The extraordinary acceleration of human population growth in recent centuries is determining a fast-increasing demand for food, mainly cereals, and other agronomic crops. In recent decades, modern farming techniques and plant breeding programs enabled a linear enhancement in agricultural productivity; however, the positive trend is expected to slow down because global climate change is leading to adverse environmental conditions and to related negative effects on crop quality and yield (Lobell and Gourdji, 2012). The increase in global average temperature recorded in the 20th century was approximately 0.74°C; even worse perspectives are predicted by simulation models with temperature rising between 2.5 and 5.4°C by 2100, coupled with a decrease in precipitation of about 15%, if no mitigation efforts take place to reduce the emission of green-house gasses (Tadross et al., 2007; Ciscar, 2012). Dramatic changes in rainfall frequency, intensity, duration, and distribution have also been forecast at regional scale (Kao and Ganguly, 2011). This combination of events could significantly modify the levels of heat and drought stresses negatively impacting growth, development, and productivity of cultivated plants (Trenberth, 2011; Vadez et al., 2011).

Soil dryness and high temperature are main limitations to the productivity of agro-ecosystems. In cultivated plants, drought and heat stresses act at a metabolic level, negatively affecting agronomic performance. Seed germination and seedling growth rates are reduced, as well as yield, quality, and safety of the final product (Hasanuzzaman et al., 2013). In cereals, for example, a decrease of the kernel number inside the ear (Sekhon et al., 2010), of the number and size of starch granules and of the amount of lipids and proteins within the kernel (Wilhelm et al., 1999) has been reported under stress conditions, besides significant variations in the bromatological composition of flours (Wardlaw et al., 2002; Balla et al., 2011).

Drought and heat are multidimensional stresses that can hit the plant at various stages of the life cycle, within short (minutes) to prolonged (weeks or even months) time periods with consequences at molecular, physiological, and morphological level, leading to reduced plant growth, improper organ development, photosynthesis inhibition, severe cellular damage, and ultimately, death (Zhang et al., 2005; Wahid et al., 2007).

Plants exposed to sub-lethal abiotic stress conditions exhibit a broad range of morphogenic responses related, at least in part, to variations in gene expression and in the action of agents involved in organ development and growth processes. Despite the diversity of stress-related phenotypes, generic “stress-induced morphogenic responses” (SIMRs) have been recognized. SIMRs comprise alterations in cell division and elongation, processes probably functioning as focal points of response regulation (Potters et al., 2007, 2009).

The plant primary cell wall is a dynamic structure that plays a fundamental role in controlling cell growth and morphogenesis. Its three-dimensional architecture is dynamically remodeled by a number of enzymatic and non-enzymatic factors, in turn controlled by phytohormone-related physiological and stress-induced processes (Wolf et al., 2012). Among them, xyloglucan endotransglucosylase/hydrolase (XTH) enzymes catalyze either the cleavage and molecular grafting of the β-(1-4)-xyloglucan backbone through endotransglucosylase (XET) activity or its irreversible shortening through the endohydrolase (XEH) activity (Fry et al., 1992; Nishitani and Tominaga, 1992; Rose et al., 2002; Shi et al., 2015), thus having a role in the modification of the load-bearing cell wall framework (Nishitani and Vissenberg, 2007; Park and Cosgrove, 2015). The two (XET and XEH) known activities of XTH proteins have been monitored in growing tissues of monocots and dicots suggesting that these enzymes are ubiquitous and essential for the development of all land plants (Van Sandt et al., 2007a; Eklöf and Brumer, 2010; Franková and Fry, 2013). The extractable XET activity and XTH gene expression have often been correlated with cell expansion (Fry et al., 1992; Hetherington and Fry, 1993; Pritchard et al., 1993; Potter and Fry, 1994; Xu et al., 1995; Palmer and Davies, 1996; Antosiewicz et al., 1997; Catalá et al., 1997; Vissenberg et al., 2000, 2003, 2005; Van Sandt et al., 2007a), and clear evidence of the role of some XTHs as cell growth promoters has been established by loss- and gain-of-function molecular approaches (Osato et al., 2006; Van Sandt et al., 2007b; Zhu et al., 2012; Hara et al., 2014).

