Introduction

Molecular chaperones are a large and diverse group of proteins with property of supporting noncovalent assembly/disassembly of other macromolecules in the cell (Kregel 2002; Burdon 1986; Frydman 2001). Heat shock proteins (HSP) also actively participate in long-term adaptations to the environmental changes and seasonal developmental patterns in invertebrates. In a number of insect species, up-regulation of hsps begins at the start of diapause and decreases back to the normal homeostatic level during the re-initiation of development (Rinehart et al. 2006b, 2007a; Gkouvitsas et al. 2008). During the last decade, involvement of large HSPs (HSP40, HSP60, HSP70, HSP90, and HSP100) in various hypometabolic processes in arthropods has been experimentally confirmed on a number of insect species. In spite of some species-based and HSP-type-related controversy, it is widely accepted that insect diapause is associated with changes in HSP expression both on transcriptional and translational levels. After an initial report in 1998 by Denlinger’s group, changes in expression of HSP associated with different types of diapause were confirmed for members of several insect taxa, including Hymenoptera, Coleoptera, Lepidoptera, and Diptera (Benoit et al. 2009; Rinehart et al. 2006b, 2007a; Danks 2000; Gkouvitsas et al. 2008). The growing amount of data suggests that association of HSP with developmental arrest is a common pattern even beyond insects, and temporal changes or continuously higher levels of HSP expression would have impact in developmental progress, thermal resistance, and general level of metabolic activity (Watanabe et al. 2003a). The main functions of HSPs in these examples of development arrest are proposed to be an interaction with other cryoprotectants to increase general stability of the protein pool and a direct action on the suppression of the development (Rinehart et al. 2007a).

Dehydration of cells is one of the most serious stresses, and is crucial for most organisms, since massive irreversible protein–protein aggregation caused by the hydrophobic effect occurs as a compensation for the loss of free water. Such changes in most cases lead to death, as the majority of organisms have limited potency to withstand water loss (Alpert 2006; Goyal et al. 2005; Sakurai et al. 2008a; Bohnert 2000). At the same time, there are examples of anhydrobiosis—the phenomena of maintaining viability for a long period of time under the absence of free water. The best-characterized examples include microorganisms, plants, rotifers, nematodes, tardigrades, crustaceans, and insects (Alpert 2006; Goyal et al. 2004; Watanabe 2006). In such organisms, all biochemical reactions and metabolism are undetectable in the dried state, but anhydrobiotes are able to revive back to the active life in a short period of time after appearance of water. In some groups of organisms, anhydrobiosis is an obligate part of the life cycle, while others continuously maintain the potential to reversibly enter the dry state (Crowe and Madin 1974; Guidetti and Jonsson 2002; Watanabe 2006; Watanabe et al. 2005; Clegg 2001, 2005). While physiological and morphological aspects of anhydrobiosis are relatively well-described, the molecular mechanisms allowing such natural dry preservation of cell organelles and macromolecules are yet to be understood in detail. HSPs have, for a long time, been thought to have high impact on the intracellular processes associated with desiccation tolerance in higher eukaryotes, but only a few members of this group of chaperones have been actually analyzed with a special focus to anhydrobiosis. Two small HSPs, p26 and artemin, have been found in high amounts (10–15% of total non-yolk proteins) in the encysted embryos of several branchiopod crustaceans, and a growing amount of data suggests that chaperone activity of these proteins is a key factor for the formation of dry cysts, viable and resistant to environmental stresses (Clegg 2001, 2005; Willsie and Clegg 2001). Accumulation of small HSPs has also been observed in plant seeds (Hoekstra et al. 2001; Kalemba and Pukacka 2008; Wehmeyer and Vierling 2000). Larger HSPs have had even less attention. Recently, Schill and co-authors have demonstrated that at least some isoforms of HSP70-coding genes are up-regulated when tardigrades enter anhydrobiosis and revive back to active metabolism, while other chaperones show no clear pattern of involvement in the process of anhydrobiosis (Schill et al. 2004, 2009; Reuner et al. 2009).

In the present study, we have focused on the African chironomid Polypedilum vanderplanki—the largest known anhydrobiotic animal. The larvae of this chironomid can withstand complete desiccation and maintain viability for years in a dry state, indicating the existence of a highly effective mechanism of long-term preservation of proteins in the dried larvae. The activity of HSPs, together with other protectants (trehalose and LEA proteins) for cells and organelles, has previously been suggested to be of significance for the larvae, enabling protection of the metabolic machinery upon anhydrobiosis (Kikawada et al. 2006, 2007; Nakahara et al. 2008; Watanabe 2006). We have conducted comparative analyses of structural and expression of genes coding six members of main HSP families and a heat shock factor (HSF) in relation to anhydrobiosis of P. vanderplanki.

