The enigmatic loss of light‐independent chlorophyll biosynthesis from an Antarctic green alga in a light‐limited environment

Chlorophyll production is a complicated, multifaceted process. Indeed, the cyanobacterial progenitor of chloroplasts bestowed eukaryotic plants and algae with two distinct nonhomologous enzymes for reducing protochlorophyllide to chlorophyllide (the penultimate step of chlorophyll a (chla) biosynthesis): light-dependent (LPOR) and light-independent protochlorophyllide oxidoreductase (DPOR) (Armstrong, 1998; Fujita & Bauer, 2003; Reinbothe et al., 2010). The former, which is encoded by the nuclear gene por, is employed by all photosynthetic eukaryotes explored to date (Hunsperger et al., 2015) and, as its name implies, is only active when its pigment substrate (protochlorophyllide) absorbs light (Griffiths et al., 1996; Shui et al., 2009). DPOR, conversely, is encoded in the chloroplast genome by the genes chlL, chlN, and chlB (Suzuki & Bauer, 1992; Li et al., 1993), has been lost multiple times independently throughout eukaryotic evolution (Fujita & Bauer, 2003; Ueda et al., 2014; Hunsperger et al., 2015; Kim et al., 2017), and can facilitate chlorophyll synthesis in the dark (Shui et al., 2009; Reinbothe et al., 2010). The evolutionary origins of LPOR and DPOR are reflected in how they function today. For instance, DPOR, which predates LPOR, is believed to have evolved from a nitrogenase-like enzyme in anoxygenic photosynthetic bacteria (Reinbothe et al., 1996; Fujita & Bauer, 2003), and is why this enzyme is oxygen sensitive (Yamazaki et al., 2006; Yamamoto et al., 2009; Stolárik et al., 2017). Conversely, LPOR first arose in cyanobacteria (Suzuki & Bauer, 1995) whose oxygenic mode of photosynthesis probably provided strong selective pressures for an enzyme that works well in oxygen-rich conditions, which it does (Reinbothe et al., 1996; Yamazaki et al., 2006; Shui et al., 2009). These two enzymes also differ in their sensitivities to light. Again, LPOR's ability to function is contingent on the absorption of light energy by protochlorophyllide (Griffiths et al., 1996), which has maximum absorbances in both the blue and red regions of the visible light spectrum (Koski & Smith, 1948). But research suggests that LPOR is three to seven times more efficient when protochlorophyllide absorbs red light (647 nm) relative to blue light (407 nm) (Hanf et al., 2012), which penetrates deeper into the water column. Unlike LPOR, the efficiency of DPOR is not impaired by differing wavelengths of visible light. DPOR, however, is dependent on iron for constructing iron–sulfur clusters (Fujita & Bauer, 2000), which is not true of the iron moiety-lacking LPOR. The various differences between LPOR and DPOR can help explain why both enzymes have been maintained in a wide range of photosynthetic eukaryotes for hundreds of millions of years. Nevertheless, DPOR has been lost on multiple occasions. For example, angiosperms and some gymnosperms have surrendered light-independent chlorophyll biosynthesis (Skribanek et al., 2008; Solymosi & Schoefs, 2010; Ueda et al., 2014), as have various algae, with examples from species with primary plastids as well as from those with complex plastids, which are derived from one eukaryote engulfing another (Hunsperger et al., 2015; Kim et al., 2017). The reasons why some lineages have forfeited DPOR while others have retained it are poorly understood. It has been hypothesized that for algae inhabiting iron-depleted environments maintenance of DPOR could be metabolically disadvantageous and, therefore, such conditions might contribute to its loss (Behrenfeld et al., 2006; Bowler et al., 2010; Hunsperger et al., 2015). But for algae living in deep or turbid waters, with limited availability of red light, or those spending extended periods in darkness, having DPOR would seem to be an asset, and could partly account for its widespread conservation across photosynthetic life (Fong & Archibald, 2008; Ueda et al., 2014). One lineage in which the maintenance of DPOR is particularly prevalent is the chlorophycean class of green algae (Hunsperger et al., 2015; Turmel & Lemieux, 2018). The ability to carry out light-independent chlorophyll production is a reoccurring theme throughout this monophyletic group of mostly freshwater flagellates, including in the model species Chlamydomonas reinhardtii, Volvox carteri, and Dunaliella salina (Turmel & Lemieux, 2018). Therefore, it was surprising when a chlorophycean that has lost DPOR was recently identified, a finding made all the more interesting