The relevance of the cell wall remodeling functions of XTH enzymes in the processes of cell division and/or expansion growth suggests a potential central role in stress responses. The Arabidopsis genome contains 33 different XTH genes showing diverse and distinct expression patterns in terms of organ specificity and response to developmental and environmental stimuli such as touch, darkness, cold/heat shock (Xu et al., 1995, 1996), wind (Antosiewicz et al., 1997), flooding (Saab and Sachs, 1995), drought and salt stresses (Schopfer and Liszkay, 2006), light quality (Sasidharan et al., 2010), or aluminum stress (Shi et al., 2015). A similar number of XTH genes was reported in other angiosperms including poplar (Populus trichocarpa Torr. & A.Gray ex. Hook., 41 members) and the monocots rice (Oryza sativa L., 29 members), and wheat (Triticum aestivum L., >57 members; Yokoyama et al., 2004; Geisler-Lee et al., 2006; Liu et al., 2007). The functional role of this enzyme family in drought stress conditions has been shown in tomato (Solanum lycopersicum L.) and Arabidopsis overexpressing CaXTH3 gene from hot pepper (Capsicum annuum L.). The transgenic plants exhibited an increased tolerance to severe water deficit, confirming that XTH is involved in drought tolerance (Cho et al., 2006; Choi et al., 2011). Furthermore, transgenic Arabidopsis plants, overexpressing AtXTH21, showed an improved frost tolerance (Shi et al., 2014). Interestingly, an enhanced cadmium tolerance was evidenced in transgenic poplar (P. alba, L.) overexpressing XTH genes, likely to be related to a decreased uptake and storage capacity of the root. This was accompanied by an increased xyloglucan degradation activity, leading to a reduction of xyloglucan content in root cell walls (Han et al., 2014). XTH enzymes also seem to be involved in triggering stem and petiole elongation, primarily through cellular expansion, in response to a reduction in perceived blue or red light (Keuskamp et al., 2012; Sasidharan et al., 2014).

High temperature and drought stress have emerged as the major constraint for durum wheat (Triticum durum Desf.) production (Kumar et al., 2013). Although, in the Mediterranean area these two stresses usually occur simultaneously, this study reports the separate effects of dehydration and heat exposure on XTH expression as well as XET in vitro activity and in vivo action in apical and sub-apical root regions and shoots excised from 3-day-old seedlings of durum wheat (T. durum Desf.) cultivars (Svevo, Creso, Ardente, and Simeto) differing in stress susceptibility/tolerance.

There is an important distinction between enzyme ‘activity’ (assayed in vitro, under optimized conditions, usually after solubilisation of the enzyme) and enzyme ‘action’ (usually monitored in vivo in the walls of living plant cells; Fry, 2004). An XTH which (when extracted) exhibits high activity in vitro might exhibit low action in vivo – for example because the enzyme may not be physically co-localized with its potential substrate(s), the endogenous xyloglucan may be ‘end-capped’ with an unusual repeat-unit such as V [α-Xyl-(1 → 4)-α-Xyl-(1 → 6)-Glc; Franková and Fry, 2012], the apoplastic pH may be non-optimal in vivo, or the enzyme may be inactivated by immobilization on cellulose (Franková and Fry, 2013). Conversely, an enzyme may exhibit higher action in vivo than would have been predicted by its activity measured in vitro – for example because the apoplast may contain cofactors that were omitted from the in vitro assay (Takeda and Fry, 2004) or simply because a proportion of the wall enzyme was inextractable. Another important difference is that enzyme action assays conducted in vivo permit a better spatial resolution than is feasible with activity assays conducted on enzyme extracts in vitro. In the present work, we use a fluorescently labeled xyloglucan oligosaccharide (XXXGol-SR) as acceptor substrate. For activity assays in vitro, the transglycosylation reaction is monitored in an artificial reaction mixture containing solubilised XTH, donor substrate (xyloglucan) and acceptor substrate (XXXGol-SR). In contrast, for action assays in vivo, only XXXGol-SR is added, and the ability of endogenous XTHs to act on endogenous xyloglucan (as donor substrate) can thereby be monitored. In short, enzyme activity assays indicate potential enzyme action, which may or may not be implemented in vivo.

The ability of plants to tolerate abiotic stress varies at both inter- and intra-specific level; so that, understanding the mechanisms underlying this variability as well as differences/similarities in molecular responses to different stresses, could identify and implement innovative strategies to develop breeding programs aimed at obtaining new high-yield productive hybrids adapted to climate future scenarios (Wahid et al., 2007).

