mM Tris base, 105 mM Tris-HCl, 0.5 mM EDTA, 2% lithium dodecyl sulfate, 10% glycerol, and 0.1 mg/mL PefaBloc inhibitor) and cells were lysed by bead beating for four cycles of 30 s with 1 min ice incubations between cycles. Crude cell lysate was centrifuged at 8,000 x g for 2 min to clear particulate material from the lysate. Protein concentration was measured according to the Bradford method (1976) using the Bio-Rad DC Kit protocol with BSA as a standard. Whole cell lysates were electrophoretically separated by SDS-

50

PAGE on a 15% (w/v) polyacrylamide resolving gel and 8% (w/v) polyacrylamide stacking gel using a Mini-Protean II apparatus (Bio-Rad, Hercules, CA). RubisCO large subunit (RbcL) standard was used to generate a standard curve for quantitation (Agrisera, Vännäs, Sweden). Following separation by SDS-PAGE, proteins were electrophoretically transferred to nitrocellulose membranes at 100 V for 1.5 hr. Nitrocellulose membranes were blocked with 5% Amersham ECL blocking agent (GE Healthcare Bio-Sciences, Piscataway, NJ) in 0.05% TBS- Tween and probed with antibody raised against RbcL (1:15,000 dilution) (Agrisera, Vännäs, Sweden). Secondary antibody conjugated with horseradish (Sigma-Aldrich, St. Louis, MO) was applied and the membrane was incubated with Amersham ECL Select detection reagents (GE Healthcare Bio-Sciences, Piscataway, NJ). Last, the membrane was developed on CL-X Posure film (Thermo Scientific, Rockford, IL). Western blots were quantified by densitometry using ImageJ software.

3.2.6. RubisCO activase gene sequencing. Primers targeting RubisCO activase in C. raudensis UWO241 were designed in CLC Main Workbench 6 using a contig sequence recovered from a UWO241 transcript library which showed high homology to C. reinhardtii rca gene (Morgan- Kiss, Kiss & Raymond, in prep). A fragment of RubisCO activase (1,259 bp) was PCR amplified with forward primer, UWORcaF1 (5’ CTC ACC CAA CGA CAC CAT 3’), and reverse primer, UWORcaR1 (5’ CCC ACC GAC TAC TGC ACT 3’). Taq DNA polymerase was used as supplied in the GoTaq Green Master Mix (Promega Corporation, Madison, WI) to amplify rca in a three-step PCR program (one cycle at 95 °C for five minutes, 30 cycles of denaturing: 95 °C for one minute, annealing: 49 °C for one minute, and elongation: 72 °C for one minute and 15 seconds, and a final cycle at 72 °C for five minutes). The PCR products were purified using a Wizard SV gel and PCR clean-up system (Promega Corporation, Madison, WI). The resultant PCR product was ligated into a pGEM vector (pGEM-T Easy vector systems, Promega) and transformed into chemically competent E. coli DH5α. The cells were plated on Luria Burtani plates supplemented with Ampicillin (100 µg/mL) for selection of resistant transformants. Recombinant E. coli colonies were identified by blue/white screening and confirmed by colony PCR using pUC/M13 sequencing forward primer (5’ GTT TTC CCA GTC ACG AC 3’) and reverse primer (5’ CAG GAA ACA GCT ATG AC 3’). Colony PCR products were run on 1.2% agarose gels and DNA bands matching the size of the target gene were cut out and gel purified. The resulting DNA concentration was measured on a Nanodrop1000 (Thermo Scientific, West

51

Palm Beach, FL) and sequenced in both directions by Sanger sequencing using the pUC/M13 sequencing primers. Sequences were translated and alignments containing closely related Rca sequences chosen from NCBI BLAST blastx results were generated in CLC Main Workbench 6. Last, a neighbor-joining protein sequence tree was generated in MEGA 6.0.

3.3. RESULTS

3.3.1. Effect of growth irradiance and temperature on algal growth. The growth physiologies of UWO241 and SAG49.72 were compared under either varying growth irradiances or temperatures. First, both strains were grown under a range of growth irradiances and temperature was kept constant at levels comparable to their natural environments (8 and 29 °C for UWO241 and SAG49.72, respectively). Despite growth at low temperatures, the psychrophilic strain exhibited the ability to achieve growth rates comparable to cultures of the mesophile which were grown at a temperature >3-fold higher (Figure 3.1). However, the two strains exhibited differential relationships between growth rate and irradiance. Four major phases can be described between growth rate and increasing light availability. At very low light levels, phase 1 represents no net growth, and the light levels where growth rate balances maintenance costs is called the growth compensation point. At irradiances below 25 µmol photons m-2 s-1, UWO241 exhibited higher growth rates compared with SAG49.72 and by extrapolation the light compensation point for UWO241 is ~30% lower compared with SAG 49.72 (0.94 vs. 1.35 µmol photons m-2 s-1, respectively). Conversely, an estimate of the maintenance respiration rate can be extrapolated from the zero intercept on the growth rate axis: UWO241 appears to require significantly higher maintenance costs compared with SAG49.72 (Fig. 3.1, arrows). Phase 2 represents a linear relationship between growth rate and irradiance. SAG49.72 exhibited almost a 2-fold higher ratio of growth rate: irradiance compared with UWO241 (0.76 vs. 0.40, respectively). Phase 3 represents the balance of gross photosynthesis and respiration rate. UWO241 exhibited a clear balance in growth at irradiance of ~50 µmol photons m-2 s-1, while growth rates of SAG49.72 are not balanced with respiration even at the highest growth irradiance (110 µmol photons m-2 s-1). The final phase in the growth rate: irradiance response in photosynthetic organisms is the onset of photoinhibition of growth. This phase is evident in psychrophile cultures grown at 110 µmol photons m-2 s-1, but is not observed in the mesophile, indicating that the UWO241 growth rates are saturated at low irradiances compared with SAG49.72 and exhibit inhibition above the light

52 saturation point. UWO241 and SAG49.72 were also grown under varying temperatures while growth irradiance was comparable to their natural environments (20 and 110 µmol photons m-2 s- 1 in UWO241 and SAG49.72, respectively; Fig. 3.1B). The growth rate of the psychrophile was lower than the mesophile across all temperatures, but this was likely an irradiance effect. Unlike the differential responses to irradiance, the response to temperature between the two algal species was comparable with both strains exhibiting similar growth rate: temperature ratios of ~0.36 oC-1 (Fig. 3.1B).

3.3.2. Effect of growth irradiance and temperature on photochemical function. Room temperature chl a fluorescence in combination with the saturation pulse method was used to analyze in vivo photosynthetic performance. Acclimation to variable growth irradiance and temperature influenced photochemical function in both strains. The parameter FV/FM (i.e., maximum quantum yield of PSII under dark adaptation) is a measure of the theoretical maximum of functional PSII reaction centers. Values as high as 0.9 are achievable in a healthy, non- stressed plant; however, values of ~0.75 are more typical in algal cultures (Flameling and Kromkamp 1998). The lower values in algal cultures reflect the occurrence of alternative electron transport pathways in algae that remain reduced even under prolonged dark adaptation.

In response to varying light intensity, FV/FM values in UWO241 remained relatively stable, but -2 -1 for SAG49.72 FV/FM decreased 18% and 14% at the lowest (8 and 14 µmol photons m s , respectively) and 33% at the highest (110 µmol photons m-2 s-1) irradiances compared to intermediate irradiances (Fig. 3.2A). The parameter ɸPSII is a measure of the PSII quantum yield and reflects more closely the true functional proportion of PSII reaction centers under a given growth condition. Therefore, ɸPSII, values are always equal to or in many cases lower than that of FV/FM. This value was greater in SAG49.72 than UWO241 at all irradiances measured, but began to decrease at irradiances higher than 45 µmol photons m-2 s-1 for -2 -1 SAG49.72 and 25 µmol photons m s for UWO241 (Fig. 3.2B). The quenching parameters qL and qN (photochemical and non-photochemical quenching, respectively) provide estimates of the proportion of energy which is used for photochemical reactions vs. lost as heat dissipation. In addition, the reciprocal of qL (i.e., 1-qL) can be used as an estimate of the reduced state of the plastoquinone (PQ) pool. Values of 1-qLwere elevated in cultures of UWO241 relative to SAG49.72 at all growth irradiances above 14 µmol photons m-2 s-1 (Fig. 3.2C). UWO241 responded to an increase in irradiance by linearly increasing 1-qL up to a maximum of ~0.6 at a

53 growth irradiance of 45 µmol photons m-2 s-1, which corresponded to the light saturation point for growth rate in UWO241 (Fig 3.1A). In contrast, while there was some variability across treatments, the PQ pools remained relatively oxidized in SAG49.72 cultures, as evidenced by 1- qL at or below 0.3 (i.e., ~30% of the PQ pool was reduced) (Fig. 3.2C). Similarly, since the majority of the photons were used to drive photochemistry, the capacity for qN was low in

SAG49.72 cultures regardless of the growth irradiance. Response of qN to light availability in the psychrophile exhibited two phases: a slow rise from ~0.2 to 0.3 at growth irradiances between 5 and 65 µmol photons m-2 s-1, followed by a rapid increase in non-photochemical quenching from 0.3 to 0.6 at light levels above 65 µmol photons m-2 s-1 (Fig. 3.2D).

For cultures grown under varying growth temperatures, FV/FM was stable across all temperatures for both algal strains (Fig. 3.3A). The quantum yield of PSII for UWO241 declined by 22% between 8 and 12 °C, but remained high at all growth temperatures in cultures of SAG49.72 (Fig. 3.3B). The PQ pool of UWO241 was more highly reduced than that of SAG49.72, consistent with the trends observed under varying growth irradiance. Complementary to the changes observed in ɸPSII, the PQ pool reduction state increased with growth temperature in UWO241; however, the PQ pool remained relatively oxidized for SAG49.72 grown at variable temperatures (Fig.3.3C). Last, despite the temperature sensitivity observed in UWO241 at the level of 1-qL, non-photochemical quenching capacity (qN) remained low in both organisms regardless of growth temperature (Fig.3.3D).

3.3.3. Effect of growth irradiance and temperature on RubisCO activity and carbamylation state. RubisCO activity was measured under varying growth irradiance and temperature conditions in order to determine the impact of these factors on the initial (i.e., an estimate of the “in vivo” carboxylase activity) and maximal RubisCO carboxylase activities, as well as the carbamylation state of the enzyme. In general, both organisms exhibited similar responses to light; although, the trends were variable across the two organisms. The highest initial activity was observed under the lowest irradiance levels and declined with increasing growth irradiance (Fig. 3.4A). Maximal RubisCO activity was generally stable across the irradiances measured, but exhibited some variability at low growth irradiances (8-25 µmol photons m-2 s-1) in both algal strains. It should also be noted that maximum activity was often lower than initial activity, particularly at lower light levels (Fig. 3.4A, B). This phenomenon occurred in UWO241 cultures grown at lower light levels and resulted in carbamylation states (i.e., ratio of initial:maximal RubisCO activity) that

54 were at or above 100%. Carbamylation state began to decrease and stabilized at 50-60% for both algal strains at 25-110 µmol photons m-2 s-1 (Fig. 3.4C). In algal cultures grown under varying temperatures, initial RubisCO activity exhibited a moderate decrease in cultures of UWO241 grown at 2 vs. 8 °C, while growth temperature did not have a significant effect on initial RubisCO activity in SAG49.72 (Fig. 3.5A). In contrast, maximal RubisCO activity declined by 36% for UWO241 grown under 8 vs. 12 °C, and remained stable in cultures of SAG49.72 across all growth temperatures (Fig. 3.5B). As a consequence of the loss in maximal activity at 12 oC, carbamylation state exceeded 100% for UWO241 at 12 °C, while growth temperature effects on carbamylation state in SAG49.72 were minimal (Fig. 3.5C).

3.3.4. Effect of growth irradiance and temperature on RubisCO abundance. Adjustments in enzyme activity in response to environmental pressures may reflect alterations in the enzyme turnover rates and/or in cellular abundance of the enzyme. RubisCO abundance was investigated in response to varying growth irradiance and temperature in the psychrophilic and mesophilic algal strains. There was a minimal effect of irradiance on RubisCO abundance (represented by the large subunit of RubisCO, RbcL) across the irradiances measured in both algal strains (Fig. 3.6A, C). In contrast, growth temperature had a marked effect on RbcL levels in UWO241. In UWO241, RubisCO content decreased 40% between 2-8 °C and 19% between 8-12 °C. In SAG49.72, RubisCO content decreased 33% from 12-20 °C and increased from 20-29 °C (Fig. 3.6B).

3.3.5. Determination of the RubisCO activase sequence for C. raudensis UWO241. The enzyme RubisCO activase plays the role of maintaining functional RubisCOs and therefore influences the carbamylation state of RubisCO. A putative rca cDNA was identified in a transcriptome of UWO241 using blastx. To validate whether this contig was a bona fide RubisCO activase gene, primers were designed to cover 1,259 bp of the rca gene. This represented about 91% of the full gene. We also attempted to amplify rca from SAG49.72 but were unsuccessful. A neighbor- joining phylogenetic tree was generated based on translated DNA. At the protein level, UWO241 activase was most closely related to that of the green alga, Chlorococcum littorale with 84% identity (Fig. 3.7), and was more distantly related to the Rca of the model alga C. reinhardtii (78% identity). Results of blastx further showed that UWO241 Rca shared 67% identity with

55

Antarctic hairgrass (D. antarctica) Rca beta form. Thus, C. raudensis UWO241 likely expresses a functional RubisCO activase.

3.4. DISCUSSION Varying irradiances affected both strains of C. raudensis as measured by their growth, photochemistry, and RubisCO carbamylation states. The growth rates for UWO241 and SAG49.72 under varying irradiances were reflective of their respective natural environments. For example, growth rates in UWO241 cultures exhibited a lower growth saturation point than SAG49.72, the latter strain being adapted to an environment with higher irradiances (Fig. 3.1A). Furthermore, the requirement for maintenance was higher for UWO241, although this may be due to its lower growth temperature (Fig. 3.1A). The range of growth irradiances did not impact maximum photochemical capacity since

FV/FM remained high in both algal strains under almost all growth irradiances (Fig. 3.2A). However, UWO241 cultures grown at 45 µmol photons m-2 s-1 exhibited a decrease in the quantum yield PSII photochemistry (ɸPSII) which more accurately reflects the in vivo functioning of PSII at a given irradiance, and this correlated with a more highly reduced PQ pool

(1-qL) (3.2B, C). These trends could reflect excess electrons saturating the PQ pool which may be a result of limitations in downstream carbon fixation reactions (Hüner et al. 1998). Excess electrons dissipated by non-photochemical quenching responded to a higher irradiance (65 µmol photons m-2 s-1). This lower capacity for non-photochemical quenching induction has been previously observed for UWO241 suggesting limited xanthophyll cycle function (Morgan et al.1998, Szyszka et al. 2007). For SAG49.72, ɸPSII was lower at 110 µmol photons m-2 s-1 compared with cultures grown at lower irradiances, however this did not correspond with an increase in PQ pool reduction nor non-photochemical quenching (Fig. 3.2B, C, D). Therefore, it is likely that the irradiances used here were never high enough to saturate PSII photochemistry in SAG49.72, as this organism is adapted to a higher irradiance environment than UWO241. Carbamylation state of RubisCO reflects the proportion of active enzyme under a particular growth condition. In response to irradiance, both organisms exhibited an inverse relationship between RubisCO carbamylation state and light (Fig. 3.4). This result was not anticipated for SAG49.72 as this organism is adapted to moderate irradiances and therefore should have a high carbamylation state at all irradiances measured. Moreover, the phenomenon observed at low irradiance levels (carbamylation state >100%) indicated that there may be

56 additional metabolic considerations under very low light levels (Fig. 3.4). There are two possible explanations for this phenomenon: an artificial inhibition of maximal activity during assay preparation or true inhibition in response to low light. Evidence for both possibilities exists: for example, maximal activity was low in the Thalassiosira pseudonana presumably due to RubisCO inactivation during assay pre-incubation, but this occurred at high growth irradiances (>180 µmol photons m-2 s-1) (MacIntyre et al. 1997). Alternatively, in P. vulgaris a carbamylation state >100% was attributed to inhibitor production and binding to RubisCO at low irradiance (Sharkey et al. 1986, Kobza and Seemann 1988) and decreased photosynthetic ATP generation which does not promote RubisCO activase activity (i.e., removal of the inhibitors) (Sharkey et al. 1986). UWO241 expresses RubisCO activase and it is presumed that SAG49.72 also possesses this enzyme, but we were unable to identify a sequence. Overall, the regulation of RubisCO in response to irradiance may not be as straightforward in aquatic microalgae as it is for plants, but our results suggested regulation of RubisCO activity and not protein abundance is important in both organisms in response to varying irradiance (Fig. 3.4 and 3.6A). Despite the low growth temperatures for UWO241 compared to SAG49.72, this organism maintained high growth rates indicating major alterations in metabolism that allow the psychrophile to grow at low temperatures (Fig. 3.1B). Evidence for this was observed in the relatively high photochemical activity (Fig. 3.3A) and comparable RubisCO turnover rates in UWO241 and SAG49.72 (Fig. 3.5A, B). Both organisms showed a strong temperature response in their growth rates which increased with temperature (Fig. 3.1B). However, UWO241 exhibited thermal sensitivity at the level of its photochemistry, RubisCO activity/carbamylation state, and RubisCO abundance. Photochemical redox poise was disrupted at high temperature grown cultures of UWO241 as indicated by the steep rise in 1-qL which could be a result of limitations in carbon cycling (Fig. 3.3B). Additionally, a decrease in functional PSII reaction centers occurred at 12 °C; however, non-photochemical quenching did not respond, similar to the delay in non-photochemical induction observed under varying irradiances (Fig. 3.3C, D). This may indicate that other energy dissipation processes were playing a role at 12 °C. For example, energy quenching by constitutive unregulated thermal dissipation and fluorescence was found to increase in UWO241 grown at 16 °C (Szyszka et al. 2007) or incubated at 24 °C (Possmayer et al. 2011). Photosynthetic electron transport did not become saturated for SAG49.72 grown at

57 suboptimal growth temperatures which indicates that downstream carbon fixation reactions were not affected by these conditions. The response in RubisCO activity to temperature was distinct compared with light- induced effects. Temperature generally did not impact carbamylation state indicating that regulation of RubisCO activity was not a target for the temperature response. One exception was observed in 12 °C grown UWO241 cultures where the ability to reach maximal activity was negatively impacted (Fig. 3.5). This could be a reflection of RubisCO inhibitors binding during the assay preparation or a true physiological response. This phenomenon may indicate that the 12 °C grown culture of UWO241 has increased abundance of inhibitors relative to RubisCO, and/or impaired function of RubisCO activase which allows for inhibitor binding to RubisCO. Previous studies have shown temperature sensitivity of RubisCO activase resulting in impaired function at elevated temperatures (Kobza and Seemann 1987, Feller et al. 1998, Salvucci and Crafts- Brandner 2004a). The 12 °C grown UWO241culture had a more highly reduced PQ pool further suggesting limitation in downstream carbon fixation reactions which could be a result of impaired RubisCO activase function (Fig. 3.3C). While RubisCO carbamylation played a minor role in modulating RubisCO activity in response to temperature, enzyme abundance played a major role. Both strains of C. raudensis increased levels of the large subunit of RubisCO in response to a decline in temperature, but this occurred to a greater extent for the psychrophile than SAG49.72. UWO241 had overall higher RubisCO levels which may account in part for its ability to maintain relatively high growth rates at low temperatures (Fig. 3.1B and 3.6B). An increased concentration of RubisCO in UWO241 at low temperatures could be important in maintaining RubisCO activity under conditions in which the catalytic rate is likely to be reduced. Although variability in concentration in UWO241 was high, there was a large increase from 12 to 2 °C (Fig. 3.6B). This variability was attributed to a high western blot signal in UWO241 compared to SAG49.72 which were exposed on the same piece of film and interference by a secondary signal in UWO241. This secondary signal was consistently present for UWO241 and may indicate post- translational modification of RbcL (Houtz et al. 2008). Modulation of RubisCO differed based on growth conditions where carbamylation state and enzyme abundance varied in response to irradiance and temperature, respectively. Carbamylation state was differentially affected at low and high growth irradiances in both

58

UWO241 and SAG49.72, although the mechanism of this response requires further analysis. In UWO241, RubisCO enzyme abundance increased in response to decreased growth temperature as was observed in multiple green plants (Yamori et al. 2009, Dahal et al. 2012) and algae (Savitch et al. 1996, Devos et al. 1998). RubisCO abundance, activity, and carbamylation state of UWO241 were more sensitive to temperature than that of SAG49.72, and may indicate the importance of the environmental parameters within the alga’s natural habitat on shaping enzyme activity (Cavanagh and Kubien 2013).

59

A B

0.6

) -1

0.4

0.2 UWO241 Growthrate (day SAG49.72

0.0 0 40 80 120 0 10 20 30 Irradiance (mol photons m-2 s-1) Temperature (C)

Figure 3.1. Effects of growth irradiance (A) and temperature (B) on growth rates of psychrophilic C. raudensis UWO241 and mesophilic C. raudensis SAG49.72. Growth temperatures were 8 °C and 29 °C (A) and growth irradiances were 20 and 100 µmol photons m- 2 s-1 (B) for UWO241 and SAG49.72, respectively. Arrows show maintenance respiration rate extrapolated for UWO241 and SAG49.72. Error bars show standard error (n=3-4).

60

A B

C D

0.8 0.8

0.6 0.6

M /F

0.4 PSII 0.4

V

 F

0.2 UWO241 0.2 SAG49.72

0.0 0.0

0.8 0.8

0.6 0.6 L

0.4 N 0.4

q 1-q

0.2 0.2

0.0 0.0 0 20 40 60 80 100 120 0 20 40 60 80 100 120 -2 -1 Growth irradiance (mol photons m s ) Growth irradiance (mol photons m-2 s-1)

Figure 3.2. Effect of growth irradiance on maximal photochemical efficiency of PSII; FV/FM (A), quantum yield of PSII photochemical efficiency; ɸPSII (B), reduction state of PQ pool; 1-qL (C), and non-photochemical quenching; qN (D) of psychrophilic C. raudensis UWO241 and mesophilic C. raudensis SAG49.72 grown at temperatures (8 and 29 °C, respectively) comparable to that of their natural environments. Error bars show standard deviation (n=3-4).

61

0.70 A 0.6 B

0.65

0.4

M 0.60

PSII

/F

V

 F 0.2 0.55 UWO241 SAG49.72

0.50 0.0 0.6 C 0.6 D

0.4 0.4

L

N

q 1-q 0.2 0.2

0.0 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Temperature (C) Temperature (C)

Figure 3.3. Effect of growth temperature on maximal photochemical efficiency of PSII; FV/FM (A), quantum yield of PSII photochemical efficiency; ɸPSII (B), reduction state of PQ pool; 1-qL (C), and non-photochemical quenching; qN (D) of psychrophilic C. raudensis UWO241 and mesophilic C. raudensis SAG49.72 grown at irradiances (20 and 100 µmol photons m-2 s-1, respectively) comparable to that of their natural environments. Error bars show standard deviation (n=4).

62

12 A ) -1 UWO241 10 SAG49.72

8 mg protein

-1 6

hr 2 4

molCO 2

Initiallyactivated activity

 ( 0

12 B )

-1 10

8

6

mg protein mg -1

hr 4 2

2

mol CO mol

 Maximally activated activity activated Maximally

( 0

160 C 140

120

100

80

60

Carbamylationstate (%) 40

0 20 40 60 80 100 120 Growth irradiance (mol photons m-2 s-1)

Figure 3.4. Effect of growth irradiance on initial (A) and maximal (B) RubisCO activity, and carbamylation state (C) of C. raudensis UWO241 and C. raudensis SAG49.72 grown under temperatures (8 and 29 °C) comparable with that of their natural environments and varying irradiance. Dotted lines indicate 10% error on either side of 100% carbamylation state. Error bars show standard error (n=3).

63

10 )

-1 A 8

6

mg protein

-1 hr

2 2 4

2 UWO241 Initiallyactivated activity (molCO SAG49.72 0 10

B

) -1 8

6

mg protein -1

hr 4 2

2

molCO

Maximallyactivated activity ( 0 160 C 140

120

100

80

60 Carbamylationstate (%)

40 0 5 10 15 20 25 30 Temperature (C)

Figure 3.5. Effect of growth temperature on initial (A) and maximal (B) RubisCO activity, and carbamylation state (C) of C. raudensis UWO241 and C. raudensis SAG49.72 grown under irradiances (20 and 100 µmol photons m-2 s-1) comparable with that of their natural environments and varying temperatures. Dotted lines indicate 10% error on either side of 100% carbamylation state. Error bars show standard error (n=4-5).