given the environment from which this alga comes. In the McMurdo Dry Valleys of Antarctica sits the perennially ice-covered Lake Bonney, which is home to a diversity of microbial life, despite the harsh conditions (Bielewicz et al., 2011; Kong et al., 2014; Dolhi et al., 2015; Li et al., 2016), including the polyextremophilic green alga Chlamydomonas sp. UWO241 (hereafter UWO241) (Possmayer et al., 2016; Cvetkovska et al., 2017). Lake Bonney is not for the faint-hearted photosynthesizer. Situated c. 17 m below its surface, UWO241 is exposed to continuous cold (c. 5°C year round), high salinity (0.7 M), reduced levels of phosphorus, seasonal extremes in photoperiod (e.g. 24-h darkness during the peak austral winter), and perpetual low irradiance (< 50 μmol photons m−2 s−1), which is biased in the blue-green spectrum (450–550 nm) (Neale & Priscu, 1995). UWO241 is an obligate cold extremophile (psychrophile) and is unable to grow ≥ 18°C. Accordingly, it has evolved an unconventional photosynthetic apparatus, tailored to work best at c. 8°C, but its photosynthetic activity is severely inhibited at moderate temperatures. It can also rapidly repair photosystem II reaction centres, therefore avoiding photoinhibition at low temperatures (Morgan et al., 1998; Pocock et al., 2007; Possmayer et al., 2011). Even more unconventional is the inability of UWO241 to undergo photosynthetic state transitions (from state 1 to state 2), which balance the energy distribution between photosystems I and II (Morgan-Kiss et al., 2002). Instead, UWO241 achieves optimal rates of photosynthesis by maintaining high cyclic electron flow via a novel PSI supercomplex (Szyszka-Mroz et al., 2015). Remarkably, given its light-restricted environment, our present work suggests that this alga has also lost the ability to carry out the light-independent synthesis of chla. Our sequencing, assembly, and annotation of the entire UWO241 chloroplast genome failed to identify the three genes encoding DPOR (chlL, chlN and chlB) from an otherwise standard and full set of chloroplast coding regions (GenBank accession no. MH590838; Supporting Information Fig. S1; Methods S1). Exhaustive searches of our draft nuclear genome and transcriptome sequences for UWO241 also failed to locate chlL, chlN or chlB (Table S1; Methods S1), indicating that these genes have not migrated to the nuclear compartment, which would have been unprecedented if it were the case. Even a search of the mitochondrial DNA (mtDNA) (GenBank accession no. MH598508; Fig. S2), which in some species can sometimes acquire chloroplast genes, came up empty for the three DPOR-encoding loci. We screened the nuclear genome for other genes encoding enzymes involved in chlorophyll biosynthesis and uncovered a conserved pathway between UWO241 and its close relatives (Table S2). In some algae that lack DPOR, including certain haptophytes and stramenopiles, the por gene is duplicated, and it has been hypothesized that this duplication might compensate for the loss of DPOR, potentially by allowing for the differential regulation of por genes (Hunsperger et al., 2015). We found no evidence in the transcriptome or genome data that por is duplicated in UWO241. The single por gene and its deduced amino acid sequence are complete and similar to their counterparts in other chlorophycean algae (Fig. S3) (e.g. 73.4% pairwise identity with Chlamydomonas eustigma). We did, however, uncover two other duplicated genes: GENOMES UNCOUPLED (GUN4), encoding a regulatory subunit of Mg-chelatase that enhances chlorophyll biosynthesis and contributes to retrograde signalling (Formighieri et al., 2012; Brzezowski et al., 2014), and chlorophyllide a oxygenase (CAO), which is responsible for the production of chlb (Tanaka et al., 1998; Bujaldon et al., 2016). To the best of our knowledge, this is the first report of duplication of these genes in a green alga. From these data, it is clear that UWO241 has a functional chlorophyll biosynthesis pathway but has lost DPOR and is solely dependent on LPOR for the enzymatic reduction of protochlorophyllide (Fig. S5). On a side note, the UWO241 ptDNA, at 174 kb, is the second smallest plastome identified from the Chlamydomonadales – a group renowned for harbouring some of the largest plastomes on record, including those of Volvox carteri (c. 525 kb) (Smith & Lee, 2010) and Haematococcus lacustris (1352 kb) (Bauman et al., 2018; Smith, 2018). The comparatively small size of the UWO241 ptDNA reflects its moderate noncoding content (< 50%) rather than the absence of chlL, chlN and chlB, which together