Photoecology of the Antarctic cyanobacterium Leptolyngbya sp. BC1307 brought to light through community analysis, comparative genomics and in vitro photophysiology

Cyanobacteria are important photoautotrophs in extreme environments such as the McMurdo Dry Valleys, Antarctica. Terrestrial Antarctic cyanobacteria experience constant darkness during the winter and constant light during the summer which influences the ability of these organisms to fix carbon over the course of an annual cycle. Here, we present a unique approach combining community structure, genomic and photophysiological analyses to understand adaptation to Antarctic light regimes in the cyanobacterium Leptolyngbya sp. BC1307. We show that Leptolyngbya sp. BC1307 belongs to a clade of cyanobacteria that inhabits near‐surface environments in the McMurdo Dry Valleys. Genomic analyses reveal that, unlike close relatives, Leptolyngbya sp. BC1307 lacks the genes necessary for production of the pigment phycoerythrin and is incapable of complimentary chromatic acclimation, while containing several genes responsible for known photoprotective pigments. Photophysiology experiments confirmed Leptolyngbya sp. BC1307 to be tolerant of short‐term exposure to high levels of photosynthetically active radiation, while sustained exposure reduced its capacity for photoprotection. As such, Leptolyngbya sp. BC1307 likely exploits low‐light microenvironments within cyanobacterial mats in the McMurdo Dry Valleys.

Despite the challenging conditions, many prokaryotic and eukaryotic microorganisms thrive in the McMurdo Dry Valleys. Amongst these are the cyanobacteria, which represent an important component of Antarctic photoautotrophic assemblages. Cyanobacteria are common in many habitats of the McMurdo Dry Valleys, including cryoconite sediment (Porazinska et al., 2004), rocks (Pointing et al., 2009) and deep lakes that remain permanently wet throughout the annual cycle (Laybourn-Parry & Wadham, 2014;Zhang et al., 2015).
The lakes of the McMurdo Dry Valleys are a collection of stratified, perennially ice-covered closed basin lakes that have long been studied in terms of their biological, chemical and physical properties (Laybourn-Parry & Wadham, 2014). They range from freshwater to saline (Green & Lyons, 2009) and exhibit a variety of geochemical properties. Within them, cyanobacteria constitute the major primary producers (Taton, Grubisic, Brambilla, Wit, & Wilmotte, 2003;Zhang et al., 2015), with community dynamics significantly influenced by light availability (Dolhi, Teufel, Kong, & Morgan-Kiss, 2015). In Lake Hoare, a freshwater lake located next to Canada glacier (Wharton et al., 1989), irradiance is 1-40 µmol photons m −2 s −1 at a depth of 13 m, tapering to <1-14 µmol photons m −2 s −1 between 13 and 23 m (Hawes & Schwarz, 1999), making the ability to effectively photosynthesize in low-light essential for cyanobacteria surviving at depths. Conversely, the ice-free moat formed by seasonal melting that surrounds the surface of Lake Hoare experiences irradiances in the range of 140-1,400 µmol photons m −2 s −1 during the austral summer, with an annual mean of 188 µmol photons m −2 s −1 (Hawes & Schwarz, 1999).
Cyanobacterial mats are found in both the depths of lakes and their shallow moats (Jungblut et al., 2016;Mohit, Culley, Lovejoy, Bouchard, & Vincent, 2017;Zhang et al., 2015). Deep mats are distinctly laminated, sometimes forming large pinnacles, while those at lake margins and in shallow pools are less structurally complex.
Microbial mats are heterogeneous at both a macro-and a microscale (Bolhuis, Cretoiu, & Stal, 2014) and generate steep irradiance gradients, with light attenuation of at least 90% occurring 1 mm below the mat surface (Jørgensen & Marais, 1988). As such, the structure of cyanobacterial mats encourages the formation of distinct light microenvironments.
In this study, we investigated the light tolerance of Leptolyngbya sp. BC1307 from the McMurdo Dry Valleys by combining community structure analysis, comparative genomics and photophysiological assessments. We sequenced the genome of Leptolyngbya sp. BC1307 and examined it within the context of known Antarctic cyanobacterial diversity from deep perennially ice-covered lakes and terrestrial and shallow, ice-free environments. To assess light tolerance inferred from genomic analysis, Leptolyngbya sp. BC1307 photophysiology and carotenoid pigment regulation under different irradiance regimes were constrained using a combination of pulse amplitude modulated (PAM) fluorimetry and high-performance liquid chromatography (HPLC). Findings provide first insight into the photoecology of this Antarctic photoautotroph.

