• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntphysLink to Publisher's site
Plant Physiol. Jun 2006; 141(2): 638–650.
PMCID: PMC1475461

Structural Investigation of Disordered Stress Proteins. Comparison of Full-Length Dehydrins with Isolated Peptides of Their Conserved Segments1


Dehydrins constitute a class of intrinsically disordered proteins that are expressed under conditions of water-related stress. Characteristic of the dehydrins are some highly conserved stretches of seven to 17 residues that are repetitively scattered in their sequences, the K-, S-, Y-, and Lys-rich segments. In this study, we investigate the putative role of these segments in promoting structure. The analysis is based on comparative analysis of four full-length dehydrins from Arabidopsis (Arabidopsis thaliana; Cor47, Lti29, Lti30, and Rab18) and isolated peptide mimics of the K-, Y-, and Lys-rich segments. In physiological buffer, the circular dichroism spectra of the full-length dehydrins reveal overall disordered structures with a variable content of poly-Pro helices, a type of elongated secondary structure relying on bridging water molecules. Similar disordered structures are observed for the isolated peptides of the conserved segments. Interestingly, neither the full-length dehydrins nor their conserved segments are able to adopt specific structure in response to altered temperature, one of the factors that regulate their expression in vivo. There is also no structural response to the addition of metal ions, increased protein concentration, or the protein-stabilizing salt Na2SO4. Taken together, these observations indicate that the dehydrins are not in equilibrium with high-energy folded structures. The result suggests that the dehydrins are highly evolved proteins, selected to maintain high configurational flexibility and to resist unspecific collapse and aggregation. The role of the conserved segments is thus not to promote tertiary structure, but to exert their biological function more locally upon interaction with specific biological targets, for example, by acting as beads on a string for specific recognition, interaction with membranes, or intermolecular scaffolding. In this perspective, it is notable that the Lys-rich segment in Cor47 and Lti29 shows sequence similarity with the animal chaperone HSP90.

Intrinsically disordered proteins have gained increasing attention during the past few years. These proteins have no fixed tertiary structure under physiological conditions, but nevertheless perform specific functions in different cellular processes (Dunker et al., 2000). Of 7,849 protein sequences from the Arabidopsis (Arabidopsis thaliana) genome, approximately 650 (8%) were predicted to lack an ordered structure. Taking into account partial disorder as well, the number increases more than 3-fold; 29% of the proteins in Arabidopsis are predicted to have sequence stretches of more than 50 amino acids that lack fixed secondary or tertiary structure (Dunker et al., 2000).

So how do disordered proteins function? Normally, protein function requires well-defined scaffolds for molecular recognition and catalysis. In some cases, disordered proteins adopt such well-ordered structures as they bind to biological targets such as other proteins (Bourhis et al., 2004), membranes (Davidson et al., 1998), RNA (Tompa and Csermely, 2004), or DNA (Love et al., 2004). Similar induction of structure is indicated in self assembly (Namba, 2001) and oligomerization as observed for the synergetic folding of the activator for thyroid hormone and retinoid receptors-CREB-binding protein complex (Namba, 2001; Demarest et al., 2002). There are also examples where induction of ordered structure is conditional (e.g. the protein structure responds to changes in temperature, pH, or ion concentration [Uversky, 2002] or the availability of water [Luo and Baldwin, 1997]). It is thus apparent that disorder in certain cases could actually provide an advantage. For example, in regulation, the function of individual proteins can be tuned by changes in the environment. In signaling, high structural flexibility could allow interactions with multiple targets and efficient control by rapid proteolytic turnover. Folding upon target interaction also has the benefit of providing high binding specificity with low affinity because their dissociated states are favored by a compensating gain in entropy (Dunker et al., 2002). Moreover, disordered proteins could provide large molecular interfaces without the drawback of a large protein scaffold. In support of this idea, the binding interfaces employed by disordered proteins are generally more extensive than for common proteins (Gunasekaran et al., 2003). If these interfaces were instead presented on folded proteins, their size would need to increase 15% to 30% with significant consequences for both cell size and intracellular crowding (Gunasekaran et al., 2003). Accordingly, disordered proteins can in several cases be seen as unstable variants of normal proteins where the biological activity is, after all, coupled to ordered three-dimensional structure in a rather conventional manner. The main difference is that a binding event is needed to pull the folding equilibrium toward the biologically active structure.

There are also instances where the functional coupling to fixed three-dimensional structure is less obvious. One example is the family of plant-specific dehydrins (Eom et al., 1996; Lisse et al., 1996; Soulages et al., 2003) that shows, on the whole, no sequence similarity to any other proteins (Uversky et al., 2000). These disordered proteins are expressed at high levels under conditions of stress, such as cold, drought, high salinity, phosphate starvation, or abscisic acid. A unique feature of the dehydrins is the presence of one or several copies of a highly conserved sequence stretch (EKKGIMDKIKEKLPG), named the K-segment (Dure, 1993; Close, 1996). The K-segment is particularly abundant in cold-induced dehydrins where it could be repeated up to 11 times (Close, 1997). Some dehydrins also contain a conserved poly-Ser stretch, called the S-segment, followed in several cases by a characteristic stretch of residues rich in Lys. Moreover, they sometimes have a 7-amino acid segment, called the Y-segment, at the N terminus. Accordingly, the nomenclature of the dehydrins is written YnSnKn (Close, 1996). The dehydrins have been found in the cytoplasm, the vacuole, and the nucleus (Heyen et al., 2002; Karlson et al., 2003) and their location is, in some cases, tissue specific (Nylander et al., 2001). Several models have been put forth for the dehydrin's poorly understood biological function. Based on the amino acid composition of the K-segment, it was suggested that an amphiphatic α-helix was formed (Dure, 1993) that could interact with membranes or proteins modulating their phase properties and conformational transitions (Danyluk et al., 1998; Koag et al., 2003). Along the same lines, it was also suggested that the dehydrins act as chaperones or protein cryoprotectants (Hara et al., 2001). Yet other ideas are that the dehydrins have a role in metal and calcium binding (Heyen et al., 2002; Alsheikh et al., 2003), bind and store water to prevent cells from complete dehydration (Close, 1996; Rinne et al., 1999), or act as antifreeze proteins by preventing intracellular ice growth (Wisniewski et al., 1999).

