RESOURCE Multi-omics atlas of combinatorial abiotic stress responses in wheat Letitia Da Ros1,2,†, Venkatesh Bollina1,†, Raju Soolanayakanahally3,* , Shankar Pahari3, Raed Elferjani3, Manoj Kulkarni1, Neha Vaid4, Eddy Risseuw1, Dustin Cram1, Asher Pasha5, Eddi Esteban5, David Konkin1, Nicholas Provart5 , Eiji Nambara5 and Sateesh Kagale1,* 1Aquatic and Crop Resource Development, National Research Council Canada, Saskatoon, Saskatchewan, Canada, 2Summerland Research and Development Centre, Agriculture and Agri-Food Canada, Summerland, British Columbia, Canada, 3Saskatoon Research and Development Centre, Agriculture and Agri-Food Canada, Saskatoon, Saskatchewan, Canada, 4Department of Biological Sciences, University of Calgary, Calgary, Alberta, Canada, and 5Department of Cell and Systems Biology, University of Toronto, Toronto, Ontario, Canada Received 1 May 2022; revised 10 May 2023; accepted 26 May 2023; published online 29 May 2023. *For correspondence (e-mail sateesh.kagale@nrc-cnrc.gc.ca; raju.soolanayakanahally@agr.gc.ca). †These authors contributed equally to this work. SUMMARY Field-grown crops rarely experience growth conditions in which yield can be maximized. Environmental stresses occur in combination, with advancements in crop tolerance further complicated by its polygenic nature. Strategic targeting of causal genes is required to meet future crop production needs. Here, we employed a systems biology approach in wheat (Triticum aestivum L.) to investigate physio-metabolic adjustments and transcriptome reprogramming involved in acclimations to heat, drought, salinity and all combinations therein. A significant shift in magnitude and complexity of plant response was evident across stress scenarios based on the agronomic losses, increased proline concentrations and 8.7-fold increase in unique differentially expressed transcripts (DETs) observed under the triple stress condition. Transcriptome data from all stress treatments were assembled into an online, open access eFP browser for visualizing gene expression during abiotic stress. Weighted gene co-expression network analysis revealed 152 hub genes of which 32% contained the ethylene-responsive element binding factor-associated amphiphilic repression (EAR) transcriptional repression motif. Cross-referencing against the 31 DETs common to all stress treat- ments isolated TaWRKY33 as a leading candidate for greater plant tolerance to combinatorial stresses. Inte- gration of our findings with available literature on gene functional characterization allowed us to further suggest flexible gene combinations for future adaptive gene stacking in wheat. Our approach demonstrates the strength of robust multi-omics-based data resources for gene discovery in complex environmental con- ditions. Accessibility of such datasets will promote cross-validation of candidate genes across studies and aid in accelerating causal gene validation for crop resiliency. Keywords: wheat (Triticum aestivum L.), abiotic stress, multi-omics, multi-environmental stresses, RNAseq, metabolome, physiological traits, transcription factors, Ethylene-responsive element binding factor-associ- ated amphiphilic repression motif. INTRODUCTION The current global challenge in wheat (Triticum aestivum L.) production is in attaining greater crop productivity under increasingly demanding environmental conditions (Thirsty work., 2021). Although wheat demonstrates high plasticity, with cultivars grown across a range of biomes from warm and humid to cool and dry (Acevedo et al., 2002), wheat has the lowest total productivity among the cereal crops (Abhinandan et al., 2018). Still, global wheat production is projected to reach 777 Mt in 2021–2022 (International Grains Council, 2021, https://www.igc.int/en/ gmr_summary.aspx) with an estimated 60% increase in � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. 1118 The Plant Journal (2023) 116, 1118–1135 doi: 10.1111/tpj.16332 https://orcid.org/0000-0002-9345-9640 https://orcid.org/0000-0002-9345-9640 https://orcid.org/0000-0002-9345-9640 https://orcid.org/0000-0001-5551-7232 https://orcid.org/0000-0001-5551-7232 https://orcid.org/0000-0001-5551-7232 https://orcid.org/0000-0002-2173-0876 https://orcid.org/0000-0002-2173-0876 https://orcid.org/0000-0002-2173-0876 https://orcid.org/0000-0002-7213-1590 https://orcid.org/0000-0002-7213-1590 https://orcid.org/0000-0002-7213-1590 mailto:sateesh.kagale@nrc-cnrc.gc.ca mailto:raju.soolanayakanahally@agr.gc.ca https://www.igc.int/en/gmr_summary.aspx https://www.igc.int/en/gmr_summary.aspx http://creativecommons.org/licenses/by-nc/4.0/ http://crossmark.crossref.org/dialog/?doi=10.1111%2Ftpj.16332&domain=pdf&date_stamp=2023-06-28 wheat production required by 2050 to feed our growing population (CIMMYT, 2016, https://www.cimmyt.org/about/ our-strategy/). Climate change is a major threat to these objectives, with wheat yield loss explained primarily by heat stress, followed by excess water and lastly drought (Zampieri et al., 2017; Zandalinas & Mittler, 2022). These stresses are not experienced in isolation, and crop resil- iency to persistent record-breaking weather events for sta- ble food production should be top of mind after the range of weather extremes experienced throughout 2021 (WMO, 2021, https://public.wmo.int/en/media/news/2021-meeting- challenge-of-extreme-weather; ECCC, 2021, https://www. canada.ca/en/environment-climate-change/services/top-ten- weather-stories/2021.html#toc0). This is in addition to the ongoing salinization of arable land, which is expected to affect 50% of cultivated land by 2050 (Mustafa et al., 2019). As a largely temperate crop, the first challenge to overcome is the susceptibility of wheat to heat, which is projected to reduce global wheat production by 6% for every 1°C of warming (Asseng et al., 2015). Yield is particu- larly