Abbreviations
DGDG
digalactosyldiacylglycerol
ESI
electrospray ionization
GIPC
glycosylinositolphosphoceramide
GlcCer
glucosylceramide
HILIC
hydrophilic interaction liquid chromatography
HPLC
high-performance liquid chromatography
HR-MS
high-resolution mass spectrometry
ITS2
internal transcribed factor 2
MGDG
monogalactosyldiacylglycerol
PC
phosphatidylcholine
PCA
principal component analysis
PG
phosphatidylglycerol
PI
phosphatidylinositol
PS
phosphatidylserine
TAG
triacylglycerol
INTRODUCTION
Cold ecosystems cover a substantial part of the Earth’s surface and are often dominated by microorganisms from various groups (bacteria, archaea, fungi, and protists) that were able to adapt to the extreme conditions of mountain, polar, or deep-sea habitats (Morgan-Kiss et al. 2006). Snow algae are a highly specialized group of microorganisms growing in the challenging habitat of melting snow in mountain or polar regions worldwide. Their habitat is characterized by many stresses, namely low temperature, high and variable irradiance, and freeze-thaw cycles. Under favorable conditions, they form intensively colored patches on snow. Most snow algae belong to the order Chlamydomonadales (Chlorophyta), and the genera Chloromonas and Sanguina are the most represented (Procházková et al. 2019a, Hoham and Remias 2020). Their life cycles include flagellates that can migrate through liquid water films between snow crystals and formation of resistant cysts on the surface during ongoing melting. Cysts, which withstand sub-zero temperatures in winter and, in contrast, higher temperatures in soil in combination with desiccation stress after complete snow melt, usually have large amounts of lipid reserves, polyols and sugars (Remias 2012).
In relation to their taxonomic affiliation and life cycle stage, snow algal cells from Chlamydomonadales can be colored green, orange, or pink to red. In green cells that are represented by flagellates, swarmer-like stages, or young cysts, chlorophylls dominate in the pigment profile. Depending on the degree of accumulation of secondary carotenoids (namely astaxanthin), snow blooms formed by cysts are colored orange, pinkish, or red. The high concentration of red cells causes the striking phenomenon of ‘red’ or ‘watermelon’ snow that is frequently reported from polar and alpine sites (Remias 2012).
Lipids are currently considered not only as energy depots and structural components of cells, but also play an important role in responding to environmental stresses (Hölzl and Dörmann 2007, Li et al. 2017). Depending on the taxonomic affiliation and environment, their composition may vary significantly. One of the main causes of variability is the ambient temperature, which causes changes in the physical properties of membranes. The detrimental effects of low temperature on membrane functions have been thoroughly documented (Los and Murata 2004). As maintenance of membrane fluidity and its integrity is essential for any organism, it represents a major adaptation of metabolic function that influences growth in permanently cold environments. In particular, the extent of unsaturation of fatty acids in membrane lipids plays a major role in avoiding membrane rigidification (Morgan-Kiss et al. 2006). In line with this, fatty acid analyses of snow algae pointed to their efficient adaptation to low temperatures by the abundant presence of polyunsaturated fatty acids (e.g., Řezanka et al. 2008, Procházková et al. 2019b). Analysis of the few sequenced genomes of cold-adapted microalgae revealed characteristics connected with adaptation to their extreme environment (Zhang et al. 2020).
In recent years, lipidomic analysis enabled to define quantitatively the lipid classes, including their molecular species. Shotgun lipidomic uses direct injection of total lipids into the mass spectrometer, which is very fast and allows the identification of hundreds of molecular species, especially when using high-resolution tandem mass spectrometry (HR-MS) (Řezanka et al. 2014). Of course, it also has limitations, e.g., identification of regioisomers of triacylglycerols (TAG) is very difficult, and enantiomers of TAG cannot be identified. The second method uses lipid separation by high-performance liquid chromatography (HPLC) and their identification by mass spectrometry (MS). However, it is more time-consuming, depending on the use of the column for HPLC; the separation of regioisomers and enantiomers is also possible (Palyzová et al. 2021).
