Abiotic stress threatens global plant functioning and results in secondary effects such as oxidative damage. Stress action is often related to suboptimal climatic changes that can hamper crop growth and production. Among the environmental stimuli that seriously affect plant productivity are drought, salinity, frost, cold, and heat (Sewelam et al., 2020; Bouabdelli et al., 2022). Abiotic stress leads to significant changes in various cellular compartments and at various physiological and molecular levels (Arora and Rowland, 2011; Rurek et al., 2018; Araújo et al., 2019; Yadav et al., 2023). Plant organelles, including mitochondria and chloroplasts, are particularly sensitive to stress, and alterations in organelle biogenesis have been observed under adverse conditions (Taylor et al., 2009). Within the cell response to stress, ROS produced by organelles, which are particularly dangerous to cellular integrity, can damage DNA and affect enzyme activity, thus modulating the metabolism (Sun and Lin, 2010). To properly maintain cellular homeostasis and organellar biogenesis, highly specific cellular and molecular responses and the coordination of diverse compartment functions in stress and recovery are indispensable (Welchen et al., 2014).

Dehydrins (DHNs), a class II protein of the late embryogenesis abundant protein family (LEA), are among the most important cellular players in the abiotic stress response. The disordered structure of DHNs allows them to be included in intrinsically disordered proteins (IDPs). Generally, in unbound form, DHNs lack defined secondary structures; only after binding to the ligand (e.g., the membrane lipid or metal ion) does their secondary structure helix appear, allowing those proteins to perform protective functions within the cell. In the last ten years, the flexibility and activities of DHNs were subjects of valuable literature reviews; however, some of these reviews focused mainly on diverse DHN functions (e.g., their involvement in membrane protection) and structural aspects rather than presenting more updated data on DHN involvement in stress response (Graether and Boddington, 2014; Liu et al., 2017; Murray and Graether, 2022).

This review provides an updated view of the segmental structure, tissue- and developmentally- specific accumulation, functions, and participation of various cellular DHNs in response to selected abiotic stress conditions. The discussed aspects are presented in the context of plant physiological and molecular alterations under stress acclimation, with an emphasis on interactions with organelles. Additionally, current opinions on the identity, subcellular localization of immunologically related dehydrin-like proteins (dlps), and their participation in the abiotic stress response are also presented. Finally, the lesser-studied points on DHNs and dlps and their future valuable applications for plant biology studies are discussed.

In the following chapter, we provide a general characterization of a class II protein from the LEA family. DHNs were initially identified in cotton (Gossypium) generative cells during embryogenesis and then in various plant species, including rice (Oryza sativa), barley (Hordeum vulgare), and maize (Zea mays) (Galau et al., 1986; Mundy and Chua, 1988; Close et al., 1989; Close, 1997). Later, they were found not only among other seed plants but also in liverworts (Marchantiophyta; Hellwege et al., 1994; Ghosh et al., 2016; Melgar and Zelada, 2021), mosses (Bryophyta; Saavedra et al., 2006; Ruibal et al., 2012; Agarwal et al., 2017), ferns (e.g., Polypodium; Layton et al., 2010), and even among Cyanobacteria (Close and Lammers, 1993). Fragmented and contradictory data exist on the presence of DHNs in algal taxa. The study by Li et al. (1998) suggests the presence of a protein immunologically related to DHNs of a broad size (17-105 kDa) in Fucus. These proteins appeared to be species- and developmentally specific and substantially thermolabile. Gasulla et al. (2009) immunodetected constitutively accumulated five proteins related to DHNs (ranging from 15 to 120 kDa in size) in the lichen chlorophytic photobiont Trebouxia erici. Becker et al. (2020) mention (without reference) a DHN in the green algae Dunaliella salina, which may be seen either as a result of a late evolutionary event or as an ancient feature lost during algae evolution. Unfortunately, none of these proteins were further characterized. On the contrary, Malik et al. (2017) could not find any algal DHNs despite an exhaustive genomic search. The striking feature of DHNs is their conformational dynamism, which is due to their primary structure and, particularly, the presence of numerous polar amino acid residues (Close, 1996). The polypeptide chain of DHNs does not contain tryptophan and cysteine, while more than 25% of their primary sequence is occupied by alanine or glycine (Close, 1997; Malik et al., 2017).

