Abstract
Mesotaenium berggrenii is one of few autotrophs that thrive on bare glacier surfaces in alpine and polar regions. This extremophilic alga produces high amounts of a brownish vacuolar pigment, whose chemical constitution and ecological function is largely unknown until now. Field material was harvested to isolate and characterize this pigment. Its tannin nature was determined by photometric methods, and the structure determination was carried out by means of HPLC-MS and 1D- and 2D-NMR spectroscopy. The main constituent turned out to be purpurogallin carboxylic acid-6-O-β-d-glucopyranoside. This is the first report of such a phenolic compound in this group of algae. Because of its broad absorption capacities of harmful UV and excessive VIS radiation, this secondary metabolite seems to play an important role for the survival of this alga at exposed sites. Attributes and abundances of the purpurogallins found in M. berggrenii strongly suggest that they are of principal ecophysiological relevance like analogous protective pigments of other extremophilic microorganisms. To prove that M. berggrenii is a true psychrophile, photosynthesis measurements at ambient conditions were carried out. Sequencing of the 18S rRNA gene of this alpine species and of its arctic relative, the filamentous Ancylonema nordenskiöldii, underlined their distinct taxonomic position within the Zygnematophyceae.
Introduction
Nowadays, we realize that alpine and polar glacier surfaces are not just bare, permanently cold surfaces devoid of any life. In fact, they represent basic ecosystems that are occupied by a number of highly specialized psychrophilic organisms like bacteria, algae and fungi (Hoham & Duval, 2001; Hodson et al., 2008; Remias, 2012). Ling & Seppelt (1990) classified algae that thrive on glaciers as causers of grey snow, because they can form dark, dust-like surface layers. Interestingly, virtually, all photoautotrophs reported to dwell on bare glacier ice are members of a single algal class, the Zygnematophyceae (Chlorophyta s.l.) (Kol, 1968). Among these, the unicellular Mesotaenium berggrenii (Wittrock) Lagerheim 1892 and the filamentous Ancylonema nordenskiöldii Berggren 1871 represent the most common members. Noteworthy, both species accumulate high amounts of a brownish, so far unidentified pigment in vacuoles. These organisms have been scarcely investigated albeit reports of their occurrence at many alpine and polar locations (e.g. Kol, 1968; Takeuchi et al., 2006; Uetake et al., 2010). Thus, they can be regarded as a cosmopolitan phenomenon of permanently frozen locations. General reports of these freshwater ice algae have been restricted to disciplines like taxonomy, cytology or the measurement of biomass abundances. In terms of physiology and biochemistry, the lack of knowledge is still striking. Two earlier studies dealt with the diversity of soluble carbohydrates in M. berggrenii from Antarctica and detected sucrose, glucose and glycerol as main components (Roser et al., 1992; Chapman et al., 1994). In a third report, Remias et al. (2009) described primary plastidal carotenoids in the same species from populations in the European Alps. Secondary carotenoids like astaxanthin, which are common in a variety of snow algae summarized under the collective taxon Chlamydomonas nivalis (Remias et al., 2005; Fujii et al., 2010), had not been found. Some of these species cause the red snow phenomenon in alpine and polar regions (Müller et al., 1998; Leya, 2004; Komárek & Nedbalová, 2007; Stibal et al., 2007), which frequently occurs at sites not far away from true glacier ice algae, but on a somewhat different substrate, namely rather wet snow.
The most striking morphological attribute of this Zygnematophycean species on glaciers are the vacuoles filled with a brownish, water-soluble compound. Ling & Seppelt (1990) suggested that this secondary pigmentation may be caused by ferric tannins, but gave no further evidence. Remias et al. (2009) proposed that the high abundances might play an ecophysiological role in protecting the chloroplast against excessive irradiation. The latter aspect is important to prevent photoinhibition, considering that open high alpine habitats are frequently subject to elevated UV and VIS radiation. Although field samples of M. berggrenii evidently contain high amounts of these brownish metabolites, structural determinations have not been performed until now. Therefore, the aim of this study was to isolate and to characterize the constituents of this pigmentation. This was realized by two consecutive steps. First, the classical way of spectrophotometric tests was used to determine roughly the nature of the pigment complex. Second, single compounds were isolated by HPLC and structurally elucidated by means of mass spectrometry (MS) and 1D- and 2D-NMR spectroscopy. Moreover, photosynthesis and respiration measurements of field material were conducted to estimate the degree of its metabolic adaptation to glacial temperatures. While many reports about photosynthesis of snow algae exist (e.g. Kvíderová, 2010; Remias et al., 2010a), freshwater ice algae hardly have been studied under this aspect before (Hoham, 1975). Molecular sequencing of the 18S rRNA gene was performed to clarify the taxonomic position of M. berggrenii within the order of the Zygnematales. For a phylogenetic tree, we also sequenced the same gene of its arctic relative A. nordenskiöldii from Spitsbergen (Svalbard, 79°N). Because of similarities in morphology and their habitat, both species have been regarded as very closely related algae, and also, some overlapping on subspecies level seems to exist. To resolve these uncertainties and to prove the taxonomic position of both taxa, a phylogenetic analysis on the basis of molecular data from field material was performed, particularly as these psychrophilic representatives seem to be unculturable in the laboratory so far over longer periods of time.
In summary, we found it important to study the strategies of how M. berggrenii is able to cope with the harsh glacier habitat from an analytical and physiological point of view.