Lipidomics of microalgae has been performed before; for an overview, see, e.g., Kato et al. (2022). Lipidomic analysis was performed on commonly cultivated green algae of the genus Chlorella (Couto et al. 2022) and Chlamydomonas, where examples include, e.g., data on a strain isolated from snow under different stresses (Lu et al. 2012, 2013). Although snow algae represent an excellent model for the study of metabolic adaptations to low temperatures, the study of their lipids is currently underrepresented. The first study to demonstrate differences between the TAG profiles of green and red snow was by Bidigare et al. (1993) and later by Spijkerman et al. (2012), who used the same classical gas chromatography / mass spectrometry protocol to screen strains of snow algae. However, there are only a few examples of complex lipid analyses of snow algae. Řezanka et al. (2014) used an integrated approach based on both lipidomic profiling using Orbitrap HR-MS and the traditional method of analysis by silver ion liquid chromatography and non-aqueous reversed-phase liquid chromatography of TAG to study flagellates of a Chloromonas species that cause green snow. Lu et al. (2012) focused on the putative snow algal strain UTEX LB2824, which was, however, wrongly identified and does not form blooms (i.e., does not represent a true snow alga). The comparison of green and red communities of polar snow algae (i.e., dominated by vegetative and cyst stages, respectively) showed significant differences in metabolomics, but the data related to lipid metabolism were limited (Lutz et al. 2015, Davey et al. 2019).
As noted above, lipidomic analysis of snow algae has been performed only on samples containing one cell type of a single species from a limited geographical area (Řezanka et al. 2014). Here, we performed a comprehensive comparison of lipidomic profiles of snow algae differing in taxonomic classification, life cycle stage, and geographic origin.
MATERIALS AND METHODS
Chemicals and standards
For a full description of all chemical compounds, see the Supplementary Text S1. All other reagents used were of analytical grade and purchased from Merck (Prague, Czech Republic).
Sample collection and characteristics
Unialgal field samples of snow algae (green, orange to pinkish, and red snow) were collected in the Jeseníky Mts., Jizera Mts., Krkonoše Mts. (Czech Republic), High Tatra Mts. (Slovakia), and European Alps (Austria) in May 2019, June 2020, and June 2021 (Table 1). Surface snow was harvested with a sterile shovel, placed in 1 L thermos bottles or 5 L buckets, and transported the same day to the laboratory. To ensure the unialgal character of the samples (i.e., only one species appeared to be present), snow bloom samples were checked using a portable microscope in the field and / or light microscope in the laboratory. For taxonomic identification and determination of the life cycle stage of snow algae, light microscopy (magnification 1,000×) was performed under oil immersion using an Olympus BX43 microscope (Olympus Corporation, Tokyo, Japan) with Nomarski contrast and an Olympus DP27 camera, and the algal biomass was kept frozen for further analysis. In addition, molecular internal transcribed factor 2 (ITS2) metabarcoding / Sanger sequencing of selected samples was performed for determination of algae from the genus Chloromonas (Lutz et al. 2019). However, identification to the species level was not possible in some cases, as some taxa have not been formally described yet.
The sampling sites were sun-exposed above the timberline (LP07, LP11, WP235, WP242, WP244, WP245, and WP266) or in the middle part of the avalanche slope (Krk19-2), partly shaded next to Pinus mugo vegetation (LP13), next to a spruce tree (Jes19-3 and Jes19-5), in broad leaf forest (still without leaves, Jiz 19-3) or shaded in spruce forest (Krk19-7 and Mil19-5). Electrical conductivity and pH of the meltwater were obtained with HANNA (Combo EC, Cluj Napoca, Romania) or with WTW Instruments (Cond 340i; Inolab, Weilheim, Germany). Green blooms were formed entirely by vegetative flagellates of various species of the genus Chloromonas, while red and orange (or pinkish) blooms consisted of the immotile cysts from the genera Chloromonas, Sanguina, and Chlainomonas (Table 1, Supplementary Fig. S1).