Materials and methods

Insect rearing

P. vanderplanki larvae were reared on milk agar under controlled light (13 h light:11 h dark) at 27 to 28°C. The procedure of desiccation to induce anhydrobiosis is as previously described (Watanabe et al. 2003b), i.e., the larvae were placed on filter paper with 0.44 ml of distilled water in a glass Petri dish (diameter 65 mm, height 20 mm), which was set in a desiccator (20 × 20 × 20 cm) with 1 kg of silica gel. Larvae for RNA and protein expression analyses were sampled according to the time (in hours) passed from the beginning of desiccation (D) and of rehydration (R), correspondingly.

Heat shock treatment

To examine the heat shock response of HSP-coding gene expression, 100 wet active larvae were kept at 42°C for 60 min in a 50-ml tube with preheated deionized water and then transferred to a tube of the same volume of the water at 25°C during 90 min for recovery. After that, total RNA was extracted from the larvae for further analysis. The control larvae were kept at 25°C continuously in deionized water until RNA extraction.

Chaperones cDNA cloning from P. vanderplanki

In a P. vanderplanki EST database (Kikawada et al. 2006), the clones showing structural similarities to known hsps were isolated, and the corresponding full-length cDNAs were obtained by 5′- and 3′-RACE using a SMART RACE cDNA amplification kit (Clontech, Mountain View, CA, USA) with specific primers (sequences are available upon request). The full length of Pv-hsps and Pv-hsf1 corresponding cDNAs were subcloned into pCR®4-TOPO® vector (Invitrogen, Carlsbad, CA, USA) and further used as templates for real-time PCR.

Quantitative real-time PCR

Total RNA from hydrated, dehydrating, rehydrated, and heat-shocked larvae was extracted using Trizol (Invitrogen) and the RNeasy Mini Kit (Qiagen, Hilden, Germany) and reverse transcribed using Ready-To-Go™ T-Prime First-Strand Kit (GE Healthcare Bio-Sciences, Piscataway, NJ, USA). The RNA samples from dehydrating and rehydrating larvae were named “D” and “R”, respectively, and numbers correspond to the hours of treatment. Real-time PCR was performed using a LightCycler® 2.0 real-time PCR apparatus (Roche Diagnostics, Basel, Switzerland) with SYBR® Green PCR Master Mix (TaKaRa, Ohtsu, Japan).

Amplifications were performed using 1× SYBR Green PCR mix (TaKaRa) and 10 pmol of each primer. PvEf1-alpha gene (AB490338.1) was used as an internal standard for data normalization and quantification. The expression of each gene was tested in triplicate in each of the three biologically independent experiments. The cycling conditions were as follows: 3 min activation at 95°C, 45 cycles of 10 s at 95°C, 20 s at 60°C, and 25 s at 72°C. Melting curves from 60 to 99°C, rising by 1°C at each step and pausing 5 s after each step, and the accompanying software were used for qPCR data normalization and quantification. The full list of primers is described in Supplementary Table S1.

Statistical analysis

Results of gene expression are reported as means ± 95% CIA (95% confidence interval) with statistical evaluation performed using a two-tailed Student t test. A difference at P < 0.05 was considered significant in Prism version 5 (GraphPad Software, San Diego, CA, USA). The full list of reference sequences used in this study is described in Supplementary Table S2.

Results

Heat shock factor (Pv-hsf1) gene expression in relation to anhydrobiosis

The full sequence for Pv-hsf1 (HM589528) was obtained by a combination of sequence data from the P. vanderplanki EST database and RACE cloning. Pv-hsf1 has a 15-bp 5′-UTR, a 571-amino acid coding ORF, and a 771-bp 3′-UTR (Supplemental Fig. S1). The deduced amino acid sequence of Pv-hsf1 shows the typical characteristics of the HSF1 family (Wu 1995), including the DNA binding domain and oligomerization domains (Supplemental Fig. S1). Phylogenetical analysis confirmed the identity of Pv-HSF1 (Fig. 1c).