Materials and methods
Field measurements and cell isolation
For the detection and on-site observation of ice algae populating glacial sites, a portable field microscope (Evolution; Pyser-SGI Ltd, UK) was used. The surface of the Gurgler glacier in the Ötztal valley in Tyrol, Austria, exhibited large populations of M. berggrenii var. alaskana Kol 1942, which were harvested for this study on 20 August 2009 (46°47.799 N, 10°59.043 E, 2829 m a.s.l.) and 7 September 2010 (46°47.728 N, 10°59.472 E, 2881 m a.s.l.). On the first date, the solar radiation was logged with a PMA2100 radiometer (Solar Light Company) using sensors for photosynthetic active radiation (PAR; Solar Light PMA2132) and ultraviolet radiation (UVB; Solar Light PMA2102). A surface ice layer of approximately 2–4 cm containing the algae was harvested into buckets, transported to the laboratory in < 4 h and allowed to melt gently at 5 °C. Larger debris, e.g. rock dust particles, remnants of moss leaves or insects, were separated from the melt water containing the suspended algae by filtering it through 400-, 200- and 100-μm stainless steel sieves (Retsch, Germany) one after the other. A part of the resulting suspension was directly used for the photosynthesis assays. The rest was purified from pollen and small cryoconite particles by density centrifugations using Percoll (SigmaAldrich, 400 g, 10 min; for details see Remias et al., 2005). While pollen remained on the surface layer, the algae formed a band at the 80% Percoll layer. The dark cryoconite accumulated mainly in the pellet. The layer of algae was withdrawn from the tube with a syringe and washed twice with distilled water by centrifugation prior to further analytical use.
Pigment extraction
For analytical purposes, approximately 250 mL of the algal suspension was filtered through glass fibre filters (Whatman GF/C) by mild vacuum filtration with a Millipore Sterifil set (Sigma Aldrich). The filters with algae were immediately frozen in liquid nitrogen and lyophilized for 48 h to total dryness. The freeze-dried cells were broken in a Mikro-Dismembrator S mill (Sartorius, Germany) with a 10-mm agate-stone grinding ball in 5-mL Teflon jars, the latter cooled for 10 min in liquid nitrogen prior to use. Afterwards, the powdery material was suspended and extracted in 2 mL of distilled water (Millipore). Preliminary tests showed that the brownish pigment was very hydrophilic and hardly dissolved in organic solvents. To remove apolar constituents from the raw extract, a phase separation with n-hexane was performed.
Pigment isolation by HPLC
The HPLC protocol was performed with an Agilent ChemStation 1100 with binary pump, a Synergy Hydro 4 μm column (150 × 2 mm; Phenomenex) at 25 °C with an injection volume of 30 μL and a flow rate of 0.3 mL min−1. Mobile phases: A: water with 1% formic acid (v/v); B: methanol with 1% formic acid (v/v); detection wavelengths: 280 and 350 nm. Linear solvent gradient; start 0% B; 40 min: 100% B; 43 min: 100% B; 45 min: 0% B and stop; post-time: 8 min. The peak of the main compound was isolated with an Agilent 1200 fraction collector. The mobile phase was evaporated, and the compound redissolved in water for further analysis. The spectral absorption was measured with a Lambda 20 photometer (Perkin Elmer). Additionally, absorption changes of aqueous raw extract were recorded at different pH levels of 4, 7 and 12.
The presence of mycosporine-like amino acids in M. berggrenii was surveyed according to the HPLC protocol of Karsten et al. (2005) using 25% methanolic extracts.
Spectrophotometric assays and determination of phenols and tannins
After purification and fractioning by HPLC, qualitative tests for phenolic substances were performed with the main compound, which was later identified as purpurogallin carboxylic acid-6-O-β-d-glucopyranoside. To get a first idea of its chemical nature, different spectrophotometric tests were made. The Prussian blue assay (Price & Butler, 1977) was performed by adding 2 mL of potassiumferri(III)cyanide (SigmaAldrich, 8 mM) to 0.5 mL of the 1 : 10 diluted main compound. After some minutes of incubation, 1 mL of ferric(III)chloride (SigmaAldrich, 0.1 M) was added and differences in the absorption were measured at 720 nm. This test was followed by a quantitative measurement of the phenolic content by the Folin–Ciocalteu (FC) method as described in Scalbert et al. (1989). To 1 mL of the 1 : 20 diluted main compound, 0.25 mL of FC reagent (SigmaAldrich, 2 M) was added, and after 5 min of incubation, 2.5 mL of an aqueous 20% Na2CO3 solution (w/v) was added. After 40 min of incubation at room temperature, the absorption was measured at 725 nm. Further tests for tannins were performed according to Makkar et al. (1993) with polyvinylpyrrolidone (PVP), which has a high affinity to precipitate tannins. The difference between total phenol values before and after the PVP treatment is an indicator for tannins and was performed with the FC test. 100 mg PVP (SigmaAldrich, p.a.) was dissolved in 1.9 mL of water and mixed with 100 μL of the main compound and incubated for 15 min at 4 °C. Afterwards, the solution was centrifuged (10 000 g, 5 min) and a FC assay was performed with the filtrate.
Several Tests for condensed tannins were performed including the HCl-4-dimethylaminocinnamaldehyde assay (DMACA) as described in Li et al. (1996), the butanol–HCl assay (Hagerman, 2002), the vanillin assay and the precipitation method of proanthocyanidins with formaldehyde (Scalbert et al., 1989). The DMACA assay was performed by adding 0.25 mL of a 0.3% DMACA-HCl solution (w/v) (SigmaAldrich) to 0.75 mL of the 1 : 15 diluted main compound. After an incubation time of 20 min, the absorption was measured at 643 nm. For the butanol–HCl assay, 50 μL of the main compound was mixed with 0.3 mL of butanol–HCl and then 10 μL of an aqueous 2% ferric(II)ammonium sulphate solution (w/v) was added. After 50 min of incubation at 100 °C, the absorption at 530 nm was measured; the blank sample was equally handled but not heated. For the vanillin assay, 0.2 mL of the main compound was mixed with 0.4 mL of freshly prepared 1% vanillin solution (w/v), incubated for 15 min at 20 °C and afterwards analysed at 500 nm.