DHNs contain characteristic sequence motifs; their presence affects the functions and subcellular localization of DHNs. These proteins can be identified by their relatively high hydrophobicity and the presence of at least a single lysine-rich K-segment of 15 residues, which is typically C-terminal and contains up to 11 copies per protein (Close, 1997). Although the consensus sequence of the K-segment [XKXGXX(D/E)KIK(D/E)KXPG] was initially thought to be entirely conserved, recent studies have showed that the most conserved amino acids are in the motif core [(D/E)KIK(D/E)K], while the regions flanking the core are more variable (Malik et al., 2017). The amphipathic properties of the K-segment, with hydrophilic and hydrophobic residues, enable it to form secondary α-helical structures only in the presence of a ligand; otherwise, the protein is fully disordered (Close, 1996). Ligand interaction plays a vital role in providing the dynamic structure of DHNs (Koag et al., 2009). The properties of the K-segment are believed to be essential for DHN interactions with the other proteins (Liu et al., 2017; Hernández-Sánchez et al., 2019) and phospholipid membranes (Koag et al., 2009; Clarke et al., 2015) and to prevent protein aggregation (Drira et al., 2015; Szabała, 2023).

DHNs also include the Y- and S-segments. The N-terminal Y-segment [(V/T)D(E/Q)YGNP] often contains up to three tandem repeats. The C-terminal part of this motif, which is enriched with asparagine and glycine, is highly conserved, while aspartic acid at the N-terminus and proline at the C-terminus are also evolutionarily conserved (Close, 1997; Malik et al., 2017). The S-segment [LHR(S/T)GS_4–6(S/D/E)(D/E)₃], which contains serine repeats, mostly precedes the K-segment (Close, 1997; Malik et al., 2017). The S-segment is often phosphorylated, which influences the interaction of DHNs with Ca²⁺ and the localization of DHNs within the cell nucleus; it only partially regulates the nuclear localization (though not the dimerization) of selected DHNs, for example, wheat (Triticum aestivum) WZY1-2 protein (Alsheikh et al., 2003; Wang et al., 2020).

One of the less conserved DHN segments is the φ-segment. It contains polar and non-polar residues and is particularly enriched with glycine, glutamic acid, and threonine (Malik et al., 2017). The φ motif may be present in all DHN regions that are not within the conserved segment. Due to its relaxed secondary structure, it allows polar amino acids to interact with water molecules (Lv et al., 2018). Hughes and Graether (2011) have suggested that it is flexible and important for the maintenance of the DHN disordered structure, which would prevent interactions of partially denatured proteins.

The recently identified 18-residue-long F-segment, like the K-segment, contains hydrophobic residues within the core [EXXDRGXFDFX(G/K)] (Richard Strimbeck, 2017; Riley et al., 2019). It is present in different DHN subtypes, known from many genera, including Rhododendron and Camelina (Wei et al., 2019). For example, the F-segment was observed among numerous SK_n DHNs in ten proteins from the K_n subclass and in the SK_nF subclass. The presence of the F-segment in the DHNs of gymnosperms and angiosperms suggests that it may be evolutionarily older than the Y-segment, which is found exclusively in angiosperms (Riley et al., 2019). The presence of the F-segment (and the K-, Y-, and S-segments) has also been confirmed among halophyte DHNs. Interestingly, the role of DHNs in salinity adaptation is poorly studied (Ghanmi et al., 2022). It has been suggested that the F-segment may play a role in the membrane and in protein binding (Richard Strimbeck, 2017); some recent findings by Ohkubo et al. (2020) have provided additional information on its cryoprotective properties.

Riley et al. (2019) mentioned another DHN motif consisting of a stretch of consecutive lysine residues separated by arginine, aspartic acid, and glutamic acid; it occurs frequently between the S and K motifs. This motif was observed only among K_nS and SK_n DHNs and in a single DHN from the K_n subclass. Other potential motifs are still described. Another newly identified segment was recently discovered only in the moss Syntrichia ruralis. It is closely related to the Y-segment and similarly to the F- and K-segments. It contains hydrophobic residues, which enable the formation of amphipathic α-helices. Their properties, which allow the self-association of DHNs, are indispensable for stress tolerance (Upadhyaya et al., 2021). Moreover, Yokoyama et al. (2020) noted ‘non-K segment 1’ in Arabidopsis HIRD11 DHN. The analysis of various plant genomes containing genes for K_nS DHNs revealed the presence of an N-terminal motif consisting of 15 residues (the H-segment). This motif was identified in angiosperms, gymnosperms, and lycophytes, indicating the ancestral origin of this subclass, potentially tracing back to the early stages of land plant evolution (Melgar and Zelada, 2021). In addition to the mentioned motifs, DHNs contain other repetitive fragments that affect their properties, such as histidine-rich sequences that affect the ability of LEA-2 proteins to bind ligands such as metals (Hara et al., 2005; Hara et al., 2016) and lipids (Eriksson et al., 2011) and to interact with the other DHNs (Hernández-Sánchez et al., 2014; Hernández-Sánchez et al., 2017). As will be indicated below, histidine-rich sequences also increase membrane affinity (Eriksson et al., 2011; Gupta et al., 2019).