64

A B

) 1.25 UWO241 -1 SAG49.72

1.00

g 

0.75

0.50

0.25 Abundance (pmol RbcL RbcL (pmol Abundance

0.00 0 40 80 120 0 10 20 30 Irradiance (mol photons m-2 s-1) Temperature (C) C UWO241 SAG49.72 8 14 25 45 65 110 8 14 25 45 65 110 Standards 55 kDa

Figure 3.6. Effect of growth irradiance (A) and growth temperature (B) on RbcL (the large subunit of RubisCO) abundance determined by densitometry. Representative growth irradiance experiment immunoblot against RbcL corresponding to the 55 kDa marker protein (C). The film was exposed for 30 s. Lanes are labeled with growth irradiances from left to right: C. raudensis UWO241 at 8, 14, 25, 45, 65, 110 µmol photons m-2 s-1, C. raudensis SAG49.72 at 8, 14, 25, 45, 65, 110 µmol photons m-2 s-1, 0.075 and 0.15 pmol standard. Error bars show standard deviation (n=3).

65

Figure 3.7. Neighbor-joining phylogenetic tree based on amino acid sequence (387 residues) of C. raudensis UWO241 RubisCO activase and nearest neighbors according to results of NCBI BLAST blastx. Values at node represent bootstrap support observed in 500 replicates. The scale bar represents a difference of 0.05 substitutions per site.

66

CHAPTER 4

Diversity and distribution of carbon fixation genes in ice-covered lakes of the McMurdo Dry Valleys, Antarctica

Jenna M. Dolhi, Amber G. Teufel, Weidong Kong, Rachael M. Morgan-Kiss

Author contributions: JD (75% contribution) and AT (25% contribution) developed qPCR methods, data collection and analyses; WK developed clone library and sequencing methods, data collection and analyses; JD and AT wrote the manuscript; JD, AT, WK and RMK edited the manuscript

Publication status: submitted to Freshwater Biology

67

CHAPTER 4

SUMMARY 1. The goal of this study was to characterize the diversity and distribution of autotrophic microorganisms in permanently ice-covered lakes located in the Taylor (east and west lobes of Lake Bonney, Lake Fryxell) and Wright (Lake Vanda) Valleys of the McMurdo Dry Valleys (MDV, Victoria Land, Antarctica) to determine whether lake physicochemical factors correlated with spatial trends among major photosynthetic and chemolithoautotrophic groups. 2. Abundance and diversity of a suite of functional genes that play key roles in carbon fixation (rbcL isoforms IA/B, ID; cbbM; nifJ) were measured throughout the water columns of MDV lakes. 3. Both lobes of Lake Bonney were dominated by Ribulose-1, 5-bisphosphate carboxylase oxygenase (RubisCO) form ID rbcL (, stramenopiles, and cryptophytes), while Lake Fryxell was dominated by form ID cryptophyte rbcL. Lake Vanda was the least productive lake and was dominated by form IA/B rbcL (cyanobacteria and chlorophytes). Chemolithoautotrophs harboring either form II RubisCO (cbbM) or pyruvate:ferredoxin (nifJ) were highly abundant at depths below the permanent chemocline in the west lobe of Lake Bonney and Lake Fryxell. 4. Depth profiles in functional gene abundance (monitored by quantitative PCR) exhibited correlations with a number of lake physicochemical parameters which showed some lake- specific variability. In general, autotrophic carbon fixation genes from photosynthetic organisms (forms IA/B and ID rbcL) were positively correlated with light availability. Abundance of genes from chemolithoautotrophic organisms (cbbM, nifJ) exhibited a negative correlation with light and were associated with the presence of sulfide (Fryxell) or dimethylated sulfur compounds (west lobe Lake Bonney).

4.1. INTRODUCTION

The McMurdo Dry Valleys (MDV) is one of the largest polar deserts on earth. Within this extreme ecosystem are numerous permanently ice-covered lakes and ponds. Ice caps (3 to 7 m thick) produce stable, highly stratified water columns, the absence of seasonal mixing, minimal gas exchange, low light penetration, and reduced rates of primary production (PPR) (Priscu 1991, Spigel and Priscu 1996). Although similar in origin, the lakes vary in their

68 physicochemical parameters, owing to differential solute input and evolutionary histories (Green and Lyons 2009). Microbial community structure and dynamics are largely influenced by ‘bottom up’ controls such as resource availability and environmental conditions (Moorhead et al. 1999). The water columns of the MDV lakes provide an oasis for phototrophic, chemolithoautotrophic, and heterotrophic microorganisms which interact within simplified, -dominated food webs (Priscu et al. 1999). Each lake harbors distinct communities which are vertically stratified within the water column based on light and nutrient availability (Bielewicz et al. 2011, Kong et al. 2012a, b, Priscu et al. 1999). Within the planktonic community, single-celled photosynthetic eukaryotes provide the majority of organic carbon through light-dependent primary production and are important for nutrient cycling (Bielewicz et al. 2011). Evidence of bacteria capable of anoxygenic photosynthesis (Karr et al. 2003), as well as chemolithoautotrophic carbon fixation, have also been reported (Kong et al. 2012a, Sattley and Madigan 2006, Vick and Priscu 2012). Microbial functional diversity has been explored in the MDV lakes with a focus on four lakes located in the Taylor Valley (east and west lobes of Lake Bonney, Lake Fryxell and Lake Hoare) which are intensively studied as part of the McMurdo Long Term Ecological Research (LTER) Program. There are extensive studies on the importance of mixotrophic phytoflagellates such as cryptophytes and heterotrophic nanoflagellates (Laybourn-Parry 2002, 2009, Laybourn- Parry and Pearce 2007, Roberts et al. 1999, Roberts et al. 2004a, b). Several studies of Lake Fryxell have reported the presence of phototrophic purple bacteria (Karr et al. 2003), sulfate- reducing bacteria (Karr et al. 2005), sulfur-oxidizing bacteria (Sattley and Madigan 2006), and methanogenic Archaea (Karr et al. 2006). Voytek et al. (1999) used the functional gene amoA to identify the presence of ammonia-oxidizing bacteria across several MDV lakes, and found evidence for beta-Proteobacteria in Lake Hoare. In contrast with the Taylor Valley lakes, there are currently no reports on the phylogenetic diversity of Lake Vanda (located in Wright Valley) microbial communities; although, microscopic analysis of phytoplankton revealed microflagellates in the upper depths (above 55 m) and cyanobacteria at the bottom of the euphotic zone (Vincent and Vincent 1982). The main carbon fixation pathway is the Calvin Benson Bassham (CBB) cycle. RubisCO catalyzes the carbon fixation step and exists as one of four isoforms in photoautotrophic and

69 chemolithoautotrophic microorganisms (Tabita et al. 2008). The large subunit, encoded by rbcL, is highly conserved and has been used to define diversity and abundance of autotrophic communities (Alfreider et al. 2012, Elsaied et al. 2007, Giri et al. 2004, John et al. 2007, Kovaleva et al. 2011, van der Wielen 2006). Form I RubisCO can be further divided into several isoforms: marine α-cyanobacteria and proteobacteria harbor form IA, chlorophytes and β- cyanobacteria harbor form IB, proteobacteria harbor form IC, and chromophytic algae and diatoms harbor form ID (Tabita et al. 2008). Form II RubisCO is encoded by the cbbM gene and has been applied as a marker for chemolithoautotrophic organisms (Campbell and Cary 2004, Giri et al. 2004). Evidence for cbbM was observed in the west lobe of Lake Bonney (Kong et al. 2012a) and past studies indicate that chemolithoautotrophs harboring cbbM are likely to be abundant in Lake Fryxell (Karr et al. 2003, Sattley and Madigan 2006). Additional gene markers have been used to describe autotrophic populations in natural environments. The reverse tricarboxylic acid (rTCA) cycle, an alternate carbon fixation pathway, has been identified in microorganisms including Chlorobium and members of delta- and epsilon-proteobacteria (Campbell and Cary 2004). The rTCA cycle can be detected using the carbon fixation enzyme, pyruvate:ferredoxin oxidoreductase (nifJ) as a molecular marker; however, this enzyme can also function in the reverse direction, catalyzing a heterotrophic oxidative reaction. Other genes encoding enzymes of the rTCA cycle such as, ATP citrate (aclBA) and 2- oxoglutarate:ferredoxin oxidoreductase (oorDABC), have been used as molecular markers (Campbell and Cary 2004). In recent years environmental monitoring of functional genes has been applied across diverse microbial habitats to define the role of microbial functional groups in nutrient cycling (e.g. Karr et al. 2003, Karr et al. 2005, Kong et al. 2012a, b, 2014, Lopez-Garcia et al. 2001, Lovejoy et al. 2006, Massana and Pedros-Alio 2008, Massana et al. 2006, Not et al. 2009, Piganeau et al. 2008, Piquet et al. 2008, Zhu et al. 2005). Three recent studies applied a combined sequencing and environmental quantitative PCR (qPCR) approach to provide new information on the spatial and temporal distribution of photosynthetic (Kong et al. 2012b, 2014) and chemolithoautotrophic (Kong et al. 2012a) microorganisms in Lake Bonney. These studies reported that three major groups of photosynthetic eukaryotes (stramenopiles, haptophytes, and chlorophytes) dominate the photic zone of the east (ELB) and west (WLB) lobes of Lake Bonney (Kong et al. 2012b). A unique chemolithoautotrophic bacterial community was discovered to be

70 living within the deep photic zone of WLB (Kong et al. 2012a). Last, the effect of light availability on gene (DNA) and transcript abundance (mRNA) of RubisCO form IA/B exhibited strong positive correlations with photosynthetically available radiation (PAR) (Kong et al. 2012b). A recent study in Lake Bonney on the diversity and expression of psbA, which encodes a major photochemistry protein, also showed that psbA transcript abundance exhibited complex spatial and temporal trends that were dependent upon both PAR and nutrient availability (Kong et al. 2014). While these studies provided evidence for phylogenetic diversity and gene activity of MDV autotrophic communities, sampling schemes were limited to the photic zone of Lake Bonney and occurred during the seasonal transition between autumn and polar winter (Feb to Apr). Here we investigated carbon fixation potential represented by the CBB and rTCA cycles in planktonic communities of MDV lakes ELB, WLB, Fryxell, and Vanda. Depth profiles for gene (DNA) abundance and diversity of communities catalyzing light-dependent vs. - independent carbon fixation pathways were monitored via a suite of autotrophic functional gene markers. Functional gene abundance was correlated with several limnological parameters to ascertain major abiotic drivers of MDV autotrophic community distribution.

4.2. MATERIALS AND METHODS

4.2.1. Site description. Lakes Bonney and Fryxell are located in Taylor Valley and Vanda is located in Wright Valley. Lake Bonney is separated into two basins (ELB, WLB) by a 13 m sill which allows mixing between the upper layers of WLB and ELB during the summer (Spigel and Priscu 1996, 1998). WLB water chemistry is influenced by a unique geochemical feature, an iron-rich subglacial outflow (Blood Falls), which drains episodically from nearby Taylor Glacier (Mikucki et al. 2009). A complex history has produced distinct water chemistry in each lobe, particularly in the anoxic waters below the chemoclines, and the two basins are considered to be separate watersheds (Lee et al. 2004a, b, Lyons et al. 2000, Priscu et al. 1996). Lake Fryxell is located between the Canada and Commonwealth Glaciers, and nine major inflow streams contribute to this relatively shallow lake. Lake Fryxell is relatively freshwater and sulfidic, and has a deep anoxic zone below depths of 9.5 m. Lake Vanda is one of the deepest MDV lakes (80 m). Its water column is extremely oligotrophic and supports very low levels of microbial production (Vincent and Vincent 1982). This lake is fed by the largest river in the dry valleys,

71 the Onyx River, which flows for 1-2 months per year and originates from the Wright Lower Glacier (Canfield and Green 1985). The ice cover of Lake Vanda is unusually smooth and transparent compared with other dry valley lakes which contributes to bottom waters reaching relatively high temperatures (Table 4.1) (Vincent and Vincent 1982). Similar to Lake Bonney, Lake Vanda is hypersaline below the chemocline and plankton productivity in both lakes is phosphorus limited (Priscu 1995, Vincent and Vincent 1982).

4.2.2. Sample collection. Samples were collected on the following dates during the field season of 2012: Fryxell - Nov 11, Vanda - Nov 19, east lobe Bonney - Nov 21, and west lobe Bonney - Nov 28. All sampling depths were measured from the piezometric water level in the ice hole (c. 30 cm below the ice surface). Water samples were collected through water columns (8 to 12 sampling depths per lake) using a 5-L Niskin bottle (General Oceanics, FL) and were filtered in triplicate (0.4 – 2 L lake water per filter) onto 47-mm, 0.45-µm Durapore polyvinylidene fluoride membrane filters (Millipore, MA) using a vacuum of 0.3 mBar. The filters were frozen immediately in liquid nitrogen before being transported to McMurdo Station where they were sent on dry ice to the US laboratory. Samples were stored at -80 °C until processing.

4.2.3. Limnological parameters. Temperature, PAR, and nutrients (inorganic nitrogen and soluble reactive phosphorus; SRP) were measured using previously described methods (Kong et al. 2012b, Priscu 1997). Depth profiles for in situ PAR were measured with a Li-Cor LI-193 spherical quantum sensor (Li-COR Biosciences, NE). Nutrient data and dissolved inorganic carbon (DIC) were collected by the McMurdo LTER program and are available on their website (http://www.mcmlter.org/). Inorganic nitrogen species were measured with a Lachat autoanalyzer (Priscu 1997). SRP was measured manually using the antimony-molybdate method (Strickland and Parsons 1972). Depth profiles of Chl a levels were monitored by measuring in situ chlorophyll fluorescence with a submersible FluoroProbe (BBE Moldaenke GmbH, Germany) (Beutler et al. 2002). The FluoroProbe estimates abundances of major pigment groups and can detect four broad groups of Chl a-containing microorganisms based on differences in their light harvesting pigments (Beutler et al. 2002). Algal groups that reportedly can be discriminated by the FluoroProbe include (i) ‘green’ (Chlorophyta and Euglenophyta), (ii) ‘mixed’ or ‘brown’ (Haptophyta, Bacillariophyceae, Chrysophyceae, Dinophyceae), (iii) ‘blue’ (Cyanobacteria), and

72

(iv) ‘red’ (Cryptophyta). The FluoroProbe has been adopted across a broad range of aquatic ecosystems for rapid estimation of phytoplankton biomass (Carraro et al. 2012, Catherine et al. 2012, Houliez et al. 2012, Perga et al. 2013, Roubeix et al. 2011). This instrument was also used to measure temperature.

4.2.4. DNA extraction. Frozen filters were cut in half with sterile scissors and then cut into smaller pieces for DNA extraction. Environmental DNA was isolated using a FastDNA® spin for soil (MP Biomedicals, OH) following the manufacturer’s protocol and according to Kong et al. (2012b).

4.2.5. Quantitative PCR. The abundance of several functional genes were determined by qPCR as markers of autotrophy, including isoforms of form I RubisCO (form IA/B and ID rbcL), form II RubisCO (cbbM), and pyruvate:ferredoxin oxidoreductase (nifJ), generated from DNA using gene specific primers (Table 4.2). A primer set targeting form ID rbcL specific to the cryptophyte, Geminigera cryophila, was also included as it was determined previously that the form ID primers do not amplify rbcL from this organism (Kong et al. 2012b). We attempted to amplify other key functional markers of the rTCA pathway (ATP citrate lyase and 2- oxoglutarate:ferredoxin oxidoreductase); however, we were unable to recover amplicons from our samples, indicating that these genes were either not present or the homology between the primers and our samples was low. Bacterial 16S and eukaryotic 18S rRNA genes were quantified as a measure of prokaryotic and eukaryotic biomass (Bielewicz et al. 2011, Muyzer et al. 1993). Specificity of each primer set was determined by running a melting curve (50-95 °C). We performed a suite of quality controls using the minimum information for publication of qPCR experiments (MiQE) guidelines (Table 4.3) (Bustin et al. 2009). Standard curves were generated using pGEM-T Easy vector (Promega, WI) containing the gene of interest amplified from environmental DNA with specific primer sets (Table 4.2) (Kong and Nakatsu 2010). A standard curve (in technical duplicates) of gene abundance covering at least five orders of magnitude was included with each qPCR run to allow for absolute quantification of each gene and to avoid interassay variation (Bustin et al. 2009). qPCR reactions were run on a CFX Connect Real-time System (Bio-Rad Laboratories, CA) in a 10 µL reaction volume containing 0.8 µL DNA, 0.6 µL of each primer (0.6 µM), and 5 µL iQ SYBR Green Supermix (Bio-rad, CA). qPCR reaction conditions were 95 °C for 3 min followed by 40 cycles of 95 °C for 10 s,

73 annealing temperature (Table 4.2) for 10 s, and 72 °C for 30 s. Fluorescence intensity was acquired in the extension step. Triplicate DNA samples from each lake depth were analyzed in technical duplicates. Mean average gene abundance and standard error were determined for each sampling depth.

4.2.6. PCR, cloning and sequencing. To broaden the perspective on the diversity of autotrophic organisms residing in MDV lakes, we complemented an existing functional gene database which focused on the photic zone of Lake Bonney (Kong et al. 2012a, b) by sequencing isoforms I (IAB rbcL, ID rbcL) and II (cbbM) RubisCO genes from below the photic zone in Lake Bonney, as well as samples collected from Lakes Fryxell and Vanda (Table 4.4). We performed a more intensive sampling strategy for nifJ sequencing since sequences of this functional gene have not been previously reported for the dry valley lakes. DNA extracted from water samples collected during the 2009-2010 field season was used in nifJ amplification. PCR products were generated for this gene using 35 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 40 s. PCR for each gene was performed in triplicate and products were pooled and purified (Wizard SV Gel and PCR Clean-up System, Promega, WI). Purified PCR products were ligated into pGEM-T easy vector and transformed into TOP-10 E. coli (Life Technologies, CA). Positive transformants were randomly selected as templates for colony PCR with M13 vector primers and amplification was verified on a 1.2% agarose gel. The remaining PCR reaction was sent for single pass Sanger sequencing (Beckman Coulter Inc., MA). Sequences generated in this study have been deposited in the National Center for Biotechnology Information (NCBI) GenBank database under accession numbers KJ848331-KJ848439.

4.2.7. Phylogenetic analysis. Representative and related sequences from GenBank were aligned using MUSCLE in MEGA 5.2. Phylogenetic trees based on DNA were constructed using the maximum likelihood method and the best fit distance model predicted by the MEGA 5.2 software (Tamura et al. 2011). For nifJ sequences, a maximum-likelihood protein tree using the WAG+G model was generated because DNA sequences had at or less than 70% identity with NCBI BLAST results. Bootstrapping was used to estimate the node support with 500 replicate trees.

74

4.2.8. Statistical analysis. Graphic figures were generated using OriginPro 8.5.1 (OriginLab Corp, MA). Pearson correlation coefficients (R) and probability values (p) were calculated on log-transformed data to determine potential relationships between lake environmental factors and abundance of functional genes using the statistical software package in OriginPro (Vila-Costa et al. 2014, Zhong et al. 2013). All data was log-transformed to fit normal distribution.

4.3. RESULTS

4.3.1. Lake chemistry and biology. The water columns of Lakes Bonney, Fryxell, and Vanda exhibited high spatial stratification in lake chemistry (Fig. 4.1), as well as additional lake- specific physical and chemical characteristics (Fig. 4.1; Table 4.1). Above the chemocline, the + - water columns were generally highly oligotrophic. Inorganic nitrogen (NH4 and NO3 ) and - phosphorus levels increased sharply in the chemocline (Fig. 4.1), with the exception of NO3 in + WLB, Fryxell and Vanda (Fig. 4.1 C, E, G). The highest nitrogen (NH4 ) and phosphorus levels were measured in Lakes Fryxell and Vanda, while both lobes of Lake Bonney exhibited extremely unbalanced N:P ratios above and below the chemocline (Fig. 4.1E, G; Table 4.1). While irradiance within the water columns was <10% of incident in all lakes, variations in the transparent characteristics of the ice covers resulted in different depths for maximum light penetration (Table 4.1). While light reached 1% of incident directly under the ice cover in Lake Fryxell, light availability was at least 5-fold higher under the ice in Vanda compared with the other lakes, and did not reach 1% surface levels until lake depths below 60 m (Fig. 4.1F, H; Table 4.1). Last, MDV lakes also exhibited significant differences in maximum conductivity levels and temperature gradients (Table 4.1).

We monitored diversity of four major phytoplankton spectral classes and total Chl a using spectral Chl a fluorometry (Fig. 4.2). As expected, algal diversity and biomass levels exhibited distinct vertical stratification between the lakes. In agreement with past reports that have used direct chlorophyll extraction as an estimator of biomass, Lake Fryxell exhibited the highest Chl a levels while phytoplankton biomass in Lake Vanda exhibited the lowest Chl a of all four lakes (Fig. 4.2C, D). Water columns of ELB and WLB were dominated by phytoplankton in the ‘green’ group. A ‘mixed’ group was also detected in both lobes of Lake Bonney which correlated with the peak for maximum Chl a abundance (Fig. 4.2A, B). Based on

75 recent phylogenetic evidence, the ‘mixed’ phytoplankton group is likely comprised primarily of haptophytes and chrysophytes (Bielewicz et al. 2011, Kong et al. 2012b). In Lake Fryxell, shallow depths harbored a mixture of these three algal groups, while at the peak of Chl a abundance (8 – 10 m sampling depth), phytoplankton communities were dominated by the ‘mixed’ and ‘cryptophyte’ groups (Fig. 4.2C). Lake Vanda exhibited significant levels of cyanobacteria in the water column above the chemocline; however, overall Chl a levels were at or close to the minimum level of detection in this lake (Fig. 4.2D).

4.3.2. Spatial distribution of rRNA genes. Abundance of eukaryotic18S and bacterial 16S rRNA genes was measured to gain a general view of spatial distribution of bacterial vs. eukaryotic microbial communities across the four study lakes (Fig. 4.3). In ELB, bacterial and eukaryotic rRNA gene abundance followed similar trends, exhibiting minor peaks at 6 m and 15 m and declining at sampling depths below 22 m (Fig. 4.3A). In contrast, the abundance of 18S rRNA gene was less variable with depth in WLB vs. ELB, while 16S rRNA gene abundance was relatively high at all sampling depths (Fig. 4.3B). In Lake Fryxell, 18S and 16S rRNA gene abundance exhibited different spatial trends, with the eukaryotic community peaking near the chemocline (9 m) and the bacterial community exhibiting maximum 16S rRNA gene levels at depths below the chemocline (Fig. 4.3C). Depth profiles for bacterial and eukaryotic rRNA gene abundance in Lake Vanda were generally lower compared with the other MDV lakes, and exhibited maximum levels at the peak for Chl a (68 m sampling depth; Fig. 4.3D).

4.3.3. Autotrophic gene abundances. We extended earlier studies on functional gene diversity in MDV lakes (Kong et al. 2012 a, b) by monitoring gene abundance of five autotrophic gene markers (photoautotrophic: IA/B rbcL, ID rbcL, ID cryptophyte rbcL and chemolithoautotrophic: IA/B rbcL, cbbM, nifJ) in four MDV lakes (Figs. 4.4-4.6). In Lake Bonney, form ID rbcL dominated the photic zones of both basins, and exhibited maximum levels at 20 m and 15 m in ELB and WLB, respectively (Fig 4.4A, C). ELB exhibited cbbM levels that were at or below the level of detection as well as peaks in nifJ abundance at 6 m and 20 m. However, WLB exhibited >1500-fold increase in both cbbM and nifJ at depths below the aphotic zone (below 20 m; Fig. 4.4B, D). In Lake Fryxell, form IA/B levels peaked under the ice while form ID cryptophyte rbcL dominated the water column and peaked at 11 m (Fig. 4.5A) which agreed with spatial trends in phytoplankton biomass and spectral class distribution (Fig. 4.2C). Spatial trends in nifJ

76 and cbbM abundance in Lake Fryxell were similar to those observed in WLB, that is, abundance of both chemolithoautotrophic functional genes increased ~1000-fold at depths below the permanent chemocline (Figs. 4.4D and 4.5B). Abundance patterns for nifJ and cbbM also matched closely with bacterial 16S rRNA gene spatial trends in Lake Fryxell (Figs. 4.3C and 4.5B). Last, in the most unproductive MDV lake, Lake Vanda, we detected forms ID and IA/B rbcL (Fig. 4.6A), as well as nifJ (Fig. 4.6B). All three functional genes exhibited peaks within the permanent chemocline (68 m; Fig. 4.6A, B).