Extraction of lipids
Lipid extraction followed a modified protocol from Palyzová et al. (2021). The dried biomass underwent cryogenic grinding with glass beads under liquid nitrogen. The ground material was then mixed with a dichloromethane-methanol solution (2 : 1 v/v) and stirred for 30 min. Following this, additional dichloromethane and water were introduced to the mixture. The dichloromethane phase was collected and concentrated to dryness using vacuum evaporation. The resulting total lipid fraction was solubilized in acetonitrile for subsequent analysis. For polar lipid extraction, we adapted the method of Markham and Jaworski (2007) with modifications as described in Vítová et al. (2022). The water phase underwent three sequential extractions using an isopropanol / hexane / water bottom phase mixture (55 : 25 : 20 v/v/v). These extracts were combined, evaporated to dryness, and redissolved in pure acetonitrile. Lipid extraction followed a modified protocol from Palyzová et al. (2021). The dried biomass underwent cryogenic grinding with glass beads under liquid nitrogen. The ground material was then mixed with a dichloromethane-methanol solution (2 : 1 v/v) and stirred for 30 minutes. Following this, additional dichloromethane and water were introduced to the mixture. The dichloromethane phase was collected and concentrated to dryness using vacuum evaporation. The resulting total lipid fraction was solubilized in acetonitrile for subsequent analysis. For polar lipid extraction, we adapted the method of Markham and Jaworski (2007) with modifications as described in Vítová et al. (2022). The water phase underwent three sequential extractions using an isopropanol / hexane / water bottom phase mixture (55 : 25 : 20 v/v/v). These extracts were combined, evaporated to dryness, and redissolved in pure acetonitrile.
Shotgun lipidomics
Lipid extraction followed a modified protocol from Palyzová et al. (2021). The dried biomass underwent cryogenic grinding with glass beads under liquid nitrogen. The ground material was then mixed with a dichloromethane-methanol solution (2 : 1 v/v) and stirred for 30 min. Following this, additional dichloromethane and water were introduced to the mixture. The dichloromethane phase was collected and concentrated to dryness using vacuum evaporation. The resulting total lipid fraction was solubilized in acetonitrile for subsequent analysis. For polar lipid extraction, we adapted the method of Markham and Jaworski (2007) with modifications as described in Vítová et al. (2022). The water phase underwent three sequential extractions using an isopropanol / hexane / water bottom phase mixture (55 : 25 : 20 v/v/v). These extracts were combined, evaporated to dryness, and redissolved in pure acetonitrile.
Hydrophilic interaction liquid chromatography
Hydrophilic interaction liquid chromatography (HILIC) was conducted following Vítová et al. (2022) using two Ascentis Express HILIC columns (2.7 μm particle size, 150 mm × 4.6 mm L × I.D.) connected in series. The separation employed a flow rate of 0.95 mL min−1 with a 60-min linear gradient between two mobile phases: solvent A (methanol : acetonitrile:aqueous 1 mM ammonium acetate, 50 : 30 : 20 v/v/v) and solvent B (methanol : acetonitrile:aqueous 1 mM ammonium acetate, 10 : 70 : 20 v/v/v). Of the total HPLC flow, 10% was directed to the electrospray ionization (ESI) source for detection, while the remaining 90% containing lipid class fractions was collected manually. Complete experimental details are available in the Supplemenary Text S1.
Lipid nomenclature, data visualization and analyses
The lipid nomenclature was applied according to Liebisch et al. (2013). To perform multivariate statistical analyses, the program CANOCO 5 (Microcomputer Power, Ithaca, NY, USA) was used. Because the gradient lengths in the data were short, a principal component analysis (PCA) was performed to visualize the variability within the dataset.