Overall, based on the arrangement of domains and their number, the subtypes K_n, SK_n, K_nS, Y_nK_n, Y_nSK_n, F_nK_n, and F_nSK_n DHN (n denotes the number of copies of a given segment; Richard Strimbeck, 2017; Riley et al., 2019; Upadhyaya et al., 2021; Figure 1 ) can be assigned to three orthologous DHN clusters containing H-, F- or Y-segments (Melgar and Zelada, 2020).

The chemical properties of amino acid residues affect the structure of proteins. Hydrophobic amino acids present in the core of the chain allow globular proteins to fold properly. On the contrary, DHNs contain an excess of polar residues that impact their hydrophobicity (Close, 1997), resulting in the disorganization of the polypeptide chain (Tompa, 2002; Tompa et al., 2005).

Intrinsically disordered proteins (IDPs), including DHNs, do not have a fixed three-dimensional structure. Because of their lack of stable structure, IDPs retain their inherent integrity even in unfavorable conditions, which makes them incapable of denaturation. In general, they can dynamically change their structure. For this reason, it is not the rigid, stable, and fixed protein structure that determines the functions of IDPs. Rather, their role is mainly based on subcellular localization and sequence motifs (Tompa et al., 2005; Tompa et al., 2006). The structural plasticity of IDPs allows them to perform various cellular functions. For example, the binding of different ligands within a region or changing the DHN conformation depends on current cellular conditions and thus alters the functions performed, which is referred to as ‘moonlighting’ (Tompa et al., 2005). As a result of the association with phospholipid-containing vesicles, DHNs can acquire a helical structure (Koag et al., 2003) in which hydrophobic residues are located on one side and hydrophilic residues are located on the other side of the polypeptide chain (Eriksson et al., 2016). However, under stressful conditions, DHNs may still be in a disordered state, which is only slightly affected by solvent alterations. DHNs maintain such a disordered structure in stress, contrary to other proteins in which the structural collapse of the unfolding state has often been observed (Mouillon et al., 2008).

DHNs are most often localized in young, still-growing, and differentiating tissues such as root cells, young stems or petioles, root meristems, phloem, pollen sacs, and germ cells (Nylander et al., 2001; Rorat, 2006). Changes in DHN accumulation under stress are not exclusive to generative tissues but also occur in vegetative organs (Close, 1997). DHN accumulation is usually induced by various adverse conditions, such as cold, drought, or frost, which affect gene expression and post-translational modifications of DHNs (Sun et al., 2021b). The level of some DHNs can be elevated with stress. Examples include LTI29 and ERD14 in Arabidopsis, which are present in the root apex, root vascular tissues, and shoots, or the RAB18 protein, localized in the protective gurd cells of the stomata apparatus (Nylander et al., 2001).

However, multiple genes and proteins play a role in the vegetative/generative phase transition (Yadav et al., 2023), and it is also expected that DHNs will participate in this tightly regulated and complex process. For example, Nazemof et al. (2016) identified EDR14 DHN (often annotated as a pollen coat protein) in the Brassica napus stigma proteome (ca. 2700 proteins detected). Furthermore, investigation of the proteome of B. rapa floral buds at meiosis, which allowed a better understanding of the relevance of meiosis and polyploidization events in vegetative/generative transition, revealed the downregulation of the ERD14 protein in the autotetraploid vs. diploid line (Yang et al., 2019a).