| Sampling and cultivation
Leptolyngbya sp. BC1307 was isolated from a 50-mL water sample

| DNA extraction, genome assembly and annotation
Full methods for DNA extraction and genome assembly are de- insert size =400 bp). Assembly was carried out using SPades v3.5 (Bankevich et al., 2012) and noncyanobacterial sequences identified and removed using BLAST and Bandage (Wick, Schultz, Zobel, & Holt, 2015) according to the methods described in Chrismas et al. (2016). Original reads were mapped to the final draft assembly using BBmap revealing an overall coverage of 181.23 ×. The draft assembly of Leptolyngbya sp. BC1307 was annotated using the JGI IMG/ ER pipeline (GOLD Analysis Project ID: Ga0078185) (Markowitz et al., 2012). This Whole Genome Shotgun project is available on GenBank (Accession no. PRJNA399838). Genome sequence and annotation data are publicly available on the JGI IMG/ER database.

| Phylogenetics
To determine the phylogenetic position of Leptolyngbya sp. BC1307

| Community analysis
Cyanobacterial community composition of Antarctic deep lakes and shallow/terrestrial habitats was compared by analysing a combination of previously published SSU rRNA gene clone library data sets.
The following locations were covered: Lake Joyce, Lake Vanda, Lake Hoare (Zhang et al., 2015); Reid lake, Heart Lake (Taton et al., 2006); Lake Fryxell (Taton et al., 2003); Alexander Island ( based on their phylogenetic position and closest named relative previously identified in Antarctica ( Figure S1). Five sequences that could not be classified in this way were excluded from further analysis.

| Comparative genomics
Genomes were investigated using online tools as part of the JGI IMG/ER pipeline. Phycobiliprotein gene clusters were located using sll1578 (CpcA) and SynWH7803_0486 (CpeA) as search queries and compared with those from genomes of closely related organisms isolated from marine pelagic (Leptolyngbya sp. PCC 7375) and intertidal (Leptolyngbya sp. PCC 7335) environments. Carotenoid biosynthesis and photoprotective genes were searched for using queries shown in Table 1. All searches were performed using BLASTp with an e-value threshold of 1e-10. Gene diagrams were generated using genoPlotR (Guy, Roat Kultima, & Andersson, 2010) and manually edited in Inkscape http://inkscape.org/en/.

| Photophysiology and carotenoid regulation
Cultures were established in BG-11 at 15°C (according to optimum growth temperature, Figure S2) in both low light (25 µmol photons m −2 s −1 ) and high light (170 µmol photons m −2 s −1 ) and incubated for a minimum period of three weeks. Cultures in exponential growth phase were examined using a combination of variable chlorophyll fluorescence measurements and HPLC. Rapid light-response curves (Perkins, 2006) were performed using a Walz Water PAM fluorometer on five replicate low-and high-light cultured samples. Given clumping of filaments, samples were gently filtered onto moist (saturated with culture water) GF/F filters prior to rapid light-response curve (RLC) assessment with a red-light fibre-optic emitter/detector unit. Following five minutes of dark adaptation, RLCs were performed using a saturating pulse of ca. 8,600 µmol photons m −2 s −1 , for 600-ms duration, with nine 20 s incrementally increasing light steps from 0 to 1,944 µmol photons m −2 s −1 . Analysis of RLCs followed Perkins, Mouget, Lefebvre, and Lavaud (2006) with iterative curve fitting and calculation of the relative maximum electron transport rate (rETR max ), the theoretical maximum light utilization coefficient (α) and the light saturation coefficient (E k ) following Eilers and Peeters (1988). Additionally, the maximum light utilization efficiency in the dark-adapted state (F v /F m , Genty, Briantais, & Baker, 1989) and Stern-Volmer NPQ were calculated from RLC fluorescence yields. Given Fm′>Fm during RLCs, NPQ was calculated after Serodio, Cruz, Vieira, and Brotas (2005) (2001) HPLC protocol was applied using a c8 column in an Agilent 1,100 HPLC equipped with a diode-array detector. Pigments were identified and quantified against analytical standards from DHI and Sigma using both retention time and spectral analysis.