Fig. 1
figure 1

Relative mRNA expression profiles for Pv-hsf1 and Pv-hsp90 in heat-shocked (a, d) and anhydrobiotic (b, e) chironomid larvae. Values for the mRNA level of each gene were corrected with PvEf1-alpha expression level. The level of expression was calculated for each gene relative to the expression in control hydrated larvae (value = 1). Error bars represent mean value ± 95% CI for three replicates. cont. control hydrated larvae. c, e Neighbor-joining tree of the Pv-HSF1 and Pv-HSP90 amino acid sequences

Full size image

Pv-hsf1 expression was examined throughout dehydration and rehydration of the larvae using real-time PCR technique. While transcripts were detected in all samples, relative Pv-hsf1 expression increased in the larvae already in D-8 (corresponding to 8 h of desiccation) and maintained a high level, up to nearly complete dehydration of the larvae (Fig. 1b). In the desiccated, rehydrating, and completely revived larvae (R 1–48), Pv-hsf1 expression was detected at low level as observed in non-stressed active larvae. The Pv-hsf1 gene was also heat shock-inducible (Fig. 1a).

Heat shock protein 90-kDa (Pv-hsp90) gene expression in relation to anhydrobiosis

The full-length sequence for Pv-hsp90 (HM589529) was obtained through combination of the sequence from the EST database and RACE cloning of P. vanderplanki cDNA. A conceptual translation of the composite cDNA revealed that the ORF encodes a 714-amino-acid protein with classical ATPase domain and HSP90 catalytic domain (Supplemental Fig. S2). The deduced amino acid sequence of Pv-hsp90 is structurally similar to cytoplasmic HSP90s and, according to the modern nomenclature, should be classified as a member of HSP90a group (Fig. 1f).

The presence of mRNA encoding Pv-HSP90 was linked to the entering to anhydrobiosis and further rehydration periods. As shown in Fig. 1e, over-expression of Pv-hsp90 was detected after 8 h of desiccation. After approximately 48 h of desiccation, its expression reached the maximum level, but in completely dried larvae (R0) and during the rehydration mRNAs gradually declined to the level of average metabolic expression pattern at 48 h after rehydration. The expression of Pv-hsp90 was also enhanced by heat shock (Fig. 1d).

Expression of genes coding two 70-kDa heat shock proteins (Pv-hsp70 and Pv-hsc70) in relation to anhydrobiosis

We have cloned two full transcripts corresponding to proteins from HSP70 family (Fig. 2). The first transcript, named Pv-hsc70 (HM589530), is a full-length clone of 2,272 bp, with a 1,962-bp ORF encoding a 654-aa protein. The deduced amino acid sequence contained the typical HSP70 protein family signatures 1–3 (Supplemental Fig. S3) and showed the highest similarity to Chironomus HSC70 (AAN14525) (Fig. 2c). No up-regulation of Pv-hsc70 was observed under heat stress, and we defined the transcript as a cognate 70-kDa HSP (Fig. 2a). Analysis of Pv-hsc70 expression showed that up-regulation of the mRNA was first detectable in the larvae subjected to 8-h drying and kept increasing with maximum level occurring at 48 h of dehydration. The mRNA level showed a lower value in completely dried and just rehydrated larvae and finally dropped to the average non-stress level comparable to the 12–24 h revived larvae (Fig. 2b).

Fig. 2
figure 2

Relative mRNA expression profiles for Pv-hsp70, Pv-hsc70, and Pv-hsp60 in anhydrobiotic (b, e, h) and heat-shocked (a, d, g) larvae. Values for the mRNA level of each gene were corrected for expression level of PvEf1-alpha, and the relative level of expression changes for each gene was calculated using that of control hydrated larvae as standard (value = 1). Error bars represent mean value ± 95% CI for three replicates. cont. control hydrated larvae. c, f, i Neighbor-joining tree of Pv-HSC70 and Pv-HSP60 amino acid sequences

Full size image

The second full-length transcript, Pv-hsp70 (HM589531), consisted of 2,441 bp with a 1,974-bp ORF, encoding 658-aa protein. The deduced amino acid sequence, with the highest similarity to Aedes hsc70 (ABF18258), showed the presence of all conservative HSP70 family structural signatures (Fig. 2f, Supplemental Fig. S4). Over-expression of the transcript in response to heat stress was observed, and we defined the gene as inducible HSP70 (Fig. 2, Supplemental Fig. S4). Though up-regulation of Pv-hsp70 was already detectable in larvae after 8 h of desiccation (Fig. 2e) and reaching at its peak in the larvae subjected to 48-h drying, the level of expression was lower than that of Pv-hsc70 (Fig. 2b). Furthermore, a second peak of up-regulation appeared in the larvae 3 h after the beginning of rehydration. The expression returned to the average metastatic level at the stage of 24 h of rehydration (R-24, Fig. 2e).