The precipitation method of proanthocyanidins was performed by measuring the total phenolic content, first with the FC reagent as described earlier, then formaldehyde with HCl and phloroglucinol was added for a quantitative precipitation. After an overnight incubation at room temperature, any unprecipitated phenols were estimated by the FC protocol again.
MS and structure determination
The aqueous raw extracts were characterized by HPLC-MS at an Agilent ChemStation 1100 equipped with autosampler, diode array detector and column oven hyphenated to an Esquire 3000plus MS (Bruker Daltonics, Bremen, Germany). The LC parameters were modified as follows: Synergy Hydro 4-μm column (150 × 4.6 mm; Phenomenex) with identical guard column. Mobile phases: A: water with 0.9% formic acid and 0.1% acetic acid (v/v/v); B: methanol; detector: 280 and 325 nm, temp.: 45 °C; injection volume: 10.0 μL; flow rate: 1.0 mL min−1; solvent composition during analysis: start 0% B; 15 min: 30% B; 20 min: 30% B; 25 min: 100% B; 35 min: stop; post-time: 10 min. MS parameter: split to MS: 1 : 5; ESI, alternating mode; spray voltage: −4.5 kV, 350 °C; dry gas: 5 L min−1; nebulizer: 40 psi; full scan mode: m/z 100–1500. MS/MS experiments: ESI, negative mode; spray voltage: −4.5 kV, 350 °C; dry gas: 5 L min−1; nebulizer: 40 psi; full scan mode: m/z 100–1500. MS3 automated mode. With the isolated main compound, 1D- and 2D-NMR experiments were performed on a Bruker Avance II 600 (Bruker Biospin Rheinstetten, Germany) at 600 MHz (1H) and 150.9 MHz (13C) at 300 K. The sample (3.5 mg) was dissolved in 0.7 mL of MeOH-d4 containing 10% D2O. With the small amounts of the further phenolic compounds available, no structure determination was performed.
Light microscopy and photosynthesis
Photomicrographs were taken with a Zeiss Axiocam MRc5 (Carl Zeiss AG) mounted on a Zeiss Axiolab 200M light microscope. The sample was studied at 1 °C in a temperature-controlled chamber (Buchner et al., 2007) to prevent heat stress artefacts during observation.
Light-dependent photosynthesis and dark respiration measurements were performed with an oxygen optode (Fibox 3; PreSens, Germany) placed in a 3-mL acryl chamber (DW1; Hansatech Instruments, UK) and thermostated at 1 °C (see also Remias et al., 2010b). The mean oxygen turnover per time (n = 3) was normalized to the amount of total chlorophyll per sample. The amount of chlorophyll a + b was determined spectrophotometrically using N-N-dimethylformamid (SigmaAldrich) for extraction.
Molecular taxonomy
To verify the identifications and phylogenetic positions of M. berggrenii var. alaskana (sample AS08) and A. nordenskiöldii (sample 7-10/00), field material of both species was phylogenetically compared with other algal strains from the Zygnematales and Desmidiales based on the 18S rRNA gene sequences. Sequence data of reference species were downloaded from the website of the National Center for Biotechnology Information (NCBI). Accession and strain numbers are stated in the phylogenetic tree in Fig. 0005. Field material of M. berggrenii var. alaskana used for the sequence determination was withdrawn from the sample collected on 20 August 2009. The algae were harvested on a glass fibre filter (GF/C, Whatman) and frozen for storage. For DNA extraction, the filter was frozen in liquid nitrogen and ground to powder in a mortar with a small amount of quartz sand. The material was transferred to a reaction tube, and genomic DNA was extracted using the DNeasy Plant Kit (Qiagen, Hilden, Germany). The sample of A. nordenskiöldii was originally collected on 26 July 2000 from the surface of the Makarovbreen glacier (Spitsbergen, Svalbard; 79.810250° N, 11.834667° E, 40 m a.s.l.), and sequence data had been submitted to NCBI in 2004. For this study, the DNA of that sample, stored at −20 °C, was used again to generate a new 18S rRNA sequence with the sequencing primers shown below and to update the original sequence.
PCR was performed on whole genomic DNA using the forward primer NS1 (5′-GTAGTCATATGCTTGTCTC-3′) and the reverse primer 18L (5′-CACCTACGGAAACCTTGTTACGACTT-3′). PCR products were purified with the QIAquick PCR Purification Kit (Qiagen) and sequenced directly using the forward primers 34F (5′-GTCTCAAAGATTAAGCCATGC-3′), E528F (5′-TGCCAGCAGCYGCGGTAATTCCAGC-3′), 920F (5′-GAAACTTAAAKGAATTG-3′), 1422F (5′-CAGGTCTGTGATGCCCTTAG-3′), and the reverse primers 536R (5′-GWATTACCGCGGCKGCTG-3′), 895R (5′-AAATCCAAGAATTTCACCTC-3′), and 1263R (5′-GAACGGCCATGCACCACC-3′). All working kits were used respective to the manufacturer's instructions. The gene sequences obtained were assembled and aligned using the CLC Main Workbench software (version 6; CLC Bio, Denmark). The alignment was imported into the phylogenetic software paup* 4.0 (beta version 10; Sinauer Associates). To choose the best-fit evolutional model for nucleotide substitution, the data were analysed using Modeltest version 3.7 (Posada & Crandall, 1998). Final data analysis, analyses of bootstrap values (100 replicates) and phylogenetic tree building were performed with the paup software.
