Understanding the role of compatible solutes in adaptive responses of plants to salinity and waterlogging

When exposed to salinity, plants increase production of various organic osmolytes, termed "compatible solutes"(CS). Their role is often assumed to be osmotic, but the expense of producing high levels of endogenous CS (compared with taking up cheap inorganic ions from soil), casts doubt on this assumption. CS have a plethora of other putative means of alleviating stress, which have proved hard to disentangle and verify, due both to their overlapping effects and the overlap with other stress relieving mechanisms. This work aimed to quantify salt stress-induced changes in plant CS profiles and link them to adaptive responses to salinity in a range of 7 species contrasting in their salt tolerance, with different anatomical features (monocots vs dicots). Previous studies have been limited either by a narrow range of studied solute types or by a lack of quantification, thus we first developed a method using UHPLC and MS/MS to efficiently identify and quantify all of the organic solutes that responded to salt stress in the studied plant species. In all species, the overall amount of Na and K in leaves (especially in the youngest leaves) was constantly proportional to the sum of CS and Cl, prompting a new explanation for the involvement of CS in osmotic adjustment. According to this explanation, CS offset the toxic effects of inorganic solutes, raising their threshold of toxicity and allowing them to be used as osmolytes. This is not mutually exclusive with the conventional explanation for the use of CS in osmotic adjustment, which is that CS are used in the cytoplasm to balance the osmolarity of the vacuole (in which osmotic adjustment is achieved with inorganic ions). However, the levels of CS synthesised in response to salt treatment were often several times higher than needed for osmotic adjustment in the cytoplasm, suggesting that this was not their sole purpose. Further, the conventional explanation implies that the ratio of CS to inorganic ions should fall with salt treatment (as the vacuole expands and the cytosol contracts) and this was not so. In contrast to some prior reports, potassium maintained constant levels in young leaves and contributed little to osmotic adjustment, but its contribution increased as leaf age did. Salt tolerant plants were distinguished by segregation of Na between old and young leaves. This essential salt tolerance trait has been very under-reported. Comparisons of species showed that while Na exclusion from shoots was important, Na segregation between old and young leaves was more important. Salt tolerant plants were also distinguished by their ability to keep low levels of K in old leaves to spare K for young leaves and by their ability to increase the percentage of osmolarity accounted for by sucrose in the strongest salt treatments (although sucrose levels may fluctuate with lower salt treatments). In sensitive plants, the percentage of osmolarity accounted for by sucrose reached a clear maximum in the 50 or 100 mM salt treatment and fell in higher salt treatments.

The ranking of species by total CS varied from one salt treatment to another, and a good correlation with salt tolerance was only achieved by using the maximum induced solute level of each species. Constitutive (control) levels did not relate to salt tolerance. Conversely, the major CS glycine betaine and proline correlated well with salt tolerance when control levels were used, implying that high constitutive levels of these solutes distinguish salt tolerant genotypes. Since some species emphasised glycine betaine and others emphasised proline, a more widely applicable correlation with salt tolerance was found by using the sum of glycine betaine and proline rather than either alone. CS often failed to correlate with salt tolerance when expressed as µmol/g of dry leaf weight; yet they correlated well when expressed as a percentage of osmolarity - which is rarely done. Fv/Fm did not correlate with salt tolerance unless measured on old leaves, which is of considerable practical importance for the use of this rapid screening tool.

Several organic solutes that have rarely or never been linked to salt tolerance were found to correlate with salt tolerance in our study. These included malate, pipecolate, GBB (gamma-butyrobetaine), adenine and adenosine. Malate is probably not a driver of salt tolerance, but rather reflects the ability of a plant to protect malate-synthesising enzymes (with high CS levels and by excluding Na from cytoplasm). Pipecolate was uniquely correlated to the major CS of every species, suggesting that it either enhances the effects of these CS or stimulates their synthesis. The latter is likely, considering that studies of biotic stress have shown pipecolate to be essential to the role of salicylate. Pipecolate also correlated with adenosine and/or GABA in most species, hinting at influence over energy metabolism. GBB correlated negatively with glycine betaine, for which reason we tentatively suggest that in some niche it may fulfil a task otherwise filled by glycine betaine. GABA correlated with Na levels only in the species expected to have the highest energy demand for Na exclusion, which were also the only species in which adenine and adenosine correlated with Na. GABA also had notably frequent correlations with adenine and adenosine, all of which points to the importance of the GABA shunt in salinity.

There was evidence that both malate synthesis and operation of PSII (quantified by Fv/Fm) were protected by inositol and sucrose, for which reason these solutes may be recommended for trials with exogenous use. GABA can also be recommended for trials with exogenous use, if feedstock for the GABA shunt would be helpful. Glycine betaine and proline were responsive to different stress levels (proline only responded to high stress, in which glycine betaine fell), suggesting that for exogenous treatments these solutes may be appropriate for different stress levels or may need to be used in specific ratios.

In addition to the UHPLC work, we used salt treatment to manipulate endogenous CS levels and demonstrated a dose-dependent cross-tolerance in leaves later exposed to UV-B radiation (oxidative stress). This is the first in plantae evidence of a dose-dependent relationship between endogenous CS and oxidative stress tolerance, and these findings are helpful for understanding plant performance in real field situations, when they are often subjected to multiple stresses.

The same methods of solute analysis were also applied to a waterlogging study of the same plant species. The dicots (all sensitive to waterlogging) suffered a fall in shoot water content while the grasses did not, pointing to different causes of stress. Thus, Fv/Fm correlated with tolerance in grasses but not dicots, while the reverse was true for shoot water content. Likewise organic solutes that correlated with tolerance in dicots were generally those that are known to relate to osmotic stress, whereas these did not correlate with tolerance in grasses. Those that correlated with tolerance in grasses correlated with dry matter accumulation rather than shoot water content. The solute with the strongest correlation with tolerance of grasses was choline, but only in young leaves. In old leaves of grasses choline did not correlate with tolerance but did correlate with shoot water content. GABA and pipecolate also correlated with tolerance in grasses. GABA also correlated with tolerance in dicots, correlating with growth rather shoot water content. Several solutes that did not correlate with tolerance were yet strongly responsive to waterlogging (malate, adenine, GBB, tyramine and inositol). Plausible physiological reasons for these responses and correlations are presented. Fv/Fm only correlated with tolerance of grasses when measured on the youngest leaf (and then did so very well). Fv/Fm is usually measured on mature leaves and gives poor correlations with waterlogging tolerance, so this finding potentially facilitates the use of Fv/Fm as a practical selection tool for plant breeders.

Overall, this work brings a new understanding of the interaction between organic and inorganic osmolytes, suggests hitherto unknown physiological roles for several organic solutes in salinity stressed or waterlogged plants, and new insights into the roles of the major solutes glycine betaine, proline and GABA. From a practical perspective, this work presents a method of identifying and quantifying a wide range of organic solutes in plants, and methods of optimising correlations between organic solutes and salt tolerance. It presents solutes that are drivers or indicators of salt tolerance and waterlogging tolerance for the purposes of study, or plant selection and breeding. It reveals the importance of segregating both Na and K between old and young leaves as an essential salt tolerance trait, and shows that the use of Fv/Fm in plant breeding may be facilitated by using the appropriate leaf age group. Some solutes have emerged as potential exogenous treatments, as did evidence that the well-known treatments with glycine betaine and proline may be better used in specific ratios.