Despite xyloglucans often being minor wall components in the Poaceae (Carpita, 1996; Popper and Fry, 2004), XET activity tends to be extremely high (Fry et al., 1992; Wu et al., 1994). The endogenous transglycanases of H. vulgare are incapable of using mixed-linkage β-glucans instead of xyloglucans as donor substrate (Mohler et al., 2013; unlike the case with the fern-ally, Equisetum). An explanation for this apparent paradox may be provided by the “biomechanical hotspot” hypothesis recently proposed by Park and Cosgrove (2015). In their model the control of wall extensibility occurs by cleavage of a limited fraction (<1%) of the total xyloglucan that is tightly interlaced with cellulose in specific regions (biomechanical hotspots) where multiple microfibrils come into close contact; thus the high XET activities registered in commelinoid monocots may have a role, along with expansins and hydrolases, in finely regulating cell wall expansion by loosening the constraints on the movement of cellulose microfibrils at these hotspots. Nonetheless, it is noteworthy that in a previous study we have reported a surprisingly high amount of xyloglucans (23–39 mol% of the total cell wall polysaccharides) in the apical root zone of Capeiti and Creso durum wheat seedlings grown under water-stressed and unstressed conditions (Leucci et al., 2008), suggesting an important role of xyloglucans as load-bearing components even in the type II primary cell wall of monocots, at least during cell expansion and differentiation. Likewise, in the walls of rapidly growing maize cell-suspension culture, ∼37% of the total xylose residues are in xyloglucans rather than β-xylans (Kerr and Fry, 2003).

The XTH expression levels detected in the root apical region of wheat seedlings subjected to dehydration gave evidence for a slight but significant down-regulation in all investigated cultivars, with the highest inhibition (log₂ effect ≈-2.6 with respect to control) found in the sensitive cultivar Simeto. In the root sub-apical region, however, dehydration did not induce any significant variation of the XTH expression, irrespective of treatment duration. A consistent down-regulation in the expression of different XTH isoform genes (XTH6. XTH9. XTH15, and XTH16) has been previously reported in early developing leaf tissue of six natural variants of Arabidopsis subjected to mild or severe drought stress by Clauw et al. (2015), whereas an up-regulation of XTH coding genes has been reported in the elongation zone of drought-stressed rice roots (Yang et al., 2006), in water-stressed leaves of hot pepper plants (C. annuum L.; Cho et al., 2006) and in osmotically stressed leaves of Leymus mollis (Trin.) Pilg., a wild relative of wheat (Habora et al., 2012). Further, the expression behavior of OsXET9 was differently affected by drought in the Pei’ai64s rice cultivar depending on tissue and development stage, being up-regulated exclusively in the panicles of heading and flowering stage plants (Dong et al., 2011). Different expression patterns of XTH genes were also found in shoots and roots of Arabidopsis plants subjected to 24 h of drought stress by microarray determination, with differential rates among the family members (Tenhaken, 2014). Thus, a highly specific spatial regulation of XTH genes could contribute to strengthening or loosening the cell wall in well-defined topological regions of the plant contributing to susceptibility/tolerance to dehydration. In support of this, using a proteomic approach, Zhu et al. (2007) revealed major changes in the abundance of a large number of cell wall proteins, including XTHs, based on their spatial distribution within the primary root elongation zone of maize seedlings subjected to water deficit. Specifically, of the four putative XTHs identified, two were quantitatively reduced in the first 3 mm of the root apical region, one decreased along the whole root and the last exhibited a slight increase in abundance mainly in the 3–7 mm root sub-apical region after drought treatment.

In spite of the significant down-regulation of XTH genes, the extractable XET activity assayed in vitro increased considerably with stress exposure time in the root apical region of all assayed cultivars. The greatest increase was recorded in the tolerant cultivar Ardente (180%), while in Simeto and Creso it was 110%. These increases in extractable activity were mirrored by increases in XET action, monitored in vivo. Major differences were observed in the root cell elongation and differentiation zones, in compliance with the cultivars’ dehydration tolerance degree. The increase of XET activity and action recorded in the root apical region may cause (or result from) an increased division rate of the meristematic cells and/or a controlled loosening of cell wall polymers in the zone of expansion growth. Both these events could support root growth and development even at low water potential, representing an important mechanism of drought tolerance which allows the stressed plants to invest available resources into soil exploration for residual water (Spollen et al., 1993; Tenhaken, 2014). Similar results were reported by Wu et al. (1994) in the apical few millimeters of maize primary roots growing at low (-1.6 MPa) water potential. The authors reported that the enhancement of the total extractable XET activity was greater immediately behind the apex (0–5 mm) than in following (5–10 mm) sub-apical region of water-stressed compared to well-watered roots, suggesting that XTH may be involved in the stress-induced enhancement of cell wall extensibility in the apical region. This is supported by the work of Sharp et al. (2004) reporting the persistence of root growth rate in the first 3 mm of maize roots subjected to drought stress, while the cells in next 3–4 mm were characterized by a progressive decrease of elongation rate. A differential inhibition of coleoptile (∼75%) and root (∼50%) growth has also been reported in durum wheat seedlings (cultivars Creso and Capeiti) incubated in the presence of a 20% solution of PEG 4000 (Ψ_w - 0.5 MPa) for 24–48 h (Piro et al., 2003). Another possible explanation for the maintenance of XET action in spite of the down-regulation of XTH genes may also be related to an increased supply of donor substrates for the smaller number of XTH enzyme molecules still able to perform endotransglucosylation reactions. This may be achieved by up-regulation of the enzymes involved in the synthesis of these precursors and/or by their increased secretion in stress conditions.