Results
Cell harvest, light microscopy and photosynthesis
The mostly flat ice surface of Gurgler glacier was populated virtually exclusively by the unicellular species M. berggrenii var. alaskana. We collected samples from the middle part of the glacier, which was already free from snow in August of both years. Light microscopy revealed that the cells harvested contained one marginal, green chloroplast with a central pyrenoid and several vacuoles filled with a brown phenolic pigment (Fig. 0001a). The cytoplasm also included colourless, putative lipid bodies that occasionally can be observed by light microscopy. When keeping field material under laboratory conditions for several weeks (5 °C, approx. 50 μmol PAR m−2 s−1), a decrease in the vacuolar coloration could be observed (Fig. 0001b). Furthermore, some cell divisions took place during this period causing short and loose filaments, which, however, never were observed in the field. The principal cellular composition is schematically depicted in Fig. 0001c. The population density reached 2 × 105 cells mL−1 melt water, which is about twice of the findings of Ling & Seppelt (1990) in samples from Antarctica. The ambient radiation values at the glacier surface during harvest (20 August 2009, 13:00–14:00 hours Central European Time) varied between 1630 and 1760 μmol photons m−2 s−1 for PAR and between 15.8 and 22.3 mW m−2 for UVB radiation.
Cellular appearance of Mesotaenium berggrenii var. alaskana collected on Gurgler glacier. Photomicrograph of a typical field sample (a) and of atypical short filaments (b) with reduced brownish pigmentation 4 weeks after collection. Schematic drawing of a cell (c) with organelles: N, nucleus; V, vacuole; L, lipid body; M, mitochondria; P, pyrenoid; C, chloroplast; G, golgi apparatus. Scale: 5 μm.
The photosynthetic activity was measured at 1 °C and at five different light levels from 48 to 1378 μmol PAR m−2 s−1. The net photosynthesis of M. berggrenii performed well under these conditions (Fig. 0002). Consequently, the alga was physiologically adapted to low temperatures as expected. Furthermore, no decline at higher irradiance levels, e.g. because of photoinhibition, was observed. The calculated light compensation point (Ik) read 35.2 μmol PAR m−2 s−1.
Light-dependent net photosynthesis and dark respiration of Mesotaenium berggrenii var. alaskana measured at 1 °C.
Characterization of the vacuolar compounds by spectrophotometric assays
To obtain a first impression of the chemical nature of the main constituent, different photometric tests were performed. Using the Prussian blue assay, it rapidly turned out to be a phenolic substance because of the characteristic blue coloration. This was indicated also by the standard assay for the determination of polyphenolic substances, the Folin–Ciocalteu (FC) assay. The presence of non-tannin phenols was checked by precipitating tannins with the phenol binding agent PVP. As in the residual solution no phenolic compounds could be detected by the FC assay, the major vacuolar constituent was suggested to be a tannin-like compound. To assess whether this compound was a condensed or hydrolysable tannin, several spectrophotometric tests were performed, including the HCl-4-dimethyaminocinnamaldehyde assay (DMACA), the butanol–HCl assay and the vanillin assay. All three photometric tests gave negative results, indicating that the main compound of the aqueous extract of M. berggrenii neither was a condensed tannin nor a proanthocyanidine-like derivate, such as oligomeric flavan-3-ols. Another method to determine structural characteristics of tannic substances relies on the precipitation of proanthocyanidins with formaldehyde, but this assay also failed, indicating that the main compound was a condensed tannin.
Pigment analysis and structure determination
The HPLC analysis of the aqueous raw extract of M. berggrenii resulted in the detection of three major compounds named in their retention order as compound 1, 2 and 3. A chromatogram monitored at 280 nm is presented in Fig. 0003a. The chromatograms of different field samples collected in 2009 and 2010 did not show significant differences, indicating the presence of a specific and stable intracellular composition of these hydrophilic substances. Compound 1 only absorbed in the UV region (Fig. 0003b; λmax of HPLC online spectrum: 278 nm). Compounds 2 and 3 were brownish vacuolar pigments possessing practically the same qualitative absorption: both HPLC online spectra showed peaks at 304 and 389 nm, and in contrast to compound 1, they also had an absorption in the VIS region (Fig. 0003c). The spectral absorption of the pigmentation was found to be pH dependent, which was tested with the aqueous raw extract (Supporting Information, Fig. S1 in Appendix S1).
(a) HPLC chromatogram (280 nm) of Mesotaenium berggrenii var. alaskana. The peaks of the three main compounds with retention times at 6.57, 17.23 and 18.07 min are labelled (1, 2, 3). Inserts: online absorption spectra of compound 1 (b) and compound 3 (c).
The alga was also screened by HPLC for the presence of mycosporine-like amino acids (MAAs); however, the result was negative. There were no characteristic UV absorptions around 310 or 330 nm.
