香京julia种子在线播放

Glycerol is well-known for its ability to mitigate environmental stresses such as freezing and hyperosmotic conditions in a wide variety of organisms, including yeast (Wang et al., 2001), insects (Bentz and Mullins, 1999; Bennett et al., 2005), and vertebrates (Raymond, 1992; Layne, 1999). It is also well-known as an osmoprotectant in some algae, including salt pond isolates of Dunaliella (Ben-Amotz and Avron, 1973; Sussman and Avron, 1981) and several marine isolates of Chlamydomonas (Ahmad and Hellebust, 1986; Miyasaka and Ikeda, 1997; Miyasaka et al., 1998). As a compatible solute, glycerol prevents osmotic water loss and thus a build-up of intracellular salt concentrations. Two psychrophilic Chlamydomonas species, Chlamydomonas sp. UWO241 (Neale and Priscu, 1995) (hereafter Chlamy-UWO) and Chlamydomonas sp. ICE-MDV (Li et al., 2016) (hereafter Chlamy-ICE) were isolated from Lake Bonney, a permanently ice-covered lake in Antarctica. The lake has a steep halocline, in which the salinity reaches about 150 PSU (equivalent to about 2.1 M NaCl, or 4.3 x the salinity of seawater) at a depth of 40 m (Spigel et al., 2018). Chlamy-UWO’s ability to survive in this environment has been intensively studied for over two decades. Among its adaptation are a photochemical apparatus well-suited for a cold, saline and low light environment (Morgan et al., 1998), more fluid membranes (Morgan-Kiss et al., 2002) and ice-binding proteins (IBPs) to prevent freeze-thaw injury (Raymond and Morgan-Kiss, 2013). Less is known about Chlamy-ICE which was isolated more recently: it also possesses IBPs (Raymond and Morgan-Kiss, 2017) but appears to have differences in acclimatory ability compared to Chlamy-UWO (Cook et al., 2019). Much more is known about the cold adaptations of a closely-related Antarctic pyschrophile Chlamydomonas sp. ICE-L (Cvetkovska et al., 2017). Because both Chlamy-UWO and Chlamy-ICE are adapted to a saline environment, we wished to know whether they also produce glycerol as an osmoprotectant or cryoprotectant.

Until recently, the pathways used by algae to produce glycerol have been a matter of speculation. Glycogen, or one of its products, dihydroxyacetone phosphate (DHAP), has been identified as the source of glycerol in a number of organisms, including yeast, (Modig et al., 2007), Arabidopsis (Caparros-Martin et al., 2007), a bacterium (Larrouy-Maumus et al., 2013), mountain pine beetle (Fraser et al., 2017), and rainbow smelt (Raymond, 1995). In the DHAP pathway, DHAP is converted to glycerol-3-phosphatate (G3P) by a nicotinamide-adenine dinucleotide (NAD⁺)-dependent glycerol-3 phosphate dehydrogenase (GPDH), which is then converted to glycerol by a phosphoserine phosphatase (PSP). The problem is that algae typically have several GPDHs and phosphatases, and attempts to identify the enzymes involved have been until recently unsuccessful.

Dunaliella salina has a bidomain enzyme that includes an N-terminal PSP and a C-terminal NAD⁺-dependent GPDH (He et al., 2007). It was proposed as a candidate for glycerol production directly from DHAP (He et al., 2007), but initial attempts to confirm its activity were unsuccessful (He et al., 2009). However, this enzyme still seemed like the most likely source of glycerol. During our study of IBPs in Chlamy-UWO (Raymond and Morgan-Kiss, 2013), we noticed that it produced glycerol as an osmoprotectant, i.e., glycerol production increased with increasing salinity. We searched the Chlamy-UWO genome for a Dunaliella-like bidomain enzyme and found at least two homologs. We then examined the expression of one of the homologs and found that it strongly increased with increasing salinity, as reported in a master’s thesis (Stahl, 2014). This suggested that this enzyme alone was capable of producing glycerol from DHAP

Recently, Morales-Sanchez et al. (2017) showed that a similar bidomain enzyme in Chlamydomonas reinhardtii was sufficient to produce glycerol directly from DHAP, and He Q. H. et al. (2020) finally confirmed that the Dunaliella salina bidomain protein was sufficient to produce glycerol. Here, we report our test of the hypothesis that Chlamy-UWO uses a similar bidomain enzyme to produce glycerol. We also show that Chlamy-ICE is capable of producing high levels of glycerol under saline conditions and has similar bidomain enzymes.

Intracellular glycerol levels in Chlamy-UWO and Chlamy-ICE increased with increasing NaCl concentration. Both species maintained glycerol levels of about 150 mM at 700 mM NaCl (Figure 1A). At 1,300 mM NaCl, Chlamy-UWO cells reached over 400 mM glycerol while Chlamy-ICE cells were unable to grow. In contrast, glycerol levels in the supernatant of centrifuged dense cultures of Chlamy-UWO were very low (0.4 mM at 700 m NaCl and 1.1 mM at 1,300 mM NaCl), indicating that the cells were maintaining a strong gradient between the intracellular and extracellular environments.