Heat shock protein 60-kDa (Pv-hsp60) gene expression during anhydrobiosis

The cDNA encoding the P. vanderplanki HSP60 (Pv-hsp60, HM589532) was 2,243 bp in length (Supplemental Fig. S5) with a 1,710-bp ORF. The deduced protein comprised 569 amino acids. The first 26 triplets downstream of the initiation codon ATG encode amino acids that together have the necessary characteristics of a mitochondrial presequence. A GGM motif was located at the carboxyl terminus; other domains indicating typical mitochondrial HSP60 (GroEL_like type I chaperonin) properties were found in the deduced protein sequence (Supplemental Fig. S5, Fig. 2i). The highest-scoring amino acid similarities were to HSP60 (chaperonins) from Anopheles (XP_318461.2), Culicoides (AAB94640.1), and Aedes (XP_001661764). The gene was not heat-inducible (Fig. 2g). Expression activity of Pv-hsp60 in the larvae was similar to that of Pv-hsp70: a two-peak pattern during entering and reviving from anhydrobiosis. We observed initial increase of the expression already in the larvae subjected to 8 h of desiccation and some further increase (2–4 folds) up to the nearly complete desiccated state of the larvae (D-48). At the same time, level of expression of Pv-hsp60 in rehydrating larvae was initially on the lower level compared with that in the dehydrating larvae, but showed the second peak in a few hours after complete revival of the larvae (Fig. 2h).

Expression of genes coding two small chaperones pv-hsp20 and pv-p23 (p26 homolog) during anhydrobiosis

The last two analyzed chaperones belong to the group of low molecular mass proteins containing alpha-crystallin domain (alpha-HSP). One of the small HSP-coding genes (Pv-hsp20, HM589533) is a full-length clone of 835 bp with a 522-bp ORF (Supplemental Fig. S6), coding a putative 20-kDa protein (173 aa) with highest similarity to Culex HSP22 (XP_001847195) and Drosophila HSP27 (ABX80641.1) (Fig. 3c).

Fig. 3
figure 3

Relative mRNA expression profiles for Pv-hsp20 and Pv-p23 in anhydrobiotic (b, e) and heat-shocked (a, d) larvae. Values for the mRNA level of each gene were corrected for expression level of EF1-alpha, and the relative level of expression changes for each gene was calculated using that of control hydrated larvae as standard (value = 1). Error bars represent mean value ± 95% CI for three replicates. cont. control hydrated larvae. c, f Neighbor-joining tree of the Pv-HSP20 and Pv-p23 amino acid sequences

Full size image

The second small HSP (Pv-p23, HM589534) is represented by full-length clone of 885 bp, with a 588-bp ORF (Supplemental Fig. S7), coding a putative 22.8-kDa protein (196 aa) that most closely resembles Belgica sHSP (ABF01017) and Artemia p26 (ABC41139) (Fig. 3f). Both HSPs showed presence of the conservative alpha-crystallin domain, assuming formation of large oligomeric complexes as a structural prerequisite for the chaperone activity. At the same time, the domain showed a low level of cross-similarities (less than 35% on the amino acid level), suggesting that activity patterns and cellular functions of the proteins would be different. In addition, translated sequence of Pv-hsp20 showed the presence of a typical WDPF motif (amino acids residues 10–30, Supplemental Fig. S6 and S7), whereas Pv-p23 did not.

We found that both Pv-hsp20 and Pv-p23 were up-regulated in the larvae subjected to heat shock (Fig. 3a, d). Furthermore, both genes showed over-expression associated with anhydrobiotic processes in the larvae, but their expression patterns were different. Under normal non-stressed conditions, the expression of Pv-hsp20 was at the low level (Fig. 3b, e) and then up-regulated already after 8 h of desiccation and showed the highest level of expression in the nearly dry and just revived larvae and followed by slow declining to the normal, non-stressed level at 48 h of rehydration. In contrast, expression of Pv-p23 was dehydration-inducible and restricted to the late stages of dehydration of the larvae showing its maximum level in the completely desiccated and just revived larvae and dropped to non-stress lowest level as early as 12 h after rehydration (Fig. 3e).

Discussion

In this study, we have demonstrated that expression of genes coding several members of the main HSP groups is tightly linked to the anhydrobiosis of P. vanderplanki larvae. Figure 4 summarizes the periods of up-regulation of HSP-corresponding genes in chironomid larvae during the dehydration–rehydration cycle.