represent only a small proportion of the DNA (< 5 kb) of plastomes. It has been argued that the energy limitations from living in a low-light environment contributed to the evolutionary reduction of chloroplast genome size (Marcelino et al., 2016), making the reduced size of UWO241 plastome all the more interesting. The UWO241 mitochondrial genome, however, is the largest and most bloated mitosome observed to date from the Chlamydomonadales (Del Vasto et al., 2015), measuring 59.9 kb and containing more than 75% noncoding DNA (Fig. S2). There are other clear-cut examples of the loss of DPOR in green algae, including from prasinophytes Ostreococcus tauri and Micromonas pusilla (Hunsperger et al., 2015) but, to the best of our knowledge, this is the first concrete case from the Chlorophyceae (Table S3). Plastome sequencing has suggested that DPOR was abandoned in the chlorophyceans Hafniomonas laevis and Neochloris aquatica (Lemieux et al., 2015; Fučíková et al., 2016), as well as certain ulvophytes and pedinophyceans (e.g. Pedinomonas minor) (Turmel et al., 2017); however, nuclear and mitochondrial genome analyses are needed to confirm these findings. It is notable that N. aquatica has been isolated in multiple places on mainland Antarctica, including frozen ponds (Campbell & Claridge, 1987). The absence of DPOR has also been documented in algae with complex plastids, such as the cryptophyte Guillardia theta, the stramenopile Phaeodactylum tricornutum, and the haptophyte Emiliania huxleyi (Fong & Archibald, 2008; Hunsperger et al., 2015; Kim et al., 2017). Most of these algae are marine phytoplankton that are often found in iron-depleted ocean environments where the costs of producing an iron-requiring DPOR protein may outweigh its benefits (Behrenfeld et al., 2006; Bowler et al., 2010). Understanding the loss of light-independent chlorophyll biosynthesis from UWO241 – an alga from a light-limited environment – is not so straightforward. To understand the absence of DPOR in this alga we need to carefully examine its extreme environment. Living in a permanently ice-covered Antarctic lake with continual shading and long periods of sustained darkness would appear to be an ideal place to have DPOR (Ueda et al., 2014). Moreover, the fact that the light penetrating the waters of Lake Bonney is skewed towards the blue-green spectrum where LPOR is thought to be less efficient (Koski & Smith, 1948; Hanf et al., 2012) would seem to make DPOR all the more valuable – not to mention that UWO241 survival depends on blue light, as it reverts to a downregulated photochemical state and is unable to grow in the presence of red light (Morgan-Kiss et al., 2005). So, why has UWO241 lost this important enzyme? The levels of iron in Lake Bonney at the depth at which UWO241 is found are potentially quite low (Ward et al., 2003; Mikucki et al., 2004). But if an iron deficiency contributed to the loss of DPOR in this species then one might also expect other iron-dependent proteins to have been lost or substituted, particularly the replacement of ferredoxin with flavodoxin, which has occurred in certain algae (La Roche et al., 1993, 1995). This is not the case here: UWO241 has two near-identical copies of the ferredoxin gene, and accumulates high amounts of functional ferredoxin protein, which is thought to be an adaptation to the cold (Cvetkovska et al., 2018). That said, UWO241 does display an iron-stressed phenotype and can show signs of iron stress even under conditions that are iron replete (Cook, 2018). Moreover, iron stress can be exacerbated by high salinity (Tripathi et al., 2018) and, as already noted, the depth at which UWO241 was isolated is hypersaline. However, Lake Bonney does have one striking feature that could shed some light on the loss of DPOR in UWO241. The dissolved oxygen concentration over the first 15 m is very high (> 1000 μM), c. 250–350% higher than would be predicted if the lake was not ice covered and saturated with air above its surface (Morgan-Kiss et al., 2006), and it remains high (c. 200% air saturation) at 17 m, where UWO241 is found. This is significant because, as noted earlier, the Achilles' heel of DPOR is its oxygen-sensitive iron−sulfur cluster (Yamazaki et al., 2006; Ueda et al., 2014; Stolárik et al., 2017), whereas LPOR is insensitive to oxygen. If the high dissolved oxygen content of Lake Bonney inhibits the functioning of DPOR, then one would presume that there would be no deleterious effect resulting from a mutation that knocked out the DPOR pathway. This nonadaptive scenario could explain why we did