A new microscopic method to analyse desiccation‐induced volume changes in aeroterrestrial green algae

Summary Aeroterrestrial green algae are exposed to desiccation in their natural habitat, but their actual volume changes have not been investigated. Here, we measure the relative volume reduction (RVRED) in Klebsormidium crenulatum and Zygnema sp. under different preset relative air humidities (RH). A new chamber allows monitoring RH during light microscopic observation of the desiccation process. The RHs were set in the range of ∼4 % to ∼95% in 10 steps. RVRED caused by the desiccation process was determined after full acclimation to the respective RHs. In K. crenulatum, RVRED (mean ± SE) was 46.4 ± 1.9%, in Zygnema sp. RVRED was only 34.3 ± 2.4% at the highest RH (∼95%) tested. This indicates a more pronounced water loss at higher RHs in K. crenulatum versus Zygnema sp. By contrast, at the lowest RH (∼4%) tested, RVRED ranged from 75.9 ± 2.7% in K. crenulatum to 83.9 ± 2.2% in Zygnema sp. The final volume reduction is therefore more drastic in Zygnema sp. These data contribute to our understanding of the desiccation process in streptophytic green algae, which are considered the closest ancestors of land plants.


Introduction
Green algae have the ability to perform photosynthesis and their major distribution is in aquatic habitats making them rather an ecological unit than a taxonomic entity. In fact, according to recent phylogenetic research, green algae can be divided in two main lineages (Leliaert et al., 2012): (1) Chlorophyta that contains the majority of described green algal species and (2) Streptophyta that includes the Charophytes, a paraphyletic group of freshwater and terrestrial green algae, as well as all land plants, the Embryophytes (Leliaert et al., Correspondence to: Andreas Holzinger, University of Innsbruck, Institute of Botany, Functional Plant Biology, Sternwartestrasse 15, 6020 Innsbruck, Austria. Tel: 0043-512-507-51028; fax: 0043-512-507-51099; e-mail: Andreas. Holzinger@uibk.ac.at 2012). Among both lineages, several groups with terrestrial representatives occur. In case of the Chlorophyta, the classes Ulvophyceae, Trebouxiophyceae and Chlorophyceae include some taxa living in terrestrial habitats, e.g. the Trentepohliales of the Ulvophyceae (Leliaert et al., 2012), the members of Trebouxiophyceae ('lichen algae group') and sparse genera in the Chlorophyceae, e.g. Fritschiella (Holzinger & Karsten, 2013). In the streptophyte lineage, terrestrial members can be found in the classes Klebsormidiophyceae, Zygnematophyceae and Coleochaetopyhceae (Graham et al., 2012;Holzinger & Karsten, 2013;Mikhailyuk et al., 2015). The closest relatives to land plants can be found among charophyte green algae (Wodniok et al., 2011;Timme et al., 2012;Delwiche & Cooper, 2015). Their genome holds primary factors for plant terrestrial adaptation (Hori et al., 2014), and transcriptional changes upon severe desiccation stress demonstrated a land plant like defence reaction in Klebsormidium . Raffinose family oligosaccharides have been found to be upregulated upon desiccation stress likely contributing to the osmotic potential . Increasing knowledge on cell wall properties of these algal groups has become available recently, demonstrating the close relationship to land plants (Sørensen et al., 2011;Mikhailyuk et al., 2014;. However, green algae are poikilhydric organisms that do not have cuticles or similar structures to protect from water evaporation (Delaux et al., 2013).
Although the hydraulic parameters (osmotic potential at full turgor and relative water content at the turgor loss point) in pokilohydric plants of the embryophytes, ferns, mosses and lichens have been well investigated (see Table 1), only little information is available in aeroterrestrial green algae (Holzinger & Karsten, 2013). This lack of data may be related to the difficulty of determining the volume or weight of terrestrial green algal samples due to their small size, mostly ranging between 5 and 25 μm in diameter (Rindi et al., 2011;. By contrast, the osmotic potential has been determined in Klebsormidium and Zygnema (Kaplan et al., 2012(Kaplan et al., , 2013. In Table 1. Osmotic potential at full turgor ( osat ) and relative water content at the turgor loss point = incipient plasmolysis (RWC TVP ) of different poikilohydric species of angiosperms, ferns, mosses and lichens. these studies, the water potential at the turgor loss point ( TLP ) was determined by means of the 'incipient plasmolysis technique'. For plasmolysis fully turgescent plant cells, like the cell filaments of the investigated green algae, were exposed to osmotically active solutions with increasing concentrations and the value where 50% of the cells plasmolyzed determined microscopically. At the equilibrium point, the cells have the same osmolarity as the external solution, which allowed determining the osmotic potential. In K. crenulatum a TLP of -2.09 MPa whereas in Klebsormidium nitens a TLP of -1.67 MPa was found (Kaplan et al., 2012). In the same way, TLP values of -1.67 and -0.8 MPa were determined in different strains of Zygnema sp. (Kaplan et al., 2013).
In this study, we developed and tested a special chamber that allows microscopic observations at several preset relative air humidity (RH) levels. To illustrate the capacities of this chamber, we analysed the desiccation-induced volume changes of the cytoplasm in streptophyte green algae. We hypothesized that green algae from distinct classes (Klebsormidiophyceae, Zygnematophycae) show different desiccation patterns. We measured the relative volume reduction (RV RED ) of the protoplast at different preset RH levels under controlled conditions. This allowed us to get information on the water loss at different RHs and to enhance our understanding of desiccation processes in the studied organisms.