Fig. 4
figure 4

Temporal representations of anhydrobiosis stages, in which genes encoding heat shock proteins are up-regulated in the larvae of the sleeping chironomid

Full size image

Large HSP-encoding genes

Experimental data related to involvement of large heat shock proteins specifically in anhydrobiosis in invertebrates is limited to several papers related to hsp expression in tardigrades, where it is shown that there are at least three hsp70-like genes (all heat shock up-regulated) with different expression profiles in relation to anhydrobiosis. All three isoforms show significant up-regulation after revival of the tardigrades to active stage (Schill et al. 2004; Jonsson and Schill 2007; Reuner et al. 2009). At the same time, after further analysis of other groups of hsp, the authors concluded that chaperones are not involved in anhydrobiosis in water bears.

In the present study, we have isolated two different isoforms of hsp70 and they both are significantly up-regulated during entire process of anhydrobiosis in the chironomid larvae (Fig. 2b, e). Pv-hsp70, a heat-shock-responsive isoform, referred to gene encoding as real HSP70, showed an initial up-regulation pattern at the beginning of dehydration of the larvae and later, an even higher peak, at the stage corresponding to 3 h after rehydration (Fig. 2e). Such patterns of inducible hsp70 expression are similar to that in tardigrades, where it was suggested that chaperone over-expression was related to refolding of the proteins damaged during formation of the “tun” and revival back to active life after anhydrobiosis (Schill et al. 2004). Pv-hsc70, a cognate isoform, instead shows only one-peak up-regulation, with a gradual increase in the expression level during desiccation, reaching a maximum level at the final step of dry larvae formation (stage D-48). Then, continuous decrease of mRNA level was observed during rehydration, approaching average non-stressed level at the stage R12 (Fig. 2b). Based on absolute copy numbers estimated by real-time PCR and EST database (data not shown) and on the expression pattern, Pv-hsc70 seems to be the dominant HSP70 isoform both in wet (active) and dry (anhydrobiotic) larvae. Activity of HSC70 (cognate form of HSP70) is usually tightly linked to de novo protein synthesis machinery and during hypometabolic processes, such as insect diapause or under stress conditions (including desiccation), where the expression of hsc70 is reported to be unchanged (Denlinger 2002; Sonoda et al. 2006a) or even down-regulated (Gkouvitsas et al. 2009). No specific studies on HSC70 in anhydrobiotic invertebrates have been reported so far, but our results suggest that protein product of Pv-hsc70 is an active participant of anhydrobiotic processes in the larvae. While in the cases of typical hypometabolic processes both transcription and translation activity drop to lower level, in the larvae of P. vanderplanki de novo synthesis of several groups of LEA proteins and conversion of glycogen to trehalose followed by its delivery to cells are critical factors defining the success of anhydrobiosis (Kikawada et al. 2006, 2008; Watanabe et al. 2007). As water leaves cells during dehydration, the synthesis of LEA proteins and other molecular factors essential for the anhydrobiotic state of larvae have to be completed under conditions of increased intracellular concentration of macromolecules. The over-expression of Pv-hsc70 probably reflects an active participation of Pv-HSC70 in this synthesis process.

Members of HSP90 family are considered to be the among the most abundant chaperones in the cytoplasm of eukaryotic cells, but, to the best of our knowledge, no studies focusing on the relation of these chaperones to anhydrobiotic processes have been reported to date (Goyal et al. 2005; Schill et al. 2009; Watanabe 2006). In spite of the fact that HSP90 is important for multiple developmental and metabolic networks, information about its involvement in insect diapause remains controversial. Depending on the type of diapause and species, the expression level of the chaperone decreases, increases, or remains unchanged (Lopez-Martinez and Denlinger 2008; Rinehart et al. 2007a; Sonoda et al. 2006b). Recently, it was found that expressions of both HSC70 and HSP90 in the larvae of flesh fly were unresponsive to dehydration (Hayward et al. 2004; Rinehart et al. 2006a, 2007a). In our study, up-regulation of Pv-hsp90 gene in the larvae of P. vanderplanki was observed throughout the entire process of entering into, and revival from, anhydrobiosis with a dynamic of expression similar to that of Pv-hsc70 (Figs. 1 and 2). The similar patterns of over-expression for both genes continued during rehydration of the larvae, assuming the reinitiation of productive protein folding pathways, in which both HSC70 and HSP90 play a fundamental role (Leung and Hightower 1997; Young et al. 2001). This might suggest the co-activity of both chaperones as a sign of maturation of the proteins involved in anhydrobiosis process (Rajapandi et al. 2000).