In this study, we investigate the structural properties of four different dehydrins from Arabidopsis, Cor47 (SK3), Lti29 (SK3), Lti30 (K6), and Rab18 (Y2SK2), by mapping out their response to temperature and solvent perturbations. To elucidate specifically the role of their constituent K-, Y-, and Lys-rich segments, these have been studied separately as synthetic peptides. The results show that none of the four dehydrins make any specific tertiary interaction in physiological buffer and that their overall content of classic secondary structure is low. Nevertheless, they reveal a significant content of poly-Pro (PII) helices consistent with earlier observations on other dehydrins and late-embryogenesis abundant (LEA) proteins (Soulages et al., 2002, 2003). This type of expanded secondary structure represents an intriguing example of structures involving water binding combined with high structural adaptability. Notably, PII content is not controlled by the conserved segments, which remain as disordered as the full-length proteins unless forced into α-helical conformations by the addition of trifluorethanol (TFE). Titration with the stabilizing salt Na2SO4, which is known to be a powerful structural promoter in other systems, disfavors the idea that the dehydrins are in equilibrium with folded high-energy states. The results suggest that the dehydrin proteins do not rely on the acquisition of fixed tertiary structure for their biological function, but more likely serve as a highly specialized dynamic linkage for the conserved segments.


Full-Length Proteins and Conserved Segments Are Disordered under Standard Conditions

The amino acid sequences of the dehydrins investigated in this study are shown in Figure 1. In physiological buffer at 4°C, the circular dichroism (CD) spectra of all four full-length dehydrins display large minima around 200 nm and an ellipticity near zero around 220 nm, indicative of poorly structured conformations with low content of secondary structure (Fig. 2A). The observation is consistent with the predominantly hydrophilic signature of the dehydrin sequences (i.e. their residue composition does not allow the formation of an extensive hydrophobic core typical for folded protein). The strongest CD signal is observed for Cor47, followed by Lti30 (K6) and Lti29 (SK3), which have overall similar amplitudes and spectral characteristics. The spectrum of Rab18 (Y2SK2) deviates from the other proteins by being flatter overall with no appreciable component in the 220-nm region. Notably, the spectral differences between the four proteins show no direct relation with their sequence homology: The highest sequence identity is between Cor47 (SK3) and Lti29 (SK3; 64%) and the lowest between Cor47 and Rab18 (22%; compare Figs. 1 and and2).2). In parallel with the full-length dehydrins, CD analysis was performed on synthetic peptides of their conserved segments. These were the K-segment (EKKGIMDKIKEKLPG), the Y-segment (VDEYGNP), charge-peptide 1 (ChP-1; EEGEDGEKKKKEKKKKI), and its shorter version charge-peptide 2 (ChP-2; EDGEKKKEK; Fig. 2B). The spectra of the isolated segments are qualitatively similar to those of the full-length proteins and indicative of coil-like structures.

Figure 1.
Amino acid sequences of Cor47 (SK3), Lti29 (SK3), Lti30 (K6), and Rab18 (Y2SK2). The K-segment is highlighted in green, the Y-segment in red, the S-segment in yellow, and the Lys-rich segment (ChP-1) in purple.
Figure 2.
CD spectra of the full-length dehydrins (top) and their conserved segments (bottom) at 4°C.

Effect of Protein Concentration

In other studies, it has been observed that some proteins need to aggregate or polymerize to obtain their physiologically active states (Johansson et al., 1998). To examine the susceptibility of the dehydrins to undergo such self association, the CD spectra of the full-length constructs and their isolated segments were measured at a series of different concentrations, ranging from 0.5 to 18 mg/mL. Under the conditions used, however, neither of the constructs showed any tendency to undergo concentration-induced structural changes. Consistently, all constructs also display an approximately linear relation between CD amplitude and protein concentration (Fig. 3).

Figure 3.
The CD signal at 222 nm versus protein concentration shows linear relations indicating that the dehydrins and their conserved segments have low propensities to aggregate. The K-segment, [filled square]; the Y-segment, ○; Lti30, X; Lti29, □; ...

Experimental Determination of Helix Propensity by TFE Titration

The formation of local secondary structure is an integral part of any protein-folding reaction. To examine the propensity of secondary-structure formation of the dehydrins, we undertook titrations with the helix-inducer TFE that is commonly used as a tool for extrapolation of quantitative helix-coil equilibrium constants for peptides with little or no apparent helical content in solution. The ability to form helical structure with TFE was tested on both the full-length dehydrins and their isolated segments. The results show that all full-length proteins, except Rab18, display a pronounced increase in their helical content in the presence of TFE (Figs. 4 and and5).5). The midpoint for the helix-coil transition of Cor47 is at 25% TFE. The corresponding midpoints of Lti29 and Lti30 are at 30% and 35%, respectively, indicating that the helix propensity is slightly different for the three proteins. Interestingly, the final extent of helical structure at high TFE is not the same for Cor46, Lti29, and Lti30 because the titration curves level off at different values. As estimated by CDPro, the maximal helix content reaches about 50% for Cor47 and 20% to 30% for Lti29 and Lti30 (Fig. 5A). By comparison, Rab18 is remarkably unaffected by TFE and is only 2% helical in 90% TFE. In the latter case, the deviating result is most likely due to the unusually high Gly content of this protein (34% of total amino acids), an amino acid with low helix propensity (Fersht, 1999). The finding is in good agreement with the results of Soulages et al. (2003), who observed that a dehydrin from soybean (Glycine max; Y2K) containing 27% Gly, adopted, at most, 12% α-helices in TFE. The different final content of α-helical structure indicates that only certain sequence regions of the dehydrins are able to adapt helical structure, whereas others remain disordered in the presence of 90% TFE. If all parts of the proteins were involved in the transition, helical content would have reached 100%.

Figure 4.
CD difference spectra showing the degree of structural induction at 90% TFE at 4°C. The full-length dehydrins (top) and their conserved segments (bottom).
Figure 5.
The increase in helical content upon titration with TFE at 4°C, as derived from CD data by the spectral analysis software CDPro. The full-length dehydrins (top) and their conserved segments (bottom).