RESULTS AND DISCUSSION
Altogether, 14 unialgal samples of snow blooms of different colors collected in five mountain ranges of Central Europe were analyzed. They were formed by nine species from the genera Sanguina, Chlainomonas, and Chloromonas (Chlamydomonadales, Chlorophyta). The altitude of the sites was 920–3,426 m a.s.l. and included both forested and open exposure habitats. The red and orange snow blooms were formed by a cyst stage, while the green snow blooms contained flagellates and swarmer-like stages (Table 1).
Shotgun lipidomic analysis
Total lipids were obtained by two-step extraction from individual snow algae samples (Table 1). First, it was a classical extraction according to Palyzová et al. (2021) followed by an extraction performed according to Markham and Jaworski (2007) with some modifications (Vítová et al. 2022). Two different extraction methods allowed to extract even very polar lipids, which are in the aqueous phase when only the first method is used. These included numerous classes of polar plant sphingolipids, such as glycosphingolipids, glucosylceramides (GlcCer), and glycosylinositol phosphoceramides (GIPC). Because the solubility of GIPC in organic solvents is poor (Vítová et al. 2022) due to high hydrophilicity of polar heads, GIPC are not extracted and consequently not determined.
The use of ESI allowed us to analyze both highly polar lipids (GIPC and / or GlcCer) and moderate polar lipids (Vítová et al. 2022) (Fig. 1). Furthermore, the use of HR-MS allowed for identification of such molecular species that differ in the values of [M+H]+ only in hundredths of Da. As an example, three molecular species of lipids, i.e., phosphatidylserine (40:3-PS), phosphatidylcholine (40:2-PC), and GlcCer (t18:1/h24:1-GlcCer) at m/z 842.5908, 842.6635, and 842.6717, were identified earlier. Tandem MS enabled the determination of the long-chain base (sphingoid bases) and acyl(s) of individual lipid classes. Although the use of shotgun lipidomic facilitates the detection of up to thousands of molecular species, as demonstrated by Lu et al. (2012), the latter authors actually identified only about one percent of the total molecular species (4,975 vs. 35). Thanks to the use of a high-resolution mass spectrometer and, at the same time, a tandem MS, more than 300 molecular species have been identified in this study (Supplementary Table S1). Below are examples of the analysis of two important classes of lipids found in snow algae (Supplementary Fig. S1).
Glucosylceramides
In positive tandem MS mode, fragmentation of [M+NH4]+ ions from GlcCer (d18:1/16:0) at m/z 810.6821 resulted in a fragment ion at m/z 792.6715 due to the loss of water [M+H-H2O]+. The precursor ion [M+H]+ of GlcCer produced also fragment ions [Glc+H]+ (often referred to as C1 ion) at m/z 169.0683 and [Glc-H2O+H]+ at m/z 181.0717 (ion B1). Fragments at m/z 298.2745 and 272.2586 (the nomenclature of ions was described by Li et al. (2017) provide information of the N-acyl chain (Ann and Adams 1992). In addition, the precursor ion [M+H]+ of GlcCer produced a fragment ion of long-chain base type at m/z 264.2657. Furthermore, the identification of head polar groups of glycosphingolipids was performed. In the tandem MS spectrum, the product ion Y0 is formed by the loss of one saccharide ([M+H-Glc]+ from the precursor ion of monosaccharide in ceramide. Likewise, water is lost from the ion Y0 to form ions Z0 at m/z 554.5147 and further loss of formaldehyde is formed by the ion Z0-CH2O at m/z 506.4936. This proved the structure of one of the major GlcCer, see tandem MS spectrum shown in Fig. 2.