4.3.4. Autotrophic gene diversity. To validate the specificity of our qPCR primers as well as gain further insight into the functional gene diversity across MDV lakes, we amplified and sequenced several autotrophic genes at various sampling sites (Table 4.4; Figs. 4.7 – 4.12). As expected, form IA/B rbcL primers amplified three major groups of autotrophic organisms that harbor this isoform of RubisCO, chlorophytes, cyanobacteria and chemolithoautotrophic proteobacteria (Figs. 4.7 – 4.9). A total of 15 phylotypes were detected in the form IA/B rbcL clone libraries amplified from environmental DNA samples extracted from Lakes Fryxell and Vanda. The majority of form IA/B rbcL chlorophyte sequences were related to organisms of the class Trebouxiophyceae, some of which were closely related to OTUs from Lake Bonney (Fig. 4.7). Form IA/B rbcL sequences from Lake Fryxell were related to the genera Chlorella and Chlamydomonas, while sequences recovered from the 68 m sampling depth in Lake Vanda were all related to a small (< 5 µm diameter), non-flagellated green alga, Nannochloris bacillaris (GenBank accession no. AB383150; 92% similarity) (Fig. 4.7). Last, the 11 m sampling depth in Lake Fryxell harbored IA/B rbcL sequences that were related to a psychrophilic Antarctic marine prasinophyte, Pyramimonas tychotreta (GenBank accession no. L34778.1; 97% similarity) (Fig. 4.7) (Daugbjerg 2000). Several form IA/B rbcL OTUs related to cyanobacteria were identified in samples from Lakes Fryxell and Vanda, the majority of which were recovered from 7 m and 20 m sampling depths, respectively (Fig. 4.8). Nearly 50% of genes from Lake Fryxell were most closely related to cyanobacteria form IA/B rbcL sequences recovered from Lake Bonney (Kong et al., 2012b). Cyanobacterial sequences from Lake Vanda were related to sequences from ELB as well as an Oscillatorian, Phormidium ambiguum (GenBank accession no. FN813329.1; 99% similarity). While Nostocales sequences were detected in Lake Bonney in our earlier study (Kong et al. 2012b), none were recovered from Lake Frxyell or Vanda in the current study (Fig. 4.8). Last, form IA/B sequences related to a chemolithoautotrophic bacterium, were detected in

77 clone libraries generated from Lake Fryxell (Fig. 4.9). This OTU was related to a sulfur- oxidizing gamma-proteobacteria, Thiomicrospira sp. (86% similarity to T. kuenenii, GenBank accession no. DQ272531.1), while our earlier study reported that chemolithoautotrophic form IA/B rbcL sequences recovered from 15 m and 20 m sampling depths in WLB were most closely related to a beta-proteobacteria, Thiobacillus denitrificans, as well as a chemolithoautotrophic bacterial endosymbiont (Fig. 4.9) (Kong et al. 2012b). We previously reported that form ID rbcL is highly abundant in Lake Bonney and that haptophytes and stramenopiles harboring this RubisCO isoform dominate the photic zone of both lobes of this lake (Kong et al. 2012b). In contrast with Lake Bonney, form ID rbcL gene abundance was often at or below the level of detection in Lake Fryxell and Vanda (Figs. 4.5 and 4.6), so we amplified and sequenced this gene from a single sampling depth in Lake Vanda at the peak of Chl a abundance (i.e., 68 m). We also sequenced a 20 m sample from ELB as it represented peak form ID rbcL abundance in this lake (Fig. 4.4A); however, in an earlier study the form ID rbcL gene abundance peak was detected at a shallower sampling depth of 15 m (Kong et al. 2012b). All form ID rbcL sequences recovered from Lake Vanda were related to stramenopiles, Heterococcus sp. (99% similarity to H. caespitosus, GenBank accession no. JX681212) and Chlorellidium sp. (99% similarity to C. tetrabotrys, GenBank accession no. AJ580947). These OTUs were distinct from those recovered from Lake Bonney which were related to a haptophyte, Isochrysis galbana, and a stramenopile, Nannochloropsis limnetica (Kong et al. 2012b) (Fig. 4.10). We amplified and sequenced the form II RubisCO gene, cbbM, from Lake Fryxell (Fig. 4.11). Form II RubisCO genes related to Thiobacillus thioparus strain DSM505 (GenBank accession no. EU746412.1; 82% similarity) as well as cbbM sequences recovered from the west lobe of Lake Bonney (Kong et al. 2012a) dominated the clone library (Fig. 4.11). For the rTCA gene, nifJ, we employed a more intensive sampling strategy as this gene had not been previously sequenced in the MDV lakes (Table 4.4). Due to the low sequence similarity (<70% similarity) at the nucleotide level between the MDV nifJ sequences and those in the NCBI database, representative sequences were translated and used to generate a protein based phylogenetic tree (Fig. 4.12). We recovered nifJ sequences that were related to a diverse lineage of bacteria. Shallow lake depths (6 m) of ELB and Lake Fryxell harbored nifJ sequences that were distributed throughout the tree, the majority of which were related to Niastella

78 koreensis (GenBank accession no. YP005007034.1), Rhodovulum sp. PH10 (GenBank accession no. WP008386030.1), and Cesiribacter andamanensis (GenBank accession no. WP009194953.1). The majority of nifJ sequences from Lake Fryxell 7 m and 9 m samples were also related to C. andamanensis. Genes from mid to deep depths in ELB (18 and 30 m) and WLB (20 m and 30 m) of Lake Bonney were related to a sulfur-oxidizing symbiont (GenBank accession no. WP005964713.1), Aeromonas sp. (GenBank accession no. WP005325732.1, WP021231117.1), and Chlorobium sp. (GenBank accession no. YP001130920.1, YP380024.1). The majority of genes from Lake Vanda sampling depths were distantly related to N. koreensis, a sulfur-oxidizing symbiont, and Methylobacter tundripaludum (GenBank accession no. WP006892638.1).

4.3.5. Structure of autotrophic communities in relation to abiotic variables. The potential relationships between functional gene abundance and lake physicochemistry were tested by linear pairwise correlations (Table 4.5). Overall, form IA/B rbcL as well as form ID cryptophyte rbcL gene abundances correlated positively with light availability (i.e., PAR), while the chemolithoautotrophic carbon fixation genes (cbbM, nifJ) were negatively correlated with PAR (Table 4.5). An exception to this trend is form IA/B rbcL levels in Lake Fryxell which exhibited a negative correlation with PAR which was likely due to the presence of chemolithoautotrophic bacteria harboring form IA/B rbcL in this lake (Fig. 4.9). Gene abundance of the three form I rbcL genes (form IA/B, form ID, and form ID cryptophyte rbcL) were generally negatively - + correlated or had no significant correlation with nitrogen (NO3 , NH4 ) and/or phosphorus (SRP) availability in ELB and WLB. Abundance of nifJ and cbbM were generally positively correlated with nutrient levels (Table 4.5). In ELB and WLB, gene abundance of form IA/B rbcL and the form ID cryptophyte rbcL were negatively correlated with conductivity levels (Table 4.5). Chemolithoautotrophic gene abundance (nifJ, cbbM) in WLB and Lake Fryxell exhibited positive correlations with both conductivity and DIC (Table 4.5). Few significant correlations between gene abundance and environmental parameters were observed in Lake Vanda, with the exception of abundance of form ID or IA/B rbcL and PAR (Table 4.5).

4.4. DISCUSSION

Antarctic lakes in the McMurdo Dry Valleys harbor stratified microbial communities which are isolated from the environment by permanent ice caps. Past studies have shown that

79 microbial productivity is limited by ‘bottom up’ controls including light availability and major nutrients (Moorhead et al. 1999; Priscu et al. 1999). Our study extends this earlier work on carbon cycling in the dry valley planktonic communities by targeting the diversity and spatial dynamics of key carbon fixation genes across four chemically distinct MDV lakes. As expected, all lakes exhibited strong stratification of carbon fixation communities. Measurements of phytoplankton biomass (Chl a) and rbcL gene abundance both supported past reports that Lake Fryxell is the most productive lake, while Lake Vanda is very unproductive owing to extreme oligotrophy. Form ID rbcL harboring phytoplankton are major primary producers in both lobes of Lake Bonney (haptophytes, stramenopiles) and Lake Fryxell (cryptophytes) (Kong et al. 2012b), while form IA/B rbcL (cyanobacteria, chlorophytes) dominates Lake Vanda. Evidence for form II RubisCO was detected in the aphotic zones of WLB and Lake Fryxell, indicating that these two lakes have additional energy sources to support chemolithoautotrophic organisms. Environmental drivers differentially impacted spatial distribution of MDV autotrophic communities. Vertical trends in the dominant form of RubisCO (i.e., form ID rbcL in Lake Bonney and cryptophyte rbcL in Lake Fryxell) did not significantly correlate with any of the physicochemical parameters measured. However, gene numbers of less dominant RubisCO isoforms in both lobes of Lake Bonney (i.e., form IA/B rbcL and ID cryptophyte rbcL) exhibited positive correlations with PAR. Spatial distribution of the dark carbon fixation genes, cbbM and nifJ was negatively correlated with PAR, reflecting the capacity for light-independent carbon fixation within organisms harboring these genes. Gene abundance of both cbbM and nifJ correlated positively with nutrients and DIC, indicating that nutrient and inorganic carbon availability appear to be positive drivers of the distribution of organisms harboring dark carbon fixation genes. Form I RubisCOs (isoforms IA, IB, IC and ID) catalyze the predominant light dependent mechanism of inorganic carbon fixation (Tabita et al. 2008). While the diversity of form I rbcL harboring organisms varied across the MDV lakes, most sequences were related to eukaryotic algae and therefore corresponded to the depths of peak Chl a as well as 18S rRNA gene abundance. In addition, depth-dependent trends in the bacterial 16S rRNA gene closely matched that of 18S rRNA gene in ELB and Lake Vanda, suggesting that bacterial abundance and production in these lakes is reliant on photosynthate derived from light dependent carbon fixation. This hypothesis is supported by earlier work by Koob and Leister (1972) who also

80 observed correlations between the vertical distribution of bacterial and phytoplankton populations in the east lobe of Lake Bonney. In contrast, vertical trends in 18S and 16S rRNA genes were uncoupled from each other in WLB and Lake Fryxell, suggesting that bacteria populations in these lakes are influenced by grazing activity of heterotrophic or mixotrophic (i.e., capable of photosynthesis and phagotrophy) protists (Roberts and Laybourn-Parry 1999) or may utilize alternative sources of carbon and/or energy. We recently reported that phytoplankton harboring form ID rbcL were dominated by sequences related to a haptophyte (Isochrysis sp.) and a stramenopile (Nannochloropsis sp.) in Lake Bonney (Kong et al., 2012b). While these phytoplankton were predicted to play an important role in primary production in Lake Bonney, it appears that cryptophytes related to a marine psychrophile, Gemingera cryophila, are the dominant phytoplankton in Lake Fryxell, which supports work by Laybourn-Parry and colleagues (1997). Lake Fryxell harbored chlorophytes (Chlorella, Chlamydomonas spp.) with form IA/B rbcL, as previously reported in Lake Bonney (Kong et al. 2012b), as well as a prasinophyte related to Pyramimonas tychotreta (Fig. 4.7). Prasinophytes appear to be wide-spread among Arctic and Antarctic marine environments (Daugbjerg 2000, Daugbjerg and Moestrup 1992, 1993, Lovejoy et al. 2007, McFadden et al. 1982). Earlier studies based on microscopy also confirmed that Pyramimonas resides at a depth of 9.5 m in Lake Fryxell (Roberts et al. 2004b, Seaburg et al. 1979), and Laybourn-Parry and Pearce (2007) reported that Pyramimonas spp. are dominant phytoflagellates in Antarctic lakes with salinities between 15-20 ‰. This may account for the absence of Pyramimonas in the hypersaline water columns of Lake Bonney and Vanda. Pyramimonas populations have been shown to graze on bacteria in the Taylor Dry Valley and Vestfold Hill lakes (Bell and Laybourn-Parry 1999, 2003, Marshall and Laybourn-Parry 2002, Roberts and Laybourn-Parry 1999), indicating that this organism is mixotrophic. It has been documented that mixotrophy is an important adaptive advantage in Antarctic lakes (Laybourn- Parry et al. 2005, Laybourn-Parry and Pearce 2007, Laybourn-Parry et al. 2000). In agreement with past reports (Vincent and Vincent 1982), Lake Vanda was dominated by cyanobacteria related to the filamentous species Phormidium ambiguum which made up >60% of the 20 m community, as well as a small non-flagellated chlorophyte related to Chlorella. However, the deep chlorophyll maximum in Lake Vanda was dominated by chlorophytes related to a marine chlorophyte, Nannochloris bacillaris. Nannochloris has been identified in other low temperature

81 ecosystems including the Taylor Dry Valley Lake Meirs (Vincent and James 1996), but among the lakes in our study, is unique to Lake Vanda. An isolate of the marine N. bacillaris exhibited high tolerance to a broad range of salinity levels typical of other euryhaline microalgae, such as the Lake Bonney chlorophyte C. raudensis UWO241 (Brown 1982, Dolhi et al. 2013, Pocock et al. 2011). Last, form ID rbcL harboring organisms related to the stramenopile, Heterococcus sp., were also unique to Lake Vanda. The genus is widespread in cold environments (Andreoli et al. 2000, Darling et al. 1987), and includes an Antarctic soil isolate which exhibited halotolerance and production of osmotica in response to growth under high salinity (Fujii et al. 1999). Our evidence indicates that Lake Fryxell and WLB harbor bacterial communities capable of chemolithoautotrophy. Steep chemical gradients in stratified lakes allow chemolithoautotrophic communities access to both electron donors and electron acceptors (Camacho et al. 2001, Jorgensen and Postgate 1982, Shively et al. 1998). The water column of Lake Fryxell maintains an extensive anoxic zone which is highly sulfidic due to dissimilatory sulfate reduction by sulfate reducing bacteria (Karr et al. 2005). Maximum numbers of cultivatable sulfur-oxidizing bacteria (SOB) were recovered at 9.5 m in Lake Fryxell, which represents the midpoint in the chemocline for oxygen and sulfide levels (Sattley and Madigan 2006). Phylogenetic analyses of 16S rRNA indicated that Lake Fryxell SOB isolates were close relatives of Thiobacillus thioparus (Sattley and Madigan 2006). Our current study supports the cultivation-based work by providing evidence for spatial distribution of the Lake Fryxell SOB community. Gene abundance of CBB (cbbM) and rTCA (nifJ) enzymes increased dramatically at sample depths below 9 m by 5,000- and 30,000-fold, respectively (Fig. 4.4), indicating that the chemolithoautotrophic communities in Lake Fryxell reside at layers where electron donors (sulfide) and acceptors (oxygen) are available (Sattley and Madigan 2006). Spatial trends in these carbon fixation genes also coincided with depths of maximum gene abundance of bacterial 16S rRNA gene. Dark carbon fixation in the oxic-anoxic interface of Lake Fryxell during the summer was estimated to contribute up to 40 % as much as PPR (Vick and Priscu 2012). One consideration is the high levels of both chemolithoautotrophic genes at depths below 10.5 m at which point the water column becomes anoxic. Fulton et al. (2004) have reported that bacteria from Lake Fryxell have the potential to use humic acids as terminal electron acceptors in the region of the oxycline. Alternative electron acceptors such as nitrate or ferric iron are unavailable for anaerobic respiration in the anoxic zone of Lake Fryxell (Green et al. 1989).

82

In WLB, abundance of both chemolithoautotrophic genes (cbbM, nifJ) exhibited similar vertical trends as was observed in Lake Fryxell. Three previous reports detected the presence of either phylogenetic (16S rRNA) or functional (cbbM, IA/B rbcL) genes related to SOB in WLB (Kong et al. 2012a, b, Vick-Majors et al. 2014). These combined data suggest that WLB harbors an SOB community that balances availability of electron donors and acceptors in the oxic-anoxic interface of the WLB water column. Bacterial 16S rRNA sequences related to SOB were also recovered from the outflow of Blood Falls which flows into WLB during the summer (Mikucki and Priscu 2007). Unlike Lake Fryxell, sulfide concentrations are very low in WLB and are therefore unavailable to the chemolithoautotrophic community as an energy source (Lee et al.2004a). However, the water columns of both lobes of Lake Bonney exhibit elevated levels of dimethylated-sulfur species (Lee et al. 2004a). Specifically, both lobes exhibit high levels of particulate and dissolved dimethylsulfoniopropionate (DMSPP and DMSPD, respectively); however, WLB exhibits a build-up of dimethyl sulfide (DMS), while ELB exhibits high levels of dimethyl sulfoxide (DMSO). Trends in DMSPP in both lobes corresponded with depths of maximum Chl a levels (Lee et al. 2004b), as well as transcript levels of rbcL ID (Kong et al. 2012b) and a major photochemistry gene, psbA (Kong et al. 2014). DMSP production occurs in many nonpolar and polar marine phytoplankton (e.g., , haptophytes, chrysophytes) where it functions in either cryo- or osmo-protection (Trevena et al. 2000). We suggest that a haptophyte related to Isochrysis galbana, is likely the major producer of DMSP production in Lake Bonney and may serve as an adaptation to high salinity low temperatures. Sequences of this organism dominate ID rbcL clone libraries derived from DNA or mRNA recovered from the chemocline of both lobes (Kong et al. 2012b). Moreover, our current study as well as a recent report which compared microbial diversity in WLB versus Lake Fryxell (Vick-Majors et al. 2014) indicate that organisms related to Isochrysis sp. were absent from MDV lakes (i.e., Lake Fryxell and Vanda) which do not accumulate DMSP (Lee et al. 2004a). In marine systems, a diverse group of bacteria possess the ability to utilize DMSP and its breakdown products (DMS, DMSO) for sources of carbon and sulfur, as well as alternative electron donors (DMS) or acceptors (DMSO) (Raina et al. 2010). Lee et al. (2004) suggested that differences in redox chemistry between ELB and WLB could influence biogeochemical cycling of dimethylated sulfur in the two lobes of Lake Bonney. Specifically, utilization of DMSO as an alternative electron acceptor was predicted to be favorable in the WLB anoxic zone which would account

83 for the build-up of DMS in this lobe (Lee et al. 2004a). Thus, it seems plausible that the chemolithoautotrophic community in WLB utilizes dimethylated compounds for carbon and/or energy metabolism. Our study indicates that spatial distribution of autotrophic organisms of the ice-covered MDV lakes depends not only on physicochemical factors, but is also influenced by the presence of other organisms. Dominant photosynthetic eukaryotes exploit availability of light and nutrients, while other autotrophic organisms are more strongly driven by light alone. Microbial eukaryotes are likely to play major roles in the spatial distribution of bacterial communities via production of photosynthate, grazing activity, or synthesis of biogenic sulfur compounds. Last, interactions between microorganisms and their environment, as well as with each other, are likely to be influenced by climate-related events, such as increased glacial stream load and thinning ice covers, which are predicted to intensify in these delicate aquatic ecosystems in the coming decades.

4.5. ACKNOWLEDGEMENTS The authors thank the McMurdo LTER limnology team, Ratheon Polar Services and PHI helicopters for logistical assistance in the field. We thank A. Kiss and the Center for Bioinformatics and Functional Genomics at Miami University for assistance with qPCR and N. Ketchum for assistance with clone library construction. This work was supported by NSF Office of Polar Programs Grant OPP-1056396.

84

Table 4.1. General physical and chemical characteristics for study sites. For N:P ratios, the integrated mean (standard error) for the mixolimnion (mixo) and the chemocline and monolimnion (chemo+mono) are reported.

Abiotic Parameter East Bonney West Bonney Fryxell Vanda GPS Coordinates 77o44’S 77o43’S 77o37’S 77o32’S 162o10’E 162o17’E 163o11’E 161o33’E Area (km2) 3.5 1.3 7.8 5.2 Max Depth 37 m 40 m 20 m 80 m Ice thickness 4.0-4.5 m 3.8-4.1 m 3.3-4.5 m 3.1-4.0 m % Transmission of 7.0 % 7.0 % 1.5 % 18.6 % PAR Max Salinity (PSU) 150 125 6.2 100 Temperature (°C) -2.8 to 7.9 °C -5.4 to 3.2 °C 0 to 3.2 °C 4 to 23.5 °C Depth of chemocline 20-25 m 17-20 m 9 m 68 m Molar N:P -mixo 316 (98) 164 (37) 2.41 (0.60) 13.93 (2.2) -chemo+mono 694 (101) 670 (94) 11.0 (0.57) 135.6 (14.1)

85

Table 4.2. Primer set sequences with annealing temperatures for qPCR and/or clone library construction.

Gene Primer Primer sequence Amplicon Annealing Refs. direction (5’→ 3’) length (bp) temp. (°C) 16S F CCTACGGGAGG 200 55 Muyzer et al. CAGCAG 1993 16S R ATTACCGCGGC 200 55 Muyzer et al. TGCTGG 1993 18S F AAGGAAGGCAG 500 50 Zhu et al. CAGGCG 2005 18S R CACCAGACTTG 500 50 Zhu et al. CCCTCYAAT 2005 Form IA/B F TCIGCITGRAACT 615 50 Paul et al. rbcL AYGGTCG 2000 Form IA/B R GGCATRTGCCAI 615 50 Paul et al. rbcL ACRTGRAT 2000 Form ID F GATGATGARAA 554 52 Paul et al. rbcL YATTAACCT 2000 Form ID R ATTTGDCCACA 554 52 Paul et al. rbcL GTGDATACCA 2000 Form ID F GCGTTTCTTATT 500 52 Kong et al. crypto rbcL CGGTATGGA 2012b Form ID R GGCCACAGTGA 500 52 Kong et al. crypto rbcL ATACCACCT 2012b cbbM F TTCTGGCTGGGB 350 58 Campbell & GGHGAYTTY Cary 2004 ATYAARAAYGA CGA cbbM R CCGTGRCCRGC 350 58 Campbell & VCGRTGG Cary 2004 TARTG nifJ F CIGGITGYGGIGA 475 55 Campbell & AACICC Cary 2004 nifJ R CCIATRTCRTAIG 475 55 Campbell & CCCAICCRTC Cary 2004

86

Table 4.3. Checklist for MiQE Guidelines

Item to check Importance Completed Comments/Location (Y/N) Expermimental Design Assay carried out by the core or D Y Investigator’s lab investigator’s laboratory? Acknowledgement of author’s D Y 18S, 16S: AGT contributions Form IA/B, ID, and ID cryptophyte rbcL, cbbM, nifJ: JMD Sample Description E Y Methods Volume/mass of sample processed D Y Methods Processing procedure E Y Methods If frozen, how and how quickly? E Y Methods If fixed, with what and how quickly? E Y N/A Sample storage conditions and E Y -80 °C: filters stored 1 month duration Nucleic acid extraction Procedure and/or instrumentation E Y Methods Name of kit and details of any E Y MP Biomedicals (Solon, OH) modifications FastDNA Spin Kit for Soil (116560-200) , Methods Source of additional reagents used D N/A N/A Details of DNase or RNase treatment E N/A N/A Contamination assessment (DNA or E N/A N/A RNA) Nucleic acid quantification E Y Instrument and method E Y NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies, Inc, Rockland, DE) Inhibition testing (Cq dilutions, spike, E Y Cq dilutions or other) Reverse transcription N/A N/A N/A qPCR target information Gene symbol E Y Introduction Sequence accession number E Y Methods Amplicon length E Y Methods In silico specificity screen (BLAST, E Y Methods and so on) qPCR oligonucleotides Primer sequences E Y Methods Location and identity of any E N/A N/A modifications Manufacturer of oligonucleotides D Y Integrated DNA Technologies (IDT) Purification method D Y Standard desalting qPCR protocol

87

Complete reaction conditions E Y Methods Reaction volume and amount of E Y Methods cDNA/DNA Primer, (probe), Mg2+, and dNTP E Y 0.6 µM each primer, 6 mM concentrations Mg2+, 0.4 mM of each dNTP, Polymerase identity and concentration E Y 50 U/mL iTaq DNA polymerase Buffer/kit identity and manufacturer E Y iQ SYBR Green Supermix (170- 8882) Bio-rad (Hercules, CA) Exact chemical composition of the D Y 40 mM Tris-HCl, pH 8.4, 100 buffer mM KCl Additives (SYBR green I, DMSO, E Y SYBR green I etc.) Complete thermocycling parameters E Y Methods Reaction setup (manual/robotic) D Y Manual Manufacturer of qPCR instrument E Y Methods qPCR validation Specificity (gel, sequence, melt or E Y Available upon request digest) For SYBR Green I, Cq of the NTC E Y Cq of NTC below that of lowest standard Calibration curves with slope and y E Y Available upon request intercept PCR efficiency calculated from slope E Y Efficiency ranged from 87-113% R2 of calibration curve E Y R2 ranged from 0.97-0.99 Evidence for LOD E Y Cq could not be determined for <102 copies Data analysis Method of Cq determination E Y Methods Results for NTCs E Y Available upon request Repeatability (intraassay variation) E Y Two technical replicates Statisitical methods for results E Y Methods significance

88

Table 4.4. Autotrophic gene clone libraries and total clones sequenced

Fryxell rbcL form I A/B rbcL form ID cbbM nifJ 6m 0 0 0 48 7m 15 0 0 48 9m 0 0 0 72 11m 15 0 18 0 14m 15 0 0 0 Total 45 0 18 168 East lobe Bonney 6m 0 0 0 39 18m 0 0 0 24 20m 0 19 0 0 30m 0 0 0 17 Total 0 19 0 80 West lobe Bonney 20m 0 0 0 15 30m 0 0 0 16 Total 0 0 0 31 Vanda 20m 15 0 0 24 35m 0 0 0 24 45m 0 0 0 24 60m 0 0 0 24 68m 15 17 0 0 Total 30 17 0 96

89

Table 4.5. Pearson’s correlation coefficients (R) for the relationship between lake biological and environmental parameters and abundance of functional genes in four MDV lakes. All data was log-transformed to fit normal distribution. *, significant correlation (p<0.05); **, significant correlation (p<0.01); N/A, correlation not shown as majority of measurements for gene abundance were below limit of detection.