To gain additional structural information of the three major compounds, LC-MS3 experiments were performed. Compound 1 showed ions with m/z 355.1 [(M+Na)+] and 687.1 [(2M+Na)+] in positive mode, while the negative mode revealed m/z 331.7 [(M-1)−] and 663.2 [(2M-1)−]. As the online UV spectrum of compound 1 was completely different compared with those of the other two other constituents, we assumed that this compound belongs to a different compound class. Compound 2 revealed ions with m/z 611.1 [(M+Na)+] and 1215.0 [(2M+K)+] in positive mode, while the negative mode showed one prominent ion with m/z 587.5 [(M-1)−]. LCMS analysis of the major compound 3 revealed ions with m/z 449.1 [(M+Na)+], 875.1 [(2M+Na)+] and 1317.1 [(3M+K)+] in positive mode, while the negative mode showed only one prominent ion at m/z 425.5. The MS2 experiments of compound 3 showed a breakdown of the initial ion of m/z 425.3 to m/z 380.9 (−CO2), 262.7 (−162 = a hexose moiety) and 218.8 (−CO2, −hexose unit). The observed fragments suggested the presence of a sugar moiety, e.g. glucose, as well as a carboxyl group.
With the pure compound 3, additional NMR experiments were performed to elucidate the structure of this major compound. As proposed by the MS data, the 1H-NMR spectrum showed not only typical signals of a hexose moiety, but also three additional broad singlets (δH 8.29, 8.04 and 7.06 ppm) in the aromatic region. In comparison, the 13C-NMR spectrum showed signals for 18 carbons. Heteronuclear single-quantum coherence (HSQC) and heteronuclear multiple bond coherence (HMBC) experiments revealed the presence of an unusual benzotropolone backbone known from oxidation products of black tea (Sang et al., 2004) bearing four hydroxyl groups in position C-2, C-3, C-4 and C-6, and a carboxyl function at position C-8. An HMBC cross-peak between the anomeric proton [δH 5.00 (d, 7.7 Hz) ppm; δC 102.89 ppm] and the carbon at position 6 (δC 153.70 ppm) established the glycosidic connectivity. The position of the sugar unit could be verified by the long-range correlations of H-1 to H-9 to H-7 to H-1′ observed in the COSY (correlation spectroscopy) spectrum (see Data S1). The NMR data are summarized in Table 0001. The sugar moiety could be identified by comparison to literature data as a glucopyranoside unit. A β-configuration was indicated by the large coupling constant (7.7 Hz) of the anomeric proton. The d-configuration of the sugar was confirmed by GC after hydrolysis and by measurement of the optical rotation of the hydrolysate (see Data S1). Thus, compound 3 was identified as purpurogallin carboxylic acid-6-O-β-d-glucopyranoside (C18H18O12), representing a new natural product (Fig. 0004).
Chemical structure of compound 3 (purpurogallin carboxylic acid-6-O-β-d-glucopyranoside) as identified by NMR in this work.
NMR data of purpurogallin carboxylic acid-6-O-β-d-glucopyranoside. The positions are labelled in Fig. 0004
| Position | 1H-NMR values (in MeOH-d4 with 10% D2O, 600 MHz; δ in ppm; mult.; J in Hz) | 13C-NMR values (in MeOH-d4 with 10% D2O; 150 MHz; δ in ppm; mult.) | HMBC-contacts |
| 1 | 7.06 (br s; 1H) | 114.93 d | 3; 4 (weak); 4a; 5 (weak); 9 |
| 2 | – | 153.78 s | |
| 3 | – | 137.85 s | |
| 4 | – | 151.70 s | |
| 4a | – | 117.78 s | |
| 5 | – | 186.21 s | |
| 6 | – | 153.70 s | |
| 7 | 8.04 (br s; 1H) | 122.24 d | 5; 6; 8; 9; 10 |
| 8 | – | 127.21s | |
| 9 | 8.29 (br s; 1H) | 142.20 s | 1; 4a; 6 (weak); 7; 8; 10 |
| 9a | – | 131.73 s | |
| 10 | – | 171.78 s | |
| Sugar moiety | |||
| 1′ | 5.00 (d, 7.7; 1H) | 102.89 d | 6 |
| 2′ | 3.65 (dd, 7.9; 8.8; 1H) | 74.51 d | 1′ 3′ |
| 3′ | 3.57 (m; 1H) | 77.23 d | 4′ |
| 4′ | 3.53 (m; 1H) | 70.65 d | 5′ |
| 5′ | 3.53 (m; 1H) | 77.95 d | |
| 6′ | Ha 3.90 (d, 12.2; 1H) Hb 3.90 (dd, 12.1; 3.0; 1H) | 61.88 t | 4′ 5′ |
| Position | 1H-NMR values (in MeOH-d4 with 10% D2O, 600 MHz; δ in ppm; mult.; J in Hz) | 13C-NMR values (in MeOH-d4 with 10% D2O; 150 MHz; δ in ppm; mult.) | HMBC-contacts |
| 1 | 7.06 (br s; 1H) | 114.93 d | 3; 4 (weak); 4a; 5 (weak); 9 |
| 2 | – | 153.78 s | |
| 3 | – | 137.85 s | |
| 4 | – | 151.70 s | |
| 4a | – | 117.78 s | |
| 5 | – | 186.21 s | |
| 6 | – | 153.70 s | |
| 7 | 8.04 (br s; 1H) | 122.24 d | 5; 6; 8; 9; 10 |