Glycosylinositol phosphoceramides
The [M+H]+ ion at m/z 1292.7857 corresponds to predominant molecular species containing t18:0 with OH-26:0 fatty-N-acylation. The most abundant fragments in positive tandem ESI were ceramide ions [Y0PO3+H]+ at m/z 792.6480 and [Z0PO3+H]+ at m/z 774.6373 and the pair of glycosylinositol phosphate ions [B3PO3+H]+ and [C3PO3+H]+ at m/z 599.1221 and at m/z 581.1115, respectively. Furthermore, minor ions of the Y2 and Y1 type are present in the tandem MS at m/z 1,116.7535 and at m/z 954.7007, respectively. The tandem MS spectrum is shown in Fig. 3 that fully confirmed the unusual structure of HexA-Hex-IPC where the ceramide part is formed t18:0 with OH-26:0 fatty acid. Tandem mass spectra of three commonly occurring lipids, namely Po/Po/Ln-TAG (Supplementary Fig. S2), L/St-MGDG (Supplementary Fig. S3), and Po/Ln-PG (Supplementary Fig. S4), all molecular species were from sample WP235, are presented in Supplementary Table S1.
HILIC
Due to the different relative hydrophobicity of lipid molecules, it is impossible to extract all kinds of lipids from plant samples (de Jesus and Filho 2020). It has already been published several times (Markham and Jaworski 2007, Shiva et al. 2018) that the chloroform / methanol mixture is not suitable, for example, for plant sphingolipids such as GIPC. Therefore, a mixture of isopropanol, hexane, and water (e.g., lower phase of isopropanol / hexane / water (55 : 20 : 25 v/v/v)) is used for the extraction of plant sphingolipids (Markham and Jaworski 2007). Therefore, after extraction according to Palyzová et al. (2021), the aqueous phase was still extracted according to Markham and Jaworski (2007).
Total lipids were extracted both with a dichloromethane-methanol mixture (2 : 1, v/v) and with an isopropanol / hexane / water bottom phase (55 : 25 : 20, v/v/v). The separation of total lipids by HILIC is shown in Fig. 4 (for the identification of peaks, see Supplementary Table S2) and the values are shown in Table 2. The content of sphingolipids, i.e., GlcCers and GIPCs reaches one-half of total lipids in WP235 sample. Similar results were published by, for example, Cacas et al. (2016), who describe in their publication that the content of sphingolipids in tobacco leaves reaches up to 70% of total lipids (see his fig. 4A).
Main lipid groups and species in snow algae
Regarding the main lipid groups, all the samples were dominated by sphingolipids and TAG, forming 39.9–50.5% and 21.9–31.8% out of total lipids, respectively. Phospholipids and glycolipids were less abundant with the ranges of 11.8–20.0% and 9.5–18.5% (Table 2, Fig. 5). This is in line with a previous lipidomic analysis of snow algae from the genus Chloromonas (Řezanka et al. 2014). In contrast, glacier ice algae from the genus Ancylonema, which represent another group of extremophilic cold-adapted microorganisms forming blooms (though on glacier surfaces), lipidomic analysis revealed a slightly different lipid composition with a higher proportion of TAG and lower abundance of sphingolipids and phospholipids (Procházková et al. 2021). Like snow algae, they belong to Viridiplantae, but to different infrakingdom Streptophyta, class Zygnematophyceae. However, more testing will be needed to confirm whether this pattern is a characteristic one, or if it is influenced by environmental conditions, sampling date, or other factors.
In line with the previously published data (Řezanka et al. 2014), TAG in snow algae were represented by species with even number of acyl carbons from 48 to 54 and 0–12 double bonds. The most abundant TAGs were 50:7, 54:8, and 50:4 that formed on average >30% of the total TAG. Among sphingolipids, GIPC were the most abundant (particularly those with sphingoid bases t18:1/h24:0, t18:1/h24:1, and t18:1/h26:0), followed by GlcCer, where t18:1/h24:1, t18:1/h24:0 and t18:1/h26:1 were the most frequent. GIPC have been detected in fungi and the green lineage (Vítová et al. 2022), and are considered the most abundant sphingolipids in the biosphere with multiple roles connected with the plasma membrane (Gronnier et al. 2016). The major glycolipids were monogalactosyldiacylglycerol (34:6-MGDG) and digalactosyldiacylglycerol (36:6-DGDG). However, a small proportion of sulfoquinovosyldiacylglycerols has been already also detected in snow algae (Řezanka et al. 2014). On the contrary, the latter lipid group was rather important in the previously mentioned glacier ice algae (Procházková et al. 2021). The main phospholipids were phosphatidylglycerols (PG) (namely 34:4) and PC (36:4, and 36:5), which was again in agreement with a previous study (Řezanka et al. 2014).