In view of the idea that the role of the K-segments is to form an amphiphatic helix in vivo, it is notable that the helical propensities of the isolated segments are relatively weak and substantially lower than for the full-length dehydrins (Fig. 5; Table I). The K-segment shows a transition midpoint around 20% TFE and plateaus at a maximum of less than 10% helical content, whereas the TFE response of ChP-1 appears less cooperative, but increases progressively to 20% helicity at 90% TFE (Fig. 5B). ChP-2, on the other hand, shows no significant response to TFE. On this basis, we conclude that dehydrin susceptibility to adopt helical structure in TFE is not primarily controlled by their conserved segments, but is mainly determined by their remaining amino acids. As will be discussed below, however, caution should be exercised in using this result for elucidating physiological function. First, the helical propensity of individual segments may be different in the context of the full-length protein (i.e. neighboring residues could modulate the helical content). Second, even low levels of flickeringly present helices could serve as critical initiation points for more complex interactions in vivo, such as template-mediated folding against membranes and interactions with low-molecular ligands or other proteins.

Table I.
Comparing experimental data with structural predictions

Effect of GdmCl

Despite the disordered nature of the dehydrins in physiological buffer, they seem to unfold further upon addition of the chemical denaturant guanidium chloride (GdmCl; Fig. 6). The CD features observed between 210 and 240 nm diminish, producing coil-like spectra at high concentrations of GdmCl. However, the unfolding transitions of the dehydrins are not sigmoidal as for archetypical proteins, but progressive, indicating low structural cooperativity, even though it cannot be excluded that the denaturation curves constitute the tails of the cooperative unfolding of species that are only partly populated in the starting material.

Figure 6.
Changes in the CD signal at 222 nm upon titration of the full-length dehydrins with GdmCl.

Effects of Na2SO4

To examine whether the titration data in Figure 6 stem from unfolding of partly populated structures, the full-length dehydrins were subjected to 0.5 m Na2SO4. In essence, Na2SO4 contributes to increased protein stability by minimizing the protein solvent-exposed surface area, and has been demonstrated to increase the occupancy of folded states and partly structured intermediates in other systems (Otzen and Oliveberg, 1999; Hedberg and Oliveberg, 2004). Even so, none of the dehydrins in this study show any response to the addition of Na2SO4, disfavoring the idea that the ambiguous spectral characteristics of dehydrins correspond to an equilibrium mixture of folded and unfolded species (data not shown). More likely, the weak structural signal in standard buffer reflects an ensemble of flickering local interactions or the presence of diffuse long-range attraction between hydrophobic side-chain moieties under poor solvent conditions. Upon addition of GdmCl, this weak structural organization is gradually lost.

Effects of Temperature

Elevated temperature increases the CD amplitudes for both the full-length dehydrins and their isolated segments (Figs. 7 and and8).8). The CD signal describes reversible changes between 4°C and 85°C, producing difference spectra typically peaking around 218 nm. Moreover, the transition amplitudes for the different dehydrins are correlated with their structural extent at 5°C (i.e. Cor47 displays the largest change and Rab18 the smallest). Interestingly, the thermal transitions display a characteristic isoelliptic point around 208 (Fig. 7A), similar to that observed for PII helices (Park et al., 1997) in equilibrium with either disordered- or β-turn conformations (Bochicchio and Tamburro, 2002, and refs. therein). The unfolding transition of PII helices would, in this case, represent the tail of a broad melting curve with a midpoint around −30°C (Park et al., 1997); that is, the occupancy of PII structure is only fractional at 5°C. A consistent spectral response to elevated temperature has been reported for a soybean dehydrin and a LEA protein, both with respect to the breadth of the thermal transition and the isoelliptic point around 208 nm (Soulages et al., 2002, 2003). However, the PII content of the dehydrins need not be family specific, but could reflect the generic properties of any disordered polypeptide chain (Shi et al., 2005; Whittington et al., 2005). As a control, we undertook the same experiment with a highly unstable mutant of the Greek key protein superoxide dismutase (SOD1A4V) that, in contrast to wild-type SOD, is predominantly unfolded under physiological conditions (i.e. ΔGd – N < 0 [gift from Mikael Lindberg]). As shown in Figure 8A, unfolded SODA4V displays melting characteristics that closely resemble the dehydrin data.

Figure 7.
Structural response to increased temperature in pure buffer and in the presence of TFE. Top, CD spectra of Lti29; spectra are reading from the top at 220 nm Lti29 at 4°C, Lti29 at 25°C after having been heated to 85°C, Lti29 at ...
Figure 8.
Temperature scans at 222 nm indicating the gradual loss of PII helices. The full-length dehydrins and a destabilized mutant of SOD1 (top) and the conserved segments (bottom).

Of the isolated peptides, the K-segment shows the most pronounced thermal transition, followed by ChP-2 and ChP-1 (Fig. 8B). The reason for these sequence-specific differences (compare with the full-length dehydrins) is still unclear, but may be related to differences in α-helical propensity; the melting of residual α-helical structure could oppose the signal from PII helices. To examine this possibility, we estimated the contribution from α-helical melting from thermal scans in the presence of TFE. As predicted, the melting of α-helical structure yields positive difference CD spectra (i.e. the ellipticity vanishes at high temperatures) with a shoulder at 221 nm (Fig. 7B). As this shoulder is also significantly shifted from the PII component at 218 nm obtained from thermal scans in the absence of TFE, it can be deduced that the K-segment describes a thermal transition that is dominated by PII melting, whereas ChP-1 describes a more balanced combination of PII and α-helical melting. Consistently, the melting transitions of the K-segment at different concentrations of TFE begin at the levels expected from the relative content of PII and α-helical structure and converge at high temperature where they monitor a common thermally unfolded state (data not shown). At around 25% TFE, the two counteracting spectral components are precisely balanced, producing a flat line. Thus, we conclude that the variation in melting characteristics of the dehydrin species is, at least partly, accounted for by differences in their intrinsic α-helical propensity.

Effects of Metals

The dehydrins in this study are known to bind metal ions, a characteristic used in the purification procedure (Svensson et al., 2000). To investigate the structural effects of such metal coordination, Cor47 and Lti29 were subjected to a series of different metal-containing salts (data not shown). To achieve high solubility, we focused our analysis on the Cl salts CuCl, NiCl, FeCl (<2 mm), MgCl, MnCl, ZnCl, and CaCl. As a control for possible Cl or ionic strength contributions, we used NaCl. However, the effect of metals on the CD spectra of Cor47 and Lti29 is not compelling. CuCl and ZnCl show minor effects on the spectra, but only at concentrations above 10 mm (i.e. outside the physiological limits [data not shown]).