We thus searched the Chlamy-UWO transcriptome for enzymes similar to the proposed PSP/GPDH that had been found in Dunaliella (He et al., 2007). Three complete isoforms could be assembled. We selected one of the isoforms to see if it could account for the increased glycerol production at increased salinity. The mRNA expression of Chlamy-UWO isoform 2 increased about four-fold as the NaCl concentration increased to 1,300 mM, regardless of the reference gene (Figure 1B). For the reference gene combination Rps10+Histone, whose points were more linear than those of the single reference genes, the increase was significant at the p <0.01 level (linear regression). This increase closely paralleled the increase in glycerol concentration. This, together with the recent findings in C. reinhardtii and D. salina, strongly supports the idea that the bidomain enzymes are a major source (and possibly the main source) of glycerol in Chlamy-UWO.

When the Chlamy-ICE transcriptome became available in 2016, it was also examined for homologs of the bidomain proteins. It appeared to have seven such homologs, but only five of them could be completely assembled (Table 1). For each of the Lake Bonney species, the expression levels of the different isoforms varied considerably (Table 1). In the approximately 600-a.a. conserved region that includes both enzyme domains, the identities of the Chlamy-UWO and Chlamy-ICE isoforms to C. reinhardtii GPD3 are about 60% (Table 1). To gain insights into the relationships of the enzymes in the three Chlamydomonas species, their exon/intron structures were compared. The numbers of introns (and their locations) differed considerably among the three Chlamydomonas species (Table 1), suggesting that their divergence occurred long ago.

The eight complete isoforms of Chlamy-UWO and Chlamy-ICE consist of an N-terminal chloroplast-targeting signal, a PSP domain and a C-terminal NAD⁺-dependent GDPH domain (Figure 2A), similar to the structure in Dunaliella (He Q. H. et al., 2020). A 3D model of isoform 1 of the Chlamy-UWO enzyme predicted from the structure of the D. salina enzyme is shown in Figure 2B. The region that could be modeled ranged from Thr60 at the start of the PSP domain (lower part of molecule in Figure 2B) to Phe639 near the end of the GPDH domain (upper part). The model is similar to the structure of the Dunaliella protein, with two nearly independent domains connected by a short link. The key residues forming the binding sites of DHAP, NAD⁺, and glycerol-3-P in Dunaliella (He Q. H. et al., 2020) are conserved in the Chlamy-UWO isoform (Figure S1). As a test of the accuracy of the model, the predicted binding sites of DHAP and NAD⁺ on the GPDH domain of the Chlamydomonas protein were compared to those in Dunaliella. The residues that form the binding sites in the two structures as well as their locations are nearly the same (Figure S2), supporting the accuracy of the model.

To better understand the evolution of the Chlamydomonas bidomain proteins, a phylogenetic tree of the conserved PSP/GPDH domains was constructed by the neighbor-joining method (Figure 3). The tree was rooted on homologous proteins from two non-chlorophytes, a rhodophyte (Porphyridium purpureum) and a primitive relative of the fungi (Sphaeroforma arctica) that were recently submitted to GenBank. Within each Chlamydomonas species, the isoforms clustered with high bootstrap values, suggesting that within each species, the isoforms diverged from a single gene. C. reinhardtii, Chlamy-UWO and Chlamy-ICE formed a group that clusters separately from Dunaliella, but otherwise the three Chlamydomonas clusters appear weakly related, in agreement with the considerable differences in exon structures of the three species. These findings, together with the finding of the P. purpureum protein, raise the possibility that the chlorophyte bidomain proteins have an ancient origin, possibly dating back to a common ancestor of the chlorophytes and rhodophytes.

In view of the importance of glycerol as a compatible solute in animal and plant physiology, it is surprising that so little was known about how it was synthesized until recently. Previous studies referred to the enzymes involved only generically without identifying specific genes, and thus not really “nailing down” the pathway. Recently, the specific enzymes involved in glycerol synthesis have been identified in a bacterium (Larrouy-Maumus et al., 2013), a fish (Raymond, 2015), and, as highlighted here, two species of algae (Morales-Sanchez et al., 2017; He Q. H. et al., 2020). The algal enzymes are especially interesting because of their novel form in which the two enzymes needed to convert DHAP into glycerol are fused into a single enzyme. Although these double enzymes are largely limited to chlorophytes, the recent finding of homologous genes in two unicellular organisms (P. purpureum and S. arctica) that have links to some of the earliest eukaryotes (Mendoza et al., 2002; Bhattacharya et al., 2013) suggests that the fused gene may have evolved early in the history of eukaryotes.

