SummarySalicylic acid (SA) is reported to protect plants from heat shock (HS), but insuf®cient is known about its role in thermotolerance or how this relates to SA signaling in pathogen resistance. We tested thermotolerance and expression of pathogenesis-related (PR) and HS proteins (HSPs) in Arabidopsis thaliana genotypes with modi®ed SA signaling: plants with the SA hydroxylase NahG transgene, the nonexpresser of PR proteins (npr1) mutant, and the constitutive expressers of PR proteins (cpr1 and cpr5) mutants. At all growth stages from seeds to 3-week-old plants, we found evidence for SA-dependent signaling in basal thermotolerance (i.e. tolerance of HS without prior heat acclimation). Endogenous SA correlated with basal thermotolerance, with the SA-de®cient NahG and SA-accumulating cpr5 genotypes having lowest and highest thermotolerance, respectively. SA promoted thermotolerance during the HS itself and subsequent recovery. Recovery from HS apparently involved an NPR1-dependent pathway but thermotolerance during HS did not. SA reduced electrolyte leakage, indicating that it induced membrane thermoprotection. PR-1 and Hsp17.6 were induced by SA or HS, indicating common factors in pathogen and HS responses. SA-induced Hsp17.6 expression had a different dose±response to PR-1 expression. HS-induced Hsp17.6 protein appeared more slowly in NahG. However, SA only partially induced HSPs. Hsp17.6 induction by HS was more substantial than by SA, and we found no SA effect on Hsp101 expression. All genotypes, including NahG and npr1, were capable of expression of HSPs and acquisition of HS tolerance by prior heat acclimation. Although SA promotes basal thermotolerance, it is not essential for acquired thermotolerance.

The growth of Arabidopsis plants in chilling conditions could be related to their levels of salicylic acid (SA). Plants with the SA hydroxylase NahG transgene grew at similar rates to Col-0 wild types at 23°C, and growth of both genotypes was slowed by transfer to 5°C. However, at 5°C, NahG plants displayed relative growth rates about one-third greater than Col-0, so that by 2 months NahG plants were typically 2.7-fold larger. This resulted primarily from greater cell expansion in NahG rosette leaves. Specific leaf areas and leaf area ratios remained similar in both genotypes. Net assimilation rates were similar in both genotypes at 23°C, but higher in NahG at 5°C. Chlorophyll fluorescence measurements revealed no PSII photodamage in chilled leaves of either genotype. Col-0 shoots at 5°C accumulated SA, particularly in glucosylated form. SA in NahG shoots showed similar tendencies at 5°C, but at greatly depleted levels. Catechol was not detected as a metabolite of the NahG transgene product. We also examined growth and SA levels in SA signaling and metabolism mutants at 5°C. The partially SAinsensitive npr1 mutant displayed growth intermediate between NahG and Col-0, while the SA-deficient eds5 mutant behaved like NahG. In contrast, the cpr1 mutant at 5°C accumulated very high levels of SA and its growth was much more inhibited than wild type. At both temperatures, cpr1 was the only SA-responsive genotype in which oxidative damage (measured as thiobarbituric acid-reactive substances) was significantly different from wild type.Salicylic acid (SA) has received much attention due to its association with economically important plant responses to disease and other stresses. Detailed evidence implicates SA in PR gene expression, systemic acquired resistance, and the hypersensitive response (Kunkel and Brooks, 2002). SA also seems to be involved in responses to abiotic stresses, such as ozone (Sharma et al., 1996;Rao and Davis, 1999;Koch et al., 2000), salt and osmotic stress (Borsani et al., 2001;Molina et al., 2002;Shim et al., 2003), UV-B (Surplus et al., 1998), drought (Senaratna et al., 2000Nemeth et al., 2002), paraquat (Kim et al., 2003, and heat (Dat et al., 1998a(Dat et al., , 1998b(Dat et al., , 2000Lopez-Delgado et al., 1998a;Senaratna et al., 2000;Larkindale and Knight, 2002; Clarke et al., 2004). Stress-influenced developmental transitions, including flowering (Hatayama and Takeno, 2003;Martinez et al., 2004), tuberization (Lopez-Delgado and Scott, 1997), and senescence (Morris et al., 2000), may also involve SA.Cold is one of the most important limitations to crop productivity and species distribution. Freezing (subzero) or chilling (low positive) temperatures can cause injury or reduced growth depending on the cold tolerance of the species (Schneider et al., 1995;Pearce, 1999;Humphreys et al., 2003). Recent studies describe potentially valuable effects of salicylate treatment on cold tolerance in maize, rice, and wheat (Janda et al., 1999;Szalai et al., 2000;Kang and Saltveit, 2002;Tasgin et al., 200...

No abstract

Real-world applications will inevitably entail divergence between samples on which chemometric classifiers are trained and the unknowns requiring classification. This has long been recognized, but there is a shortage of empirical studies on which classifiers perform best in 'external validation' (EV), where the unknown samples are subject to sources of variation relative to the population used to train the classifier. Survey of 286 classification studies in analytical chemistry found only 6.6% that stated elements of variance between training and test samples. Instead, most tested classifiers using hold-outs or resampling (usually cross-validation) from the same population used in training. The present study evaluated a wide range of classifiers on NMR and mass spectra of plant and food materials, from four projects with different data properties (e.g., different numbers and prevalence of classes) and classification objectives. Use of cross-validation was found to be optimistic relative to EV on samples of different provenance to the training set (e.g., different genotypes, different growth conditions, different seasons of crop harvest). For classifier evaluations across the diverse tasks, we used ranks-based non-parametric comparisons, and permutation-based significance tests. Although latent variable methods (e.g., PLSDA) were used in 64% of the surveyed papers, they were among the less successful classifiers in EV, and orthogonal signal correction was counterproductive. Instead, the best EV performances were obtained with machine learning schemes that coped with the high dimensionality (914-1898 features). Random forests confirmed their resilience to high dimensionality, as best overall performers on the full data, despite being used in only 4.5% of the surveyed papers. Most other machine learning classifiers were improved by a feature selection filter (ReliefF), but still did not out-perform random forests.