| 8 | – | 127.21s | |
| 9 | 8.29 (br s; 1H) | 142.20 s | 1; 4a; 6 (weak); 7; 8; 10 |
| 9a | – | 131.73 s | |
| 10 | – | 171.78 s | |
| Sugar moiety | |||
| 1′ | 5.00 (d, 7.7; 1H) | 102.89 d | 6 |
| 2′ | 3.65 (dd, 7.9; 8.8; 1H) | 74.51 d | 1′ 3′ |
| 3′ | 3.57 (m; 1H) | 77.23 d | 4′ |
| 4′ | 3.53 (m; 1H) | 70.65 d | 5′ |
| 5′ | 3.53 (m; 1H) | 77.95 d | |
| 6′ | Ha 3.90 (d, 12.2; 1H) Hb 3.90 (dd, 12.1; 3.0; 1H) | 61.88 t | 4′ 5′ |
NMR data of purpurogallin carboxylic acid-6-O-β-d-glucopyranoside. The positions are labelled in Fig. 0004
| Position | 1H-NMR values (in MeOH-d4 with 10% D2O, 600 MHz; δ in ppm; mult.; J in Hz) | 13C-NMR values (in MeOH-d4 with 10% D2O; 150 MHz; δ in ppm; mult.) | HMBC-contacts |
| 1 | 7.06 (br s; 1H) | 114.93 d | 3; 4 (weak); 4a; 5 (weak); 9 |
| 2 | – | 153.78 s | |
| 3 | – | 137.85 s | |
| 4 | – | 151.70 s | |
| 4a | – | 117.78 s | |
| 5 | – | 186.21 s | |
| 6 | – | 153.70 s | |
| 7 | 8.04 (br s; 1H) | 122.24 d | 5; 6; 8; 9; 10 |
| 8 | – | 127.21s | |
| 9 | 8.29 (br s; 1H) | 142.20 s | 1; 4a; 6 (weak); 7; 8; 10 |
| 9a | – | 131.73 s | |
| 10 | – | 171.78 s | |
| Sugar moiety | |||
| 1′ | 5.00 (d, 7.7; 1H) | 102.89 d | 6 |
| 2′ | 3.65 (dd, 7.9; 8.8; 1H) | 74.51 d | 1′ 3′ |
| 3′ | 3.57 (m; 1H) | 77.23 d | 4′ |
| 4′ | 3.53 (m; 1H) | 70.65 d | 5′ |
| 5′ | 3.53 (m; 1H) | 77.95 d | |
| 6′ | Ha 3.90 (d, 12.2; 1H) Hb 3.90 (dd, 12.1; 3.0; 1H) | 61.88 t | 4′ 5′ |
| Position | 1H-NMR values (in MeOH-d4 with 10% D2O, 600 MHz; δ in ppm; mult.; J in Hz) | 13C-NMR values (in MeOH-d4 with 10% D2O; 150 MHz; δ in ppm; mult.) | HMBC-contacts |
| 1 | 7.06 (br s; 1H) | 114.93 d | 3; 4 (weak); 4a; 5 (weak); 9 |
| 2 | – | 153.78 s | |
| 3 | – | 137.85 s | |
| 4 | – | 151.70 s | |
| 4a | – | 117.78 s | |
| 5 | – | 186.21 s | |
| 6 | – | 153.70 s | |
| 7 | 8.04 (br s; 1H) | 122.24 d | 5; 6; 8; 9; 10 |
| 8 | – | 127.21s | |
| 9 | 8.29 (br s; 1H) | 142.20 s | 1; 4a; 6 (weak); 7; 8; 10 |
| 9a | – | 131.73 s | |
| 10 | – | 171.78 s | |
| Sugar moiety | |||
| 1′ | 5.00 (d, 7.7; 1H) | 102.89 d | 6 |
| 2′ | 3.65 (dd, 7.9; 8.8; 1H) | 74.51 d | 1′ 3′ |
| 3′ | 3.57 (m; 1H) | 77.23 d | 4′ |
| 4′ | 3.53 (m; 1H) | 70.65 d | 5′ |
| 5′ | 3.53 (m; 1H) | 77.95 d | |
| 6′ | Ha 3.90 (d, 12.2; 1H) Hb 3.90 (dd, 12.1; 3.0; 1H) | 61.88 t | 4′ 5′ |
The molecular weight of compound 2 (588 g mol−1) differs by 162 g mol−1 in comparison to compound 3 (426 g mol−1). As the online UV spectra of both compounds were identical, we suggest that compound 2 might be a derivative of compound 3, differing by the presence of an additional hexose moiety. The structure of compound 1 remains unclear, but it might be a biosynthetic precursor of compounds 2 and 3.
Molecular taxonomy
The taxonomic position of the alpine M. berggrenii var. alaskana and the arctic A. nordenskiöldii within the Zygnematophyceae and their phylogenetic relationship to each other were investigated on the basis of 18S rRNA sequence data. The analysis for the best-fit evolutional model chose the transitional model TIM + I + G, for both, the hierarchical likelihood ratio test (hLRT) and the Akaike information criterion (AIC), with a proportion of invariable sites (I) of 0.5749 and a gamma distribution shape parameter (G) of 0.6262 (base frequencies = 0.2540 0.2097 0.2706 ACG). Maximum parsimony (MP) as well as distance (D) analysis gave a robust tree topography, while a maximum likelihood (ML) analysis placed the Desmidiales within the top clade of the genus Mesotaenium. Initially also strain sequences of the genus Cylindrocystis were used in the analysis, but again, all models placed it within the genus Mesotaenium; thus, it was omitted from further analyses.
The phylogenetic analysis (Fig. 0005) placed both species, as expected, within the Zygnematophyceae. Mesotaenium berggrenii var. alaskana appeared close to Mesotaenium caldariorum with a moderate support for the branching. Ancylonema nordenskiöldii was grouped together with another strain of M. caldariorum (UTEX 41), but with very strong support from the bootstrap values to form a separate clade. Both clades were separated from each other by other Mesotaenium spp. and also by two Desmidiales, the genera Penium and Staurastrum.