Dehydrins in Relation to Other Disordered Proteins

Disordered proteins can be categorized according to their CD spectra by plotting the ellipticity at 222 nm versus 200 nm (Uversky, 2002). The plot shows that the proteins cluster into two groups, where one group represents intrinsic coils and the other represents the so-called premolten globular proteins. Even though Uversky's structural assignment of the two clusters is yet tentative, the distinct clustering of the CD spectra provides a simple classification of disordered states. Comparison with data obtained in this study shows that the positions of the dehydrins are relatively scattered in terms of their CD characteristics and span both the coil and the premolten globular regions of Uversky's plot (Fig. 9). At low temperatures, the dehydrins tend to fall outside the reference groups by having a 222-nm signal typical for intrinsic coils, but a 200-nm signal in the region for premolten globules. At elevated temperatures, however, their spectra shift uniformly toward those of the premolten globules. From the corresponding data of the isolated ChP-1, Y-, and K-peptides, it is further apparent that these characteristics of the full-length dehydrins are not determined by their conserved segments (Fig. 9). Cor47 and Lti29 have similar composition and overall sequence content of conserved segments, but still behave differently, both in terms of their location in the 222-nm versus 200-nm plot and their thermal shift. Lti30, on the other hand, which comprises six K-segments that together comprise nearly 45% of the Lti30 sequence, shows, not surprisingly, some resemblance to the isolated K-segment. Two conclusions can be drawn from these results. First, the thermal malleability of the dehydrins exceeds the apparent structural variability of the disordered (nondehydrin) proteins in Uversky's reference set. Second, the plots of the full-length dehydrins and their conserved segments are not strongly coupled, indicating that their structures behave as separate entities in the native protein.

Figure 9.
Plot of the CD signals at 222 nm versus 200 nm for a set of disordered proteins (○; adapted from Uversky, 2002). The dataset indicates two clusters, one representing intrinsic coils (top left) and the other premolten globules (right). Data for ...

Comparing Experiments with Structural Prediction

To examine in more detail the structural content of the dehydrins, we calculated the α-helical and PII content from CD data with CDPro software and compared these data with results from structure prediction using the following freely available programs: AGADIR, which predicts the α-helical propensity of peptides, PONDR, which calculates overall disorder from sequence data, DISEMBL, which combines a set of algorithms for disorder prediction (Linding et al., 2003), TANGO, which predicts β-aggregation, and, finally, CAST, which predicts sequence complexity. The results are listed in Table I. In good agreement with the experiment, AGADIR calculated an average α-helix propensity for the dehydrins and peptides in this study that is low overall. Even so, it is notable that when the full-length proteins are screened by AGADIR for local maxima in α-helical propensity, these coincide in most, but not all, instances with the position of a conserved segment. It is easy to envisage that these scattered stretches with helix-forming capacity could have functional importance even though they remain largely unfolded under the conditions explored here, for example, by constituting inducible docking sites to external ligands, like membrane surfaces or other biomolecules (Uversky et al., 2005). As pointed out above, such heterogeneous distribution of the α-helical propensity explains why the total helical content of the dehydrins does not reach 100% at high TFE concentrations: Only the portions of the sequence that are able to adopt helical structure are affected by the helix inducer. With respect to the random coil properties, PONDR predicts a relatively high degree of disorder (>60%) for Rab18, Cor47, and Lti29, whereas the corresponding value for Lti30 turns out to be unrealistically low at only 21% (Fig. 2; Table I). The reason for this low value is that PONDR assigns full helical content for the K-segment. Consistently, DISEMBL calculates a loop/coil content that is similar overall across the different proteins and with only small indications of ordered structure. Following the experimental observation that the dehydrins are soluble even at very high protein concentrations (Fig. 3), TANGO predicts low aggregation propensities for all proteins in Table I. In contrast, the unfolded states of structured proteins typically aggregate in physiological buffers if prevented from adopting the folded conformation (Silow and Oliveberg, 2003). Underlining this difference, the only dehydrin that displayed a weak tendency to aggregate was Cor47, but only under the extreme conditions of 90°C in 90% TFE where most other proteins precipitate badly (Chiti et al., 2000). Independent support for the finding that dehydrins possess low propensity to aggregate was recently presented in a TANGO-based comparison of the β-aggregation propensity of globular and disordered proteins in general (Linding et al., 2004). As a rule, aggregation occurs when the amino acid sequences have high hydrophobicity, high β-sheet propensity, a low net charge, and are readily exposed to solvent. Disordered proteins that typically lack these sequence characteristics were found, as a group, to have 3 times less aggregation nucleating regions than globular proteins (Linding et al., 2004). The likelihood of dehydrins forming either specific tertiary or oligomeric structures is thus very small: Besides the high charge repulsion that intrinsically opposes chain collapse, there seems not to be enough hydrophobic residues to conglomerate a sufficiently large hydrophobic cluster. Even so, the high complexity of the dehydrin sequences predicted by CAST indicates that the amino acid composition of these proteins is critically functional and under tight biological control.

Structural Impact of the Conserved K-Segment

Although the biological role of the highly conserved K-segment is not yet established, it has been suggested on the basis of its amino acid composition that it forms an amphiphatic α-helix that could interact with membranes and thereby stabilize them under stresses (Dure, 1993). The results presented here show that, in the absence of such biological targets, the K-segment is disordered both as an isolated peptide and as part of the full-length dehydrins. This low helical propensity is particularly evident in the case of Lti30 with a total of six K-segments (comprising approximately one-half of its sequence) and still retains the archetypical CD spectrum indicative of disordered/PII structures. Moreover, it is evident that temperature changes alone cannot be used to increase significantly the α-helical content of the K-segment; the experimental estimate of its α-helical propensity through TFE titration yields only 10% α-helical structure in 90% TFE. In the absence of TFE, the α-helical content drops to 0.7%, in good agreement with predictions by AGADIR (Table I). Thus, the results provide strong evidence that any α-helical structure of the K-segments needs to be promoted through binding to specific biological ligands. One possible role for the conserved sequences could be to constitute hooks on a string interacting with particular targets/anchoring points for spatial localization, organization, or support of the cellular interior.