Test chamber
A test chamber that allowed: (1) maintaining a constant RH for desiccation experiments and (2) microscopic observation of the samples was constructed (Fig. 1). The test chamber

Experimental setting of the RH
The precise setting of the RH inside the hermetically sealed test chamber was achieved by different nonsaturated lithium chloride (LiCl) solutions (ࣙ98%, Sigma-Aldrich, Vienna, Austria) according to Hay et al. (2008), see Table 2. Alternatively, a saturated K 2 CO 3 solution (ࣙ99.0%, Sigma-Aldrich) was used. The lowest RH generated by saturated LiCl solution is 11.2% at 20°C (Hay et al., 2008). Therefore, silica gel (with moisture indicator, Sigma-Aldrich) was used to generate RH values of 3.7-4.2% (Table 2).

Experimental procedure
The test organisms (K. crenulatum, Zygnema sp.) were placed in a volume of 20 μL culture medium in the 'lower storey' of the test chamber. In the 'upper storey', the respective salt solution or silica gel was filled. The cell filaments were spread and microscopic images were taken immediately. The cells were then allowed to completely equilibrate for up to 12 h at 20°C to the respective RH before 10-15 random images were taken. The measurements of the diameters at individual protoplasts were performed only on filaments that were found solitary after desiccation, where the protoplasts as well as the cell lumen diameter (CLD) were clearly visible. In K. crenulatum, an average of 22 protoplasts per RH step and in Zygnema sp. an average of 16 protoplasts per RH step were measured.

Determination of the RV RED
The desiccation process was monitored by an inverted Zeiss Axiovert 200M microscope (20x, NA = 0.50; 40x, NA = 0.75; Carl Zeiss AG, Oberkochen, Germany). To determine the protoplast volume, the CLD and the diameter of the desiccated protoplast (PD d ) were measured in individual cells and the reduction of the protoplast volume was quantified (Zeiss AxioVision 4.7.1 software). In this way, assuming a cylindrical shape, the protoplast volume at saturation (V s in Eq. (1)) and after desiccation (V d in Eq. (2)) was calculated by the following equations (for abbreviations, see Fig. 1): The RV RED (%) of the PD d related to the protoplast volume at the initial, saturated state was determined according to Eq. (3).
RV RED were then plotted against RH values of both studied species.

Statistical evaluation
Significant differences between mean values of RV RED [%] for each investigated genus (K. crenulatum, Zygnema sp.) were calculated with one-way ANOVA followed by Games-Howell's post hoc test; p < 0.01) using SPSS software. Significant differences of RV RED [%] between the two genera at a certain RH level were calculated by Student's t-test (p < 0.05) by SPSS software.

Test chamber
We first tested our newly constructed test chamber for functionality at different RH. The chamber was filled with the respective LiCl solutions of different concentrations or the saturated K 2 CO 3 solution or with silica gel to generate different levels of RH (Table 2). At high levels of RH (>80%), the RH inside the test chamber reached the saturation point within 3-4 h (Fig. 3), over silica gel, the saturation point was achieved in less than 3 h (Fig. 3). The recorded RH differed only slightly between experiments ( Table 2).

Measurements of volume reduction
After transfer of algal suspensions into the test chamber, an initial phase was observed, where only culture medium evaporated and the cells did not show any change in volume or shape. Then, the desiccation process started, in which the The diameter of the protoplasts experienced a drastic reduction during the desiccation process, whereas the length of the protoplasts usually did not change and the protoplasts remained attached to the cross-walls (Figs. 2 and 4A). Only in Zygnema sp., protoplasts occasionally were detached from the cross-walls, which lead to shrinkage in length during the desiccation process (Fig. 4B). As in these cases, severe damage was expected; these cells were not included for volume calculations. In Zygnema sp., the strongest shrinkage of the protoplasts was observed in the central area and therefore a mean value was used to calculate the volume reduction (see Figs. 5A and B). In some cases, the PD d even disintegrated into portions, resembling the chloroplast centres and pyrenoids (see Figs. 5C-E); such protoplasts were not used for measurements.
Calculations of the RV RED were performed by measuring the protoplast diameter after a constant level of RH inside the test chamber was achieved and cells did not show further changes in protoplast volume or shape. In K. crenulatum, the RV RED values (mean ± SE) ranged from 46.4 ± 1.9% at 95.4% RH to 75.9 ± 2.7% at 3.7% RH (Fig. 6). In Zygnema sp., the RV RED values of 34.3 ± 2.4% at the highest RH was significantly higher (p < 0.01) when compared to K. crenulatum. At the lowest RH the value of 83.9 ± 2.2% was significantly (p < 0.05) lower in comparison to K. crenulatum. The development of volume reductions differed between the two genera. Although in K. crenulatum a drastic water loss was observed already at highest RH tested, in Zygnema sp. the initial water loss was less pronounced. However, in Zygnema sp. the water loss at RH values <ß80% was more drastic indicated by higher RV RED values.