| Genome statistics and growth characteristics of Leptolyngbya sp. BC1307
Genome statistics for Leptolyngbya sp. BC1307 compared to Leptolyngbya sp. PCC 7335 and Leptolyngbya sp. PCC 7375 are shown in Table S1. Only small differences in % proline content (Leptolyngbya were found between genomes (Table S2). Copies of genes implicated in cold shock response (Barria, Malecki, & Arraiano, 2013;Varin, Lovejoy, Jungblut, Vincent, & Corbeil, 2012) were present in all three genomes at similar numbers (Table S2). Growth was detected as low as 4°C and as high as 24°C with maximum growth rates observed ~15°C-24°C in BG-11 cyanobacterial growth media at pH 7.1.

| Phylogenetics and composition of Antarctic cyanobacterial communities
Maximum-likelihood phylogenetic analysis found Leptolyngbya sp.

| Comparative genomics of phycobiliproteins and carotenoids
While both Leptolyngbya sp. BC1307 and Leptolyngbya sp. PCC 7335 included cpcBA of inducible phycocyanin and genes for associated linker proteins, in Leptolyngbya sp. BC1307 the genes for regulation of CCA (rcaE, rcaF, rcaC and pcyA) (Li et al., 2008) were either truncated or absent (Figure 3). Several core genes involved in phycobiliprotein structure and biosynthesis were shared by all three genomes (Figure 3), and all contained at least one copy of the phycocyanin genes cpcBA and associated structural genes. Leptolyngbya sp. BC1307 did not contain the phycoerythrin genes cpeBA, pebBA or the phycoerythrin linker-polypeptide operon cpeCDESTR (Cobley et al., 2002).

| Photophysiology and carotenoid pigment regulation
Rapid light-response curves revealed significant differences in

Group I Group II
high-light cultured samples (Figure 4a-b). Significantly decreased rates of relative maximum electron transport rate (rETR max ) over RLCs compared to high-light samples (Figure 4d-f), consistent with increased capacity for light-harvesting, electron transport and ability to photoregulate to short-term high PAR exposures.
In contrast, high-light samples achieved comparatively minimal levels of electron transport over RLCs, with significant down-turn in rETR at PARs > 580 µmol photons m −2 s −1 indicating significant photoinhibition (Figure 4a). Though patterns in the light saturation coefficient (E k ) reflected light treatment, that is, lower E k observed for low-light cultured samples, E k did not differ statistically between treatments.
Quenching analysis revealed the mechanisms underlying the ability of low-light samples to rapidly photoacclimate to increasing PAR (Figure 4b). Declines in low-light sample NPQ at the onset of RLCs, which reached a minimum at ~ E K , were consistent with redirection of excitation pressure from PSI to PSII (state II to state I transition) (Campbell et al., 1998). Subsequent steady increase in low-light sample NPQ at PAR>E K , driven by a quenching of maximum fluorescence in actinic light (Fm′, data not shown), was further indicative of induction of OCP-driven NPQ (Kirilovsky & Kerfeld, 2012

H E A T c p c V Y j b I c p e U p e b A p e b B c p c B c p c A c p c E c p c F c p c F c p c E c p c A c p c B c p c V H E A T c p c U c p c T c p c S n b 1 A r c a E r c a F r c a C p c y A c p c B c p c A c p c C c p c C c p c D c p c G r c a E r c a C p c y A c p c B c p c A c p c C c p c C c p c D
use of state transitions in photoacclimation (Figure 4b). However, this was not sufficient to maintain rETR above 580 µmol photons m −2 s −1 (Figure 4a). Given the gradual plateau in high-light NPQ above this PAR, the capacity of state transitions to balance excess excitation between photosystems was likely exhausted, resulting in photoinhibition. No increase in NPQ (or decrease in Fm′) over RLCs for high-light samples further suggested the absence of OCP-driven NPQ.
In addition to Chla, several carotenoid pigments were isolated from Leptolyngbya sp. BC1307 (Figure 5), including antheraxanthin, ß-carotene, canthaxanthin, echinenone, 3-hydroxylechinenone, myxoxanthophyll and zeaxanthin. Chla showed significant difference between high-and low-light cultures ( Figure 5b); therefore, all other pigments are expressed as a ratio to Chla concentrations (Table 2). While no significant difference in ß-carotene, echinenone or myxoxanthophyll:Chla ratios was apparent between F I G U R E 4 Photophysiology of Leptolyngbya sp. BC1307 cultured under low-and high-light conditions. Showing (a) relative electron transport rates (rETR) and (b) nonphotochemical quenching (NPQmax) determined over rapid light-response curves (RLCs), and parameters derived from RLCs, including (c) the maximal quantum efficiency in the dark-adapted state, (d) the maximum relative electron transport rate (rETR max ), (e) the maximum light utilization coefficient (a), and (f) the light saturation coefficient (Ek) (mean ± SE, n = 5). Asterisks denote significant differences in parameters between low-and high-light cultured samples as determined from t tests (see Table 2