In the root sub-apical region a slight increase in the XET activity in response to dehydration was observed only in the cultivar Creso, while in the cultivar Simeto the 30% transient decline measured after 4 h from stress induction recovered to control values at 24 h. In the cultivar Ardente, stress did not seem to induce any change in XET activity. The maintenance of XET activity at the level of root sub-apical region subjected to dehydration may not be correlated to the rate of root growth, which is greatly reduced in this zone (Sharp et al., 2004), but to the initiation and tip growth of root hairs where XTHs are greatly involved (Vissenberg et al., 2001). Root hairs play an important role in water and mineral absorption and create resistance between the root and the soil particles, facilitating deep penetration by the root apex toward better hydrated areas (Peterson and Farquhar, 1996).

With respect to thermal treatment, in the root apical region, XTH expression was down-regulated in all wheat cultivars and particularly in the thermo-sensitive cultivar Ardente. A strong down-regulation was also noted in the root sub-apical region of the sensitive and moderately tolerant cultivars Ardente and Creso but not in the tolerant cultivar Svevo. In a different approach, a transcriptomic study performed on Chinese cabbage (Brassica rapa, L.) showed that several genes, encoding proteins from the XTH family, were up-regulated following heat treatment (Yang et al., 2006), indicating that plant response to stress is highly complex and often depends on species, genotype, plant age, and organ, as well as on timing and intensity of stress application. The results of XTH expression in Ardente and Creso are consistent with the decreased XET in vitro activity values measured at the level of root apical region, while they partially clash in the cultivar Svevo. A large number of XTH isoforms with different distribution and physiological roles are known to be differentially affected by stress (Campbell and Braam, 1999b; Sharmin et al., 2012). This means that, in stress conditions, specific XTH isoforms are up-regulated while other are down-regulated according to the proposed redundant function of XTH genes, whereby the repression of one member may be compensated for by the up-transcription of the others (Matsui et al., 2005).

In the cultivar Ardente (thermo-sensitive), the in vitro XET activity was significantly inhibited at both root apical (>45%) and sub-apical (>50%) level after 24 h at 42°C. Such inhibition is reduced to 30 and 45%, respectively, in the root apical and sub-apical regions in the cultivar Creso (moderately tolerant) and is not significant, or even increased (19%), in the thermo-tolerant cultivar Svevo. No difference in XET activity was observed in the shoots of all investigated cultivars. It is well established, in fact, that roots are more sensitive to heat than the aerial part of the plant, with high soil temperature more detrimental than high air temperature for whole plant growth (Xu and Huang, 2000; Le Gall et al., 2015).

The in vivo experiments showed a general reduction of XET action in the walls of living cells of thermal treated root apical segments, mirroring the observed change in extractable XET activity assayed in vitro, and highlighted interesting differences at microscale level. While in the cultivar Svevo XET action was not affected by thermal treatment all along the root apical region, in Ardente and Creso it was almost stable in the meristematic zone, increased in the elongation zone (25 and 60%, respectively) and decreased (50 and 20%, respectively) in the zone of cell differentiation after 24 h at 42°C, indicating an intraspecific variability in the sensitivity of XET action to high temperatures, also depending on the particular topological region of the root. Further, the localized increase/decrease of XET action in the root apical region of the heat sensitive and moderately tolerant cultivars could represent an acclimation response to high temperature realized through the redirection of root growth to diminish stress exposure (Potters et al., 2007). Thus, the increase observed within the cell expansion zone may support root growth to reach deep soil, while the decrease within the zone of cell differentiation may reduce the root surface exposed to high temperature which could potentially interfere with root hair development. This agrees with the observation that living epidermal cells of Arabidopsis subjected to short-term heat shock show a temporary disruption of cytoskeletal elements that are important for plant cell growth and morphogenesis including trichome development and tip growth of root hairs (Müller et al., 2007). The stable, or even slight increase, of XET activity and action observed in thermo-tolerant cultivar Svevo could instead support the growth of the root system ensuring the correct development of whole plant even at high temperature, constituting a fundamental mechanism in the process of heat-tolerance (Huang et al., 2012). An increased level of XTH transcripts has been reported in ripening grapevine (Vitis vinifera L.) fruits following heat stress and related to the need for more flexible cell walls to allow an adaptation of berry volume to temperature (Rienth et al., 2014). It has also been speculated that, after cell elongation has ceased, the functional significance of XTHs could be linked to the strengthening of cell walls by increasing xyloglucan polymerization, which might help plants to adapt to high temperature by reinforcing of connections between primary and secondary walls (Yang et al., 2006).