not find the chlL, chlN or chlB genes in UWO241: their sequences have slowly eroded through the steady accumulation of neutral mutations. Of course, if this hypothesis is correct, then the DPOR from not just UWO241 but from other photosynthetic species within Lake Bonney should also be rendered nonfunctional by the high oxygen concentrations of the water above the chemocline and therefore be susceptible to knockout. This reasoning fits well with previous hypotheses arguing that certain land plants and algae have lost DPOR because present-day atmospheric oxygen levels are incompatible with the oxygen-sensitive DPOR enzyme (Reinbothe et al., 1996; Schoefs & Franck, 2003; Hunsperger et al., 2015). Finally, there also is some evidence that DPOR might be cold-sensitive, at least in land plants (Muramatsu et al., 2001). It is also highly possible that the loss of DPOR was not brought about by the conditions of Lake Bonney and that it did not specifically occur in UWO241 or in the ancestral 'Lake Bonney' lineage that gave rise to UWO241. Rather, DPOR may have already been absent in the ancestral 'high-light' Chlamydomonas population that colonised Antarctica and eventually led to the present-day UWO241 strain inhabiting Lake Bonney. Support for this hypothesis comes from phylogenetic analyses showing that the closest known relatives of UWO241 are marine species (e.g. Chlamydomonas parkeae) known to exist in shallow water (Possmayer et al., 2016), which is typically an environment rich in red light and, therefore, one favoring LPOR over DPOR. Currently, there are no available chloroplast genome sequences for any close marine relatives of UWO241. The closest relative for which there are ptDNA data is Chlamydomonas moewusii, a freshwater species encoding DPOR (Boudreau et al., 1994). Lake Bonney harbours a diversity of photosynthetic eukaryotes, including chlorophytes, cryptophytes, haptophytes, and stramenopiles, which have been shown to be vertically stratified within the water column (Bielewicz et al., 2011; Kong et al., 2014; Dolhi et al., 2015; Li et al., 2016). Cryptophytes, for instance, dominate the nutrient-deficient shallower water (6–10 m), whereas haptophytes and stramenopiles occupy the mid-depths (c. 13 m), and chlorophytes reside in the deepest layers (15–20 m) of the photic zone (Bielewicz et al., 2011), below which the lake becomes anoxic. UWO241 is currently the only photosynthetic protist from Lake Bonney to have its chloroplast genome completely sequenced. Therefore, it is not known if other eukaryotic algae in the lake (or any of the other lakes in the McMurdo Dry Valleys) have lost DPOR. One of the most prolific stramenopiles within Lake Bonney belongs to the genus Nannochloropsis (Kong et al., 2012). It is noteworthy, in this context, that the six available plastome sequences from Nannochloropsis species all contain the genes chlL, chlN, and chlB (Wei et al., 2013), but it should be stressed that the presence/absence of DPOR can occur even among members of the same genus. UWO241 is not the only chlamydomonadalean in Lake Bonney. Chlamydomonas sp. ICE-MDV is, in fact, the dominant chlorophyte in the lake (Li et al., 2016). This psychrophile resides at a depth of 13–15 m where the dissolved oxygen concentration is even higher and the iron levels lower than in the deeper photic zone where UWO241 is located. Although data on Chlamydomonas sp. ICE-MDV are limited, it has been shown that it can grow under a broad range of light intensities but cannot grow in the dark in the presence of organic carbon (Li et al., 2016). Given all of this, it will be especially interesting to see if Chlamydomonas sp. ICE-MDV has forfeited DPOR, but this information will only be useful alongside detailed phylogenetic data on this species and on its relationship to UWO241. Its namesake, Chlamydomonas ICE-L (arguably the best studied psychrophilic green alga) does encode DPOR, but unlike UWO241 and ICE-MDV, it is found on Antarctic sea ice where the oxygen concentrations are not extremely high, and it hails from a different chlamydomonadalean clade (the Monadinia) than that of UWO241 (the Moewusinia) (Zhang et al., 2018). Whatever the evolutionary explanation for jettisoning DPOR, surely its absence in UWO241 has impacted this alga's ability to produce chlorophyll and efficiently perform photosynthesis in a light-limited environment. In addition to the loss of DPOR, our genomic characterization of the chlorophyll biosynthesis pathway in UWO241 revealed the duplication of CAO (Figs S4, S5), encoding an indispensable regulatory gene