The Content of PII Helices

Even though the static CD spectra of the dehydrins may appear ambiguous in terms of structural content (Fig. 2), their changes to perturbations (Figs. 6–8)) reveal the presence of the nonclassic secondary structure PII helices (Bochicchio and Tamburro, 2002). Compared to α-helices, the PII helices are quite short, usually spanning over less than five residues (Stapley and Creamer, 1999). According to CDPro, the level of PII helices for the full-length dehydrins and the isolated segments in Table I is very similar at 11.5% to 15%. As the segments constitute only between 17% and 46% of the dehydrin sequences, it is clear that their contribution alone is insufficient to explain the total PII content of the full-length proteins. The result is consistent with the observations of Soulages et al. (2002), who were the first to identify PII helices in disordered plant proteins, one LEA protein group 1 (Soulages et al., 2002), and the dehydrin Y2K showing a PII content of 27% at 12°C (Soulages et al., 2003). Interestingly, the two disordered proteins in the Soulages measurements show a thermal response that is very similar to the dehydrins analyzed in this study despite considerable differences in sequence and conserved-segment composition. An intriguing as well as characteristic feature of the PII structure is that it is kept together and stabilized by water molecule bridges (Rucker and Creamer, 2002). It is also this lack of direct intramolecular hydrogen bonds that makes the PII structure intrinsically hard to detect spectroscopically. Nevertheless, PII structures have been linked to several specific functions in animal and bacterial systems. For example, in transcription, signal transduction, cell motility, and the immune response (Ma et al., 2001), the flexible PII structure is believed to allow an induced fit to the target molecules (Bochicchio and Tamburro, 2002). The PII helices here seem to constitute a general structural motif that could be recognized by Pro recognition domains, such as the SH3 domain (Li, 2005), or to be involved in binding to lipids (Kanai et al., 2001) and possibly also to nucleic acids (Hicks and Hsu, 2004). Moreover, PII helices have been implicated in self assembly of fibrous structures through their enrichment in elastomeric proteins (Bochicchio and Tamburro, 2002). In particular, the arabinogalactan proteins contain long uninterrupted sequences of PII-inducing hydroxyprolines that assemble into large fibrous suprastructures. These proteins are anchored to the plasma membrane via lipids and have been suggested to form a periplasmic polymer cushion that stabilizes the plasma membrane against hydrostatic pressure. In contrast to the PII helices in dehydrins, however, the arabinogalactan protein structures are controlled by glycosylation (Ferris et al., 2001; Tan et al., 2004).

Putative Role of PII Helices in Water Binding

Interestingly, the PII structure also seems to be a characteristic feature of unfolded globular proteins (Shi et al., 2005; Whittington et al., 2005). This ubiquitous occurrence of the PII structure could thus reflect a biological preference for maintaining unfolded proteins in a well-defined state. In the case of the dehydrins, it has been speculated that such aggregation-resistant and extensively hydrated states are also taken advantage of in an alternative way, namely, as reservoirs or buffers for water (Soulages et al., 2003; Bokor et al., 2005). The PII structure increases the solvent-accessible area of the polypeptide chain by 50% to 60% compared with other types of secondary structures (Hicks and Hsu, 2004). In the simplest case, the water molecules are envisaged to be coordinated to the hydrogen bond acceptors and donors of the backbone in the fully disordered states. These coordinated water molecules are then released upon dehydration-induced collapse into more compact structures. A good example of such water-mediated collapse is provided by a wheat (Triticum aestivum) seed storage protein that expels its backbone water upon dehydration into a solid by becoming trapped in ordered α-helical or β-rich structures with hydrogen bonds that are predominantly intramolecular (Gilbert et al., 2000). Similarly, dehydration-induced collapse into α-helical structures have been reported for two LEA proteins from groups 3 (Goyal et al., 2005) and 7 (Wolkers et al., 2001), and yet two other LEA proteins from groups 1 and 5 have recently been observed to undergo drought-induced structural changes that increase both their β-sheet and α-helical content (Boudet et al., 2006). Based on these observations, Boudet et al. (2006) put forth the idea that the LEA proteins, besides their intrinsic modulation of coordinated water, could also have other functions that are regulated by the hydration status of the cells. More detailed studies of the hydration shells around disordered protein conformations have recently been possible through the use of solid-state NMR relaxation experiments (Bokor et al., 2005). In general, Bokor et al. (2005) observe that disordered proteins coordinate larger amounts of water per solvent-exposed surface area than folded proteins and, in particular, the dehydrin LTI29 is observed to bind more water than the other disordered proteins covered by the study (e.g. calpastatin and MAP2c). With respect to coordination strength, it is indicated that the more extensive hydration shells of the disordered proteins are also more heterogeneous than for globular proteins by having inner hydration layers that are more tightly bound and outer layers that are easily lost upon elevated temperature.

As an elaboration of the water reservoir model, the specific role of the PII helices could then be to act as an intermediate phase, holding on more strongly to the coordinated water, and thus contributing to the extension of the water-buffering range during drought-related stress. In the first step, the coil-to-PII transition would transfer water molecules from the dynamic hydration shell into specific positions in the protein backbone matrix. These tightly bound, backbone-bridging water molecules will thus preserve a reservoir of residual water under conditions where the coil states would already have given off their hydration shells. Finally, the collapse of the PII helices into more compact α-helical or β-rich structures would release and provide water in severely dehydrated environments. From this perspective, the dehydrins would emerge as a protein family that is highly optimized by taking maximal advantage of the structural properties of denatured proteins without compromising the specific and highly conserved sequence signals at the level of the primary structure.

The Lys-Rich Segment Shows Sequence Similarities with a Linker Region in the Animal Chaperone HPS90

According to AGADIR, the only one part of the dehydrins that stands out in terms of α-helical propensity is the charged stretch of amino acids that follows the S-segments in Lti29 and Cor47 (Fig. 10). From a functional perspective, it is notable that this conserved segment also has a high sequence similarity with a linker region in the animal heat shock protein HSP90. BLAST analysis shows that the dehydrin ChP-1 segment has 72% identity and is 83% positive to Hsp90 from chicken and also has high sequence identity to other animal Hsp90s. The common feature of this sequence motif is that several negative residues, like Asp and Glu, cluster before a group of positive Lys (Table II). In eukaryotic HSP90, this sequence motif forms, in turn, part of a highly conserved linker region of about 40 residues (Csermely et al., 1998). Interestingly, the corresponding homology with HSP90 from plants is weaker, hinting at the possibility of a functional overlap between certain dehydrins and heat shock proteins. Taken together, these results point to the possibility that plant dehydrins have sophisticated, multifunctional roles during conditions of environmental stress, despite their diffuse structural characteristics encountered in vitro. One may be as a water reservoir buffering over a wide range of dehydration levels, working in concert with specific anchoring points that stabilize or maintain the spatial architecture of cellular components, such as membranes, that would otherwise suffer negative effects from decreased water content in the living cell.