| D ISCUSS I ON
Genome analysis of Leptolyngbya sp. BC1307 revealed an absence of clear signals of psychrophily, similar to that found in the Arctic strain Phormidesmis preistleyi BC1401 (Chrismas et al., 2016). At 4.92 Mb, it had the smallest genome of the strains included here (Table S2), although since the genome is currently in draft format this size may be an underestimation. The G-C content of 52.93% was the highest G-C content of all three genomes included in this study. There was no clear distinction in amino acid-related indicators of genomic adaptation to the cold that allow protein flexibility at low temperatures (Table S2) (Feller, Arpigny, Narinx, & Gerday, 1997). However, the carotenoid biosynthesis gene cruP was identified, which has been shown to be upregulated in cold temperatures and has only been found in cyanobacteria in habitats characterized by large temperature fluctuations (Maresca, Graham, Wu, Eisen, & Bryant, 2007). The absence of cruA and presence of crtL suggests that Leptolyngbya sp. BC1307 uses the same pathway for the conversion of lycopene to ß-carotene as picocyanobacteria. Laboratory growth experiments revealed that Leptolyngbya sp. BC1307 is tolerant of a broad range of temperatures ( Figure S2), well above the maximum threshold for psychrophily (15°C). Together, these findings suggest that Leptolyngya sp. BC1307 is not a true psychrophile, in line with the majority of other polar cyanobacteria that have previously been investigated (Chrismas et al., 2016;Tang, Tremblay, & Vincent, 1997). However, other factors remain important in determining the ecology of polar cyanobacteria, with light being key to driving niche differentiation.

| Differences in community composition
Our comparison of clone library data sets (Figure 2) revealed that different lineages of cyanobacteria do not exhibit a uniform distribution between near-surface and deep lake environments. We found a clear switch in dominance between Group I and Group II Leptolyngbya in habitats exposed to the surface and deep perennially ice-covered lakes, respectively (Figure 2). Group I was completely absent from the surface and Group II was completely absent from below the ice, with some co-occurrence of these two groups in shallow seasonally ice-free habitats. Variation in community structure between moat and perennially ice-covered environments has also been identified in Arctic cyanobacterial lakes (Mohit et al., 2017), and these differences are likely driven at least in part by the different irradiance regimes in each habitat, with above and below the ice representing distinct ecological niches.

BC1307 was lost subsequent to its establishment in the McMurdo
Dry Valleys, its absence may have a metabolic benefit by allowing for redistribution of resources away from the phycobilisome complex, thereby assisting survival in the cold. Tang and Vincent (1999) proposed that, like in eukaryotic algae, changes in resource allocation within the photosynthetic apparatus could be a potential adaptive strategy in polar cyanobacteria; by decreasing the size of the phycobilisome complex (Davison, 1991;Geider, 1987) and channelling saved resources to RubisCO (Li & Morris, 1982), cyanobacteria could help to increase RubisCO activity at low temperatures. Lack of phycoerythrin may also have a photophysiological benefit. In the red alga Rhodella violacea, high irradiance leads to a reduction in the phycobilisome through loss of the terminal phycoerythrin hexamer (Bernard, Etienne, & Thomas, 1996;Ritz, Thomas, Spilar, & Etienne, 2000). Therefore, the absence of phycoerythrin from Leptolyngbya sp. BC1307 could help to reduce the potential for photoinhibition by limiting the wavelengths absorbed, thus reducing the potential excitation energy. Future studies examining the down-stream utilization of photochemically derived energy products are required to determine the mechanisms by which Leptolyngbya sp. BC1307 balances light capture with temperature-dependent metabolic processes.