Discussion
In this study, a new test chamber allowed the monitoring of desiccation effects at certain preset RH during microscopic observation. This device was used with two green algae, K. crenulatum and Zygnema sp., for which the RV RED was experimentally determined. A distinct desiccation behaviour was   found by quantitative analysis of the RV RED at different RHs. The RV RED values at the highest RH tested (ß95% RH) indicated a stronger water loss in K. crenulatum. By contrast, at lower RHs Zygnema sp. showed a stronger water loss. There are also qualitative differences in the shape of the retracted chloroplasts between the two studied green algal genera.
In general, these observations correlate well with physiological parameters, e.g. reduction of the effective quantum yield as a consequence of desiccation in K. crenulatum (Karsten et al., 2010) and Zygnema sp. .  investigated these two genera in a parallel setup and monitored the desiccation and recovery effects of individual filaments by imagining PAM. In both genera, desiccation at ambient air (RH ß65%) leads to a rapid decrease of the effective quantum yield of photosystem II (YII). Upon rehydration for 180 min, the Y II is re-established, in Klebsormidum cenulatum to nearly the initial value and in Zygnema sp. to only half of the initial value . In this study, we found clear differences between the two examined genera concerning the range of the RV RED -RH relationship. Most important, RV RED values below ß80% RH are considerably higher in Zygnema sp. than K. crenulatum. This means that the protoplasts in Zygnema sp. experience a clearly stronger reduction in volume under similar RH levels. This correlates well with the observations by Kaplan et al. (2012Kaplan et al. ( , 2013, and may be attributed to the higher vacuolization status of young Zygnema sp. protoplasts (Kaplan et al., 2013;. Moreover, this could also explain the weaker recovery of YII in Zygnema sp. . In Zygnema sp., older cultures form preakinetes, which are in general more stress tolerant (e.g. Pichrtová et al., 2014), and likely have different osmotic potentials (e.g. Fuller, 2013;Kaplan et al., 2013). In this study, we used young cultures, to exclude this phenomenon. However, the measurement of the Zygnema sp. protoplasts was more difficult to perform due to the shape of the desiccated cytoplasm.
In both genera, only individual cells from solitary filaments were measured. Bundles of filaments showed a slower desiccation process and so a prolonged moistness was observed. This is a well-described strategy for filamentous, soil crust inhabiting organisms to enhance the water holding capacities in nature (e.g. . In the experiment, however this leads to complications with the measurements (overlapping of filaments, protoplast shape) and was therefore avoided. Moreover, we used for the calculation only cells where the protoplast remained attached to the cross-walls, assuring that the cells were not damaged too badly had the potential to recover .
Summarizing the results of this study, here we introduce a functional test chamber suitable to follow the desiccation process during microscopic observation. We demonstrated that the test chamber could create and maintain a constant RH during the desiccation experiments. These conditions allowed to subsequently calculate the volume reduction of the test organism in situ. The more rapid water loss observed in K. crenulatum might in nature be compensated by their habitat in soil crusts, giving protection against desiccation . By contrast, Zygnema sp. collected from the sandy litoral zone of a river  might be naturally exposed more frequently to higher RHs, which they are capable to tolerate. The stronger water loss at lower RHs resulted in severe damage, and it has been shown that particularly young vegetative Zygnema sp. cells cannot recover from desiccation over silica gel, whereas they tolerate desiccation at ß83% RH with good recovery success according to measurements of the effective quantum yield (Pichrtová et al., 2014). We are aware that comparing genera after investigating only one strain per genus might lead to misinterpretations; therefore, future studies should include more strains.
The chamber presented here will allow monitoring desiccation kinetics under defined RHs. Repeated desiccation and rehydration cycles could be created in order to study possible legacy effects. Future studies should elucidate which osmotic compounds are responsible for water holding capacities that enable aeroterrestrial green algae to withstand desiccation. Additionally, this chamber might also be valuable for further studies measuring the photosynthetic activity of single cells during desiccation, which is possible with new-generation chlorophyll fluorimeters or the observation of other physiological responses employing fluorescent probes, e.g. H2DCF-DA to detect the formation of reactive oxygen species during the desiccation process.