+ - PAR Temp Cond NH4 NO3 SRP DIC east lobe Lake Bonney (ELB) IA/B rbcL 0.873** -0.466** -0.875** -0.826** -0.880** -0.765** -0.675** ID Crypto rbcL 0.892** -0.427* -0.821** -0.878** -0.849** -0.905** -0.668** IDrbcL rbcL rb rbcL 0.050 0.323 0.093 0.019 0.068 -0.640* 0.259 cbbM N/A nifJ 0.222 0.218 0.032 0.090 0.019 -0.338 0.318 west lobe Lake Bonney (WLB) IA/B rbcL 0.625** 0.472* -0.652* -0.791** 0.581** -0.754** -0.704** ID Crypto rbcL 0.831** 0.574** -0.795** -0.816** 0.848** -0.895** -0.712** IDrbcL rbcL -0.210 -0.136 0.591** 0.439 0.294 0.290 0.665** cbbM -0.685* -0.469* 0.949** 0.946** -0.410 0.843** 0.922** nifJ -0.733** -0.466* 0.921** 0.896** -0.543* 0.853** 0.865** Lake Fryxell IA/B rbcL -0.676** 0.119 0.335 0.765** 0.707** 0.745** 0.409 ID Crypto rbcL 0.046 0.217 0.018 0.057 -0.182 -0.057 0.035 ID rbcL N/A cbbM -0.784** 0.188 0.601** 0.907** 0.731** 0.836** 0.647** nifJ -0.804** 0.212 0.700** 0.903** 0.747** 0.848** 0.704** Lake Vanda IA/B rbcL 0.740** -0.059 -0.388 -0.388 -0.256 -0.430 -0.483* ID Crypto rbcL N/A ID rbcL 0.606** -0.140 -0.266 -0.266 -0.322 -0.336 -0.385 cbbM 0.440 0.338 -0.265 -0.500 -0.221 -0.472* -0.129 nifJ 0.046 0.343 0.368 -0.064 0.381 0.360 0.431

90

SRP PAR SRP PAR 0 1 2 0 10 20 1 2 0 10 20 0 0 A - ELB B - ELB C - WLB D - WLB SRP NH + 4 10 10 - NO3

20 20

Lake(m) depth Lake(m) depth 30 30

40 40 0 100 200 300 0 100 200 300 + - NH +, NO - NH4 , NO3 4 3

SRP PAR SRP PAR 0 50 100 0 10 20 0 10 20 0 100 200 0 0 E - FRX F - FRX G - VAN H - VAN 10

5 20

30

10 40

50

Lakedepth (m) Lake(m) depth 15 60

70

20 80 0 250 500 750 0 250 1500 1750 NH +, NO - + - 4 3 NH4 , NO3

Figure 4.1. Depth profiles of physical and chemical characteristics of east lobe Bonney (ELB; A, B), west lobe Bonney (WLB; C, D), Frxyell (FRX; E, F), and Vanda (VAN; G, H). Units: -2 -1 - photosynthetically active radiation (PAR; µmol m s ), soluble reactive phosphorus (SRP), NO3 + and NH4 (µM). Nutrient data was kindly provided by the McMurdo LTER.

91

Green Cyanobacteria Mixed Cryptophytes Chl a (g/L) ___ Chl a (g/L) ___ 0 10 20 0 10 20

5 5 A - ELB B - WLB

10 10

Depth (m) Depth Depth (m) Depth 15 15

20 20

0 25 50 75 100 0 25 50 75 100 Relative Abundance (%) Relative Abundance (%) ___ Chl a (g/L) Chl a ( g/L) ___ 0 10 20 30 40 0 5 10 4 C - FRX D - VAN 15

6

30

8

Depth (m) Depth 45 Depth (m) Depth

10 60

0 25 50 75 100 0 25 50 75 100 Relative Abundance (%) Relative Abundance (%)

Figure 4.2. Depth profiles for phytoplankton diversity and biomass within east lobe Bonney (ELB; A), west lobe Bonney (WLB; B), Fryxell (FRX; C), and Vanda (VAN; D). Distribution of major spectral algal classes and total chlorophyll a (Chl a) were monitored using a diving spectral fluorometer (bbe FluoroProbe) (n=1). Green: Chlorophyta and Euglenophyta, Mixed: Haptophyta, Bacillariophyceae, Chrysophyceae and Dinophyceae.

92

0 0 ELB-A WLB-B 5 5

10 10

15 15

20 20

25 25 Lake depth (m) depth Lake Lake depth (m) depth Lake 16S rRNA gene 30 18S rRNA gene 30

102 104 106 108 1010 102 104 106 108 1010 0 0 2 FRX-C VAN-D 10

4 20

6 30

8 40

10 50

12 60 (m) depth Lake Lake depth (m) depth Lake

14 70

102 104 106 108 1010 102 104 106 108 1010 -1 -1 16S or 18S DNA (copies L ) 16S or 18S DNA (copies L ) Figure 4.3. Depth profiles of 16S and 18S rRNA genes in east lobe Bonney (A), west lobe Bonney (B), Fryxell (C), and Vanda (D). rRNA gene abundance was quantified using quantitative PCR as described in the methods (n=3 with exception of Lake Fryxell 8-9 m where n=2).

93

0 0 A - ELB B - ELB

10 10

20 20

30 30 (m) depth Lake Lake depth (m) depth Lake IAB rbcL ID rbcL cbbM ID Crypto rbcL nifJ 40 40 102 104 106 108 1010 102 104 106 108 1010 0 0 C - WLB D - WLB

10 10

20 20

30 30 (m) depth Lake Lake depth Lake (m)

40 40 102 104 106 108 1010 102 104 106 108 1010 -1 rbcL DNA (copies L-1) cbbM or nifJ DNA (copies L )

Figure 4.4. East lobe (A, B) and West lobe (C, D) Bonney Lake depth profiles of autotrophic functional gene abundance including three types of form I RubisCO large subunit (rbcL) (A, C), form II RubisCO (cbbM), and pyruvate:ferredoxin oxidoreductase (nifJ) (B, D). Gene abundance was quantified using quantitative PCR as described in the methods (n=3).

94

0 0 A - FRX B - FRX

5 5

10 10

Lake depth (m) depth Lake Lake depth (m) depth Lake 15 15 IAB rbcL ID rbcL cbbM ID Crypto rbcL nifJ 20 20 102 104 106 108 1010 102 104 106 108 1010 rbcL DNA (copies L-1) cbbM or nifJ DNA (copies L-1)

Figure 4.5. Lake Fryxell depth profile of autotrophic functional gene abundance including three types of form I RubisCO large subunit (rbcL) (A), form II RubisCO (cbbM), and pyruvate:ferredoxin oxidoreductase (nifJ) (B). Gene abundance was quantified using quantitative PCR as described in the methods (n=3).

95

0 0 A - VAN IAB rbcL B - VAN cbbM 10 ID rbcL nifJ 10 ID Crypto rbcL 20 20

30 30

40 40

50 50 Lakedepth (m) 60 60 (m) Lake depth

70 70

80 80 102 104 106 108 1010 102 104 106 108 1010 rbcL DNA (copies L-1) cbbM or nifJ DNA (copies L-1)

Figure 4.6. Lake Vanda depth profile of autotrophic functional gene abundance including three types of form I RubisCO large subunit (rbcL) (A), form II RubisCO (cbbM), and pyruvate:ferredoxin oxidoreductase (nifJ) (B). Gene abundance was quantified using quantitative PCR as described in the methods (n=3).

96

Figure 4.7. Maximum likelihood phylogenetic tree of representative form IA/B rbcL, chlorophyte subgroup, DNA sequences retrieved from Lakes Bonney, Fryxell, and Vanda. Bold, underlined genes indicate sequences from this study. Clones were named by lake, depth, gene, clone number, and year sample was collected (C=2012). Bold genes indicate sequences from previous study (Kong et al., 2012b) retrieved from GenBank. Bar shows 0.05 substitutions per nucleotide position. Bootstrap replicates = 500.

97

Figure 4.8. Maximum likelihood phylogenetic tree of representative form IA/B rbcL, cyanobacteria subgroup, DNA sequences retrieved from Lakes Bonney, Fryxell, and Vanda. Bold, underlined genes indicate sequences from this study. Clones were named by lake, depth, gene, clone number, and year sample was collected (C=2012). Bold genes indicate sequences from previous study (Kong et al., 2012b) retrieved from GenBank. Bar shows 0.05 substitutions per nucleotide position. Bootstrap replicates = 500.

98

Figure 4.9. Maximum likelihood phylogenetic tree of representative form IA/B rbcL, chemotroph subgroup, DNA sequences retrieved from Lakes Bonney, Fryxell, and Vanda. Bold, underlined genes indicate sequences from this study. Clones were named by lake, depth, gene, clone number, and year sample was collected (C=2012). Bold genes indicate sequences from previous study (Kong et al., 2012b) retrieved from GenBank. Bar shows 0.05 substitutions per nucleotide position. Bootstrap replicates = 500.

99

Figure 4.10. Tamura 3-parameter phylogenetic tree of representative form ID rbcL DNA sequences retrieved from Lakes Bonney, Fryxell, and Vanda. Bold, underlined genes indicate sequences from this study. Clones were named by lake, depth, gene, clone number, and year sample was collected (C=2012). Bold genes indicate sequences from previous study (Kong et al., 2012b) retrieved from GenBank. Bar shows 0.1 substitutions per nucleotide position. Bootstrap replicates = 500

100

Figure 4.11. Maximum likelihood phylogenetic tree of representative form II rbcL DNA sequences retrieved from Lakes Bonney, Fryxell, and Vanda. Bold, underlined genes indicate sequences from this study. Clones were named by lake, depth, gene, clone number, and year sample was collected (C=2012). Bold genes indicate sequences from previous study (Kong et al., 2012a) retrieved from GenBank. Bar shows 0.1 substitutions per nucleotide position. Bootstrap replicates = 500.

101

Van45mNifJ21B Van60mNifJ6 100 ELB6mnifJ4B 100 Van20mNifJ9B Van35mNifJ20B Frx6mnifJ22A YP 005007034.1 pyruvate ferredoxin/flavodoxin oxidoreductase Niastella koreensis ELB18mnifJ18B Van45mNifJ12B 100 Van35mNifJ24B Van60mNifJ22B WP 005964713.1 pyruvate-flavodoxin oxidoreductase sulfur-oxidizing symbiont WLB 20 NifJ 10A 100 ELB 18m nifJ 14B ELB 30m nifJ 23B YP 113272.1 pyruvate ferredoxin/flavodoxin oxidoreductase Methylococcus capsulatus 83 WP 008177414.1 pyruvate-flavodoxin oxidoreductase Marinobacter manganoxydans WP 010595626.1 pyruvate-flavodoxin oxidoreductase Rhodococcus sp. P14 WP 019602074.1 pyruvate-flavodoxin oxidoreductase Teredinibacter turnerae 53 WP 005325732.1 pyruvate-flavodoxin oxidoreductase Aeromonas media WP 021231117.1 pyruvate-flavodoxin oxidoreductase Aeromonas veronii 95 ELB 30m nifJ 21B 100 ELB18mnifJ10B ELB18mnifJ23B YP 130193.1 pyruvate oxidoreductase Photobacterium profundum SS9 Frx 6m nifJ 21A 100 Frx 6m nifJ 3A Frx6mnifJ6A 73 YP 076993.1 pyruvate flavodoxin dehydrogenase Symbiobacterium thermophilum IAM 14863 WP 006892638.1 pyruvate-flavodoxin oxidoreductase Methylobacter tundripaludum 100 ELB18mnifJ15B Van45mNifJ16B Frx9mnifJ 20A Frx6mnifJ6B Van 20m NifJ 3B Van35mNifJ6B 99 Van45mNifJ22B Van60mNifJ10B 95 YP 003442261.1 pyruvate ferredoxin/flavodoxin oxidoreductase Allochromatium vinosum YP 006415780.1 pyruvate:ferredoxin (flavodoxin) oxidoreductase homodimeric Thiocystis violascens 100 ELB6mnifJ3B YP 001817495.1 pyruvate flavodoxin/ferredoxin oxidoreductase Opitutus terrae WP 008386030.1 pyruvate-flavodoxin oxidoreductase Rhodovulum sp. PH10 52 Frx6mnifJ14A 100 ELB6mnifJ6B ELB6mnifJ8B YP 008547068.1 pyruvate:ferredoxin oxidoreductase Sulfuricella denitrificans 100 YP 001130920.1 pyruvate flavodoxin/ferredoxin oxidoreductase Chlorobium phaeovibrioides YP 380024.1 pyruvate:ferredoxin oxidoreductase Chlorobium chlorochromatii 90 ELB18mnifJ13B 62 ELB18mnifJ17B 90 ELB 30m nifJ 16B WLB 20 NifJ 9A 100 WLB 30m NifJ 9A YP 003495806.1 pyruvate:ferredoxin oxidoreductase Deferribacter desulfuricans YP 003862487.1 ferredoxin oxidoreductase Maribacter sp. HTCC2170 WP 009194953.1 porA Cesiribacter andamanensis 100 Van 20m NifJ 23B 52 Frx 9m nifJ 17B Frx 9m nifJ 32B 98 ELB 6m nifJ 20B Frx 6m nifJ 21B Frx 7m nifJ 26B Frx 9m nifJ 9A 100 Frx 6m nifJ 1A 100 ETI97861.1 Veillonella dispar WP 009351094.1 Veillonella sp. oral taxon 158

0.2

102

Figure 4.12. Maximum likelihood phylogenetic tree of representative nifJ protein sequences retrieved from Lakes Bonney, Fryxell, and Vanda. Bold, underlined genes indicate sequences from this study. Clones were named by lake, depth, gene, clone number, and year sample was collected (A=2009, B=2010). Bar shows 0.2 substitutions per amino acid position. Bootstrap replicates = 500.

103

CHAPTER 5 Functional characterization of autotrophic and protist-associated heterotrophic activity in permanently ice-covered lakes of the McMurdo Dry Valleys, Antarctica

Jenna M. Dolhi, Wei Li, Rachael M. Morgan-Kiss

Author contributions: JD and WL (equal contribution) developed methods, performed data collection and analyses, and wrote the manuscript related to RubisCO and βGAM activities, respectively

Publication status: Manuscript in preparation for Polar Biology

104

CHAPTER 5

5.1. INTRODUCTION

Permanently ice-covered lakes situated in the Taylor Valley of South Victoria Land, Antarctica have closed basins and extremely stable strata which vary physically, chemically, and biologically. These unique systems are part of the McMurdo Long Term Ecological Research (LTER) site and have been monitored annually since 1993 (Lyons et al. 2000, Foreman et al. 2004). The 3-4 m ice-cover largely prevents water column mixing; however, minimal mixing occurs during the austral summer when ice melt forms a moat around the lakes. Glacier melt and stream formation is the primary mechanism of water and nutrient inputs to the lakes (Lyons et al. 2000) and is predicted to increase in future years as a result of climate change (Doran et al. 2008). Stratification occurs in MDV lakes Bonney, Fryxell, and Vanda due to the lack of mixing as well as salinity gradients that form distinct haloclines. The lower depths of these lakes contain dense saline water while shallow waters are akin to freshwater. Lake Bonney is divided into an east and west lobe (ELB and WLB, respectively) by a 13 m bedrock sill which allows mixing of the surface waters. All four lakes are oligotrophic in the shallow, photic layers and increase - - + 3- sharply in nutrients (NO2 , NO3 , NH4 , and PO4 ) at and below the chemoclines. Despite the general trends in lake chemistry, the chemical composition varies for each lake (Lyons et al. 2000). Although all of the MDV lakes are light limited, Lake Vanda is unique as it has a relatively transparent ice-cover that allows light to penetrate to the lake bottom (Vincent and Vincent1982). Irradiance is the main energy source used for production of organic carbon in the lakes however, it does not saturate photosynthesis and this has implications on the food webs (Priscu 1995). Due to the extreme conditions, MDV lake food webs are microbially dominated and truncated (lack metazoans). Microbial eukaryotes (i.e., protists) play dual roles in the food webs as they act as primary producers (i.e., light driven CO2 fixation into simple sugars) as well as tertiary consumers (i.e., predation on bacteria and small phytoplankton) (Priscu et al. 1999). Phytoplankton and chemolithoautotrophs fix inorganic carbon (i.e., autotrophy) using light- dependent and -independent pathways, respectively, and form the base of the food web in the MDV lakes. Recent applications of 18S rRNA sequencing in Lake Bonney provided the first

105 insight into the phylogenetic diversity of the microbial eukaryote populations, 85% of which was related to photosynthetic eukaryotes (Bielewicz et al. 2011). ELB and WLB phytoplankton communities are stratified into shallow (below the ice, 6 m), middle (within the permanent chemocline, 13-15 m) and deep (bottom of the photic zone, 18-20 m) photosynthetic populations. Cryptophytes dominate shallow populations while haptophytes and stramenopiles occupy the chemocline. Shade-adapted chlorophytes such as Chlamydomonas raudensis UWO241 (Dolhi et al. 2013) occupy the bottom of the photic zone (Bielewicz et al. 2011). Kong et al. (2012b) corroborated these data by tracking transcript abundance and gene copy number of multiple isoforms of the major carbon fixation gene (RubisCO; Ribulose-1,5-bisphosphate carboxylase oxygenase). This study showed that organisms related to Nannochloropsis (a stramenopile) were active (based on mRNA abundance) throughout the water column, while a haptophyte related to Isochrysis dominated the chemocline communities at the levels of DNA and mRNA (Kong et al. 2012b). Consumers which take up particulate organic carbon (i.e., heterotrophic activity via ) in the lakes include heterotrophic , , and rotifers (Priscu et al. 1999). Heterotrophic protists play important roles in energy and nutrient cycling (Strom et al. 1997, Azam and Malfatti 2007). In oligotrophic systems, a significant amount of biomass is grazed and eventually transferred to higher trophic levels by phagotrophic protists (Sanders et al. 2000, Sherr and Sherr 2002, Moorthi et al. 2009, Caron et al. 2009). Heterotrophic grazing activity by nanoflagellates and phytoplankton capable of heterotrophy (i.e., mixotrophy), including cryptophytes, was determined by feeding samples from MDV lakes Hoare, Fryxell, and Bonney fluorescently labeled bacteria (Roberts and Laybourn-Parry 1999, Marshall and Laybourn-Parry 2002, Thurman et al. 2012). In this study we functionally characterize the trophic mode of protists through the water columns of ELB, WLB, Fryxell, and Vanda by measuring enzyme activities representative of autotrophy (RubisCO) and heterotrophy (N-acetyl- beta-glucosaminidase; βGAM). Phytoplankton (eukaryotic algae and cyanobacteria) and some chemolithoautotrophic bacteria fix inorganic carbon via the Calvin-Benson-Bassham (CBB) cycle in which the enzyme RubisCO catalyzes the first and rate-limiting reaction. This enzyme binds a 5-carbon sugar, ribulose-1,5-bisphosphate (RuBP) with either CO2 (carboxylation) or O2 (oxygenation or photorespiration). Products of RubisCO carboxylation are reduced to sugars which are

106 incorporated into cellular biomass (Tabita et al. 2008). The oxygenation reaction is counter- + productive in that it results in the release of CO2 and NH4 (Peterhansel et al. 2008). The likelihood of each reaction being catalyzed depends on the enzyme specificity for CO2 versus O2 (Spreitzer and Salvucci 2002) and the availability of each substrate (Peterhansel et al. 2008). Phytoplankton, α, β, and γ-proteobacteria, and archaea contain various forms of RubisCO: cyanobacteria, proteobacteria, and green and red eukaryotic algae harbor form I, Proteobacteria and dinoflagellates harbor form II, and archaea harbor form III. Form IV RubisCO, or RubisCO- like-proteins (RLP), occur in a variety of organisms and do not catalyze a carboxylation reaction. The function of this enzyme is not yet apparent (Tabita et al. 2008). Phagotrophic protists engulf bacteria or particulate organic matter into their food vacuoles, subsequently a large amount of enzyme is released from lysosomes into the food vacuoles to rapidly hydrolyze the organic compounds (Vrba et al. 1996). Planktonic heterotrophic activity in aquatic environments can be detected and quantified based on specific activity (i.e., βGAM) (Zubkov and Sleigh 1998). Some heterotrophic bacteria utilize similar enzymes to breakdown complex organic carbon particles; however, the bacterial enzymes are excreted into their environment, while protist βGAM enzymes are restricted to the acidic environment of their food vacuoles. Thus, the optimal pH condition of protist βGAM tends to be lower (pH 4.6) compared to that of heterotrophic bacteria (neutral or slightly basic) (Gonzalez et al. 1993, Vrba et al. 1993, 1996). Methods to assess bacterivory activity are based on the measurement of glucosaminidase activity of phagotrophic protists in pure cultures and environmental samples (Gonzalez et al. 1993, Zubkov and Sleigh 1998). As phagotrophic protists synthesize enzymes and store them in lysosomes, activity assays require the release of enzymes by cell lysing, and activation at low pH resulting in cleavage of a fluorogenic substrate. Activity is calculated based on the fluorescence intensity. Therefore, the hydrolytic activity of phagotrophic protists can be used as a diagnostic tool for estimating both heterotrophic protist activity and biomass (Zubkov and Sleigh 1998). The stable, low-temperature MDV lake systems are particularly sensitive to climate change (Morgan-Kiss and Dolhi 2011, Wilkins et al. 2012, Lyons et al. 2001). In order to understand how such changes will affect the lake biota, it is necessary to frame studies of organism diversity, distribution, and function in the context of environmental parameters; however, few studies have considered MDV lake protists in this context (Kong et al. 2012b,

107

Roberts et al. 2004b). In this study, we investigate and protist-specific distribution using RubisCO and acid-βGAM enzyme activities, respectively, through the physically and chemically distinct MDV lakes and correlate activity with environmental parameters.

5.2. METHODS

5.2.1. Field sampling. Water column samples were collected during the Nov.-Dec. 2012 field season with the exception of samples for Lake Vanda RubisCO activity which were collected in Dec 2011. All sampling depths were measured from the piezometric water level in the ice hole using a depth-calibrated hand winch. Water samples were collected with a 5 L Niskin bottle and stored under low temperature and dark conditions until processing. Water samples for RubisCO carboxylase activity (2-5 L) and βGAM activity (0.75-2 L) were vacuum filtered (0.3 mBar) onto 47 mm GF/C filters (Whatman, UK) and Nylon membrane filters (Millipore, MA) with a 0.45

µm pore size, respectively. Filters were immediately flash frozen in liquid N2 and shipped to the US laboratory on dry ice. Samples were stored at -80 °C for less than two months before determining enzyme activity.

5.2.2. Limnological parameters. Photosynthetically active radiation (PAR), temperature, chl a, + - 3- primary productivity (PPR), and nutrient (NH4 , NO3 , and PO4 ) concentrations were determined through the water columns of ELB, WLB, and FRX during the Nov.-Dec. 2012 field season. For Lake Vanda only PAR and chl a were determined during the 2012 field season. PAR was measured with a Li-Cor LI-193 spherical quantum sensor (Li-Cor Biosciences, NE). Temperature was measured with a Seabird model 25 profiler (Spigel and Priscu 1998). In situ chl a fluorescence was determined with a bbe Moldaenke profiling spectrofluorometer. Light 14 mediated PPR was determined by measuring NaH CO3 incorporation in duplicate over a 24 h in situ incubation. Nutrients were measured as part of the NSF-funded McMurdo LTER program according to the methodology outlined in the McMurdo LTER manual (http://www.mcm.lter.org). Briefly, inorganic nitrogen species were determined with a Lachat autoanalyzer and soluble reactive phosphorus (SRP) was measured manually using the antimony-molybdate method (Strickland and Parsons 1972).

108

5.2.3. Lysate extraction and RubisCO carboxylase activity assay. GF/C filters were thawed on ice, cut in half and transferred to two tubes. All subsequent extraction steps were carried out on ice. Each cut half filter was transferred to a 1.5 mL screw cap tube one fifth the way full of 0.1 mm zirconia/silica beads and 650 µL ice cold RubisCO extraction buffer (50 mM bicine, 10 mM -1 MgCl2, 1 mM EDTA, 5 mg mL BSA, and 0.1% triton X-100; add fresh: 20 mM NaHCO3, 10 mM DTT, 0.1 mg mL-1, pH 7.8) was added. Filters were disrupted with a Minibead beater (Biospec Products, OK) for three cycles of 30 s at speed setting 48 with alternating 1 min on ice incubations. The lysate was transferred to a 1.5 mL microcentrifuge tube and cleared by centrifugation at 4 °C for 2 min at 15,000 x g. The soluble lysate was used in RubisCO carboxylase assays. RubisCO carboxylase activity was measured in the soluble cell lysates (200 µL) within 30 min of extraction. Soluble cell lysates were incubated at 25 °C for 1 min before RubisCO 14 activity was measured using a standard NaH CO3 assay in duplicate reactions as described previously (Dolhi et al. 2012, Tortell et al. 2006). Briefly, RubisCO was activated with MgCl2 (8 14 -1 mM) and NaH CO3 (ViTrax, Placentia, CA, specific activity in final reaction: 0.03 µCi mmol ) for 5 min at 25 °C. The carboxylase assay was initiated upon addition of 20 µL of the substrate, Ribulose-1, 5-bisphosphate (RuBP; Sigma-Aldrich, USA; 15 mM) and for 5 min at 25 °C. The reaction was stopped with 100 µL propionic acid and unincorporated 14C was exhausted by centrifugation for 1.25 hr at 2,000 x g. A multipurpose scintillation counter LS6500 (Beckman Coulter, FL) was used to determine cpm of acid stable end products. RubisCO activity was calculated per liter of lake water filtered.