Phylogram of conjugating green algae from the Zygnematales and Desmidiales based on a maximum parsimony (MP) analysis of the nuclear-encoded 18S rRNA gene sequences with 1590 nucleotides. Three embryophytes were used as outgroup taxa. Bootstrap values on the branches were determined from 100 replicates for maximum parsimony (top left), minimum evolution distance (top right) and maximum likelihood (bottom). New sequences from this study are printed in bold and refer to the two unculturable species of Mesotaenium berggrenii var. alaskana (sample AS08, accession number JF430424) and Ancylonema nordenskiöldii (sample 7-10/00, accession number AF514397). The holotype species of the genus, Mesotaenium endlicherianum Naegeli 1849, is marked with an asterisk.
Discussion
Ling & Seppelt (1990) stated that Antarctic populations of M. berggrenii cause grey snow; however, we experienced that dense alpine populations rather cause a faint, dark purple impression on glacier surfaces, especially when viewed during sunshine. On Gurgler glacier, we found populations strictly growing on bare surface ice, frequently together with dark grey, macroscopic cryoconite particles. Usually, the latter are held together mainly by cryptic photoautotrophic filamentous prokaryotes (Hodson et al., 2010). We found that only a very low number of M. berggrenii cells were associated with cryoconite. Cryoconite holes, which are small glacial ecosystems of their own, were practically absent, but M. berggrenii occasionally has been reported from there, too (Mueller et al., 2001). Moreover, we never found M. berggrenii on snow substrates on top of the ice layers, as this is typical for snow algae from the Chlamydomonadaceae, nor on deglaciated adjacent localities. As the Zygnematophyceae are a class of algae without any flagellated stages, they cannot move actively upwards from bottom ice into the melting snow layer above as some snow algae such as Chloromonas nivalis do (Hoham & Duval, 2001; Remias, 2012). The general characteristics of M. berggrenii used in this study were identical to those reported by Remias et al. (2009).
Our photosynthesis assay shows that M. berggrenii is capable to perform well at temperatures close to the freezing point. As the populations are permanently frozen for the most part of the year, the active metabolic season is restricted to a few weeks during late summer when the glacier surfaces become wet. During this short period, an efficient metabolism is necessary for maintaining a local population through reproduction or cell division, as the cells will be partly washed away from the habitat by effluent melt water. We additionally tested the performance at 10 °C (data not shown) and found that respiration and photosynthesis both were higher at this temperature. This goes along with another psychrophilic member of the Mesotaeniaceae, Cylindrocystis brebissonii, isolated from snow, which was investigated by Hoham (1975). He found that it had its best growth at a temperature of 10 °C, compared with lower values at 1 and 20 °C.
The identification of the major secondary metabolite as purpurogallin carboxylic acid-6-O-d-glucopyranoside was quite unexpected, as related compounds with a benzotropolone backbone only have been reported from higher plants like Quercus crispula (Imai et al., 2009) or Camellia sinensis so far. Such compounds can be found in infected sapwood or in fermented plant material, for example in black tea. It is assumed that they are reaction products of smaller precursor molecules, such as catechin or gallic acid. Two patents describe the synthesis and use of benzotropolone derivatives (Wagner et al., 2009, 2011). In 1993, Bailey and Nursten were able to show that purpurogallin carboxylic acid can be synthesized by chemical oxidation of gallic acid or a mixture of gallic acid and pyrogallol. Alternatively, this compound also can be formed via an enzymatic coupling reaction of catechin and gallic acid, using horseradish peroxidase in the presence of H2O2 (Sang et al., 2004). This fact might contribute to elucidate the nature of compound 1, as the molecular weight of 332 g mol−1 and the HPLC online UV absorption maximum of the peak at 278 nm suggest a gallic acid glycoside. The latter could act as a precursor for compound 3 and also for substance 2 with the postulated structure of a purpurogallin carboxylic acid diglycoside.
The most striking ecological question concerns the reason why M. berggrenii actually synthesizes the brownish purpurogallin carboxylic acid-6-O-β-d-glucopyranoside and stores it in high amounts in the vacuoles. Considering the broad absorption capacity of this pigment and its congener, which covers the entire UV A and B range and also a large part of the VIS region, an important role as a cellular protectant against excessive irradiation is obvious. On the one hand, this phenolic compound can effectively shield chloroplasts against a surplus of VIS radiation, which otherwise would result in photoinhibition. Shading the chloroplast too much from essential light quantities, and thus unfavourably increasing the physiological light compensation point, may not play a significant role as M. berggrenii only thrives in open habitats commonly exposed to high irradiances. On the other hand, the high spectral absorption of purpurogallin carboxylic acid-6-O-β-d-glucopyranoside in the UV can be seen as an adaptation to the increased UV irradiation levels at high alpine locations. However, we observed that the vacuoles of similar populations of M. berggrenii living on glaciers in arctic Svalbard, where considerably lower ambient UV radiation is common, virtually were pigmented to the same extent (data not shown). Therefore, further ecological functions of the brownish tannins become arguable; for example, in glacial or snow algal populations, herbivorous glacial springtails are very common. The compound presented in this study might act as a chemical defence against grazers as, for example phenolic compounds in marine kelp are active against herbivores (Steinberg, 1984). These arguments concerning the attributes of benzotropolone derivatives are supported by two patents, which describe their use as UV absorbers and antioxidants in sunscreen or cosmetic compositions, as well as the application as antimicrobial agents (Wagner et al., 2009, 2011). Other phenolic derivatives, for example the flavan group in higher plants, have comparable properties. Finally, the synthesis of such secondary metabolites may also represent a ‘physiological sink’ for a surplus of reductive energy, which otherwise cannot be invested arbitrarily in other metabolic pathways, for example in cell growth and division, because of limitations in temperature or nutrient availability in the harsh habitat.