DHNs belong to numerous desiccation-tolerance proteins identified in plant seeds; furthermore, other LEA, seed maturation, and embryonic maturation proteins participate in seed maturation and imbibition. However, the expression of DHNs in mature seeds and tissues (e.g., leaves) under drought relies on common molecular mechanisms induced during water shortage; moreover, differences in DHN abundance among diverse seed types should be assigned to the maturation drying (Radwan et al., 2014). The DHN level may differ in B. napus lines with a high or low oil content (Gan et al., 2013; Gu et al., 2016). For example, Rahman et al. (2021) identified Rab18-like and Xero1-like DHNs in the B. rapa seed proteome, and Gan et al. (2013) showed a markedly increased level of Rab18 in the rapeseed low-oil-containing genotype; however, few LEA proteins appeared to be more expressed in the line with high oil content. Rihan et al. (2017) analyzed, inter alia, the temporal profiles of DHNs in cauliflower (B. oleracea var. botrytis) developing seeds. Proteins of 12, 17, and 26 kDa representing small and medium-sized DHNs increased in abundance 60 days after pollination.

Legume DHNs enriched with the Y-segment accumulate primarily in seeds (Mota et al., 2019). The DHN present in pea (Pisum sativum), known as DHN-COG, accumulates during plant embryogenesis in the developing cotyledons of seedlings under water-deficient conditions. As the embryo matures, DHN-COG accounts for approximately 2% of all seed proteins (Robertson and Chandler, 1994). The Arabidopsis DHN protein LTI130 also accumulates under stress conditions and is expressed in vegetative tissues only under cold conditions (Nylander et al., 2001). In turn, Arabidopsis ERD14 and ERD10 proteins under control conditions of plant growth and development are found mainly in stem tissues, leaves, flowers, and root cells, but under cold stress they are detectable in all studied plant tissues (Abdul Aziz et al., 2021).

Lipid membranes allow the functional compartmentalization of various metabolic processes within the membrane continuum: cell membrane—inner membranes and vesicles—nuclear envelope. Membranes are particularly sensitive to temperature fluctuations and dehydration, which cause changes in the lipid structure and composition. In addition, low temperature causes membrane lipid peroxidation, an elevated formation of reactive oxygen species (ROS), and a reversible transition from a lamellar to a hexagonal membrane, which affects membrane functions (Thomashow, 1999). In contrast, as the temperature increases, the content of unsaturated fatty acids within the lipid membranes decreases (Falcone et al., 2004), resulting in reduced fluidity and increased membrane rigidity.

DHNs are not only cytoplasmic proteins; they also interact with various cellular membranes, including plasma, nuclear, and tonoplast membranes (Houde et al., 1995; Danyluk et al., 1998; Heyen et al., 2002; Zhang et al., 2020), and the physicochemical effects of such interactions have been subject to various studies. Although plant DHNs are localized primarily in the cytoplasm and nucleus, they can also be found near the cell membrane, mitochondria, and plastids due to their involvement in protecting these organelles (Candat et al., 2014; Kalemba et al., 2015). The effects of the binding of DHN to membranes on the secondary structure of DHNs were investigated using dodecylate-containing micelles; the circular dichroism (CD) revealed changes in the structure of the studied DHNs and the presence of α-helical structures (Ismail et al., 1999).

An essential factor in the interaction between DHNs and membranes is the positively charged K-segment, which facilitates the interaction of DHNs with negatively charged membranes (Koag et al., 2009). However, it should be underlined that not only the K- or Y-segment properties are crucial for the DHN-membrane binding. The cellular functioning of DHNs also depends on the composition of less conserved regions rich in polar and charged residues. In addition, post-translational modifications also affect DHN biological activity; all these factors can differentiate various DHNs even from the same species (Vazquez-Hernandez et al., 2021). LTI30 is a DHN that contains six K-segments, five of which are flanked by histidine-rich sequences. When the helical conformation occurs within the K-segment, the structure of LTI30 changes, allowing the histidine-rich regions to interact with the lipid ‘heads’ (Eriksson et al., 2016; Gupta et al., 2019; Andersson et al., 2020). Lipid phase transitions significantly affect membrane function. The LTI30 protein has been shown to reduce the temperature of the liposome phase transitions, allowing membrane fluidity to be maintained at reduced temperatures (Eriksson et al., 2011). Studies of the binding of Arabidopsis ERD10 and ERD14 to liposomes have shown that they mainly interact with lipid ‘heads’ of the membrane surface and do not directly affect membrane fluidity (Kovacs et al., 2008). However, some domains engaged in partner binding and other regions of unstructured ERD14 are indispensable, to a diverse extent, for its functioning (Murvai et al., 2021).