5.2.4. Lysate extraction and βGAM activity assay. Nylon filters were thawed on ice and extraction was carried out on ice. Filters were disrupted by bead beating in 2 mL of ice cold extraction buffer (0.1 M Acetate, 0.1% Brig 35, pH 4.6 with glacial acetic acid). βGAM enzyme activity was measured in soluble lysates (250 µL) using a β-N-Acetyl- glucosaminidase Assay Kit (Sigma-Aldrich, USA). Cell lysates were incubated with the fluorogenic substrate 4-methylumbelliferyl-n-acetyl-beta-D-glucosaminide in the dark at 20 ˚C for 4 hrs. Nano-pure water and purified β-N-Acetyl-glucosaminidase from Jack beans (came with assay kit) were used as negative and positive controls respectively. Fluorescence intensity was quantified in a Lambda-35 spectrofluorometer (Perkin-Elmer, MA). A standard curve was

109 made with known amounts of the product of the enzymatic reaction, 4-MUF (Vrba et al. 1993, Zubkov and Sleigh 1998). In order to test the assay and choose filters with optimal enzyme recovering capability, filters of various materials (glass fiber, polyethersulfone, and nylon) were tested using environmental samples from Acton Lake in College Corner, OH. The test results indicated that nylon filters had the highest enzyme yield (data not shown).

5.2.5. Protein concentration determination. Protein concentration was determined in the soluble cell lysate extracted for RubisCO and βGAM assays according to the Bradford method using BSA as a standard (Bradford 1976).

5.2.6. Bacterial enumeration. Quantitation of free-living bacteria in the lake water columns was carried out using a method modified from Lebaron et al. (1994). Lake water (2 mL) from each depth was fixed with paraformaldehyde at a final concentration of 2 % v/v for 30 min. Fixed samples were filtered on 25 mm black polycarbonate membrane with a pore size of 0.2 µm. Bacteria cells were stained with a fluorochrome (4',6-diamidino-2-phenylindole, DAPI) followed by epifluorescent microscopic enumeration. Filters were examined using an Olympus AX-70 Multi-mode System with a specific filter set (EX 360/40 nm, EM 460/50 nm) for DAPI staining, and digital images of at least 15 random views were taken. Images were then calibrated, and bacterial counts from each image were recorded using ImageJ software (V1.47, by National Institutes of Health). The bacterial concentration of each depth was derived according to the average bacterial counts of each image.

5.2.7. Statistical analyses. Graphic figures were generated in OriginPro 8.5.1 (OriginLab Corp., Northampton, MA). Relationships between physical, chemical, and biological parameters were determined using Pearson correlation (R) analysis. All data were log transformed prior to analysis. βGAM activity data represents the mean average of three replicates measured from one environmental sample. RubisCO activity data represents the mean average of 3-5 replicates, so average RubisCO activity was used in Pearson correlation analysis to incorporate all measurements for all lake depths measured.

5.3. RESULTS

110

5.3.1. Limnological parameters. The MDV lakes are physicochemically stratified. A steep increase in nutrient concentration occurs within the chemoclines, that is, at mid-depths in Lake 3- Bonney and Fryxell, and deep depths in Vanda (Fig. 5.1). Lake Bonney was PO4 limited, 3- ranging from 0.03-1.3 µM in ELB and 0.02-0.68 µM in WLB, whereas PO4 sharply increased below 11 m in Fryxell reaching concentrations which were on average more than 50 times + greater than those of Lake Bonney. NH4 concentrations were greater than (WLB, Fryxell, and - Vanda) or equivalent to (ELB) NO3 concentrations. Lake Vanda had the greatest concentrations + - of NH4 and NO3 , 1468.2 µM and 222.0 µM, respectively. PAR values under the ice-cover of the lakes were <1% PAR surface values at 15 m Lake Bonney, 9 m Lake Fryxell, and below 60 m Lake Vanda. Temperature peaked at mid-depths in ELB, WLB, and Fryxell. WLB had the largest temperature range of these three lakes: from -2.7 to 2.6 °C. Temperature increased with depth to a maximum of 29.5 °C at the bottom of the water column in Lake Vanda (Fig. 5.1).

5.3.2. MDV lake community RubisCO carboxylase activity. RubisCO and βGAM enzyme activity assays provided a measure of MDV lake microbial community trophic mode. Preliminary testing of the RubisCO carboxylase activity method was performed on a pure culture of a green algal isolate, mixed protist cultures, and environmental samples from a local reservoir, Acton Lake in College Corner, OH. Figure 5.2 shows representative data from the preliminary tests during enzyme assay optimization. RubisCO activity was measurable in all samples and activity of lake water samples were comparable to published data from the subarctic Pacific Ocean (Tortell et al. 2006). In Chapter 4, as well as two other studies (Kong et al. 2012a, b), the diversity and distribution of autotrophic organisms residing in the four study lakes was described. However, these studies were based on DNA or mRNA levels and therefore do not truly reflect the carbon fixation activity of the autotrophic communities. Therefore, RubisCO activity was used as a proxy for photoautotrophic and chemolithoautotrophic activity through the water columns of MDV lakes, and compared to other measurements of phytoplankton biomass (chl a) and carbon incorporation (primary productivity; PPR). In ELB and WLB, chl a and PPR peaks corresponded: chl a peaked at 5 m and 15 m and PPR peaked at 15 m, whereas RubisCO activity, indicating photoautotrophic or chemolithoautotrophic carbon fixation had a less defined peak at 15-20 m (Fig. 5.3). RubisCO activity in WLB was also greatest at 20 m, but was 3.5-fold lower than that of ELB. Conversely, PPR was greater in WLB compared to ELB and peaked at 15 m.

111

Chl a decreased sharply below 15 m in WLB, but in ELB the decrease was steady with a secondary chl a peak at 20 m (Fig. 5.3). RubisCO activity and chl a were greater in Lake Fryxell than the other MDV lakes, reflecting higher productivity in this lake (Fig. 5.4). RubisCO activity peaked at 11 m in Fryxell while chl a peaked at 8-9 m and PPR was greatest at 4.5 and 7 m (Fig. 5.4). In Lake Vanda RubisCO activity was greatest at 72 m and a secondary peak occurred at 54 m. Chl a peaked at 68 m in Lake Vanda, but was low compared to the other lakes (Fig. 5.5). PPR was not measured in Lake Vanda during the 2012 field season.

5.3.3. MDV lake community βGAM activity. Acidic βGAM activity was used as an indicator of protist heterotrophy through the water columns of the MDV lakes. βGAM activity was generally greatest at depths that contained peaks in total protein, with the exception of ELB. For ELB and WLB, βGAM activity increased 1.4 fold and 1.6 fold at the depths of peak activity (22 and 15 m, respectively) compared to that of 5 m (Fig. 5.3). Protein concentration peaked at 18 m with a secondary peak at 22 m in ELB and at 13 m in WLB. Protein concentration followed the same decreasing trend with depth as chl a in both lobes of Lake Bonney, further supporting differences in biomass between the two lobes. Bacteria counts were greatest at 5 m in ELB and 10 m in WLB (Fig. 5.3). The range in βGAM activity for Lake Frxyell, 0.9-5.3 nmol hr-1 L-1, was similar to that of Lake Bonney (Fig. 5.4). βGAM activity peaked at 14 m, while protein concentration peaked at 11 m. As was observed for Lake Bonney, bacteria counts peaked at more shallow depths, 8 m Lake Fryxell, compared to βGAM activity and protein concentration (Fig. 5.4). Similar to RubisCO activity in Lake Vanda, βGAM activity and protein concentration were greatest at 72 m, while bacteria counts were greatest at 68 m (Fig. 5.5).

5.3.4. Correlations of autotrophic and heterotrophic enzyme activity with physicochemical lake parameters. In order to determine if the depth stratification of RubisCO and βGAM activity correlated to physical and chemical parameters in the lakes, Pearson correlation analysis between enzyme activity and a suite of biological and environmental parameters was carried out for individual lakes and all lakes combined. When all data points were combined across the four study lakes, several environmental and biological parameters exhibited significant positive correlations with RubisCO and βGAM (Table 5.1). Spatial trends in RubisCO activity corresponded positively with the environmental parameter PAR, as well as the biological parameters chl a, bacterial counts, protein concentration, and abundance of both haptophytes and

112 cryptophytes (Table 5.1). Most of these relationships were also observed in both lobes of Lake Bonney (Table 5.2); however, notable differences in these overall trends were observed in Lake Fryxell. While RubisCO and PAR trends exhibited a negative correlation, statistically significant positive correlations occurred between RubisCO activity and nutrients (inorganic N and P) and cyanobacteria in this lake (Table 5.2). Dissolved inorganic carbon and conductivity were negatively correlated to RubisCO activity (Table 5.1). The relationship between spatial distribution of heterotrophic activity (i.e., βGAM) and various parameters was distinct from that of RubisCO. Statistically significant positive 3- correlations across all lakes for βGAM activity include those with PO4 , bacteria counts, protein concentration, chl a, as well as green algae abundance (Table 5.1). As was observed for RubisCO activity, nutrients and βGAM activity were positively correlated in Lake Fryxell (Table 5.2), as well as in Vanda, whereas chl a and PAR had no or a negative correlation with βGAM activity (Table 5.2).

5.4. DISCUSSION

It has been known that MDV lakes vary in their distribution, diversity and abundance of plankton communities, which is driven by physicochemical differences within and between the lakes (Laybourn-Parry et al. 1997, Roberts et al. 2004b, Kong et al. 2012b). Two previous studies applied combined sequencing and qPCR approaches to describe the diversity and vertical distribution of the autotrophic communities residing in Lake Bonney (Kong et al. 2012a, b). These studies predicted that light availability was a strong environmental driver for both gene copy number and transcript level of form IA/B RubisCO (large subunit encoded by rbcL) as well as cryptophyte rbcL. However, abundance of the major RubisCO isoform, ID rbcL, did not correspond with light. Here, we build upon these studies of Lake Bonney by directly measuring functional RubisCO enzyme activity in comparison with βGAM activity to gain an overall view of protist trophic mode across four chemically distinct MDV lakes. Autotrophic (RubisCO) and heterotrophic (βGAM) enzyme activities were highest at the chemocline (ELB, WLB) or began increasing at the chemocline (Lakes Fryxell and Vanda). RubisCO activity was lowest in WLB corresponding with low levels of rbcL mRNA in this lake (Kong et al. 2012b). Enzyme activities in Lake Bonney were low, whereas Lake Fryxell showed high activity corresponding to its status as the most productive lake in this study. Lake Vanda

113

(72 m) had greatest βGAM activity among all of the lakes, as well as relatively high RubisCO activity at depths below the chemocline. The warm water temperature at this depth in Lake Vanda may contribute to optimal enzyme function. Pearson correlation analysis of biological and environmental parameters provided insight on distribution of auto- and heterotrophic groups through the lake depths. In considering RubisCO activity, positive correlations were found with, chl a and PAR for ELB, WLB, and Vanda (Table 5.2) suggesting that the bulk of carbon fixation in our samples was carried out by light dependent photosynthetic organisms. This fits well with conclusions from recent molecular studies that showed that Lake Bonney is dominated by photosynthetic eukaryotes (Bielewicz et al. 2011; Kong et al. 2012a, b). The transcriptionally active RubisCO harboring community at 13 m included haptophytes dominated by an Isochrysis sp. and a stramenopile related to Nannochloropsis (Kong et al. 2012b). Presumably, these organisms contribute the majority of light-dependent carbon fixation observed in Lake Bonney. In addition, gene copy abundance and transcript levels of the RubisCO isoform harbored by these organisms, form ID rbcL, correlated with chl a and PPR, while form IA/B rbcL DNA and cDNA abundance correlated positively with PAR, indicating that photosynthetic populations are influenced by different environmental drivers (Kong et al. 2012b). In Lake Fryxell, light availability played a minor role in community carbon fixation as evidenced by a negative correlation; however RubisCO activity corresponded positively with cyanobacteria biomass and nutrients (Table 5.2). The lack of a dependence on light in Lake Fryxell is likely partly due to the low levels of PAR in this lake. Similarly, maximum phytoplankton biomass (chl a) was observed at the chemocline in Nov-Dec.1997 further indicating an influence of high nutrients on phytoplankton distribution (Roberts 2004b). This relationship suggests that inorganic N and P are the driving factors of phytoplankton distribution in Fryxell. Priscu (1995) reported nitrogen limitation and stimulation of photosynthesis upon N and P addition in samples from surface water (5 m) of Lake Fryxell. Organisms existing at and below the chemocline may be alleviated of nutrient limitation and contribute to the RubisCO activity measured, especially cryptophytes capable of mixotrophy (i.e., combined photosynthesis and heterotrophy) which have been previously reported (Chapter 4, Laybourn-Parry et al. 1997, Roberts and Laybourn-Parry 1999, Marshall and Laybourn-Parry 2002). Last, previous measurements of chl a and photosynthesis in Lake Vanda were greatest at 57.5 m and this region corresponded to a phytoplankton population dominated by cyanobacteria

114

(Vincent and Vincent 1982). The cyanobacteria may contribute to the RubisCO activity measured here and support the positive relationship between RubisCO activity, PAR, and chl a (Table 5.2). The peak in RubisCO activity observed at 72 m may be due to RubisCO harboring organisms including a stramenopile identified at 68 m in Lake Vanda (Chapter 4). RubisCO activity was measureable at depths below the photic zones in all MDV lakes studied, indicating that there is the potential for carbon fixation using alternative energy sources, such as reduced organic or inorganic compounds (i.e., chemolithoautotrophic metabolism) (Fig. 5.3-5.5). Evidence for chemolithoautotrophic bacteria in the MDV aquatic ecosystems has been reported. The anoxic, sulfidic bottom waters of Lake Fryxell have yielded culture-dependent (Sattley and Madigan 2006) and independent (Chapter 4) evidence for sulfur-oxidizing bacteria (SOB) related to Thiobacillus thioparus. This region is suitable for carbon fixation by chemolithoautotrophs which may utilize sulfide and oxygen as an electron source and acceptor, respectively (Sattley and Madigan 2006). The importance of carbon fixation via chemolithoautotrophs in Fryxell was corroborated by the negative correlation of RubisCO activity with PAR and chl a (Table 5.2). Evidence for the existence of a SOB community below the photic zone was also found for WLB (Chapter 4, Kong et al. 2012a, b, Vick-Majors et al. 2014). It is predicted that these organisms utilize dimethyl sulfoxide (DMSO), a breakdown product of dimethylsulfoniopropionate (DMSP) which is produced by algae including haptophytes, chrysophytes, and dinoflagellates as a cryo- or osmo-protectant (Trevana et al. 2000), as an electron acceptor. A second breakdown product, dimethylsulfide (DMS) may be utilized as an electron donor (Raina et al. 2010). The potential for dark carbon fixation also exists in ELB and Lake Vanda. PCR based detection of ammonia oxidizing bacteria (AOB) was found at depths above 25 m in ELB, and the lack of detection at deeper depths was likely due to inhibition of the PCR reaction (Voytek et al. 1999, Voytek and Ward 1995). AOB were detected by in situ nitrification assays at 50-57 m in Lake Vanda, where reduced nitrogen compounds (electron donor) begin to increase and oxygen is available as an electron acceptor (Vincent et al. 1981, Voytek et al. 1999). While the RubisCO enzyme assay measured the potential for a community to fix inorganic carbon, βGAM activity is a proxy for community respiration, specifically breakdown of complex organic carbon sources by acidic food vacuoles. While βGAM activity was generally low in Lake Bonney, the peaks occurred at or below the chemocline for WLB and ELB,

115 respectively (Fig. 5.3). The positive correlations between βGAM activity, chl a, and green algae in ELB and WLB suggest a relationship between algal produced dissolved organic matter which may be directly consumed by heterotrophic protists or by the bacteria on which they feed. Other sources of organic carbon include upward diffusion from a deepwater pool or fall out of particulate organic matter that exists within the ice cover (Priscu et al. 1999). In Lake Bonney ciliates and dinoflagellates were reported to make up the heterotrophic community. However, mixotrophic phytoflagellates were shown to have higher grazing rates than heterotrophic flagellates in Lake Bonney especially in the months leading up to the austral winter (Thurman et al. 2012). In Lake Fryxell, βGAM activity increased from 9-11 m and was positively correlated with nutrients (Fig. 5.4, Table 5.2). This corresponded with the high abundance of cryptophytes related to Geminigera cryophila which was observed in Chapter 4. Cryptophytes have previously been shown to dominate Lake Fryxell and are which graze on bacteria (Kong et al. 2012b, Roberts and Laybourn-Parry 1999, Marshall and Laybourn-Parry 2002). Bacterial prey abundance was highest at 8 m and is likely the major food source in this lake (Fig. 5.4); however excretion from benthic algal mats may also serve as a source of carbon (Wharton et al. 1983, Priscu et al. 1999). βGAM activity increased at the chemocline in Lake Vanda and correlated positively with nutrients, bacteria counts, protein concentration, and green algae suggesting that heterotrophic protists may graze on bacterial prey or utilize algal photosynthate as a carbon source (Fig. 5.5, Table 5.2). Ciliates and heterotrophic nanoflagellates, which may include mixotrophic Ochromonas sp., were reported at 60-64 m in Lake Vanda (James et al. 1998) and may contribute to the heterotrophic activity measured in this lake. As PPR activity is limited to light dependent carbon fixation, and RubisCO activity may be attributed to light dependent or independent carbon fixation, the two parameters were not necessarily expected to correlate. That PPR was greater in WLB, yet RubisCO activity was lower compared to ELB (Fig. 5.3) could indicate limitations in measuring RubisCO. Although the RubisCO activity assay was designed to measure the maximum potential activity, it is unlikely that the assay was optimized for all RubisCO harboring organisms as various forms of RubisCO enzymes differ in their substrate specificities (Spreitzer and Salvucci 2002). Last, extrapolating in vitro enzyme activity assay results to the natural system was complicated by mixotrophic organisms as these organisms were likely detected in both RubisCO and βGAM assays which may not be reflective of their trophic status in the lake. Considering these complications,

116

RubisCO activity should not be considered a replacement for primary productivity, but is useful in studies of complementary trophic modes (i.e., autotrophy and heterotrophy). Using functional enzyme activity, this study showed that phytoplankton, chemolithoautotrophs and protist-specific were distributed according to environmental parameters including PAR, bacteria prey or algae (i.e., source of DOM), and inorganic nutrient concentration. Further laboratory studies would be required to determine which environmental parameter(s) are responsible for driving the distribution of the protists and other lake biota. Previous studies of RubisCO gene diversity and distribution (Kong et al. 2012a, b) helped to attribute enzyme function to groups of organisms. Pairing diversity and function studies with environmental parameters will improve understanding of microbial community structure and how this may be impacted by climate change, an area of research with many unanswered questions (Caron and Hutchins 2013). Studies of polar microbial communities on the cusp of environmental change will be important for predicting how microbial communities in low latitude aquatic systems will respond.

117

Table 5.1. Pearson correlation coefficient values (R) for average RubisCO and βGAM activity with lake physical, chemical, and biological parameters for all lakes (n= 57-93). *, significant correlation (p<0.05); **, significant correlation (p<0.01). PAR; photosynthetically available radiation, Temp; temperature, Cond; conductivity, DIC; dissolved inorganic carbon, Bac; bacteria counts, Chl; chlorophyll a, Prot; protein, G. AL.; green algae, Cyan; cyanobacteria, Hapt; haptophytes, Crypt; cryptophytes.

PAR Temp Cond NH4 NO3 PO4 N:P DIC BAC CHL PROT G. AL. CYAN HAPT CRYPT

B-GAM -0.20 0.16 -0.23 0.17 0.06 0.06** 0.01 0.11 0.31** 0.44** 0.43** 0.26** 0.15 0.05 0.09

RubisCO 0.44** -0.29* -0.69** -0.65** -0.38** -0.22 -0.80** -0.30* 0.41** 0.57** 0.35** 0.05 0.18 0.68** 0.85**

118

Table 5.2. Pearson correlation coefficient values (R) for average RubisCO and βGAM activity with lake physical, chemical, and biological parameters for east lobe Bonney (ELB; n= 18-30), west lobe Bonney (WLB; n=13-21), Fryxell (n=15-21), and Vanda (n=11-21). *, significant correlation (p<0.05); **, significant correlation (p<0.01). PAR; photosynthetically available radiation, Temp; temperature, Cond; conductivity, DIC; dissolved inorganic carbon, Bac; bacteria counts, Chl; chlorophyll a, Prot; protein, G. AL.; green algae, Cyan; cyanobacteria, Hapt; haptophytes, Crypt; cryptophytes.

PAR Temp Cond NH4 NO3 PO4 N:P DIC BAC CHL PROT G. AL. CYAN HAPT CRYPT

ELB

B-GAM 0.40* -0.07 -0.20 -0.11 -0.20 -0.35 -0.06 0.07 0.34 0.56** 0.24 0.39* -0.52** 0.54** 0.49**

RubisCO 0.60** 0.37 -0.25 -0.35 -0.34 -0.52* -0.24 0.40 0.19 0.60** 0.57* 0.42 -0.16 0.54* 0.43

WLB

B-GAM 0.72** 0.45* -0.65** -0.62** 0.80** -0.75** -0.65** -0.54* 0.43 0.82** 0.63** 0.79** 0.26 0.59** 0.31

RubisCO 0.85** 0.47 -0.74** -0.79** 0.73** -0.93** -0.76** -0.81** 0.82** 0.80** 0.81** 0.79** -0.49 0.74** 0.39

Fryxell

B-GAM -0.61** 0.28 0.57** 0.57** 0.57** 0.60** 0.53* 0.58** -0.41 -0.53* 0.49* -0.42 0.50* -0.45* -0.38

RubisCO -0.64** 0.04 0.17 0.91** 0.98** 0.88** 0.94** 0.24 -0.73** -0.63* 0.77** -0.13 0.90** -0.63* -0.12

Vanda B-GAM -0.91** 0.13 -- 0.78** -0.28 0.81** 0.57** 0.76** 0.50* -0.20 0.90** 0.47* -0.30 -0.73** --

RubisCO 0.99** 0.02 -- -0.91** -0.61 -0.59 -0.81** -0.94** 0.85** 0.99** -0.77* 0.89** 0.47 0.61 --

119

SRP (M) Temperature (C) 0 10 20 30 40 50 -2 0 2 4 6 8 10 0 0 ELB-A ELB-B 5 5 SRP 10 + 10 NH4 15 NO - 15 3 20 20 25 25 Temperature

30 PAR 30 Lakedepth (m) Lakedepth (m) 0 0 WLB-D 5 WLB-C 5 10 10

15 15

20 20 25 25

30 30

Lakedepth (m) Lakedepth (m) 0 0 2 FRX-E FRX-F 2 4 4 6 6 8 8 10 10 12 12 14 14 16 16

18 Lakedepth (m) Lakedepth (m) 18 Temperature (C) 0 10 20 30 0 0 10 VAN-G VAN-H 10 20 20 30 30

40 40 50 50 60 60

70 70 Lakedepth (m) Lakedepth (m) 0 5 10 15 20 150 300 450 600 0 10 20 30 40 50 + - PAR (mol photons m-2 s-1) NH4 , NO3 (M)

Figure 5.1. Depth profiles of physical and chemical characteristics of East lobe Bonney (ELB-A, B), West lobe Bonney (WLB-C, D), Frxyell (FRX-E, F), and Vanda (VAN-G, H). Scale on Y- axis varies depending on lake depth. Note the difference in temperature scale on x-axis in H. Photosynthetically active radiation (PAR) and temperature data were collected during the 2012 field season. Nutrient data was kindly provided by the McMurdo LTER and was collected during the 2009-2010 field season for A, C, and E. Nutrient data was collected during the 2007-2008 field season for G. SRP; soluble reactive phosphorus.