Figure 10.
α-Helical propensity per amino acid of Cor47, Lti29, Lti30, and Rab18 as predicted from their sequence composition by AGADIR. The maxima in α-helical propensity match overall the positions of the conserved segments.
Table II.
Part of amino acid sequences of dehydrins (top) and HSP90s of different origin


Expression and Heat Fractionation

Expression of the recombinant Arabidopsis (Arabidopsis thaliana) dehydrin proteins was as described by Svensson et al. (2000), with minor changes. Glycerol stocks of the different Escherichia coli strains were made and 200 μL were spread on LA plates with 150 μg ampicillin. The plates were kept at 37°C and grown overnight. The cells were suspended and added to 1 L of Luria-Bertani medium containing 50 μg/mL ampicillin. Expression was induced at an OD of 0.3 by adding isopropyl β-d-thiogalactopyranoside to a final concentration of 1 mm and kept at 37°C until grown to an OD of 0.8 to 1.0 (i.e. 2–4 h). Cells were harvested by centrifugation at 4,000g for 45 min and the pellet stored at −20°C. The thawed cells from 1-L cultures were resuspended in 25 mL of 20 mm Na2HPO4, pH 7.2, and 150 mm NaCl. One millimolar (final concentration) of phenylmethylsulfonyl fluoride and 0.1 mg/mL lysosyme were added and left for 30 min on ice. Lysated cells were sonicated six times for 15 s and centrifuged at 9,000g for 30 min. The supernatants were placed in a water bath at 70°C for 20 min, at which time the samples had reached a temperature of approximately 55°C. If the samples reached a higher temperature (i.e. 70°C), the recovered amount of protein or the degree of purity of the proteins (data not shown) was not changed. To precipitate heat-denatured proteins, the sample was centrifuged for 30 min at 9,000g and the supernatant stored at −80°C.


Dehydrins were purified by metal ion affinity chromatography and ion-exchange chromatography, according to Svensson et al. (2000). The supernatants from heat precipitation were diluted 1:2 with 20 mm NaHPO4, pH 7.2, 1.85 m NaCl, and 1 mm phenylmethylsulfonyl fluoride. The samples were loaded on a 5-mL HiTrap IDA-Sepharose column (Pharmacia) charged with 2 mL of 3 mg/mL CuSO4 and connected to a FPLC system (Pharmacia). The column was equilibrated with 5 volumes of 20 mm Na2HPO4, pH 7.2, and 1.0 m NaCl. Elution was performed with 2 m NH4Cl in 20 mm Na2HPO4, pH 7.2, and 1.0 m NaCl in one step. The column was then equilibrated with 10 volumes of 20 mm Na2HPO4, pH 7.2, followed by elution with 10 mm EDTA in 20 mm Na2HPO4, pH 7.2. Precipitation of protein was done with 80% (NH4)2SO4. Proteins were collected by centrifugation at 18,000g for 30 min. The different dehydrins were resuspended in 2 mL of the following buffers: Lti29 and Cor47 in 20 mm Bistris, pH 6.0, Rab18 in 20 mm Tris-HCl, pH 8.0, and Lti30 in 50 mm Gly, pH 9.0. The dehydrins were desalted in their resuspension buffers on a PD-10 column (Pharmacia). In the case of Cor47, Lti29, and Rab18, the achieved 3-mL fractions were put on an anionic exchange column (1-mL Mono Q HR 5/5; Pharmacia) or, in the case of Lti30, a cationic exchange column (1-mL Mono S HR 5/5; Pharmacia) connected to an FPLC and absorbance read at 280 nm. The columns were equilibrated with the same buffer in which the sample is desalted and elution performed by a NaCl gradient 0 to 0.5 m in the same buffer over 30 volumes. Fractions of 0.5 mL were collected during the runs for analyses. The purity was tested by SDS-PAGE. Protein quantification was accomplished by the bicinchoninic acid assay.

CD Analysis

CD measurements were carried out using a JASCO J-810 spectropolarimeter (JASCO) or, in the case of the temperature scans, a π*-180 spectrophotometer (Applied Photophysics). Scan rate was 20 nm/min at 0.2-nm resolution and 20-mdeg sensitivity.

All samples were mixed 1 h prior to the CD run and centrifuged at 12,000g for 2 min before filling the cuvettes. Protein concentration was 1 mg/mL (0.2-mm cuvette) and, in the case of temperature scans, 0.1 mg/mL (1 mm cuvette) if not stated otherwise. All runs were performed at 4°C if not stated otherwise. All CD spectra are presented as mean ellipticity per residue.


We thank Dr. Jan Svensson for giving us the four Arabidopsis dehydrin clones used in this study. We also thank Dr. Gerard Gröbner for the introduction to CDPro, and Dr. Mikael Lindberg for the kind gift of SOD protein.


1This work was supported by the Swedish Research Council for Environment, Agricultural Sciences, and Spatial Planning (P.H.) and Carl Tryggers Stiftelse (P.H.).

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Pia Harryson (es.us.bbd@nosyrrah.aip).

Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.106.079848.