The way in which DHNs interact with lipid membranes (also organellar ones) is primarily determined by the membrane content. Phosphatidylcholine (PC) is the most prevalent lipid in plant membranes, while phosphatidic acid (PA) is less common, accounting for only 1-2% of membrane compounds (Graether and Boddington, 2014). DHNs exhibit the strongest affinity for PA and the weakest affinity for PC (Koag et al., 2003), which becomes significant under stress conditions where changes occur in the lipid content of the membrane. During stress, such as water deficiency, the PA content of the membrane increases, enabling DHN to bind and protect membranes from damage (Munnik, 2001; Koag et al., 2003). The diversity of lipids that build biological membranes is relevant for the specific organelles with which DHNs interact. For example, an experiment analyzing the conformation of the TsDHN-1 dehydrin of Thellungiella salsuginea suggested that it can acquire the β-sheet structure when it interacts with the membrane lipids. The study used membranes that mimic the properties of organellar membranes, which are mainly composed of monogalactosyl- (MGDG) and digalactosyl-diacylglycerols (DGDG), two neutral galactolipids that differ from other cellular membranes (Rahman et al., 2011; Rahman et al., 2013).

To ensure proper organellar biogenesis, it is vital to maintain membrane stability and permeability. Nagaraju et al. (2019) analyzed in silico sorghum (Sorghum bicolor) DHNs and predicted plastid targeting for nearly 60% of these proteins. Using GFP fusions, Li et al. (2017) experimentally showed that the PpDHNC protein from Physcomitrella patens is targeted to plastids and suggested that it may help stabilize thylakoid and other plastid membranes while maintaining electron flow between photosystems. COR15a and COR15b are also among the DHNs that interact with plastid membranes in cold and will be discussed in the following paragraphs (Thalhammer et al., 2014).

DHN and dehydrin-like proteins (dlps; recognized with antisera against the K-segment but more thermolabile than DHNs; see Chapter 7 for more details) have also been found to interact with mitochondrial membranes. An example is the Citrus unshiu CuCOR15 protein, whose subcellular localization was examined by the centrifugation fractionation; it likely binds to the mitochondrial membrane and inhibits lipid peroxidation (Hara et al., 2003). This is particularly important because mitochondrial membranes are sensitive to ROS. When studying the impact of drought on the biogenesis of cauliflower (B. oleracea var. botrytis) mitochondria, an increase in the expression level of the ERD14 gene that encodes DHN interacting with membranes of purified mitochondria was observed under water deprivation (Rurek et al., 2018). Borovskii et al. (2005) investigated whether proteins detected by K-segment-specific antisera bind to the outer mitochondrial membrane and if they can be imported into the mitochondrial matrix or bind to the inner membrane. Their findings revealed that certain thermostable DHNs interact with the outer mitochondrial membrane and do not penetrate the organelle. Meanwhile, Romanenko et al. (2010) observed DHNs in wheat seedlings near both the inner and outer mitochondrial membranes. Szabala et al. (2014) used immunogold labeling to indicate the precise binding site of DHN24, and their study of the subcellular targeting of DHNs of the SK₃ subclass demonstrated that this DHN binds solely to the outer mitochondrial membrane on the cytosolic side but does not undergo translocation into the matrix.

Many stress stimuli, for example, temperature treatments, also result in developmental delays, reduced membrane fluidity, and even cell death, as well as in the protein biosynthesis, production of solutes, and altered plant morphological, physiological, biochemical, and molecular responses. In numerous plant species, proteomic adjustments to abiotic stress also cover defense and protective proteins (Yadav et al., 2023); the level of these proteins affects the tolerance of the plant to a given stress stimuli, thus increasing stress acclimation (Arumingtyas et al., 2013; Graether and Boddington, 2014). Together with antioxidant proteins, metabolism enzymes, and heat shock proteins (HSPs), DHNs belong to proteins whose abundance is significantly modulated under stress (Schlesinger, 1990; Tamburino et al., 2017). DHNs have been noted in various cellular compartments, including plastids and mitochondria, under a multitude of adverse conditions (NDong et al., 2002; Mueller et al., 2003; Grelet et al., 2005; Hundertmark and Hincha, 2008). Some DHNs (e.g., DHN1 from Arachis duranensis) are involved in contradictory responses from various stress stimuli and can increase plant tolerance only to the particular treatment (Mota et al., 2019). The data below will summarize recent advancements in elucidating the participation of various cellular DHNs (including organellar proteins) in responses to temperature and drought stress. Table 1 also presents a summary of selected organellar DHNs (and the immunologically related dlps) whose accumulation increased or was induced in response to various abiotic stress conditions.