120

0 350 A B 1 300

)

2 -1 250

hr -1 3 a 200

g chl g 150

4 

2 100

Lakedepth (m) 5 RubisCOactivity 50

6 (nmol CO 0 0 2 4 6 8 10 12 UWO 4 5 6 10 11 12 RubisCO activity Culture identity (nmol CO g chl a-1 hr-1) 2

Figure 5.2. Preliminary test of RubisCO carboxylase activity assay from filtered lake water (A) or protist enriched cultures (B). Depth profile of autotrophic community RubisCO activity measured in crude soluble lysates from filtered Acton Lake water (A). RubisCO activity in pure green algae culture (UWO241) and Antarctic lake water (4-6; east lobe Lake Bonney, 10-12 Lake Fryxell) enriched for protists (B). N=1

121

0 0 ELB-A ELB-B ELB-C ELB-D ELB-E ELB-F

5 5

10 10

15 15

20 20

Depth(m) Depth(m) 25 25

30 30 0 0 WLB-G WLB-H WLB-I WLB-J WLB-K WLB-L 5 5

10 10

15 15

20 20

Depth(m) Depth(m) 25 25

30 30

0 3 6 9 0 3 6 9 0 2 4 6 0 3 6 9 0 3 6 9 0 10 20 30 40 Chl a RubisCO PPR GAM Protein Bacteria counts -1 -1 (g L-1) 4 -1 (nmol L hr ) (g chl L-1 day-1) (nmol L-1 hr-1) (g L-1) (x10 cells mL )

Figure 5.3. Depth profiles of autotrophic community RubisCO activity measured in crude soluble lysates extracted from filtered east lobe Bonney (ELB-A) and west lobe Bonney (WLB- G) lake water, total chlorophyll a (chl a) measured in situ for ELB-B and WLB-H, and light dependent primary productivity (PPR) during Nov. for ELB-C and WLB-I. Heterotrophic community βGAM activity and protein concentration measured in crude soluble lysates extracted from filtered ELB-D, E and WLB-J, K lake water and bacterial counts for ELB-F and WLB-L. For RubisCO activity and βGAM activity, error bars show standard error of 3-5 replicates and 3 replicates, respectively (error not determined for WLB 30 m for which there is only 2 replicates). One replicate for chl a, PPR, protein, and bacterial counts.

122

0 0

2 2

4 4

6 6

8 8 LakeDepth (m) LakeDepth (m) 10 10

12 12

14 14

0 10 20 30 40 50 0 10 20 30 40 50 0 2 4 6 0 3 6 9 0 10 20 30 40 50 0 10 20 30 40 RubisCO Chl a PPR GAM Protein Bacteria counts (nmol L-1 hr-1) (g L-1) (g chl L-1 day-1) (nmol L-1 hr-1) (g L-1) (x104 cells mL-1)

Figure 5.4. Depth profiles of Lake Fryxell autotrophic community RubisCO activity measured in crude soluble lysates extracted from filtered lake water (A), total chlorophyll a (chl a) measured in situ (B), and light dependent primary productivity (PPR) during Nov. (C). Depth profiles of Lake Fryxell heterotrophic community βGAM activity (D) and protein concentration (E) measured in crude soluble lysates extracted from filtered lake water. Depth profiles of bacterial counts for Lake Fryxell (F). Error bars show standard error of 3-4 replicates. One replicate for chl a, PPR, protein, and bacterial counts.

123

0 0 A B C D E

10 10

20 20

30 30

40 40 LakeDepth (m) 50 50 LakeDepth (m)

60 60

70 70

0 3 6 9 12 0 3 6 9 12 0 3 6 9 12 0 3 6 9 12 0 10 20 RubisCO Chl a GAM Protein Bacteria counts (nmol L-1 hr-1) (g L-1) (nmol hr-1 L-1) (g L-1) (x104 cells mL-1)

Figure 5.5. Depth profiles of Lake Vanda autotrophic community RubisCO activity measured in crude soluble lysates extracted from filtered lake water (collected in Dec 2011) (A), total chlorophyll a (chl a) measured in situ (B). Depth profiles of Lake Vanda heterotrophic community βGAM activity (C) and protein concentration (D) measured in crude soluble lysates extracted from filtered lake water. Depth profiles of bacterial counts for Lake Fryxell (E). Error bars show standard error of 3-4 replicates. One replicate for chl a, protein, and bacterial counts.

124

CHAPTER 6

Conclusion

Jenna Dolhi, Rachael Morgan-Kiss

Author contributions: JD wrote the manuscript

125

CHAPTER 6

Aquatic photosynthetic organisms are responsible for a significant proportion of global net primary production annually and therefore have a large influence on the global (Field et al. 1998). One of the crucial enzymes in photosynthesis, RubisCO, catalyzes the incorporation of CO2 to organic compounds as well as initiates photorespiration. The importance of this enzyme in the global carbon cycle is reflected by its long history of scientific investigation including over 5,000 publications from 1947-2007 (Portis and Parry 2007). While many achievements have been made in studying this complex and highly regulated enzyme, there is still much to be discovered, especially in non-plant organisms, such as green algae. Future research in this field should focus on RubisCO regulation for which a gap in knowledge exists. For example, a crystal structure of tobacco (green type) RubisCO activase has been solved, however the mechanism of inhibitor release from RubisCO by activase, which is the main function of this enzyme, remains to be determined (Stotz et al. 2011). Differences have been found between green-type activase and that from (red-type) (Stotz et al. 2011, Mueller-Cajar et al. 2011). Determination of the crystal structure and function of other algal (i.e., chlorophytes, cryptophytes, chrysophytes) RubisCO activases and the comparison of these to plant activase would further the understanding of RubisCO regulation and may provide insight as to how RubisCO and activase interact. The research presented here expands upon current knowledge of RubisCO by investigating activity, carbamylation state, and abundance in non-model green algae (Chapters 2 and 3). Such studies targeting RubisCO from organisms found in extreme environments will likely lead to discoveries of natural variation in enzyme structure and function (Parry et al. 2013). Additionally, by investigating diversity and distribution of McMurdo Dry Valley (MDV) lake autotrophic communities (Chapter 4), as well as protist autotrophic and heterotorphic activities (Chapter 5), we provide insight on whole lake carbon cycling. This combination of laboratory and lake scale approaches provided a novel and comprehensive view of carbon fixation by green algae in the permanently ice-covered MDV lakes of Antarctica during the austral summer. As we continue to understand the role of in microbial carbon cycling in these environmentally sensitive Antarctic lakes, this information may be useful for predicting the response of microbial autotrophs to climate change in other low temperature aquatic systems.

126

Summary and future directions

The potential for a cold active RubisCO isolated from a uniquely adapted Antarctic green algae, C. raudensis UWO241, was investigated in Chapter 2. To our knowledge, only one other cold adaptation study of RubisCO exists (Devos et al. 1998). Consistent with the Devos et al. study, results for partially purified RubisCO were not in support of a cold adapted enzyme and indicated activity and stability at temperatures as high as 50-60 °C (Fig. 2.3-2.5). Interestingly, and contrary to the extracted enzyme results, whole cell carbon fixation was enhanced in UWO241 at 1.5-11 °C compared to the closely related mesophilic alga, SAG 49.72 (Fig. 2.7). The discrepancy between the results of partially purified RubisCO and whole cell carbon fixation experiments indicates that there are additional factor(s) (e.g., RubisCO activase, carbon concentrating mechanism associated enzymes) affecting the activity of the key carbon fixation enzyme and provides avenues for further research. Additionally, investigation of activity of other CBB cycle enzymes should be carried out. Preliminary studies tracking CBB cycle intermediates (and other cellular metabolites) are underway in a collaborator’s laboratory and will identify potential bottlenecks in carbon fixation at different temperatures. In addition, follow up studies on the abundance of other CBB cycle enzymes may provide further insight into carbon fixation capability at low temperature, as increased RubisCO abundance was observed in UWO241 grown at 2 °C (Chapter 3). These analyses will reveal the mechanism(s) utilized by UWO241 to achieve enhanced growth rates at low temperature and high light conditions when compared to the green alga, Chlorella vulgaris (Morgan-Kiss et al. 2006). Furthermore, this line of research will enhance our understanding of how photosynthetic organisms carry out carbon fixation at low temperature and maintain balanced photostasis, an important consideration as cold aquatic systems, including the oceans, cover over 71% of Earth (Margesin and Miteva 2011). Studies of RubisCO in UWO241 compared to SAG49.72 were extended in Chapter 3 to include regulation via carbamylation, an indirect measurement of RubisCO activase activity, at a range of growth temperature and light conditions. Additionally, RubisCO abundance was determined under varying temperature and light conditions. Results indicated differential modulation of RubisCO (carbamylation state or abundance) depending on growth conditions. Under elevated temperatures (UWO241) and very low light (UWO241 and SAG49.72), the carbamylation state of RubisCO was affected (Figs. 3.4 and 3.5). However, further studies are required to determine if the response in carbamylation state is due to RubisCO inhibition and

127 limited activity of the regulatory enzyme RubisCO activase. RubisCO abundance was not affected by growth irradiance, however abundance decreased with increasing temperature in both strains of C. raudensis, but to a greater extent in UWO241 (Fig. 3.6). The modulation of RubisCO abundance in the psychrophile could be an acclimation of this organism to low temperature. As carbamylation activity is an indirect measurement of RubisCO activase activity, further investigation of this enzyme should be pursued. RubisCO activase has previously been over-expressed in E. coli and purified for activity measurements (Salvucci and Crafts-Brandner 2004b, van de Loo 1996). While the sequence of UWO241 RubisCO activase has been determined (Fig. 3.7), efforts to overexpress this enzyme were not successful in cloning stages and resulted in sequence errors (data not shown). Ultimately, purification of activase would allow for direct measurements of activity when assayed with inhibitor bound RubisCO. Measurement of RubisCO activase activity with varying ATP concentration amendments would help to determine if an ATP limitation could result in reduced activity of RubisCO activase, a scenario predicted for low irradiance grown cultures. In order to connect physiological studies of the isolate UWO241 with carbon cycling in the natural environment, environmental scale studies were conducted for the McMurdo Dry Valley lakes including investigating the effects of environmental factors on autotrophic (i.e., both photosynthetic and chemolithoautotrophic organisms) community composition (Chapter 4) and RubisCO activity (Chapter 5). Targeting of genes encoding key carbon fixation genes (rbcL, ccbM, nifJ) and quantification using real-time qPCR through the water columns of Lakes Bonney, Fryxell, and Vanda, showed a community shift between the eukaryotic and bacterial communities capable of inorganic carbon fixation. A similar investigation was previously carried out in Lake Bonney (Kong et al. 2012a, b); however the current study provided the first analysis of carbon fixation potential in Lakes Fryxell and Vanda, while also providing data through the entire water column of all study lakes (Chapters 4 and 5). Organisms harboring form ID RubisCO including haptophytes and cryptophytes were in greatest abundance in Lakes Bonney and Fryxell, respectively (Fig. 4.3-4.4). In addition, the high abundance of form II RubisCO (cbbM) genes (found exclusively in chemolithoautotrophic bacteria) below the chemocline of west lobe Lake Bonney and Fryxell suggested the importance of dark carbon fixation in these lakes (Fig. 4.3-4.4). The phytoplankton community in Lake Vanda is the least studied of the lakes and was found to contain an abundance of genes related to cyanobacteria harboring form

128

IA/B RubisCO large subunit (rbcL) and stramenopiles harboring ID rbcL at mid- and deep depths, respectively (Fig. 4.5). The environmental factors that play a key role in driving the distribution of the carbon fixing organisms varied between the lakes. While PAR was important in the distribution of form IA/B rbcL in Lakes Bonney and Vanda, as well as form ID cryptophyte rbcL in Bonney, nutrients were important drivers of distribution of form IA/B in Fryxell (Table 4.3). This is likely due to the form IA/B rbcL harboring community being composed of more chemolithoautotrophic bacteria, particularly in depths at and below the chemocline of Lake Fryxell, compared to that of Lake Bonney (Chapter 4). These organisms along with form ID rbcL harboring organisms related to Geminigera cryophila were the major carbon fixers in Lake Fryxell (Chapter 4). G. cryophila is a mixotrophic organism and may survive the austral winter by utilizing heterotrophic metabolism (Laybourn-Parry et al. 1997, Priscu et al. 1999). In the future, haptophytes and cryptophytes, which are likely to be active in carbon fixation (Kong et al. 2012b), should be isolated from the lakes for laboratory scale studies to target environmental factors driving distribution and photosynthetic activity. Additionally, food web interactions between these organisms with heterotrophic organisms should be investigated to determine the effects of grazing activity and photosynthate production by the mixotrophic cryptophytes and autotrophic haptophytes, respectively. As this microbial food web exists in all aquatic systems, understanding the microbial interactions and how they are affected by climate change in the simplified and closed MDV lake system has implications for other low temperature aquatic systems. Efforts should be made to identify the organisms composing the understudied chemolithoautotrophic community. In addition, if isolates of chemolithoautotrophic organisms harboring different forms of RubisCO (i.e., cbbM or form IA/B rbcL) are obtained, they could be used to determine their relative contributions to carbon fixation in each of the MDV lakes. Not only will this enhance the picture of whole lake carbon cycling, but will further knowledge of how these Antarctic lake food webs are sustained during the austral winter. Further analyses at the functional level in the dry valley lake microbial communities included measurements of RubisCO activity compared to heterotrophic enzyme (βGAM) activity within the water columns (Chapter 5). This is the first investigation of these enzyme activities in the MDV lakes. Additionally, no prior study measuring these enzyme activities concomitantly has been attempted. We found that RubisCO activity positively correlated with PAR and chl a for Lakes Bonney and Vanda, and to nutrients for Lake Fryxell (Table 5.2), corroborating results

129 in Chapter 4. βGAM activity positively correlated to chl a and green algae in Lake Bonney, nutrients in Lake Fryxell, and green algae, nutrients, and bacteria counts in Lake Vanda (Table 5.2). RubisCO activity was highest at mid-depths in Lake Bonney and at deep depths in Lakes Fryxell and Vanda. This suggests that dark carbon fixation was occurring below the photic zone, 12 m and 72 m, in Lakes Fryxell and Vanda, respectively (Fig. 5.4 and 5.5), and to a lesser extent in east and west lobes Lake Bonney (Figure 5.3). Future studies to quantify dark carbon fixation by 14C incubations would provide a deeper understanding of carbon cycling in the lakes, as this process may be critical during the austral winter. Isolation of other photosynthetic microbial eukaryotes from the lake communities, such as, cryptophytes, chrysophytes, stramenopiles, and haptophytes, would provide new sources for studying RubisCO and may lead to new discoveries in the natural variation of this enzyme (Parry et al. 2003). RubisCO carbamylation assays should be carried out under a variety of growth temperature and light conditions as was carried out for UWO241 in Chapter 3. RubisCO activity and regulation will likely differ between the organisms (MacIntyre et al. 1997, Andersson and Backlund 2008) and these experiments would contribute to the body of knowledge of RubisCO and its regulation. This study has provided a foundation on which we can spatially and temporally model carbon fixation in the environmentally sensitive MDV lakes. We have gained insight on carbon fixation in the natural environment of a green algal isolate, UWO241, as well as the microbial eukaryote community via in situ RubisCO activity. With this information, a spatial model of community-level carbon fixation through the Lake Bonney water column was developed. Moreover, targeted studies of UWO241 have allowed for a seasonal or temporal model of carbon fixation. UWO241 belongs to the community of organisms detected by the marker gene, form IA/B rbcL, which was found to be positively correlated with PAR in Lake Bonney (Chapter 4). However, dark carbon fixation is likely to occur at depths below the photic zone in this lake (Chapter 5) (Fig. 6.1A and B). The extent of dark carbon fixation, potentially carried out by chemolithoautotrophs harboring cbbM or nifJ, is not currently known. Additionally, whether the spatial distribution of the form IA/B rbcL harboring organisms and chemolithoautotrophic organisms that was observed during the austral summer holds true for the winter is unknown. However, qPCR studies of samples taken during the summer to winter transition show similar trends in form IA/B rbcL and cbbM depth profiles to those observed in the current study (Kong et al. 2012a, b). We predict that green algae are photosynthetically functional under the low light

130 and low temperature environment of Lake Bonney during the austral summer (Fig. 6.1A). This prediction is based on detectable and unaltered RubisCO activity from UWO241 cultures grown at irradiance (8 µmol photons m-2 s-1) and temperature (8 °C) conditions consistent with its natural environment (17 m) (Chapter 3). Although RubisCO sugar inhibitors are likely present within algal cells, they are predicted to be removed from RubisCO by activity of RubisCO activase (Fig. 6.1A). Increased RubisCO abundance is likely an acclimation of this organism to permanent low temperature. Whether other organisms making up the phytoplanktonic community of this lake have also responded to low temperature by increasing RubisCO abundance remains to be determined. While UWO241 RubisCO initial and maximal activity of 8 °C grown cultures is stable, activity of RubisCO from cultures grown at more moderate temperatures (12 °C), is potentially inhibited by sugar inhibitor binding due to temperature sensitivity of RubisCO activase and its subsequent decrease in function (Chapter 3). In this case, RubisCO activase does not keep up with the removal of sugar inhibitors, the production of which may be elevated at moderate temperatures due to catalytic misfire. Similarly, it is tempting to consider a regulatory limitation on RubisCO activity due to inhibitor binding under minimal to no light, and this has been observed in Phaseolus vulgaris (Sharkey et al. 1986). As the removal of sugar inhibitors from RubisCO by RubisCO activase is ATP dependent, this process would be limited under low light conditions (Fig. 6.1B). This would be an effective regulatory mechanism for carbon fixation during dark conditions such as the austral winter and is in line with the proposed down- regulation of photochemistry in UWO241 (Morgan-Kiss et al. 2006). It is unlikely that UWO241 is actively photosynthesizing during the winter; however, further experiments are required to test this prediction and the underlying mechanisms. As this alga is strictly photoautotrophic, it may enter a resting stage during the winter (Morgan-Kiss et al. 2006). This is not likely to have a major impact on the carbon cycle in Lake Bonney as organisms harboring form ID RubisCO are the more abundant and active population in terms of carbon fixation (Kong et al. 2012b). Alternatively, UWO241 may be supplementing carbon fixation by the CBB cycle with β- carboxylation reactions. β-carboxylation involves the fixation of inorganic carbon with phosphoenolpyruvate (PEP) or pyruvate via the enzymes PEP carboxylase and , respectively to generate C4 compounds: malate and aspartate (Appleby et al. 1980). These reactions function to replenish tricarboxylic acid (TCA) cycle intermediates consumed

131

+ during biosynthesis and become activated during NH4 assimilation (Huppe and Turpin 1994, Schuller et al. 1990). Homologues for PEP and pyruvate carboxylases as well as all TCA cycle enzymes were identified in a UWO241 cDNA sequence library (Morgan-Kiss, Kiss & Raymond, in prep) (Table 6.1). Establishing correlations between RubisCO harboring organisms, RubisCO activity, and environmental lake parameters resulted in a better understanding of drivers of carbon fixation. As the lakes showed some differences in their environmental drivers, particularly Lake Fryxell, future climate change effects will vary in their impact on the communities carrying out carbon fixation. Predicted effects of climate change on the MDV lakes include increased nutrient loading due to glacier melt, and subsequent stream formation, and increased light intensity due to thinning ice covers (Foreman et al. 2004, Dolhi et al. 2013). These effects are predicted to have general impacts on the MDV lake communities including increased growth of light dependent carbon fixing organisms and potentially a loss in the stratification of microbial autotrophic communities as incoming nutrients enter at shallow depths increasing the area of overlapping nutrient and light availability. Organisms which are capable of adjusting to the new conditions via acclimation or adaptation are predicted to take over the communities and out-compete specialist organisms (Dolhi et al. 2013). How these changes will affect higher trophic levels and the lake carbon cycle is not yet known.

132

Table 6.1. UWO241 cDNA sequence library (library 1_Parse August08) contigs coding for homologs of β-carboxylation and TCA cycle enzymes verified by blastn.

Contig blastn hit description BLAST accession Identity number number (%) 736 Pyruvate carboxylase C. reinhardtii XM_001696296 73 856 Pyruvate carboxylase Volvox carteri f. XM_002946045.1 75 nagariensis 2243 Phosphoenolpyruvate carboxylase AM689877.2 79 Cyrtococcum patens 2671 , glyoxysomal form C. XM_001695519.1 81 reinhardtii 649 Citrate synthase Coccomyxa subellipsoidea XM_005647609.1 75 859 ATP citrate lyase subunit B C. reinhardtii XM_001701903.1 81 1593 hydratase C. reinhardtii XM_001689650.1 83 321 NAD-dependent C. XM_001694805 79 reinhardtii 931 NAD-dependent isocitrate dehydrogenase C. XM_001697229 80 reinhardtii 1806 NAD-dependent isocitrate dehydrogenase C. XM_001694805 82 reinhardtii 3641 NADP-specific isocitrate dehydrogenase, XM_001698652 82 mitochondrial C. reinhardtii 1374 Dihydrolipoamide succinyltransferase XM_001692487 77 oxogluterate dehydrogenase C. reinhardtii 2786 2-oxogluterate ketogluterate E1 subunit C. XM_001692818.1 79 reinhardtii 1427 Succinate CoA beta chain C. XM_001691530 79 reinhardtii 106 subunit A C. XM_001689790.1 82 reinhardtii 1502 Succinate dehydrogenase subunit D C. XM_001689900 84 reinhardtii 1678 Succinate dehydrogenase subunit b560 XM_001689455.1 79 1721 Fumarate hydratase Volvox carteri f. XM_001696188.1 75 nagariensis 1058 NADP-dependent malic enzyme C. XM_001696363.1 75 reinhardtii 1367 NADP-dependent C. XM_001696734.1 80 reinhardtii 1377 Malate dehydrogenase, cytoplastmic XM_005844687.1 77 Chlorella variabilis 2882, 3310 NAD-dependent malate dehydrogenase C. XM_001696188.1 80, 77 reinhardtii 3576 NAD-dependent malate dehydrogenase C. XM_001702534.1 79 reinhardtii

133

A B

Figure 6.1. Diagram of Lake Bonney spatial trends of form IA/B RubisCO harboring organisms (green algae, cyanobacteria; green line) and those capable of dark carbon fixation (chemolithoautotrophs; blue line) including form II RubisCO harboring organisms and pyruvate:ferredoxin oxidoreductase harboring organisms during the austral summer. The distribution of green algae and cyanobacteria was positively correlated with light availability while the correlation to light was negative for chemolithoautotrophs. Photosynthetic organisms, including green algae and cyanobacteria, are likely to be important in the carbon cycle during the austral summer (A), however dark carbon fixing organisms may play a greater role in the lake carbon cycle in the winter season (B). The green alga C. raudensis UWO241 was isolated from 17 m east lobe Lake Bonney and harbors form IB RubisCO (represented by black shape). Light and moderate temperature may affect the carbamylation state of RubisCO activity in UWO241. The light driven mechanism of regulation is shown here to illustrate seasonal differences in carbon fixation by UWO241. When light is available during the austral summer, RubisCO and the regulatory enzyme RubisCO activase function normally in the presence of sugar phosphate inhibitors (red triangles) (A). During the dark austral winter, ATP-dependent RubisCO activase function is limited and sugar phosphate inhibitors bind tightly to RubisCO inhibiting its carboxylation and oxygenation activity (B).

134

REFERENCES Alexander E, Stock A, Breiner HW, Behnke A, Bunge J, Yakimov MM, and Stoeck T. 2009. Microbial eukaryotes in the hypersaline anoxic L'Atalante deep-sea basin. Environ. Microbiol. 11:360-381. Alfreider A, Schirmer M, and Vogt C. 2012. Diversity and expression of different forms of RubisCO genes in polluted groundwater under different redox conditions. FEMS Microbiol. Ecol. 79:649-660. Alley RB, Marotzke J, Nordhaus WD, Overpeck JT, Peteet DM, Pielke Jr. RA, Pierrehumbert RT, Rhines PB, Stocker TF, Talley LD, and Wallace JM. 2003. Abrupt climate change. Science 299:2005-2010.

Andersson I. 2008. Catalysis and regulation in . J. Exp. Bot. 59:1555-1568.

Andersson I, Backlund A. 2008. Structure and function of rubisco. Plant Physiol. Biochem. 46:275-291. Andreoli C, Moro I, La Rocca N, Valle LD, Masiero L, Rascio N, and Vecchia FD. 2000. Ecological, physiological, and biomolecular surveys on microalgae from Ross Sea (Antarctica). Ital. J. Zool. 67:147-156.

Appleby G, Culbeck J, Holdsworth ES, and Wadman H. 1980. β carboxylation enzymes in marine phytoplankton and isolation and purification of pyruvate carboxylase from Amphidinium carterae (Dinophyceae). J. Phycol. 16:290-295.

Atkin OK, Tjoelker MG. 2003. Thermal acclimation and the dynamic response of plant respiration to temperature. Trends Plant Sci. 8:343-351.

Azam F, Malfatti F. 2007. Microbial structuring of marine ecosystems. Nat. Reivews Microbiol. 5:782-791.

Badger MR, Lorimer GH. 1981. Interaction of sugar phosphates with the catalytic site of ribulose-1,5-bisphosphate carboxylase. 20:2219-2225.

Bar-Even A, Noor E, and Milo R. 2012. A survey of carbon fixation pathways through a quantitative lens. J. Exp. Bot. 63:2325-2342.

Bell EM, Laybourn-Parry J. 1999. Annual plankton dynamics in an Antarctic saline lake. Freshwat. Biol. 41:507-519. Bell EM, Laybourn-Parry J. 2003. Mixotrophy in the Antarctic phytoflagellate Pyramimonas gelidicola. J. Phycol. 39:644-649. Bennett J, Steinback KE, and Arntzen CJ. 1980. Chloroplast phosphoproteins: regulation of excitation energy transfer by phosphorylation of thylakoid membrane polypeptides. Proc. Nat. Acad. Sci. U.S.A. 77:5253-5257.

135

Beuf L, Kurano N, and Miyachi S. 1999. Rubisco activase transcript (rca) abundance increases when the marine unicellular green alga Chlorococcum littorale is grown under high-CO2 stress. Plant Mol. Biol. 41:627-635.