for chlb production. Previous work showed that UWO241 has a normal complement of photosynthetic pigments but exhibits a low chla : b ratio (c. 1.8–2.2) compared with other green algae (> 3). Moreover, chlb is exclusively associated with the light-harvesting antenna around PSII (Pocock et al., 2007; Szyszka et al., 2007). Chlb absorbs blue light efficiently, and higher amounts of this pigment are typically associated with shade adaptation (Falkowski & Owens, 1980). It is tempting to speculate that the duplication of CAO could be responsible for the constitutively increased levels of chlb in UWO241 and could be an adaptation to maximize light absorption in the depths of Lake Bonney, but a detailed functional analysis of the CAO enzymes will be needed to make such assertions. Recent field experiments on the acclimation of natural algal populations and transplanted samples of UWO241 within Lake Bonney demonstrated that seasonal chla accumulation trends were uncoupled from light availability during the polar night transition and continued to increase as light diminished (Morgan-Kiss et al., 2016). The accumulation of chlorophyll at the end of the growing season may be indicative of photoacclimation to extreme shade (i.e. increase in chlorophyll amounts per cell), but exactly how UWO241 achieves this, particularly without the use of DPOR, remains to be determined. This work led to the development of a model suggesting that during light−dark transitions, UWO241 undergoes a cascade of physiological and molecular alterations to the photosynthetic apparatus, keeping it in a downregulated but functional form that can be rapidly reactivated by sunlight (Morgan-Kiss et al., 2006, 2016). Such a strategy could be advantageous in polar environments where the growing season is short. Developing methodologies to study chlorophyll production, retention and degradation in UWO241 and related algae from Lake Bonney during transition to complete darkness would provide the experimental evidence to support this theory. These seasonal trends in chla abundance will become all the more interesting once there are data on the presence−absence of DPOR from additional Lake Bonney phytoplankton. Further work may show that UWO241 is not unique in its inability to carry out light-independent chlorophyll biosynthesis, but there should be some caution in assuming that what is true for UWO241 is also true for its photosynthetic counterparts in Lake Bonney and beyond, especially with respect to photosynthesis. As already mentioned, UWO241 is unique among all explored natural photosynthetic eukaryotes in that it cannot undergo photosynthetic state transitions (i.e. it is permanently locked in state 1) (Morgan-Kiss et al., 2002). Undoubtedly, the more we learn about this cold-loving biflagellate, the more atypical it turns out to be, even compared with other extremophilic Chlamydomonas species. For now, we will have to wait and see just how unconventional the loss of DPOR is in the context of the oxygen-rich lakes of the McMurdo Dry Valleys and to close marine relatives of UWO241. But, hopefully, we will not be in the dark for long. NPAH and DRS are funded by Discovery Grants from NSERC. NPAH is grateful for support through the Canadian Foundation for Innovation and The Canada Research Chairs Program. DRS, MC and NPAH conceived and planned the experiments and wrote the initial draft of the manuscript. SO helped assemble the chloroplast genome. All authors contributed to writing and revising the manuscript. Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Fig. S1 Genetic map of the Chlamydomonas sp. UWO241 chloroplast genome. Fig. Genetic map of the Chlamydomonas sp. UWO241 mitochondrial genome. Fig. of the predicted DPOR amino acid sequences from Chlamydomonas sp. UWO241 and other related green Fig. sequence of the two predicted chlorophyllide a oxygenase proteins from Chlamydomonas sp. UWO241 and related green Fig. model for the chlorophyll biosynthesis pathway in Chlamydomonas sp. UWO241. Methods S1 genome and transcriptome sequencing, and genomic analyses of Chlamydomonas sp. UWO241. S1 of Chlamydomonas sp. UWO241 and sequencing encoding for proteins involved in chlorophyll biosynthesis present in the nuclear genome of Chlamydomonas sp. UWO241. of chlL, chlN, and chlB in complete or genome sequences from chlorophycean green Please note: The is not responsible for the content or functionality of any information supplied by the authors. Any queries (other than missing should be directed to the for the

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.