The disordered structure of DHNs is thought to allow them to have many cellular functions, such as the protection of the intracellular membrane system, including organelles. DHN functions depend on their subcellular localization, protein structure, and type of stress stimuli. In subarctic-grown, cold-acclimated Thellungiella salsuginea species, two DHNs stabilize their own secondary structure by interaction with Zn²⁺ both freely and interacting with membranes; however, this is not a rule under cold stress for other DHNs. Generally, DHNs specifically interacting with some macromolecules will induce secondary structures. However, at low temperature and in the presence of Zn²⁺, secondary structure redistribution occurs for both studied DHNs rather than induction (Rahman et al., 2011). As mentioned in Chapter 4, DHN proteins lower the temperature of the membrane phase transition and thus protect cell organelles under various stress conditions ( Figure 2 ) (Eriksson et al., 2011; Clarke et al., 2015; Andersson et al., 2020).

Two particular DHNs—COR15A and COR15B DHNs—protect plastids when under cold stress from electrolyte leakage by assuming specific secondary structures, allowing them to bind to the outer membrane (Thalhammer et al., 2014); alone, they are highly unstructured. Figure 3 presents the mechanism of interactions of COR15 DHN with plastid membranes under freeze stress proposed by Thalhammer and Hincha (2014). The Arabidopsis COR15a and COR15b proteins contain an N-terminal chlorophyll a-b-binding domain, similar to the PSII light harvesting chlorophyll-binding protein, and are largely unstructured proteins in the absence of a bound membrane ( Figures 3A, B ). Plastid membranes contain monogalactosyl-diacylglycerol glycolipids (MGDG), which are unstable under stress conditions. The secondary structure of the two DHNs mentioned above enables them to bind to MGDG and provide membrane stabilization (details on the membrane binding by COR15 are shown in Figure 3C ). Binding of DHNs to organelle membranes has also been confirmed for the TsDHN-1 and TsDHN-2 proteins of Thellungiella salsuginea. TsDHN-1 binds to the membrane at 4°C by hydrogen bonds, stabilizing the lipid monolayer. On the contrary, TsDHN-2 interacts with the membrane at reduced temperature, using electrostatic interactions, and participates in the maintenance of its structure (Rahman et al., 2011; Rahman et al., 2013).

Diverse stress conditions often result in oxidative burst, which develops mainly in chloroplasts and mitochondria, where electron leakage and ROS formation occur. ROS pose a threat to the intracellular membrane continuum (Juan et al., 2021). Cell protection against ROS can also be provided by DHNs, with attributed antioxidant properties ( Figure 2 ), due to the high content of amino acid residues such as glycine, histidine, and lysine, which are targets in ROS-mediated protein oxidation (Sun and Lin, 2010). CuCOR19 DHN from Citrus unshiu mitochondria treated at low temperature showed an inhibitory effect on membrane peroxidation, which was stronger than other antioxidants used (Hara et al., 2003). ROS scavenging has also been attributed to PpDHNA and PpDHNC proteins of Physcomitrella patens (Li et al., 2017). Protection of cells from adverse ROS effects can be achieved through the interaction of DHN with heavy metals. Metal atoms, because of stress, can be a source of ROS. Therefore, metal binding by DHN can prevent ROS formation and protect cellular structures. In addition, histidine-rich regions are thought to be responsible for DHN interactions with metals (Hara et al., 2005). To investigate the ability of DHN to bind metals, Hara et al. (2005) studied CuCOR15 interactions with various divalent metals and confirmed the highest affinity of the protein for Cu²⁺ and the involvement of histidine-rich motifs in metal ion binding.