| Photophysiology and carotenoid regulation
The photoecology of Leptolyngbya sp. BC1307 inferred from genomic analysis was further confirmed here through in vitro photophysiological assessment. Typical responses to ambient light intensity were manifest through differential regulation of i) Chla, maximiz-  (Figure 4b), demonstrating the capacity for OCP-driven NPQ in Leptolyngbya sp. for low-and high-light cultured samples (mean ± SE, n = 5). Asterisks denote significant differences in parameters between low-and highlight cultured samples as determined from t tests (see Table 2 M y x o x a n t h o p h y l l Z e a x a n t h i n C a n t h a x a n t h i n Overall, our photophysiology data suggested Leptolyngbya sp. BC1307 to be adapted to a low, variable light regime, rather than a sustained, high irradiance environment such as that experienced in the McMurdo Dry Valleys during the Austral summer. Under lowlight conditions (25 µmol m −2 s −1 ), Leptolyngbya sp. BC1307 was found to harvest light efficiently, and typical photoregulatory mechanisms (i.e., state transitions, OCP-driven NPQ and regulation of light-harvesting and photoprotective pigments) were capable of preventing photoinhibition upon exposure to high irradiances (up to 1922 µmol m −2 s −1 ) for short periods of time. In contrast, when Leptolyngbya sp.
BC1307 was cultured at 170 µmol m −2 s −1 , it could not maintain effective electron transport when exposed to irradiances>580 µmols, with significant photoinhibition occurring above this threshold. This inability to photoregulate after exposure to higher irradiances for extended periods of time is likely due to exhaustion of key carotenoid pigment pools. The carotenoid 3-hydroxyechinenone, which is essential for OCP-driven NPQ (Wilson et al., 2008), was almost entirely absent from high-light cultured samples, resulting in a significant reduction in the photoregulatory capacity of Leptolyngbya sp. BC1307 during sustained periods of high light. In contrast, 3-hydroxyechinenone was found in all low-light cultured samples. Similarly, other key carotenoids known to be upregulated in cyanobacteria in response to high light, for example, myxoxanthophyll (Millie, Ingram, & Dionigi, 1990), were decreased in high-compared to low-light cultures, indicating potential exhaustion of these important pigment pools with sustained high-light conditions.

| Ecological implications
Genome and photophysiological investigations of the present study indicated high-light conditions to be detrimental to Leptolyngbya sp. BC1307. Given the sustained high-light regime apparent during the austral summer in the McMurdo Dry Valleys, Leptolyngbya sp. BC1307 must therefore occupy a low-light microhabitat in the ice-free moat surrounding Lake Hoare. Cyanobacterial mats in these environments are known to provide microhabitats for microbial eukaryotes shielded from their environment  with self-shading an important characteristic of these mats. Antarctic cyanobacterial mats often consist of a carotenoid pigmented surface layer and a deeper phycocyanin rich layer in which most production occurs (Hawes & Schwarz, 1999;Vincent, Downes, Castenholz, & Howard-Williams, 1993). Leptolyngbya sp.
BC1307 may be capable of contributing to both of these layers; organisms at the surface could take on a photoprotective role by producing carotenoids, allowing those in lower layers to retain optimum capacity for photosynthesis with phycocyanin utilizing the red-orange light that remains after shorter wavelengths have been screened. Movement within the mat itself may also be important, and while no motility was observed in culture, it may be possible for Leptolyngbya sp. BC1307 to migrate deeper into the mat, thus selecting for dark conditions. Alternatively, shading may be provided by other members of the community. Microcoleus, which are also present in the type of environment from which Leptolyngbya sp. BC1307 was isolated, have been shown to be capable of migrating within microbial mats (Yallop, Winder, Paterson, & Stal, 1994). The potential therefore exists for these organisms to move to the surface of the mat where they are exposed to most of the light, thus providing shading for more light-sensitive organisms such as Leptolyngbya deeper in the mat. Further investigation of Leptolyngbya and their associated community within active mats is therefore required to fully examine these community effects.

| CON CLUS ION
In this study, we show that genomic information from an organism can be used to infer their information about their ecological niche; by linking comparative genomics, photophysiology and microbial community analysis, ecological insights can be obtained. While within the mats of which they are a component. A huge diversity of Antarctic and Arctic cyanobacteria remains to be explored in this manner, and further investigation of other organisms using the integrated approach applied here will of great benefit to our overall understanding of polar ecology and the evolution of polar cyanobacteria in a changing world.

DATA ACC E S S I B I L I T Y
The genome of Leptolyngbya sp. BC1307 is available on GOLD (GOLD Analysis Project ID: Ga0078185) and GenBank (Accession no.