Challenges arise when trying to compare our results with existing lipidomic data from related algal species. Unfortunately, the data for the model chlamydomonadacean alga Chlamydomonas (C.) reinhardtii do not include a quantitative analysis of the relevant molecular species. For example, Yang et al. (2013) identified 35 molecular species of lipids, without any quantitative analysis. They reported the presence of e.g. 16:1/18:2-MGDG, 16:1/18:3-PG, or 16:0/18:1-PI. Tietel et al. (2020) described the effect of hyperosmotic stress on metabolomic changes in the biosynthesis of complex lipids. Again, many tens of molecular species of lipids were identified, especially TAG (18:2/18:2/18:3-TAG, 16:0/16:1/18:3-TAG, or 18:1/22:1/22:1-TAG). Polar lipids of different classes such as 18:3/18:4-DGDG, 16:3/18:5-MGDG, 18:2/20:4-PG were also identified. Unfortunately, this work did not provide the primary data. Jüppner et al. (2017) were able to quantify 212 lipid species from nine lipid classes. In contrast to our results, glycolipids (MGDG and DGDG) were dominant and TAG formed only a minor proportion of total lipids in mesophilic C. reinhardtii, which points to high metabolic variability and thus ecological flexibility within Chlamydomonadaceae. Lu et al. (2012, 2013) described the content of molecular species of lipids, e.g., MGDG, DGDG, phosphatidylethanolamines, PG, PI in order to identify lipid biomarkers in a chlamydomonadacean alga isolated from snow. They focused on the trend followed by the selected lipid biomarkers as a result of cultivation at different conditions, making any comparison with our data difficult.
Comparison of different types of snow blooms
The field samples were classified into three groups based on the pigmentation of the snow bloom (green, orange to pinkish, and red) that reflected the life cycle stage of these algae (flagellates or swarmer-like stages vs. cysts) and their taxonomic affiliation. The most common genera of snow algae (Chloromonas, Sanguina, and Chlainomonas) were covered (Table 1). In general, the diversity of snow algae below the genus level is still poorly known and many species await their formal taxonomic description, which is the case for samples labeled ‘sp.’ in Table 1. The use of ITS2 ribosomal DNA metabarcoding enables to discern even among closely related species with very similar morphologies (Lutz et al. 2019). In this study, green snow was caused exclusively by flagellates from the genus Chloromonas, while cysts of this genus were found in orange to pinkish and red snow. Most of the red snow samples were made of cysts of the recently established genus Sanguina that is widespread in exposed high mountain habitats including the Alps (Procházková et al. 2019a), but the other two genera (Chloromonas and Chlainomonas) were also represented in red pigmented snow.
Comparison of the main lipid groups showed significant differences among bloom types (Fig. 5). Green snow made of flagellates and swarmer-like stages was characterized by a higher proportion of phospholipids (mean 18.0% vs. 13.5%. in orange snow cysts and 14.4% in red snow cysts) and glycolipids (mean 17.3% vs. 13.2% and 10.3%), while TAG were less represented (mean 22.3%) compared to the other two groups, where they formed on average 30.2% and 28.3%, respectively. The proportion of sphingolipids also varied, and significant differences were found among the three sample groups (means: red 47.0%, orange 43.1, and green snow 42.4%) (Fig. 5).