Pv-hsp60 also was the only gene coding a HSP60 family member found in P. vanderplanki EST database and analyzed in this study (our unpublished data). In contrast to that of other chaperones, the mRNA expression of Pv-hsp60 showed a much lower degree of change during entire anhydrobiosis (Fig. 2h). Another member of the HSP60 family, TCP-1, involved in the folding of cytoskeleton proteins was recently found to be tightly linked with diapause and low-temperature survival in insects and brine shrimp (Rinehart et al. 2007a; Wang et al. 2007), but at present we have no data about the activity of the homologous gene in the larvae of P. vanderplanki.

Genes encoding small (crystallin-like) HSPs

Here we have cloned and analyzed expression of two genes coding small heat shock proteins (Pv-hsp20 and Pv-p23; Fig. 3). Both proteins possess crystallin domains but show differences in primary structure and expression pattern during anhydrobiosis.

Pv-HSP20 has all the typical structural features of the sHSP family and is phylogenetically close to HSP27 of Drosophila. Expression of this gene is also up-regulated by heat shock (Fig. 3). The expressional pattern of Pv-hsp20 resembles those of small HSPs appearing in developing plant seeds ahead of the processes associated with desiccation (Coca et al. 1994; Wehmeyer et al. 1996). To date, there have been no studies focusing on the participation of HSP27 homologs in the anhydrobiosis of invertebrates. At the same time, activity of this gene is tightly linked to the resistance to environmental changes and to metabolic suppression during developing and diapausing stages of insects and other invertebrates (Rinehart et al. 2007b; Hayward et al. 2004; Saravanakumar et al. 2008). Up-regulation of Pv-hsp20 was already observed in the larvae 8 h after the start of desiccation, and a high level of the corresponding mRNA was maintained for at least 24 h after revival from anhydrobiosis. We assume that Pv-hsp20 encodes a member of sHSPs and that its initial up-regulation in dehydrating larvae may be related to chaperone activity in the larvae suffering from different factors associated with water deficit, including molecular crowding and oxidative stress. The dynamic of Pv-hsp20 mRNA abundance is similar to that of Pv-hsp90 and Pv-hsf1, but at the present step we have no data on joint activity of these chaperones in the larvae (Fig. 4).

In contrast, expression of the second small HSP-coding gene, Pv-p23, was undetectable in wet and semi-dry larvae (Fig. 3e). Based on phylogenetic analysis, we consider Pv-P23 to be a desiccation-inducible ortholog of Artemia's p26—a small chaperone functionally involved in the process of anhydrobiosis in anoxic cysts of lower crustaceans (Willsie and Clegg 2001). In Artemia's cysts, p26 increases in concentration, occupying 10–15% of total non-yolk protein biomass, and migrates to the nucleus in dehydrating cysts. Artemia's p26 has been proposed to act as a chaperone, stabilizing nuclear acids (Clegg 2005; Liang and MacRae 1999; Willsie and Clegg 2001).

Drastic increase of Pv-p23 mRNA in the larvae occurred between stages D-24 and D-48, i.e., during the period preceding vitrification of the cells and tissues. Such a dehydration-specific activity of the Pv-p23-coding gene might suggest a direct link between this protein and anhydrobiosis. As a member of the crystallin family, Pv-p23 would also form an oligomeric structure and possibly be involved in the vitrification of larval tissues on the late stages of desiccation (Sakurai et al. 2008b). In summary, we have shown that the initiation of anhydrobiosis-related processes in P. vanderplanki is associated with up-regulated expression of the genes coding major groups of HSP. We suggest that HSPs constitute an important part of anhydrobiosis-related changes in the cells of P. vanderplanki larvae. At the same time, while our discussion is based on the assumption that mRNA expression reflects the activity of mature HSPs, it should be noted that a temporal shift between gene expression and the activity of mature stress proteins was observed in some cases (Kostal and Tollarova-Borovanska 2009). Consequently, more data are needed to make a conclusion on the activity of HSPs in anhydrobiotic larvae. Thus, further studies with a special focus on cellular and tissue-specific localization, defining the role of each member of this family, including careful analysis of protein expression of each isoform, will need to be carried out for a better understanding of the origin and machinery of long-term water-free preservation of biomolecules during anhydrobiosis in the sleeping chironomid.