Metabolomic insights into salt stress responses of commercial Turkish wheat varieties

Abstract

Soil salinity constrains early wheat growth and is accompanied by broad metabolic adjustment. This study characterized shoot-growth and 1H-NMR metabolomic responses to graded NaCl exposure in the Turkish wheat genotypes Altay and Kunduru and identified candidate salt-responsive metabolites associated with genotype-dependent responses. Seedlings were grown in quarter-strength Hoagland solution containing 0–1.0% NaCl. Shoot and root lengths were measured. The quantified metabolomics dataset comprised shoot (gövde; sample code G) extracts with three biological samples per genotype × NaCl condition (n = 3), analyzed by one-dimensional 1H NMR spectroscopy. Metabolite data were cube-root transformed and Pareto scaled. Unsupervised PCA, validated PLS-DA, VIP analysis, and FDR-corrected univariate tests were applied. Replicate-level shoot growth was analyzed using a two-factor model based on experimental-pack means. NaCl concentration significantly affected shoot length (F4,15 = 11.19, p = 0.00021), whereas the genotype main effect (F1,15 = 0.45, p = 0.512) and genotype × NaCl interaction (F4,11 = 0.74, p = 0.586) were not significant. PCA of Altay and Kunduru metabolite profiles explained 75.5% and 8.6% of variance in PC1 and PC2, respectively. Within-genotype PLS-DA models for 0, 0.5 and 1.0% NaCl showed good cross-validated performance (Altay: R 2 Y = 0.942, Q 2 = 0.821, accuracy = 100%; Kunduru: R 2 Y = 0.952, Q 2 = 0.692, accuracy = 88.9%) and significant exact permutation tests. Pairwise Altay-versus-Kunduru models at individual salt levels showed high apparent separation but did not reach significance in exact permutation testing ( p = 0.10), reflecting the small sample size. FDR-supported metabolite differences included asparagine, proline, alanine, lysine, serine, fructose and dimethylamine depending on salinity level. Salinity produced a clear concentration-dependent reduction in shoot growth, but the available growth data did not demonstrate a significant genotype × salinity interaction. Altay and Kunduru nevertheless exhibited distinct salt-associated metabolite profiles. The highlighted compounds should therefore be regarded as candidate salt-responsive metabolites and hypotheses for further physiological validation rather than validated biomarkers of salt tolerance.