  • Alsheikh MK, Heyen BJ, Randall SK (2003) Ion binding properties of the dehydrin ERD14 are dependent upon phosphorylation. J Biol Chem 278: 40882–40889 [PubMed]
  • Bochicchio B, Tamburro AM (2002) Polyproline II structure in proteins: identification by chiroptical spectroscopies, stability, and functions. Chirality 14: 782–792 [PubMed]
  • Bokor M, Csizmok V, Kovacs D, Banki P, Friedrich P, Tompa P, Tompa K (2005) NMR relaxation studies on the hydrate layer of intrinsically unstructured proteins. Biophys J 88: 2030–2037 [PMC free article] [PubMed]
  • Boudet J, Buitink J, Hoekstra FA, Rogniaux H, Larre C, Satour P, Leprince O (2006) Comparative analysis of the heat-stable proteome of radicles of Medicago truncatula seeds during germination identifies late embryogenesis abundant proteins associated with desiccation tolerance. Plant Physiol 140: 1418–1436 [PMC free article] [PubMed]
  • Bourhis JM, Johansson K, Receveur-Brechot V, Oldfield CJ, Dunker KA, Canard B, Longhi S (2004) The C-terminal domain of measles virus nucleoprotein belongs to the class of intrinsically disordered proteins that fold upon binding to their physiological partner. Virus Res 99: 157–167 [PubMed]
  • Chiti F, Taddei N, Bucciantini M, White P, Ramponi G, Dobson CM (2000) Mutational analysis of the propensity for amyloid formation by a globular protein. EMBO J 19: 1441–1449 [PMC free article] [PubMed]
  • Close TJ (1996) Dehydrins: emergence of a biochemical role of a family of plant dehydration proteins. Physiol Plant 97: 795–803
  • Close TJ (1997) Dehydrins: a commonality in the response of plants to dehydration and low temperature. Physiol Plant 100: 291–296
  • Csermely P, Schnaider T, Soti C, Prohaszka Z, Nardai G (1998) The 90-kDa molecular chaperone family: structure, function, and clinical applications. A comprehensive review. Pharmacol Ther 79: 129–168 [PubMed]
  • Danyluk J, Perron A, Houde M, Limin A, Fowler B, Benhamou N, Sarhan F (1998) Accumulation of an acidic dehydrin in the vicinity of the plasma membrane during cold acclimation of wheat. Plant Cell 10: 623–638 [PMC free article] [PubMed]
  • Davidson SW, Jonas A, Clayton DF, George JM (1998) Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J Biol Chem 273: 9443–9449 [PubMed]
  • Demarest SJ, Martinez-Yamout M, Chung J, Chen H, Xu W, Dyson HJ, Evans RM, Wright PE (2002) Mutual synergistic folding in recruitment of CBP/p300 by p160 nuclear receptor coactivators. Nature 415: 549–553 [PubMed]
  • Dunker AK, Brown CJ, Lawson JD, Iakoucheva LM, Obradovic Z (2002) Intrinsic disorder and protein function. Biochemistry 41: 6573–6582 [PubMed]
  • Dunker AK, Obradovic Z, Romero P, Garner EC, Brown CJ (2000) Intrinsic protein disorder in complete genomes. Genome Inform Ser Workshop Genome Inform 11: 161–171 [PubMed]
  • Dure L III (1993) A repeating 11-mer amino acid motif and plant desiccation. Plant J 3: 363–369 [PubMed]
  • Eom J, Baker W, Kintanar A, Wurtele A (1996) The embryo-specific EMB-1 protein of Daucus carota is flexible and unstructured in solution. Plant Sci 115: 17–24
  • Ferris PJ, Woessner JP, Waffenschmidt S, Kilz S, Drees J, Goodenough UW (2001) Glycosylated polyproline II rods with kinks as a structural motif in plant hydroxyproline-rich glycoproteins. Biochemistry 40: 2978–2987 [PubMed]
  • Fersht AR (1999) Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Protein Folding. W.H. Freeman and Company, New York
  • Gilbert SM, Wellner N, Belton PS, Greenfield JA, Siligardi G, Shewry PR, Tatham AS (2000) Expression and characterisation of a highly repetitive peptide derived from a wheat seed storage protein. Biochim Biophys Acta 1479: 135–146 [PubMed]
  • Goyal K, Pinelli C, Maslen SL, Rastogi RK, Stephens E, Tunnacliffe A (2005) Dehydration-regulated processing of late embryogenesis abundant protein in a desiccation-tolerant nematode. FEBS Lett 579: 4093–4098 [PubMed]
  • Gunasekaran K, Tsai CJ, Kumar S, Zanuy D, Nussinov R (2003) Extended disordered proteins: targeting function with less scaffold. Trends Biochem Sci 28: 81–85 [PubMed]
  • Hara M, Terashima S, Fukaya T, Kuboi T (2001) Characterization and cryoprotective activity of cold-responsive dehydrin from Citrus unshiu. Plant Physiol 158: 1333–1339
  • Hedberg L, Oliveberg M (2004) Scattered Hammond plots reveal second level of site-specific information in protein folding: phi′ (beta++). Proc Natl Acad Sci USA 101: 7606–7611 [PMC free article] [PubMed]
  • Heyen BJ, Alsheikh MK, Smith EA, Torvik CF, Seals DF, Randall SK (2002) The calcium-binding activity of a vacuole-associated, dehydrin-like protein is regulated by phosphorylation. Plant Physiol 130: 675–687 [PMC free article] [PubMed]
  • Hicks JM, Hsu VL (2004) The extended left-handed helix: a simple nucleic acid-binding motif. Proteins 55: 330–338 [PubMed]
  • Johansson J, Gudmundsson GH, Rottenberg ME, Berndt KD, Agerberth B (1998) Conformation-dependent antibacterial activity of the naturally occurring human peptide LL-37. J Biol Chem 273: 3718–3724 [PubMed]
  • Kanai F, Liu H, Field SJ, Akbary H, Matsuo T, Brown GE, Cantley LC, Yaffe MB (2001) The PX domains of p47phox and p40phox bind to lipid products of PI(3)K. Nat Cell Biol 3: 675–678 [PubMed]
  • Karlson DT, Fujino T, Kimura S, Baba K, Itoh T, Ashworth EN (2003) Novel plasmodesmata association of dehydrin-like proteins in cold-acclimated Red-osier dogwood (Cornus sericea). Tree Physiol 23: 759–767 [PubMed]
  • Koag MC, Fenton RD, Wilkens S, Close TJ (2003) The binding of maize DHN1 to lipid vesicles: gain of structure and lipid specificity. Plant Physiol 131: 309–316 [PMC free article] [PubMed]
  • Lacroix E, Viguera AR, Serrano L (1998) Elucidating the folding problem of alpha-helices: local motifs, long-range electrostatics, ionic-strength dependence and prediction of NMR parameters. J Mol Biol 284: 173–191 [PubMed]