Due to their high content of polar amino acids, DHNs can also bind water molecules ( Figure 2 ). This interaction prevents water loss and allows regulation of the ion concentration as it increases in drought (Tompa et al., 2006). DHNs can bind to partially dehydrated regions of other proteins and thus protect them from denaturation (Drira et al., 2013; Drira et al., 2015; Yang et al., 2019b). Water retention during drought also serves to stabilize proteins and cellular compartments (Close, 1996; Shekhawat et al., 2011; Liu et al., 2015). Furthermore, some DHNs bind Ca²⁺ ions, as noted for acidic Arabidopsis DHNs including ERD14, ERD10, and COR47 proteins. Ca²⁺ binding can often occur simultaneously with DHN phosphorylation ( Figure 2 ). Therefore, Alsheikh et al. (2005) proposed that DHN phosphorylation occurs first, thus allowing subsequent Ca²⁺ ion binding.

DHN multifunctionality also includes their interactions with the cytoskeleton ( Figure 2 ). ERD10 DHNs have been shown to bind to actin filaments and thus participate in the inhibition of actin polymerization (Abu-Abied et al., 2006). Another important role for DHNs is nucleic acid protection ( Figure 2 ; Boddington and Graether, 2019). An experiment conducted on the CuCOR15 protein showed that it bound DNA in the presence of Zn²⁺ ions (Hara et al., 2005). In the electrophoretic mobility shift assay (EMSA), Y₂K DHN of bean (Vigna radiata) bound only shorter DNA fragments (3 to 4 kb) non-specifically without ion addition; furthermore, ion binding stimulated DNA binding, and some ions (e.g., Ni²⁺) enhanced DHN binding with shorter DNA fragments (0,7-1 kb; Lin et al., 2012).

DHNs can also protect organelles from freezing and subsequent thawing by preventing membrane fusion (Clarke et al., 2015), protein aggregation, and decreased enzyme activities ( Figure 2 ; Hughes et al., 2013). Multiple studies characterizing various DHNs (including those from plant biomass, e.g., Liu et al., 2022) employed known enzymes (for instance, LDH) for in vitro tests. Oxidative damage often accompanies the stress action, and DHNs have also been shown to protect LDH against hydroxyl radicals (Halder et al., 2018). In vitro analyses of PCA60 DHN examined the responses of the protein subjected to freezing and subsequent thawing stress. Analysis of ice crystals and retained LDH activity indicated the antifreeze properties of the PCA60 protein (Wisniewski et al., 1999). Cryoprotective properties were also shown, for instance, by the interaction of COR15A with the chloroplast envelope (Steponkus et al., 1998). Bozovic et al. (2013) identified a few DHNs (ERD14, LTI29, and COR47) that bind to thylakoid membranes under freezing stress and exhibit cryoprotective effects. Szabała (2023) also showed the cryoprotective activity of purified DHN24 in the enzymatic assay and inhibition of bacterial growth under DHN24 overexpression. Interestingly, wheat DHN-5 showed weaker LDH protective effects than the heat-stable protein fraction from Opuntia ficus-indica seeds (Drira et al., 2013; Drira et al., 2023). Some vacuolar Ca²⁺-binding proteins also showed cryoprotective activity against LDH (Osuda et al., 2021).

The cryoprotection of enzymes can be shown comparably both by the K- and F-segments and depends on hydrophobic residues (Ohkubo et al., 2020); however, in the Arabidopsis HIRD11 protein, its disordered structure, the K-segments, and the non-K-segment 1 showed cryoprotective characteristics by inhibiting PC liposome aggregation even better than cryoprotective agents (Yokoyama et al., 2020; Kimura et al., 2022). Hughes et al. (2013) showed that the number of K-segments affects the hydrodynamic radius of DHNs and the LDH protection activity and that the intrinsic disorder of DHNs is necessary for their cryoprotective properties. Hydrophobic residues of the K-segment participate in the prevention of the cryoinactivation, cryoaggregation, and cryodenaturation of LDH (Hara et al., 2017).

Recently, Smith and Graether (2022) proposed a different mode of cryoprotection for Vitis riparia YSK₂ dehydrin towards the cold-labile yeast frataxin homolog-1 (Yfh1); it relies on presumably weak interactions employing positively charged residues within the YSK₂ sequence that keep the helical structure of Yfh1 at a low temperature. In general, it is different from the LDH protection model, which is primarily dependent on the hydrodynamic properties of the enzyme. The results of the Smith and Graether (2022) study thus support the idea that DHNs may show cryoprotective activities by multiple mechanisms.