Striking pigmentations caused by secondary metabolites, besides in the close relative A. nordenskiöldii, have been reported from further Zygnematophyceae, which, however, do not live on glaciers. This indicates that ecological reasons other than low temperatures can also serve as a trigger for an abundant synthesis of special pigments; for example, the soil crust alga Zygogonium ericetorum has vacuoles filled with a purple pigment (Holzinger et al., 2010), but the structure of the compound is still unknown. Pinkish vacuoles are also known from Mesotaenium macrococcum (Ettl & Gärtner, 1995). Another phenolic from a Zygnematophyceae, pentagalloylglucose isolated from Spirogyra varians, was reported to be antibiotically active (Cannell et al., 1988a). The tannin-like precipitation capability of phenols from Zygnema cruciatum was depicted by Han et al. (2007), who reported that the pigments effectively interact with RNA during grinding, thus denaturising not only proteins but also nucleotides, despite using established extraction protocols.
The extent of secondary pigmentation may vary. Mesotaenium berggrenii, when transferred to laboratorial conditions with lower irradiance levels and without any freezing events, can reduce or even lose its vacuolar pigmentation after some weeks. We made similar observation with field material of A. nordenskiöldii from Svalbard, whose filaments were originally dark pigmented.
It is likely that the Zygnematophyceae generally evolved distinct patterns of certain phenolics and phenolic pigments and that the chemotaxonomical and ecophysiological recognition of this fact has been little until now. This is remarkable, as these constituents seem to play an important role in stress avoidance just like secondary metabolites such as mycosporine-like amino acids, carotenoids or scytonemins of other microorganisms do (Mueller et al., 2005). Thus, certain phenolics occurring in the Zygnematophyceae could have an ecophysiological function similar to other secondary metabolites, especially concerning UV protection.
This assumption could explain the lack of mycosporine-like amino acids in M. berggrenii investigated this study, but which do occur in some Streptophycean green algae (Karsten et al., 2005). Also secondary carotenoids, which are common in several freshwater microalgae from exposed habitats, such as C. nivalis or Trentepohlia iolithus, have not been detected. Finally, the occurrence of distinct phenolics in several Zygnematophyceae goes along with their taxonomical relationship to higher plants, as both are part of the Streptophyta (Friedl, 1997).
From our phylogenetic analysis, it also becomes evident that the brownish pigment is not a species specific marker, but can be produced by a range of species among the Zygnematophyceae under special environmental conditions. Our analysis places M. berggrenii var. alaskana close to M. caldariorum, a species with a clearly different cell morphology and no visible secondary pigmentation, at least as observed in laboratory cultures. Our arctic sample of A. nordenskiöldii is grouped together with another strain of M. caldariorum (strain UTEX 41). None of the culture collections holding strains of M. caldariorum reported a secondary pigmentation. Interestingly, however, the UTEX culture collection states an ‘antibacterial activity’ in their online catalogue for strain UTEX 41 (= CCAP 648/1A) and cites Cannell et al. (1988b), who screened 300 algal strains and found antimicrobial activity in a number of Conjugales (Desmidiales and Zygnematales) including, for example. Staurastrum gracile. In another publication, Cannell et al. (1988a) suggest that pentagalloylglucose, a precursor of gallotannins, may be the metabolite with antimicrobial activity. This suggests that the activity of the purpurogallin carboxylic acid-6-O-β-d-glucopyranoside that we have found in M. berggrenii, and which may also be responsible for the brownish pigmentation in A. nordenskiöldii, could be typical for this class of algae and that it may have an antimicrobial function in addition to its protective role against high radiation. Why strain UTEX 41 of M. caldariorum groups together with our sample of A. nordenskiöldii remains unclear. Taking the robustness of this clade in account with bootstrap values of 100% under all models, one might have to take a taxonomic incorrect assignment of strain UTEX 41 into account. On the other hand, McCourt et al. (2000) investigated the phylogeny of conjugating green algae on the basis of rbcL sequences, and they discuss the complex relationships found among the different taxa of the Desmidiales and Zygnematales in detail. As a consequence, our tree may not reflect the position of the Desmidiales correctly; however, the phylogenetic relationship of M. berggrenii var. alaskana and A. nordenskiöldii to the other Mesotaeniaceae becomes evident. As a consequence, an earlier suggestion of Hau U. Ling (Australian Antarctic Division, pers. commun. in 1999) that A. nordenskiöldii might be a diploid or polyploid stage of M. berggrenii, can be neglected.
By the identification of a secondary purpurogallin-derived pigment new to algae, we could demonstrate that such microorganisms from extreme habitats like glacier surfaces can be promising sources of special metabolites. The ecological relevance of this phenolic in terms of light protection can be assumed because of its extensive absorption capacity over a wide range from UV B to VIS radiation. We may add that this substance might also be of commercial interest as an UV-protective agent for human skin (Wagner et al., 2009).
Acknowledgements
The authors thank Dr Judith Rollinger for providing the FT-IR spectrum of compound 3 and the Austrian Science Fund (FWF) for funding (P20810 to C.L.). D.R., S.A. and S.S. made equal contributions to the work.
References
Supporting Information
Additional Supporting Information may be found in the online version of the article:
Appendix S1. The spectral absorption shifts of the raw extract at different pH levels (Fig. S1), the HPLC-MS spectra of compounds 1–3 (Figs S2–S5), 1D- and 2D-NMR spectra of compound 3 (Figs. S6–S11), physico-chemical properties of compound 3 (melting point, optical rotation, IR spectrum) and GC analysis of the sugar component of compound 3 (Fig. S12).
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
Author notes
Editor: Riks Laanbroek