  • Li SS (2005) Specificity and versatility of SH3 and other proline-recognition domains: structural basis and implications for cellular signal transduction. Biochem J 390: 641–653 [PMC free article] [PubMed]
  • Li X, Romero P, Rani M, Dunker AK, Obradovic Z (1999) Predicting protein disorder for N-, C-, and internal regions. Genome Inform Ser Workshop Genome Inform 10: 30–40 [PubMed]
  • Linding R, Jensen LJ, Diella F, Bork P, Gibson TJ, Russell RB (2003) Protein disorder prediction: implications for structural proteomics. Structure 11: 1453–1459 [PubMed]
  • Linding R, Schymkowitz J, Rousseau F, Diella F, Serrano L (2004) A comparative study of the relationship between protein structure and beta-aggregation in globular and intrinsically disordered proteins. J Mol Biol 342: 345–353 [PubMed]
  • Lisse T, Bartels D, Kalbitzer HR, Jaenicke R (1996) The recombinant dehydrin-like desiccation stress protein from the resurrection plant Craterostigma plantagineum displays no defined three-dimensional structure in its native state. Biol Chem 377: 555–561 [PubMed]
  • Love JJ, Li X, Chung J, Dyson HJ, Wright PE (2004) The LEF-1 high-mobility group domain undergoes a disorder-to-order transition upon formation of a complex with cognate DNA. Biochemistry 43: 8725–8734 [PubMed]
  • Luo P, Baldwin RL (1997) Mechanism of helix induction by trifluoroethanol: a framework for extrapolating the helix-forming properties of peptides from trifluoroethanol/water mixtures back to water. Biochemistry 36: 8413–8421 [PubMed]
  • Ma K, Kan L, Wang K (2001) Polyproline II helix is a key structural motif of the elastic PEVK segment of titin. Biochemistry 40: 3427–3438 [PubMed]
  • Namba K (2001) Roles of partly unfolded conformations in macromolecular self-assembly. Genes Cells 6: 1–12 [PubMed]
  • Nylander M, Svensson J, Palva ET, Welin BV (2001) Stress-induced accumulation and tissue-specific localization of dehydrins in Arabidopsis thaliana. Plant Mol Biol 45: 263–279 [PubMed]
  • Otzen DE, Oliveberg M (1999) Salt-induced detour through compact regions of the protein folding landscape. Proc Natl Acad Sci USA 96: 11746–11751 [PMC free article] [PubMed]
  • Park SH, Shalongo W, Stellwagen E (1997) The role of PII conformations in the calculation of peptide fractional helix content. Protein Sci 6: 1694–1700 [PMC free article] [PubMed]
  • Promponas VJ, Enright AJ, Tsoka S, Kreil DP, Leroy C, Hamodrakas S, Sander C, Ouzounis CA (2000) CAST: an iterative algorithm for the complexity analysis of sequence tracts: complexity analysis of sequence tracts. Bioinformatics 16: 915–922 [PubMed]
  • Rinne PL, Kaikuranta PL, van der Plas LH, van der Schoot C (1999) Dehydrins in cold-acclimated apices of birch (Betula pubescens ehrh.): production, localization and potential role in rescuing enzyme function during dehydration. Planta 209: 377–388 [PubMed]
  • Rucker AL, Creamer TP (2002) Polyproline II helical structure in protein unfolded states: lysine peptides revisited. Protein Sci 11: 980–985 [PMC free article] [PubMed]
  • Shi Z, Chen K, Liu Z, Sosnick TR, Kallenbach NR (2006) PII structure in the model peptides for unfolded proteins: studies on ubiquitin fragments and several alanine-rich peptides containing QQQ, SSS, FFF, and VVV. Proteins 63: 312–321 [PubMed]
  • Silow M, Oliveberg M (2003) High concentrations of viscogens decrease the protein folding rate constant by prematurely collapsing the coil. J Mol Biol 326: 263–271 [PubMed]
  • Soulages JL, Kim K, Arrese EL, Walters C, Cushman JC (2003) Conformation of a group 2 late embryogenesis abundant protein from soybean: evidence of poly (l-proline)-type II structure. Plant Physiol 131: 963–975 [PMC free article] [PubMed]
  • Soulages JL, Kim K, Walters C, Cushman JC (2002) Temperature-induced extended helix/random coil transitions in a group 1 late embryogenesis-abundant protein from soybean. Plant Physiol 128: 822–832 [PMC free article] [PubMed]
  • Sreerama N, Woody RW (2004) Computation and analysis of protein circular dichroism spectra. Methods Enzymol 383: 318–351 [PubMed]
  • Stapley BJ, Creamer TP (1999) A survey of left-handed polyproline II helices. Protein Sci 8: 587–595 [PMC free article] [PubMed]
  • Svensson J, Palva ET, Welin B (2000) Purification of recombinant Arabidopsis thaliana dehydrins by metal ion affinity chromatography. Protein Expr Purif 20: 169–178 [PubMed]
  • Tan L, Qiu F, Lamport DT, Kieliszewski MJ (2004) Structure of a hydroxyproline (Hyp)-arabinogalactan polysaccharide from repetitive Ala-Hyp expressed in transgenic Nicotiana tabacum. J Biol Chem 279: 13156–13165 [PubMed]
  • Tompa P, Csermely P (2004) The role of structural disorder in the function of RNA and protein chaperones. FASEB J 18: 1169–1175 [PubMed]
  • Uversky VN (2002) Natively unfolded proteins: a point where biology waits for physics. Protein Sci 11: 739–756 [PMC free article] [PubMed]
  • Uversky VN, Gillespie JR, Fink AL (2000) Why are “natively unfolded” proteins unstructured under physiologic conditions? Proteins 41: 415–427 [PubMed]
  • Uversky VN, Oldfield CJ, Dunker AK (2005) Showing your ID: intrinsic disorder as an ID for recognition, regulation and cell signaling. J Mol Recognit 18: 343–384 [PubMed]
  • Whittington SJ, Chellgren BW, Hermann VM, Creamer TP (2005) Urea promotes polyproline II helix formation: implications for protein denatured states. Biochemistry 44: 6269–6275 [PubMed]
  • Wisniewski M, Webb R, Balsamo R, Close TJ, Yu XM, Griffith M (1999) Purification, immunolocalization, cryoprotective, and antifreeze activity of PCA60: a dehydrin from peach (Prunus persica). Physiol Plant 105: 600–608
  • Wolkers WF, McCready S, Brandt WF, Lindsey GG, Hoekstra FA (2001) Isolation and characterization of a D-7 LEA protein from pollen that stabilizes glasses in vitro. Biochim Biophys Acta 1544: 196–206 [PubMed]

Articles from Plant Physiology are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...