Studies carried out on two species of forage plants of the genus Agrostis with different heat stress susceptibility showed that, when exposed to high temperature, the tolerant species (A. scabra Willd.) maintained better root growth than the thermo-sensitive (A. stolonifera L.; Lyons et al., 2007). The differential change of XET activity and action observed in the Svevo, Creso, and Ardente seedlings incubated 42°C could, therefore, be related to mechanisms similar to those described in Agrostis.

Several XTH isoforms with different temperature optima in a range between 18 and 37°C have been identified in Arabidopsis (Campbell and Braam, 1999a), and an isoform exhibiting continued activity at 4°C was also reported (Purugganan et al., 1997). Similar results were obtained by Steele and Fry (2000), who measured the XET activity of various XTH isoforms isolated from seedlings of bean (Vigna radiata L. Wilczek) and shoots of cauliflower (Brassica oleracea L.) and subsequently exposed in vitro to 42°C, showing an inhibition between 55 and 80% compared to their maximal efficiency. Thus, the XTHs extracted from the various topological regions of the seedling of the three durum wheat cultivars have an optimal temperature stability different from 42°C, at which a decrease of over 40% in XET activity was measured. An exception was the extract obtained from the shoot of the cultivar Creso, where XET activity was not affected by heat. The higher heat-sensitivity of the enzyme in the extracts of the thermo-tolerant cultivar Svevo than in those of the moderately tolerant and sensitive cultivars Creso and Ardente, in contrast with the in vitro determinations, are probably due to response mechanisms adopted by living cells to improve the thermal stability of XTHs. Soto et al. (1999) reported a direct correlation between the heterologous expression of a plant small heat shock protein (HSP17.5 from Castanea sativa Miller) and the thermo-stability of soluble proteins in heat-stressed transgenic Escherichia coli. Heat shock proteins are mainly intraprotoplasmic, but some were also detected in the cell wall of barley coleoptiles and roots subjected to heat stress, and this has suggested a role in wall protein protection (Leone et al., 2000). An HSP90 was localized by confocal microscopy in the cell wall of Aspergillus fumigatus subjected to heat stress conditions, indicating that these proteins, in non-physiological conditions, can be released into the apoplast (Lamoth et al., 2012).

Our results indicate that dehydration and heat stress, applied singularly, differentially influence the XTH expression profiles and the activity and action of XET in wheat seedlings, depending on the degree of susceptibility/tolerance of the cultivars. In all investigated cultivars the root apical region was the district mainly affected by both stresses. Significantly, at shoot level, neither dehydration nor heat treatment induced a change in XET activity or action irrespective of the degree of susceptibility/tolerance. This indicates that the variations in XTH expression level and XET activity and action are implemented mainly at root apical region, confirming the pivotal role of this organ in stress perception and signaling to the whole organism. In addition, the data related to the cultivars Ardente and Creso, common to both dehydration and heat treatments, showed conflicting effects depending on stress type: dehydration determined an overall increase, at least in the apical region of the root, of XET activity and action, while heat generally induced a decrease. This suggests a different involvement of XTH enzymes in the mechanisms of abiotic stress responses. Consequently, further research on the effects of the simultaneous exposure to dehydration and heat is required to expand our knowledge of the role of XTHs in adaptation and acclimation of crop plants to stress conditions naturally occurring in open fields.

AI: Performed most of the biochemical experiments and contributed to the first draft of the manuscript. MDC: performed the microscopical observations. ES and LD: Contributed to the expression experiments and critically discussed the obtained results. MDP, PR, and CP performed the experiments and critically discussed the results on wheat susceptibility/tolerance to drought and high temperatures. ML: Conceived and designed the experiments and wrote the manuscript with the fundamental help of GD, GP, and SF. All authors contributed to the discussion and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.