Beutler M, Wiltshire KH, Meyer MA, Moldaenke C, Luring C, Meyerhofer M, Hanesen U- P, and Dau H. 2002. A fluorometric method for the differentiation of algal populations in vivo and in situ. Photosyn. Res. 72:39-53. Bielewicz S, Bell EM, Kong W, Friedberg I, Priscu JC, and Morgan-Kiss RM. 2011. Protist diversity in a permanently ice-covered Antarctic lake during the polar night transition. ISME J. 5:1559-1564.

Bradford MM. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.

Bredemeijer G, Esselink G. 1995. Glucose 6-phosphate dehydrogenase during cold-hardening in Lolium perenne. J. Plant Physiol. 145:565-569.

Brooks A, Portis AR. 1988. Protein-bound ribulose bisphosphate correlates with deactivation of ribulose bisphosphate carboxylase in . Plant Physiol. 87:244-249.

Brown LM. 1982. Photosynthetic and growth responses to salinity in a marine isolate of Nannochloris bacillaris () J. Phycol. 18:483-488. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, TN Pfaffl MW, Shipley GL, Vandersompele J, and Wittwer CT. 2009. The MIQE guidlines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55:611-622. Camacho A, Erez J, Chicote A, Florín M, Squires MM, Lehmann C, and Backofen R. 2001. Microbial microstratification, inorganic carbon photoassimilation and dark carbon fixation at the chemocline of the meromictic Lake Cadagno (Switzerland) and its relevance to the food web. Aquat. Sci. 63:91-106. Campbell BJ, Cary SC. 2004. Abundance of reverse tricarboxylic acid cycle genes in free- living microorganisms at deep-sea hydrothermal vents. Appl. Environ. Microbiol. 70:6282-6289. Canfield DE, Green WJ. 1985. The cycling of nutrients in a closed-basin Antarctic Lake: Lake Vanda. Biogeochem. 1:223-256.

Caron DA, Hutchins DA. 2012. The effects of changing climate on microzooplankton grazing and community structure: drivers, predictions and knowledge gaps. J. Plankton Res. 35:235-252.

Carraro E, Guyennon N, Hamilton D, Valsecchi L, Manfredi EC, Viviano G, Salerno F, Tartari G, and Copetti D. 2012. Coupling high-resolution measurements to a three-dimensional lake model to assess the spatial and temporal dynamics of the cyanobacterium Planktothrix rubescens in a medium-sized lake. Hydrobiol. 698:77-95.

136

Catherine A, Escoffier N, Belhocine A, Nasri A, Hamlaoui S, Yéprémian C, Bernard C, and Troussellier M. 2012. On the use of the FluoroProbe®, a phytoplankton quantification method based on fluorescence excitation spectra for large-scale surveys of lakes and reservoirs. Water Res. 46:1771-1784.

Cavanagh AP, Kubien DS. 2014. Can phenotypic plasticity in rubisco performance contribute to photosynthetic acclimation? Photosynth. Res. 119:203-214.

Chapman W, Walsh J. 2007. A synthesis of Antarctic temperatures. J. Clim. 20:4096-4117.

D'amico S, Collins T, Marx JC, Feller G, and Gerday C. 2006. Psychrophilic microorganisms: challenges for life. EMBO Rep. 7:385-389.

Dahal K, Kane K, Gadapati W, Webb E, Savitch LV, Singh J, Sharma P, Sarhan F, Longstaffe FJ, Grodzinski B, and Hüner NPA. 2012. The effects of phenotypic plasticity on photosynthetic performance in winter rye, winter wheat and Brassica napus. Physiol. Plantarum. 144:169-188.

Darling RB, Friedmann EI, and Broady PA. 1987. Heterococcus endolithicus sp. Nov.(xanthophyceae) and other terrestrial Heterococcus species from Antarctica: morphological changes during life history and response to temperature. J. Phycol. 23:598-607. Daugbjerg N. 2000. Pyramimonas tychotreta, sp. nov. (Prasinophyceae), a new marine species from antarctica: light and electron microscopy of the motile stage and notes on growth rates. J. Phycol. 36:160-171. Daugbjerg N, Moestrup Ø. 1992. Fine structure of Pyramimonas cyclotreta sp. nov. (Prasinophyceae) from Northern Foxe Basin, Arctic Canada, with some observations on growth rates. Eur. J. Protistol. 28:288-298. Daugbjerg N, Moestrup Ø. 1993. Four new species of Pyramimonas (Prasinophyceae) from arctic Canada including a light and electron microscopic description of Pyramimonas quadrifolia sp. nov. Eur. J.Phycol. 28:3-16. Demmig-Adams B, Cohu C, Muller O, and Adams W. 2012. Modulation of photosynthetic energy conversion efficiency in nature: from seconds to seasons. Photosynth. Res. 113:75-88. Depège N, Bellafiore S, and Rochaix J-D. 2003. Role of chloroplast Stt7 in LHCII phosphorylation and state transition in Chlamydomonas. Science 299:1572-1575.

Devos N, Ingouff M, Loppes R, and Matagne RF. 1998. Rubisco adaptation to low temperatures: a comparative study in psychrophilic and mesophilic unicellular algae. J. Phycol. 34:655-660.

Dolhi JM, Ketchum N, and Morgan-Kiss RM. 2012. Establishment of microbial eukaryotic enrichment cultures from a chemically stratified Antarctic lake and assessment of carbon fixation potential. J. Vis. Exp. 1-5.

137

Dolhi JM, Maxwell DP, and Morgan-Kiss RM. 2013. Review: the Antarctic Chlamydomonas raudensis: an emerging model for cold adaptation of photosynthesis. Extremophiles 17:711-722.

Doran PT, McKay CP, Fountain AG, Nylen T, McKnight DM, Jaros C, and Barrett JE. 2008. Hydrologic response to extreme warm and cold summers in the McMurdo Dry Valleys, East Antarctica. Antarct. Sci. 20:499-509.

Doyle S, Dieser M, Broemsen E, and Christner BC. 2011. General characteristics of cold- adapted microorganisms. In Polar Microbiology: Life in a Deep Freeze. Miller RV, and Whyte L (eds.), ASM Press, Washington, D.C. pp. 103-125.

Duff AP, Andrews TJ, and Curmi PMG. 2000. The transition between the open and closed states of rubisco is triggered by the inter-phosphate distance of the bound bisphosphate. J. Mol. Biol. 298:903-916.

Edmondson DL, Badger MR, and Andrews TJ. 1990. Slow inactivation of ribulosebisphosphate carboxylase during catalysis is caused by accumulation of a slow, tight- binding inhibitor at the catalytic site. Plant Physiol. 93:1390-1397.

Elsaied HE, Kimura H, and Naganuma T. 2007. Composition of archaeal, bacterial, and eukaryal rubisco genotypes in three Western Pacific arc hydrothermal vent systems. Extremophiles 11:191-202. Ensminger I, Busch F, and Huner NPA. 2006. Photostasis and cold acclimation: sensing low temperature through photosynthesis. Physiol. Plant 126:28-44. Falkowski PG, Owens TG. 1980. Light – shade adaptation. Plant Physiol. 66:592-595.

Feller U, Crafts-Brandner SJ, and Salvucci ME. 1998. Moderately high temperatures inhibit ribulose-1,5-bisphosphate carboxylase/oxygenase (rubisco) activase-mediated activation of rubisco. Plant Physiol. 116:539-546.

Ferrara M, Guerriero G, Cardi M, and Esposito S. 2013. Purification and biochemical characterisation of a glucose-6-phosphate dehydrogenase from the psychrophilic green alga Koliella antarctica. Extremophiles 17:53-62.

Field CB, Behrenfeld MJ, Randerson JT, and Falkowski P. 1998. Primary production of the biosphere: integrating terrestrial and oceanic components. Science 281:237-240.

Flameling IA, Kromkamp J. 1998. Light dependence of quantum yields for PSII charge separation and oxygen evolution in eucaryotic algae. Limnol. Oceanogr. 43:284-297.

Foreman CM, Wolf CF, and Priscu JC. 2004. Impact of episodic warming events on the physical, chemical and biological relationships of lakes in the McMurdo Dry Valleys, Antarctica. Aquat. Geochem. 10:239-268.

138

Fritsen CH, Adams EE, McKay CP, and Priscu JC. 1988. Permanent ice covers of the McMurdo Dry Valley Lakes, Antarctica: liquid water content. In Ecosystem Dynamics in a Polar Desert: The McMurdo Dry Valleys, Antarctica. Priscu JC (ed.), American Geophysical Union, Washington, D.C. pp. 269-280. Fujii S, Yamamoto R, Nakayama S, Yasui S, and Broady PA. 1999. Effects of salinity on growth and content of intracellular solutes in Heterococcus sp. (Tribonennatales, Xanthophyceae) from Antarctica. Phycol. Res. 47:65-69.

Fujiwara S, Fukuzawa H, Tachiki A, and Miyachi S. 1990. Structure and differential expression of two genes encoding in Chlamydomonas reinhardtii. Proc. Natl. Acad. Sci. U.S.A. 87:9779-9783.

Fulton JR, Mcknight DM, Foreman CM, Cory RM, Stedmon C, and Blunt E. 2004. Changes in fulvic acid redox state through the oxycline of a permanently ice-covered Antarctic lake. Aquat. Sci. 66:27-46.

Galmés J, Aranjuelo I, Medrano H, and Flexas J. 2013. Variation in rubisco content and activity under variable climatic factors. Photosynth. Res. 117:73-90.

Giordano M, Beardall J, and Raven JA. 2005. CO2 concentrating mechanisms in algae: mechanisms, environmental modulation, and evolution. Annu. Rev. Plant Biol. 56:99-131.

Giri BJ, Bano N, and Hollibaugh JT. 2004. Distribution of rubisco genotypes along a redox gradient in Mono Lake, California. Appl. Environ. Microbiol. 70:3443-3448.

Gonzalez JM, Sherr BF, and Sherr EB. 1993. Digestive enzyme activity as a quantitative measure of protistan grazing: the acid lysozyme assay for bacterivory. Mar. Ecol. Prog. Ser. 100:197-206.

Gudynaite-Savitch L, Gretes M, Morgan-Kiss RM, Savitch L V, Simmonds J, Kohalmi SE, and Hüner NPA. 2006. Cytochrome f from the Antarctic psychrophile, Chlamydomonas raudensis UWO 241: structure, sequence, and complementation in the mesophile, Chlamydomonas reinhardtii. Mol. Genet. Genomics 275:387-398. Green W, Lyons W. 2009. The saline lakes of the McMurdo Dry Valleys, Antarctica. Aquat. Geochem. 15:321-348. Green WJ, Gardner TJ, Ferdelman TG, Angle MP, Varner LC, and Nixon P. 1989. Geochemical processes in the Lake Fryxell Basin (Victoria Land, Antarctica). In High Latitude Limnology. Vincent WF, Ellis-Evans JC (eds.), Springer, Netherlands. pp. 129-148.

Hikosaka K, Ishikawa K, Borjigidai A, Muller O, and Onoda Y. 2006. Temperature acclimation of photosynthesis: mechanisms involved in the changes in temperature dependence of photosynthetic rate. J. Exp. Bot. 57:291-302.

139

Holaday AS, Martindale W, Alred R, Brooks AL, and Leegood RC. 1992. Changes in activities of enzymes of carbon metabolism in leaves during exposure of plants to low temperature. Plant Physiol. 98:1105-1114.

Honjoh K-I, Mimura A, Kuroiwa E, Hagisako T, Suga K, Shimizu H, Dubey RS, Miyamoto T, Hatano S, and Iio M. 2003. Purification and characterization of two isoforms of glucose 6- phosphate dehydrogenase (G6PDH) from Chlorella vulgaris C-27. Biosci. Biotech. Bioch. 67:1888-1896.

Houliez E, Lizon F, Thyssen M, Artigas LF, and Schmitt FG. 2012. Spectral fluorometric characterization of haptophyte dynamics using the FluoroProbe: an application in the eastern English Channel for monitoring Phaeocystis globosa. J. Plankton Res. 34:136-151.

Houtz RL, Magnani R, Nayak NR, and Dirk LMA. 2008. Co- and post-translational modifications in rubisco: unanswered questions. J. Exp. Bot. 59:1635-1645.

Hüner NPA, Bode R, Dahal K, Busch FA, Possmayer M, Szyszka B, Rosso D, Ensminger I, Krol M, Ivanov AG, and Maxwell DP. 2013. Shedding some light on cold acclimation, cold adaptation, and phenotypic plasticity. Botany. 91:127-136.

Hüner NPA, Bode R, Dahal K, Hollis L, Rosso D, Krol M, and Ivanov AG. 2012. Chloroplast redox imbalance governs phenotypic plasticity: the “grand design of photosynthesis” revisited. Front. Plant Sci. 3:255.

Hüner NPA, Öquist G, and Sarhan F. 1998. Energy balance and acclimation to light and cold. Trends Plant Sci. 3:224-230. Hüner NPA, Öquist G, and Melis A. 2003. Photostasis in plants, green algae and cyanobacteria: the role of light harvesting antenna complexes. In Advances in Photosynthesis and Respiration Light Harvesting Antennas in Photosynthesis. Green BR, Parson WW (eds.), Kluwer Academic Publishers, Dordrecht, pp. 401-421.

Huppe HC, Turpin DH. 1994. Integration of carbon and nitrogen metabolism in plant and algal cells. Annu. Rev. Plant Physiol. Plant Mol. Biol. 45:577-607.

Inman MS. 2013. Characterization of the from the psychrophilic green alga Chlamydomonas sp. UWO241. Western University.

James MR, Hall JA, and Laybourn-Parry J. 1998. Protozooplankton and microzooplankton ecology in lakes of the dry valleys, southern Victoria Land. In Ecosystem Dynamics in a Polar Desert: the McMurdo Dry Valleys, Antarctica. Priscu JC (ed.), American Geophysical Union, Washington, D.C. pp. 255-267.

Jentsch A, Kreyling J, and Beierkuhnlein C. 2007. A new generation of climate-change experiments: events, not trends. Front. Ecol. Environ. 5:365-374.

140

Joanisse D, Storey K. 1994. Enzyme activity profiles in an overwintering population of freeze- tolerant larvae of the gall fly, Eurosta solidaginis. J. Comp. Physiol. B 164:247-255. John DE, Patterson SS, and Paul JH. 2007. Phytoplankton-group specific quantitative polymerase chain reaction assays for rubisco mRNA transcripts in seawater. Mar. Biotechnol. 9:747-759.

Jordan DB, Chollet R. 1983. Inhibition of ribulose bisphosphate carboxylase by substrate ribulose 1,5-bisphosphate. J. Biol. Chem. 258:13752-13758.

Jordan DB, Ogren WL. 1984. The CO2/O2 specificity of ribulose 1,5-bisphosphate carboxylase/oxygenase. Planta 161:308-313.

Jorgensen B, Postgate J. 1982. Ecology of the bacteria of the sulphur cycle with special reference to anoxic-oxic interface environments. Philos. T. Roy. Soc. B. 298:543-561. Jungblut AD, Vincent WF, and Lovejoy C. 2012. Eukaryotes in Arctic and Antarctic cyanobacterial mats. FEMS Microbiol. Ecol. 82:416-428.

Kane HJ, Wilkin J-M, Portis Jr. AR, and Andrews TJ. 1998. Potent inhibition of ribulose- bisphosphate carboxylase by an oxidized impurity in ribulose-1,5-bisphosphate. Plant Physiol. 117:1059-1069.

Karlsson J, Clarke AK, Chen Z-Y, Hugghins SY, Park Y-I, Husic HD, Moroney JV, and Samuelsson G. 1998. A novel alpha-type carbonic anhydrase associated with the thylakoid membrane in Chlamydomonas reinhardtii is required for growth at ambient CO2. EMBO J. 17:1208-1216. Karr EA, Ng JM, Belchik SM, Sattley WM, Madigan MT, and Achenbach LA. 2006. Biodiversity of methanogenic and other archaea in the permanently frozen Lake Fryxell, Antarctica. Appl. Environ. Microbiol. 72:1663-1666. Karr EA, Sattley WM, Jung DO, Madigan MT, and Achenbach LA. 2003. Remarkable diversity of phototrophic purple bacteria in a permanently frozen Antarctic lake. Appl. Environ. Microbiol. 69:4910-4914. Karr EA, Sattley WM, Rice MR, Jung DO, Madigan MT, and Achenbach LA. 2005. Diversity and distribution of sulfate-reducing bacteria in permanently frozen Lake Fryxell, McMurdo Dry Valleys, Antarctica. Appl. Environ. Microbiol. 71:6353-6359. Kelly GJ, Latzko E, and Gibbs M. 1976. Regulatory aspects of photosynthetic carbon metabolism. Annu. Rev. Plant Physiol. 27:181-205.

Kobza J, Seemann JR. 1988. Mechanisms for light-dependent regulation of ribulose-1,5- bisphosphate carboxylase activity and photosynthesis in intact leaves. Proc. Natl. Acad. Sci. U.S.A. 85:3815-3819.

141

Kong W, Dolhi JM, Chiuchiolo A, Priscu J, and Morgan-Kiss RM. 2012a. Evidence of form II rubisco (cbbM) in a perennially ice-covered Antarctic lake. FEMS Microbiol. Ecol. 82:491- 500.

Kong W, Li W, Prášil O, Romancova I, and Morgan-Kiss RM. 2014. An integrated study of photochemical function and expression of a key photochemical gene (psbA) in photosynthetic communities of Lake Bonney (McMurdo Dry Valleys, Antarctica). FEMS Microbiol. Ecol. In press. Kong W, Nakatsu CH. 2010. Optimization of RNA Extraction for PCR quantification of aromatic compound degradation genes. Appl. Environ. Microbiol. 76:1282-1284. Kong W, Ream DC, Priscu JC, and Morgan-Kiss RM. 2012b. Dynamics of rubisco gene number and expression during the transition to polar night in Antarctic lake. Appl. Env. Microbiol. 78:4358-4366. Koob DD, Leister GL. 1972. Primary productivity and associated physical, chemical and biological characteristics of Lake Bonney: A perennially ice-covered lake in Antarctica. Antar. Res. S. 20:51-68. Kovaleva OL, Tourova TP, Muyzer G, Kolganova TV, and Sorokin DY. 2011. Diversity of rubisco and ATP citrate lyase genes in soda lake sediments. FEMS Microbiol. Ecol. 75:37-47. Laybourn-Parry J. 2002. Survival mechanisms in Antarctic lakes. Philos. T. Roy. Soc. B 357:863-869. Laybourn-Parry J. 2009. No place too cold. Science 324:1521-1522.

Laybourn-Parry J, James MR, McKnight DM, Priscu J, Spaulding SA, and Shiel R. 1997. The microbial plankton of Lake Fryxell, southern Victoria Land, Antarctica during the summers of 1992 and 1994. Polar Biol. 17:54-61.

Laybourn-Parry J, Marshall WA, and Marchant HJ. 2005. nutritional versatility as a key to survival in two contrasting Antarctic saline lakes. Freshwat. Biol. 50:830-838. Laybourn-Parry J, Pearce DA. 2007. The biodiversity and ecology of Antarctic lakes: models for evolution. Philos. T. Roy. Soc. B 362:2273-2289. Laybourn-Parry J, Roberts EC, and Bell EM. 2000. Mixotrophy as a survival strategy in Antarctic lakes. In Antarctic Ecosystems: Models for Wider Ecological Understanding. Davidson W, Howard-Williams C, and Broady PA (eds.), The Caxon Press, Christchurch, pp. 33-40.

Lebaron P, Troussellier M, and Got P. 1994. Accuracy of epifluorescence microscopy counts for direct estimates of bacterial numbers. J. Microbiol. Methods 19:89-94.

Lee PA, Mikucki JA, Foreman CM, Priscu JC, Ditullio GR, Riseman SF, Demora SJ, Wolf CF, and Kester L. 2004a. Thermodynamic constraints on microbially mediated processes in lakes of the McMurdo Dry Valleys, Antarctica. Geomicrobiol. J. 21:1-17.

142

Lee PA, Priscu JC, Ditullio GR, Riseman SF, Tursich N, and Demora SJ. 2004b. Elevated levels of dimethylated-sulfur compounds in Lake Bonney, a poorly ventilated Antarctic lake. Limnol. Oceanogr. 49:1044-1055. Liska AJ, Shevchenko A, Pick U, and Katz A. 2004. Enhanced photosynthesis and redox energy production contribute to salinity tolerance in Dunaliella as revealed by homology-based proteomics. Plant Physiol. 136:2806-2817. Lizotte MP, Priscu JC. 1992. Photosynthesis-irradiance relationships in phytoplankton from the physically stable water column of a perennially ice-covered lake (Lake Bonney, Antarctica). J Phycol. 28:179-185. Lizotte MP, Sharp TR, and Priscu JC. 1996. Phytoplankton dynamics in the stratified water column of Lake Bonney, Antarctica. I. Biomass and productivity during the winter-spring transition. Polar Biol. 16:155-162. Lopez-Garcia P, Rodriguez-Valera F, Pedros-Alio C, and Moreira D. 2001. Unexpected diversity of small eukaryotes in deep-sea Antarctic plankton. Lett. Nat. 409:603-607. Loppes R, Devos N, Willem S, Barthelemy P, and Matagne RF. 1996. Effect of temperature on two enzymes from a psychrophilic Chloromonas (Chlorophyta). J Phycol. 32:276-278. Lovejoy C, Massana R, and Pedros-Alio C. 2006. Diversity and distribution of marine microbial eukaryotes in the Arctic Ocean and adjacent seas. Appl. Environ. Microbiol. 72:3085- 3095. Lovejoy C, Vincent WF, Bonilla S, Roy S, Martineau M-J, Terrado R, Potvin M, Massana R., and Pedrós-Alió C. 2007. Distribution, phylogeny, and growth of cold-adapted picoprasinophytes in arctic seas. J. Phycol. 43:78-89.

Lyons WB, Andrew F, Doran P, Priscu JC, Neumann K, and Welch KA. 2000. Importance of landscape position and legacy: the evolution of the lakes in Taylor Valley, Antarctica. Freshw. Biol. 43:355-367.

Lyons WB, Welch KA, Priscu JC, Laybourn-Parry J, Moorhead D, McKnight DM, Doran PT, and Tranter M. 2001. The McMurdo Dry Valleys Long-Term Ecological Research Program: new understanding of the biogeochemistry of the Dry Valley Lakes: a review. Polar Geogr. 25:202-217.

MacIntyre HL, Sharkey TD, and Geider RJ. 1997. Activation and deactivation of ribulose 1, 5-bisphosphate carboxylase/oxygenase (rubisco) in three marine microalgae. Photosynth. Res. 51:93-106.

Margesin R, Miteva V. 2011. Diversity and ecology of psychrophilic microorganisms. Res. Microbiol. 162:346-361.

Marshall W, Laybourn-Parry J. 2002. The balance between photosynthesis and grazing in Antarctic mixotrophic cryptophytes during summer. Freshw. Biol. 47:2060-2070.

143

Massana R, Pedrós-Alió C. 2008. Unveiling new microbial eukaryotes in the surface ocean. Curr. Opin. Microbiol. 11:213-218. Massana R, Terrado R, Forn I, Lovejoy C, and Pedrós-Alió C. 2006. Distribution and abundance of uncultured heterotrophic flagellates in the world oceans. Environ. Microbiol. 8:1515-1522.

Mataloni G, Tesolín G, and Tell G. 1998. Characterization of a small eutrophic Antarctic lake (Otero Lake, Cierva Point) on the basis of algal assemblages and water chemistry. Polar Biol. 19:107-114.

Maxwell DP, Falk S, and Hüner NPA. 1995. Photosystem II excitation pressure and development of resistance to photoinhibition. Plant Physiol. 107:687-694.

Maxwell DP, Falk S, Trick CG, and Hüner NPA. 1994. Growth a low temperature mimics high-light acclimation in Chlorella vulgaris. Plant Physiol. 105:535-543. McFadden G, Moestrup Ø, and Wetherbee R. 1982. Pyramimonas gelidicola sp. nov. (Prasinophyceae), a new species isolated from Antarctic sea ice. Phycologia 21:103-111.

Meehl GA, Covey C, Taylor KE, Delworth T, Stouffer RJ, Latif M, McAveny B, and Mitchell JF. 2007. The WCRP CMIP3 multimodel dataset: a new era in climate change research. Bull. Am. Meteorol. Soc. 88:1383-1394.

Mikucki JA, Pearson A, Johnston DT, Turchyn AV, Farquhar J, Schrag DP, Anbar AD, Priscu JC, and Lee PA. 2009. A contemporary microbially maintained subglacial ferrous "ocean". Science 324:397-400. Mikucki JA, Priscu JC. 2007. Bacterial diversity associated with Blood Falls, a subglacial outflow from the Taylor Glacier, Antarctica. Appl. Environ. Microbiol. 73:4029-4039.

Mitra M, Lato SM, Ynalvez RA, Xiao Y, and Moroney J V. 2004. Identification of a new chloroplast carbonic anhydrase in Chlamydomonas reinhardtii. Plant Physiol. 135:173-182.