The present results show that multiple copies of similar enzymes are found in two glycerol-producing Chlamydomonas species from Antarctica. Furthermore, the expression of a representative isoform from Chlamy-UWO was strongly upregulated by high salinity and the upregulation was associated with an increase in glycerol production. This result was supported by a recent proteomic study of Chlamy-UWO in which enzymes classified as NAD⁺-dependent GPDHs showed a six-fold increase when NaCl concentration was increased from near zero to 700 mM (Kalra et al., 2020). These results indicate that the bidomain enzyme is a major source of glycerol in Chlamy-UWO, and most likely in Chlamy-ICE as well. Using glycerol as an osmoprotectant, Chlamy-UWO and Chlamy-ICE can survive in salinities greater than the salinity of seawater (at least 700 and 1300 mM NaCl for Chlamy-ICE and Chlamy-UWO, respectively; equivalent to about 1.4x and 2.5x the salinity of seawater, respectively). The highest glycerol levels observed in Chlamy-ICE and Chlamy-UWO were 150 and 420 mM, respectively (Figure 1A). (Chlamy-ICE may not need higher levels of glycerol since it appears to live in shallower and thus less saline waters than Chlamy-UWO). For comparison, C. reinhardtii, a non-halotolerant species, can tolerate a maximum salinity of 200 mM NaCl (León and Galván, 1994) (but no more than 100 mM NaCl in our laboratory) and accumulates only about 26 mM glycerol in medium containing 100 mM KCl. (Husic and Tolbert, 1986). On the other hand, the marine coastal C. pulsatilla can produce much higher levels of glycerol, 1,450 mM when grown in double-strength seawater (Ahmad and Hellebust, 1986).

It should be noted that the glycerol levels in the Lake Bonney species are not enough to balance the osmolarity of the external medium. For example, at 1,300 mM NaCl (~2.6 Osm), the glycerol level in Chlamy-UWO was only about 0.4 Osm. Thus, the cells appear to be increasing the concentrations of additional osmolytes to maintain osmotic equilibrium. In addition to glycerol, proline and sucrose were found to accumulate at high levels in high-salt-grown Chlamy-UWO cultures (Kalra et al., 2020). In C. pulsatilla, in which glycerol contributed only about 57% to the intracellular osmolality, large increases in sodium and chloride ions made important contributions to maintaining osmotic equilibrium, but sugars and amino acids did not appear to have a significant role (Ahmad and Hellebust, 1986).

C. reinhardtii has four other GPDHs designated GPD1, GPD4, GPD5, and the mitochondrial mtGPD (Morales-Sanchez et al., 2017). tBlastn analyses of contigs assembled from the Chlamy-UWO and Chlamy-ICE transcriptomes indicate that both species have close homologs of each of these proteins (data not shown). We cannot rule out the possibility that these GPDHs also have a role in glycerol production, although GPD4 in C. reinhardtii was not upregulated by 125 mM NaCl (Morales-Sanchez et al., 2017).

That all the Lake Bonney PSP/GPDH isoforms have a chloroplast targeting signal suggests that they function in the chloroplast, as was concluded for the bidomain enzymes in C. reinhardtii (Morales-Sanchez et al., 2017) and D. salina (He Q. H. et al., 2020). This seems reasonable in view of the fact that the enzymes could act directly on DHAP produced by the Calvin cycle in the chloroplast. However, several chlorophyte bidomain proteins in the databank do not appear to have chloroplast signals as judged by ChloroP, which suggests they function in the cytoplasm.

Because Chlamy-UWO and Chlamy-ICE are living in an environment that is constantly exposed to freezing, as demonstrated by their expression of numerous ice-binding proteins (Raymond and Morgan-Kiss, 2013; Raymond and Morgan-Kiss, 2017), glycerol, by lowering the freezing point of the intracellular medium, could also act as a cryoprotectant. These might be the only organisms to use glycerol as both an osmoprotectant and a cryoprotectant.

Most of the known bidomain glycerol enzymes belong to chlorophytes, which in addition to Chlamydomonas and Dunaliella, include Gonium, Raphidocelis, Micractinium, Chlorella, Chloropicon, Volvox, and Tetrabaena. Many of these species are freshwater species and so it remains to be seen whether these species also produce glycerol and for what purpose. It also remains to be seen whether other marine Chlamydomonas species that make glycerol (Ahmad and Hellebust, 1986; Miyasaka et al., 1998) also have these enzymes. Finally, little is known about how Chlamydomonas spp. sense changes in salinity. Recent studies on osmosensing in microalgae (Suescún-Bolívar and Thomé, 2015; Charneco et al., 2018; He Q. et al., 2020) have implicated the possible involvement of mitogen-activated protein (MAP) kinases. The behavior of these genes in response to increases in salinity should thus be interesting.

In summary, we show that two Antarctic extremophiles, in addition to having several adaptations to low light and low temperature, also have adapted to high salinity by producing glycerol as an osmoprotectant (and possibly a cryoprotectant), and at least one of them (Chlamy-UWO) achieves this by using an unusual bidomain enzyme that can make glycerol directly from DHAP.

Publicly available datasets were analyzed in this study. The Chlamydomonas sp. UWO241 transcriptome is available under GenBank accession PRJNA575885.

All authors contributed equally to this study.

This study was partially funded by NSF grant 1637708 and DOE Grant DE-SC0019138 to RM-K.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.