DHNs, which allow plant resistance to unfavorable stimuli, can be seen as the ultimate protein effectors responsible for prompt phenotypic changes in stress (Yadav et al., 2023). Because of their disordered structure, DHNs interact with various molecules, including membrane lipids, allowing the stabilization of organelle membranes and protecting them against adverse conditions. The protective status of DHNs also includes some other functions: these proteins can protect cells from ROS damage, stabilize DNA and enzymes, and affect cell biogenesis in numerous ways. With the progressive research on the structure of these proteins, we have learnt more and more about the mechanisms of their interactions; however, a few points regarding DHN characterization should be elucidated soon.

First, the cellular functions of recently described DHN subtypes, including F_nK_m and F_nSK_m proteins, and their subcellular localization should be further investigated. Additionally, it would be revealing to find more orthologs of plant DHNs/dlps in other taxa, including fungi, animals, or even unicellular organisms. Organelle functions are important in the development of tolerance to selected stress conditions, especially to cold (Maryan et al., 2021); therefore, the participation of DHNs/dlps in organellar biogenesis should be investigated more intensively. Furthermore, data on organellar DHNs/dlps should more broadly encompass agronomically important species as this knowledge can potentially be used in the future to construct novel or genetically modified crop genotypes that are resistant to adverse environmental stimuli, especially because systemic analyses have been deepened recently for some of these species, including Brassica (Yadav et al., 2023).

Challenges with dlp categorization indicate that it is necessary to find more reliable criteria for it and to characterize dlps in more detail. In parallel, the cloning of genes/cDNAs for most dlps and detailed tissue and organ-specific expression profiles of dlps must be investigated in detail. Special attention should be paid to elucidating the expression of DHNs/dlps in floral tissues and the potential role of these proteins in the biogenesis of certain genetic traits, including pollen sterility. This goal may be achieved relatively easily as complete transcriptomic and proteomic data of numerous plant species (including model ones) are already available.

However, not much is known about the interactions of DHNs with other organellar proteins and genomes, and detailed interactomic analyses employing DHNs/dlps should thus be performed in order to reveal the protein partners interacting with DHNs/dlps. In the near future, the expression patterns of DHNs/dlps should be tested in a plethora of abiotic and biotic stresses in a more systemic scale; however, focusing those investigations on the most detrimental conditions in agriculture, including drought, may be especially conducive to increasing crop productivity as drought adaptation is a particularly complex phenomenon (Salekdeh et al., 2009). It would also be worthwhile to further study the interconnections between melatonin treatment and the DHNs/dlps level. This concept raises the possibility of the construction of transgenic crop plants with in planta overexpression of DHNs that can be grown in the presence of exogenously sprayed melatonin (Bajwa et al., 2014). The participation of dlps in non-temperature treatments has been largely less investigated; therefore, this field should be exhaustively broadened.

Finally, detailed elucidation of the less studied DHN functions in plant cells should continue. In particular, novel ideas on the participation of new organellar DHNs and dlps in stress acclimation and deacclimation should be investigated soon using more intensive omics tools, thus broadening our knowledge on the fascinating adaptability of these proteins to numerous adverse conditions. Such data are especially important in view of the changing and affecting climate conditions of agronomically important species. Therefore, the broadened characterization of DHNs and dlps represents a vibrant and important field in plant research that may be used in the future for the successful development of stress-resistant lines/genotypes (Zhang et al., 2019). For example, recent studies such as those by Ganguly et al. (2020), who show the increased tolerance of transgenic rice (O. sativa) to drought under overexpression of the Rab16A protein, and Aduse Poku et al. (2020), who demonstrate the engineering of the LEA-S protein of watermelon (Cucumis melo) in drought and salinity, are representative examples. Another excellent study provides Arabidopsis Rab18 DHN (AT5G66400) and a few other LEA proteins, which have been shown to be affected in abundance under triple stress treatments and are therefore good candidates for stress markers (Sewelam et al., 2020). However, multiple and effective attempts to use DHN in plant agriculture should also include other methodological tools, including genomics-assisted breeding, plant gene editing, and plant metabolic engineering.

ZS partially deposited and analyzed the literature data, prepared Table 1 , and partially wrote the manuscript. MR designed the general concept of the review, partially wrote the manuscript and analyzed the literature data, prepared all the figures, reviewed the data quality, and updated the whole manuscript for publication. Both authors have accepted the final version of the manuscript and have agreed to be responsible for all aspects of the work.