Moorhead DL, Doran PT, Fountain AG, Lyons WB, Mckight DM, Priscu J, Virginia RA, and Wall DH. 1999. Ecological legacies: impacts on ecosystems of the McMurdo Dry Valleys. Biosci. 49:1009-1019.

Moorthi S, Caron DA, Gast RJ, and Sanders RW. 2009. Mixotrophy: a widespread and important ecological strategy for planktonic and sea-ice nanoflagellates in the Ross Sea, Antarctica. Aquat. Microb. Ecol. 54:269-277.

Morgan-Kiss RM, Dolhi JM. 2011. Microorganisms and Plants: a Photosynthetic Perspective. In Nature at Risk: Temperature Adaptation in a Changing Climate. Tanino K, Storey K (eds.), CABI, Cambridge, MA. pp. 24-44.

144

Morgan-Kiss RM, Ivanov AG, and Hüner NPA. 2002a. The Antarctic psychrophile, Chlamydomonas subcaudata, is deficient in state I-state II transitions. Planta 214:435-445.

Morgan-Kiss RM, Ivanov AG, Pocock T, Król M, Gudynaite-Savitch L, and Hüner NPA. 2005. The Antarctic psychrophile, Chlamydomonas raudensis Ettl (UWO241) (Chlorophyceae, Chlorophyta), exhibits a limited capacity to photoacclimate to red light. J. Phycol. 41:791-800.

Morgan RM, Ivanov AG, Priscu JC, Maxwell DP, and Hüner NPA. 1998. Structure and composition of the photochemical apparatus of the Antarctic green alga, Chlamydomonas subcaudata. Photosyn. Res. 56:303-314. Morgan-Kiss R, Ivanov AG, Williams J, Mobashsher K, and Hüner NPA. 2002b. Differential thermal effects on the energy distribution between photosystem II and photosystem I in thylakoid membranes of a psychrophilic and a mesophilic alga. Biochim. Biophys. Acta 1561:251-265. Morgan-Kiss RM, Priscu JP, Pocock T, Gudynaite-Savitch L, and Hüner NPA. 2006. Adaptation and acclimation of photosynthetic microorganisms to permanently cold environments. Microbiol. Mol. Biol. Rev. 70:222-252.

Morris I, Glover HE. 1974. Questions on the mechanism of temperature adaptation in marine phytoplankton. Mar. Biol. 24:147-154.

Mouget J, Beeson RC, Legendre L, and Noüe J. 1993. Inadequacy of rubisco initial and total activities to account for observed rates of photosynthetic assimilation by Scenedesmus ecornis. Eur. J. Phycol. 28:99-106.

Mueller-Cajar O, Stotz M, Wendler P, Hartl FU, Bracher A, and Hayer-Hartl M. 2011. Structure and function of the AAA+ protein CbbX, a red-type rubisco activase. Nature 479:194- 199.

Muyzer G, Waal ECD, and Uitterlinden AG. 1993. Profiling of complex microbial populations by denaturing gradient analysis of polymerase chain reaction amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59:695-700.

Napolitano MJ, Shain DH. 2004. Four kingdoms on glacier ice: convergent energetic processes boost energy levels as temperatures fall. Proc. R Soc. Lond. B 271:273-276.

Napolitano MJ, Shain DH. 2005. Distinctions in adenylate metabolism among organisms inhabiting temperature extremes. Extremophiles 9:93-98. Neale PJ, Priscu JC. 1995. The photosynthetic apparatus of phytoplankton from a perennially ice-covered Antarctic lake: acclimation to an extreme shade environment. Plant Cell Physiol. 36:253-263.

Nichols HW, Bold HC. 1965. Gen. et Sp. Nov. J. Phycol. 1:34-38.

145

Not F, Del Campo J, Balague V, De Vargas C, and Massana R. 2009. New insights into the diversity of marine picoeukaryotes. Plos One 4:e7143. Osmond CB. 1981. Photorespiration and photoinhibition some implications for the energetics of photosynthesis. Biochim. Biophys. Acta 639:77-98. Palmqvist K, Sundblad LG, Wingsle G, and Samuelsson G. 1990. Acclimation of photosynthetic light reactions during induction of inorganic carbon accumulation in the green alga Chlamydomonas reinhardtii. Plant Physiol. 94:357-366.

Papaleo E, Tiberti M, Invernizzi G, Pasi M, and Ranzani V. 2011. Molecular determinants of enzyme cold adaptation: comparative structural and computational studies of cold- and warm- adapted enzymes. Curr. Protein Pept. Sci. 12:657-683.

Parry MAJ, Andralojc PJ, Scales JC, Salvucci ME, Carmo-Silva AE, Alonso H, and Whitney SM. 2013. Rubisco activity and regulation as targets for crop improvement. J. Exp. Bot. 64:717-730.

Parry BR, Shain DH. 2011. Manipulations of AMP metabolic genes increase growth rate and cold tolerance in : implications for psychrophilic evolution. Molec. Biol. Evol. 28:2139-2145. Parry MA, Keys AJ, Madgwick PJ, Carmo-Silva AE, and Andralojc PJ. 2008. Rubisco regulation: a role for inhibitors. J. Exp. Bot. 59:1569-1580. Perga M-E, Domaizon I, Guillard J, Hamelet V, and Anneville O. 2013. Are cyanobacterial blooms trophic dead ends? Oecologia 172:551-562. Peterhansel C, Maurino VG. 2011. Photorespiration redesigned. Plant Physiol. 155:49-55.

Peterhansel C, Niessen M, and Kebeish RM. 2008. Review metabolic engineering towards the enhancement of photosynthesis. Photochem. Photobiol. 84:1317-1323.

Piquet AM, Bolhuis H, Davidson AT, Thomson PG, and Buma AG. 2008. Diversity and dynamics of Antarctic marine microbial eukaryotes under manipulated environmental UV radiation. FEMS Microbiol. Ecol. 66:352-366. Pocock T, Lachance M-A, Proschold T, Priscu JC, Kim S, and Hüner NPA. 2004. Identification of a psychrophilic green alga from Lake Bonney Antarctica: Chlamydomonas raudensis ETTL. (UWO 241) (Chlorophyceae). J. Phycol. 40:1138-1148. Pocock T, Vetterli A, and Falk S. 2011. Evidence for phenotypic plasticity in the Antarctic extremophile Chlamydomonas raudensis Ettl. UWO 241. J. Exp. Bot. 62:1169-1177.

Portis AR. 1981. Evidence of a low stromal Mg2+ concentration in intact in the dark. Plant Physiol. 67:985-989.

Portis Jr. AR. 2003. Rubisco activase-rubisco’s catalytic chaperone. Photosynth. Res. 75:11-27.

146

Portis Jr. AR, Parry MAJ. 2007. Discoveries in rubisco (ribulose 1,5-bisphosphate carboxylase/oxygenase): a historical perspective. Photosynth. Res. 94:121-143.

Possmayer M, Berardi G, Beall BFN, Trick CG, Hüner NPA, and Maxwell DP. 2011. Plasticity of the psychrophilic green alga Chlamydomonas raudensis (UWO241) (Chlorophyta) to supraoptimal temperature stress. J. Phycol. 47:1098-1109.

Priscu JC. 1991. Variation in light attenuation by the permanent ice cap of Lake Bonney during spring and summer. Antarct. J. U.S. 26:223-224. Priscu JC. 1995. Phytoplankton nutrient deficiency in lakes of the McMurdo Dry Valleys, Antarctica. Freshw. Biol. 34:215-227. Priscu JC. 1997. The biogeochemistry of nitrous oxide in permanently ice-covered lakes of the McMurdo Dry Valleys, Antarctica. Glob Change Biol. 3:301-305. Priscu JC, Downes MT, and McKay CP. 1996. Extreme supersaturation of nitrous oxide in a poorly ventilated Antarctic lake. Limnol. Oceanogr. 41:1544-1551. Priscu JC, Palmisano AC, Priscu LR, and Sullivan CW. 1989. Temperature dependence of inorganic nitrogen uptake and assimilation in Antarctic sea-ice microalgae. Polar Biol. 9:442- 446.

Priscu JC, Wolf CF, Takacs CD, Fritsen CH, Laybourn-Parry J, Roberts EC, Sattler B, and Lyons B. 1999. Carbon transformations in a perenially ice-covered Antarctic lake. Biosci. 49:997-1008.

Raina J-B, Dinsdale E, Willis B, and Bourne D. 2010. Do the organic sulfur compounds DMSP and DMS drive coral microbial associations? Trends Microbiol. 18:101-108.

Raines CA. 2003. The Calvin cycle revisited. Photosynth. Res. 75:1-10.

Raven JA, Beardall J. 2003. Carbon Acquisition Mechanisms of Algae: Carbon Dioxide Diffusion and Carbon Dioxide Concentrating Mechanisms. In Advances in Photosynthesis and Respiration Photosynthesis in Algae. Larkum AWD, Douglas SE, Raven JA (eds.), Kluwer Academic Publishers, Dordrecht, pp. 225-244.

Raven JA, Geider RJ. 2003. Adaptation, Acclimation and Regulation in Algal Photosynthesis. In Advances in Photosynthesis and Respiration Photosynthesis in Algae. Larkum AWD, Douglas SE, Raven JA (eds.), Kluwer Academic Publishers, Dordrecht, pp. 385-412.

Read BA, Tabita RF. 1994. High substrate specificity factor ribulose bisphosphate carboxylase/oxygenase from eukaryotic marine algae and properties of recombinant cyanobacterial rubisco containinig “algal” residue modifications. Arch. Biochem. Biophys. 312:210-218.

147

Rigano M, Vona V, Lobosco O, Carillo P, and Rigano C. 2006. Temperature dependence of in the psychrophilic unicellular alga Koliella antarctica and the mesophilic alga Chlorella sorokiniana. Plant Cell Environ. 29:1400-1409.

Roberts EC, Laybourn-Parry J. 1999. Mixotrophic cryptophytes and their predators in the Dry Valley lakes of Antarctica. Freshw. Biol. 41:737-746.

Roberts EC, Priscu JC, and Laybourn-Parry J. 2004a. Microplankton dynamics in a perennially ice-covered Antarctic lake-Lake Hoare. Freshwat. Biol. 27:238-249.

Roberts EC, Priscu JC, Wolf C, Lyons WB, and Laybourn-Parry J. 2004b. The distribution of microplankton in the McMurdo Dry Valley Lakes, Antarctica: response to ecosystem legacy or present-day climatic controls? Polar Biol. 27:238-249.

Robinson SP, Portis Jr. AR. 1989. Ribulose 1,5-bisphosphate carboxylase/oxygenase activase protein prevents the in vitro decline in activity of ribulose 1,5-bisphosphate carboxylase/oxygenase. Plant Physiol. 90:968-971.

Roesler KR, Ogren WL. 1990. Primary structure of Chlamydomonas reinhardtii ribulose 1,5- bisphosphate carboxylase/oxygenase activase and evidence for a single polypeptide. Plant Physiol. 94:1837-1841.

Roubeix V, Mazzella N, Schouler L, Fauvelle V, Morin S, Coste M, Delmas F, and Margoum C. 2011. Variations of periphytic diatom sensitivity to the herbicide diuron and relation to species distribution in a contamination gradient: implications for biomonitoring. J. Environ. Monitor. 13:1768-1774.

Rozen S, Skaletsky H. 2000. Primer3 on WWW for general users and for biologist programmers. In Bioinformatics Methods and Protocols: Methods in Molecular Biology. Krawetz S, Misener S (eds.), Humana Press, Totowa, NJ. pp. 365-386.

Rundle SJ, Zielinskis RE. 1991. Alterations in barley ribulose 1,5-bisphosphate carboxylase/oxygenase activase gene expression during development and in response to illumination. J. Biol. Chem. 266:14802-14807.

Ruuska SA, Andrews TJ, Badger MR, Price GD, and von Caemmerer S. 2000. The role of chloroplast electron transport and metabolites in modulating rubisco activity in tobacco. Insights from transgenic plants with reduced amounts of cytochrome b/f complex or glyceraldehyde 3- phosphate dehydrogenase. Plant Physiol. 122:491-504. Sage RF, Way DA, and Kubien DS. 2008. Rubisco, rubisco activase, and global climate change. J. Exp. Bot. 59:1581-1595.

Salvucci M, Crafts-Brandner SJ. 2004a. Mechanism for deactivation of rubisco under moderate heat stress. Physiol. Plant. 122:513-519.

148

Salvucci ME, Crafts-Brandner SJ. 2004b. Relationship between the heat tolerance of photosynthesis and the thermal stability of rubisco activase in plants from contrasting thermal environments. Plant Physiol. 134:1460-1470.

Salvucci ME, Osteryoung KW, Crafts-Brandner SJ, and Vierling E. 2001. Exceptional sensitivity of rubisco activase to thermal denaturation in vitro and in vivo. Plant Physiol. 127:1053-1064.

Sanders RW, Berninger U-G, Lin LE, Kemp PF, and Caron DA. 2000. Heterotrophic and mixotrophic nanoplankton predation on picoplankton in the Sargasso Sea and on Georges Bank. Mar. Ecol. Prog. Ser. 192:103-118.

Sattley WM, Madigan MT. 2006. Isolation, characterization, and ecology of cold-active, chemolithotrophic, sulfur-oxidizing bacteria from perennially ice-covered Lake Fryxell, Antarctica. Appl. Environ. Microbiol. 72:5562-5568.

Savitch LV, Leonardos ED, Krol M, Jansson S, Grodzinski B, Hüner NPA, and Ӧquist G. 2002. Two different strategies for light utilization in photosynthesis in relation to growth and cold acclimation. Plant Cell Environ. 25:761-771.

Savitch LV, Maxwell DP, and Hüner NPA. 1996. Photosystem II excitation pressure and photosynthetic carbon metabolism in Chlorella vulgaris. Plant Physiol. 111:127-136.

Schrader SM, Kane HJ, Sharkey TD, and von Caemmerer S. 2006. High temperature enhances inhibitor production but reduces fallover in tobacco rubisco. Funct. Plant Biol. 33:921- 929.

Schuller KA, Plaxton WC, and Turpin DH. 1990. Regulation of phosphoenolpyruvate carboxylase from the green alga Selenastrum minutum. Plant Physiol. 93:1303-1311.

Seaburg KG, Parker BC, Prescott GW, and Whitford LA. 1979. The algae of southern Victorialand, Antarctica. A taxonomic and distributional study. In Cramer J. (ed.), Bibliothecia Phycologica. p. 166.

Seemann JR, Berry JA, Freas SM, and Krump MA. 1985. Regulation of ribulose bisphosphate carboxylase activity in vivo by a light-modulated inhibitor of catalysis. Proc. Natl. Acad. Sci. U.S.A. 82:8024-8028.

Seemann JR, Kobza J, and Moore BD. 1990. Metabolism of 2-carboxyarabinitol 1-phosphate and regulation of ribulose 1,5-bisphosphate carboxylase activity. Photosynth. Res. 23:119-130.

Sharkey TD. 1985. Photosynthesis in intact leaves of C3 plants: physics, physiology and rate limitations. Bot. Rev. 51:53-105.

149

Sharkey TD, Seemann JR, and Berry JA. 1986. Regulation of ribulose 1,5-bisphosphate carboxylase activity in response to changing partial pressure of O2 and light in Phaseolus vulgaris. Plant Physiol. 81:788-791.

Sherr EB, Sherr BF. 2002. Significance of predation by protists in aquatic microbial food webs. 81:293-308.

Shively JM, Van Keulen G, and Meijer WG. 1998. Something from almost nothing: carbon dioxide fixation in chemoautotrophs. Annu. Rev. Microbiol. 52:191-230.

Siddiqui KS, Cavicchioli R. 2006. Cold-adapted enzymes. Annu. Rev. Biochem. 75:403-433.

Siddiqui KS, Williams TJ, Wilkins D, Yau S, Allen MA, Brown MV, Lauro FM, and Cavicchioli R. 2013. Psychrophiles. Annu. Rev. Earth Planet. Sci. 41:87-115.

Siglioccolo A, Gerace R, and Pascarella S. 2010. “Cold spots” in protein cold adaptation: Insights from normalized atomic displacement parameters (B′-factors). Biophys. Chem. 153:104- 114. Spalding MH, Critchley C, Govindjee, and Orgren WL. 1984. Influence of carbon dioxide concentration during growth on fluorescence induction characteristics of the green alga Chlamydomonas reinhardii. Photosynth. Res. 5:169-176. Spigel RH, Priscu JC. 1996. Evolution of temperature and salt structure of Lake Bonney, a chemically stratified Antarctic lake. Hydrobiology 321:177-190.

Spigel RH, Priscu JC. 1998. Physical limnology of the McMurdo Dry Valley lakes. In Ecosystem Dynamics in a Polar Desert: the McMurdo Dry Valleys, Antarctica. Priscu JC (ed.), Am. Geophys. Union, Washington, D.C.

Spreitzer RJ, Salvucci ME. 2002. Rubisco: structure, regulatory interactions, and possibilities for a better enzyme. Annu. Rev. Plant Biol. 53:449-475.

Stec B. 2012. Structural mechanism of rubisco activation by carbamylation of the active site lysine. Proc. Natl. Acad. Sci. U.S.A. 109:18785-18790.

Stotz M, Mueller-Cajar O, Ciniawsky S, Wendler P, Hartl FU, Bracher A, and Hayer- Hartl M. 2011. Structure of green-type rubisco activase from tobacco. Nat. Struct. Mol. Biol. 18:1366-1370.

Strand A, Hurry V, Henkes S, Hüner N, Gustafsson P, Gardeström P, and Stitt M. 1999. Acclimation of Arabidopsis leaves developing at low temperatures. Increasing cytoplasmic volume accompanies increased activities of enzymes in the Calvin cycle and in the sucrose- biosynthesis pathway. Plant Physiol. 119:1387-1397.

Strickland J, Parsons T. 1972. A practical handbook of seawater analysis. Fisheries Research Board of Canada, Ottawa.

150

Strom SL, Benner R, Ziegler S, and Dagg MJ. 1997. Planktonic grazers are a potentially important source of marine dissolved organic carbon. Limnol. Oceanogr. 42:1364-1374.

Szyszka, B, Ivanov, AG, and Hüner, NPA. 2007. Psychrophily induces differential energy partitioning, photosystem stoichiometry and polypeptide phosphorylation in Chlamydomonas raudensis. Biochim. Biophys. Acta 1767:789-800.

Tabita FR, Satagopan S, Hanson TE, Kreel NE, and Scott SS. 2008. Distinct form I, II, III, and IV rubisco proteins from the three kingdoms of life provide clues about rubisco evolution and structure/function relationships. J. Exp. Bot. 59:1515-1524.

Tamura K, Peterson D, Peterson N, Stecher G, Nei M, and Kumar S. 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28:2731-2739. Takizawa K, Takahashi S, Hüner NPA, and Minagawa J. 2009. Salinity effects the photoacclimation of Chlamydomonas raudensis Ettl UWO241. Photosyn. Res. 99:195-203.

Taylor TC, Backlund A, Bjorhall K, Spreitzer RJ, and Andersson I. 2001. First crystal structure of rubisco from a green alga, Chlamydomonas reinhardtii. J. Biol. Chem. 276:48159- 48164.

Thomas DN, Baumann MEM, and Gleitz M. 1992. Efficiency of carbon assimilation and photoacclimation in a small unicellular Chaetoceros species from the Weddell Sea (Antarctica): influence of temperature and irradiance. J. Exp. Mar. Bio. Ecol. 157:195-209.

Thurman J, Parry J, Hill PJ, Priscu JC, Vick TJ, Chiuchiolo A, and Laybourn-Parry J. 2012. Microbial dynamics and flagellate grazing during transition to winter in Lakes Hoare and Bonney, Antarctica. FEMS Microbiol. Ecol. 82:449-458.

To K-Y, Suen D-F, and Chen S-C. 1999. Molecular characterization of ribulose 1,5- bisphosphate carboxylase/oxygenase activase in rice leaves. Planta 209:66-76.

Tortell PD, Martin CL, and Corkum ME. 2006. Inorganic carbon uptake and intracellular assimilation by subarctic Pacific phytoplankton assemblages. Limnol. Oceanogr. 51:2102-2110.

Trevena A, Jones G, Wright S, and Van Den Enden R. 2000. Profiles of DMSP, algal pigments, nutrients and salinity in pack ice from eastern Antarctica. J. Sea Res. 43:265-273.

Van de Loo F, Salvucci M. 1996. Activation of ribulose-1,5-biphosphate carboxylase/oxygenase (rubisco) involves rubisco activase Trp16. Biochemistry 35:8143-8148.

Van Der Wielen PW. 2006. Diversity of ribulose 1,5-bisphosphate carboxylase/oxygenase large-subunit genes in the MgCl2-dominated deep hypersaline anoxic basin discovery. FEMS Microbiol. Lett. 259:326-331.

151

Vick T, Priscu J. 2012. Bacterioplankton productivity in lakes of the Taylor Valley, Antarctica, during the polar night transition. Aquat. Microb. Ecol. 68:77-90.

Vick-Majors TJ, Priscu JC, and Amaral-Zettler LA. 2014. Modular community structure suggests metabolic plasticity during the transition to polar night in ice-covered Antarctic lakes. ISME J. 8:778-789.

Vila-Costa M, Bartrons M, Catalan J, and Casamayor EO. 2014. Nitrogen-cycling genes in epilithic biofilms of oligotrophic high-altitude lakes (Central Pyrenees, Spain). Microb. Ecol. 68:60-69.

Vincent W, Downes MT, and Vincent CL. 1981. Nitrous oxide cycling in Lake Vanda, Antarctica. Nature 292:618-620.

Vincent W, James M. 1996. Biodiversity in extreme aquatic environments: lakes, ponds and streams of the Ross Sea sector, Antarctica. Biodivers. Conserv. 5:1451-1471.

Vincent W, Vincent CL. 1982. Factors controlling phytoplankton production in Lake Vanda (77S). Can. J. Fish. Aquat. Sci. 39:1602-1609.

Voytek MA, Priscu JC, and Ward BB. 1999. The distribution and relative abundance of ammonia-oxidizing bacteria in lakes of the McMurdo Dry Valley, Antarctica. Hydrobiologia 401:113-130.

Voytek MA, Ward BB. 1995. Detection of ammonium-oxidizing bacteria of the beta-subclass of the class proteobacteria in aquatic samples with the PCR. Appl. Environ. Microbiol. 61:1444- 1450.

Vrba J, Simek K, Nedoma J, and Hartman P. 1993. 4-Methylumbelliferyl-B-N- acetylglucosaminide hydrolysis by a high-affinity enzyme, a putative marker of protozoan bacterivory. Appl. Environ. Microbiol. 59:3091-3101.

Vrba J, Simek K, Pernthaler J, and Psenner R. 1996. Evaluation of extracellular, high- affinity-B-N-acetylglucosaminidase measurements from freshwater lakes: an enzyme assay to estimate protistan grazing on bacteria and picocyanobacteria. Microb. Ecol. 32:81-97.

Walsh J. 2009. A comparison of Arctic and Antarctic climate change, present and future. Antarct. Sci. 21:179–188.

Werneke JM, Chatfield JM, and Ogren WL. 1988. Catalysis of ribulosebisphosphate carboxylase/oxygenase activation by the product of a Rubisco activase cDNA clone expressed in Escherichia coli. Plant Physiol. 87:917-920.

Wharton RAJ, Parker BC, and Simmons GMJ. 1983. Distribution, species composition and morphology of algal mats in Antarctic dry valley lakes. Phycologia 22:355-365.

152

Wilkins D, Yau S, Williams TJ, Allen MA, Brown M V, DeMaere MZ, Lauro FM, and Cavicchioli R. 2012. Key microbial drivers in Antarctic aquatic environments. FEMS Microbiol. Rev. 37:303-335.

Wilson KE, Ivanov AG, Ӧquist G, Grodzinski B, Sarhan F, and Hüner NPA. 2006. Energy balance, organellar redox status and acclimation to environmental stress. Can. J. Bot. 84:1355- 1370. Wollman, FA. 2001. State transitions reveal the dynamics and flexibility of the photosynthetic apparatus. EMBO J. 20:3623-3630. Yamori W, Noguchi K, Hikosaka K, and Terashima I. 2009. Cold-tolerant crop species have greater temperature homeostasis of respiration and photosynthesis than cold-sensitive species. Plant Cell Physiol. 50:203-215. Zhang N, Portis AR. 1999. Mechanism of light regulation of rubisco: a specific role for the larger rubisco activase isoform involving reductive activation by thioredoxin-f. Proc. Natl. Acad. Sci. U.S.A. 96:9438-9443. Zhong X, Berdjeb L, and Jacquet S. 2013. Temporal dynamics and structure of picocyanobacteria and cyanomyoviruses in two large and deep peri-alpine lakes. FEMS Microbiol. Ecol. 86:312-326. Zhu F, Massana R, Not F, Marie D, and Vaulot D. 2005. Mapping of picoeucaryotes in marine ecosystems with quantitative PCR of the 18S rRNA gene. FEMS Microbiol. Ecol. 52:79- 92.

Zubkov MV, Sleigh MA. 1998. Heterotrophic nanoplankton biomass measured by a glucosaminidase assay. FEMS Microbiol. Ecol. 25:97-106.

153