A more detailed insight into the lipid variability among bloom types was provided by the principal component analyses that were performed based on the data representing 17 lipid classes (see Table 2, Fig. 6), and also based on the entire lipidomic dataset that contained 303 lipid species (Supplementary Table S1 & Fig. S5). Both analyses revealed a clear separation of the three bloom types represented in the samples with green snow having a very distinct lipid composition compared to the other two types based on the x-axis sample scores. Orange and red snow samples also formed separate groups with significant differences in y-axis scores but were closer to each other than to the green snow samples, as this axis explained only 17.7% and 14.1% of the total variability in the two analyses, respectively. This pattern is in line with the assumed physiological difference between the two key stages of the life cycle in chlamydomonadacean snow algae (flagellates vs. cysts) (Remias 2012).
Vegetative flagellates forming green snow are considered to represent the actively growing life cycle stage, which actively migrates in the melt water film surrounding the snow crystals, reproduce and develop large populations that usually appear (on or under the snow surface) as the first type of snow coloration during the season (Hoham and Remias 2020). As the period suitable for the growth of snow algae is usually rather short in the middle latitudes (April–May at low altitudes and shifted to later periods at high altitudes), it is expected (although never measured in the field) that the growth rates of flagellates must be high, which requires a high amount of energy gained via intense photosynthesis. The higher proportion of phospholipids and especially glycolipids (MGDG and the DGDG) and lower proportion of TAGs compared to the orange and red snow samples are in accordance with this metabolic setting.
The results showed that cysts accumulate more TAGs than flagellates. A high content of TAGs in the cysts of Sanguina nivaloides was interpreted by Ezzedine et al. (2023) as a response of the alga to a phosphate depleted environment. Although the three sample types were clearly separated based on lipid composition, the differences among them were only moderate (Fig. 5). This is possibly linked to the fact that (in contrast to common expectations) these cysts remain photosynthetically very active (Hoham and Remias 2020) and therefore membrane lipids must still be present in substantial amounts. The differences between both cyst categories (orange and red) could be associated with the taxonomic differences between the two groups (Table 1).
Another significant difference between flagellates and cysts has been represented by a contrasting pattern in the composition of GIPC, as the samples of green snow contained much more dihydrosphingosine compared to the other two groups. Regarding phytosphingosine, the pattern was inverted, although not so pronounced (Supplementary Table S3). The link of these patterns to the metabolic setting of both stages of the life cycle needs to be elucidated, however, it must be associated with shifts in membrane functions (Gronnier et al. 2016).
Carbon chain saturation within the main classes of lipids was determined by further analysis. The PCA of phospholipids as the key components of cell membranes based on the degree of unsaturation of fatty acids revealed a clear separation of green vs. orange and red snow (Fig. 7). Red snow samples showed more unsaturation, indicating better adaptation of membranes to low temperatures (Morgan-Kiss et al. 2006, Ezzedine et al. 2023). Generally, fatty acid composition and extent of unsaturation were shown to vary considerably in snow algae depending on species, life cycle stage, and nutritional regime (Spijkerman et al. 2012). However, a high proportion of polyunsaturated fatty acids has often been found reflecting the adaptation to low temperatures in snow.
CONCLUSION
Snow algae are a unique group of photoautotrophic microorganisms adapted to the extreme environment of melting snow. Although the phenomenon of colored snow has been known for a very long time, there are still major gaps in our understanding of their physiology in relation to the abiotic stresses that characterize their habitat. Shotgun lipidomic analysis can provide valuable insights into the metabolic adaptations of these extremophiles, but it has only been applied rarely so far. An extensive data set of 303 molecular species of snow algae lipids, based on the analysis of 14 samples of different types of snow blooms originating from mountains in Central Europe, is presented (details see the Supplementary Table S1). Samples characterized by different degree of secondary carotenoid accumulation (red, orange, or green blooms) were clearly separated on the basis of their lipid composition. For the first time, lipidomic profiles of the two main life cycle stages of chlamydomonadacean snow algae (flagellates and swarmer-like stages vs. cysts) were compared. Significant differences in the proportion of the main lipid classes and lipid species composition were found, and they are in accordance with the metabolic demands of the two cell types (growth vs. reproductive inactivity).