impacted when high temperatures occur during grain filling, resulting in accelerated leaf senescence, reduced grain weight and affected grain quality (Spiertz et al., 2006; Zhao et al., 2007). Exposure to 34°C during anthesis effec- tively reduced pollen germination, grain weight and grain number (Bheemanahalli et al., 2019). Mechanisms for heat tolerance in planta involve the initiation of expression of heat shock proteins (HSPs), antioxidant systems, and main- tenance of photosynthetic capacity (Farooq et al., 2011). Heat stress in field conditions, however, is often coupled with some form of osmotic stress, either drought or salin- ity. The resulting effects of either stress are an intensity- dependent reduction in yield. Drought tolerance involves minimization of water loss, osmotic adjustment, increased rooting depth and effective mobilization of carbon stores (Ashe et al., 2017; Ashraf, 2010; Yang & Zhang, 2006). Known key regulators during drought stress include TaER genes for determining plant transpiration efficiency (Zheng et al., 2015) and TaMOR genes for enhanced root architec- ture (Li, Liu, et al., 2016). Salinity, in addition to osmotic adjustments, requires sodium exclusion or tissue tolerance to reduce disruptions to nutrient acquisition through selec- tive ion absorption (Davenport & Tester, 2000; Genc et al., 2007; Khatkar & Kuhad, 2000). Genes specific to salt tolerance include TaNIP aquaporin genes and TaHKT trans- porter genes, which are involved in sodium homeostasis (Gao et al., 2010; Kumar et al., 2017). The unique effects of individual stresses are important to negate under high- intensity stress conditions, but several common cellular signaling pathways are initiated early on for all three stres- ses. Adjustments to these common pathways could improve baseline tolerance for crops in mild multi-stress environments. Concurrent targeting of stress-specific genes based on geographical requirements could provide the flexibility necessary to respond to the many complex multi-stress environments around the globe. Unfortu- nately, the tools currently available to conventional breed- ing programs are not sufficient in meeting these challenges and new approaches, such as gene stacking, have yet to be validated in wheat. Progress has been made on stress signaling pathways with the availability of multiple annotated reference genome sequences for wheat (IWGSC, 2018; Walkowiak et al., 2020). A broad range of transcription factors have demonstrated roles in plant tolerance to the aforemen- tioned stresses of heat, drought and salt. These have been assessed either directly in wheat or by functional charac- terization in model plant species. Differential expression of these transcription factors can affect ABA-associated path- ways (TaNAC47, TaNF-YB3;1, TaOPR1, TaCPK34, TaSIM, TabHLH1), ethylene signal transduction (TaERF1) or reac- tive oxygen species scavenging (TaFBA1) or result in improved stress-related phenotypes through multiple or unelucidated mechanisms (TaWRKY1, TaWRKY33, TaDREB3, TabZIP) (Agarwal et al., 2019; Dong et al., 2013; He et al., 2016; Li et al., 2018; Li et al., 2020; Niu et al., 2020; Xu et al., 2007; Yang et al., 2016; Yang et al., 2017; Zhang et al., 2016). Transcription factors are compelling candidates when gene stacking for complex multi-genic stress tolerance (Shailani et al., 2020), as they initiate differential expression of many stress-related and homeostasis-associated genes for plant-level tolerance. Following activity modulation of a chosen transcription factor, additional genes further down the metabolic path- ways that impart tolerance could then be stacked into culti- vars as required. Careful selection of transcription factors and downstream effector genes thus becomes imperative for efficient and effective improvement of high-yielding wheat cultivars. This study aims to identify gene targets that could improve wheat performance under combinatorial stress conditions for gene stacking using conventional selection and/or transformation techniques, as has been the case in rice (Oryza sativa) (Shailani et al., 2020). Here we apply a systems biology approach to study physiological, metabo- lomic, hormonal and transcriptomic responses associated with heat, drought, salinity and their possible combina- tions. Our objectives were to (i) rank stress treatments based on the overall physiological and growth impacts, (ii) identify the core sets of genes common to a particular stress type, (iii) examine pathways that are uniquely expressed in the various stress combinations, (iv) create an open access electronic fluorescent pictograph (eFP) browser to visualize multi-stress transcriptome data in wheat, (v) detect associations between phenotypic and transcriptomic responses and (vi) suggest possible tran- scription factors for further characterization and validation in improving wheat performance in multi-stress � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1119 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://www.cimmyt.org/about/our-strategy/ https://www.cimmyt.org/about/our-strategy/ https://public.wmo.int/en/media/news/2021-meeting-challenge-of-extreme-weather; https://public.wmo.int/en/media/news/2021-meeting-challenge-of-extreme-weather; https://www.canada.ca/en/environment-climate-change/services/top-ten-weather-stories/2021.html#toc0 https://www.canada.ca/en/environment-climate-change/services/top-ten-weather-stories/2021.html#toc0 https://www.canada.ca/en/environment-climate-change/services/top-ten-weather-stories/2021.html#toc0 environments. Datasets used in constructing the multi- omics atlas, covering all physiological, metabolic, hor- monal and transcriptomic data, will be made available for validation of the identified candidate genes with roles in combinatorial stress responses. RESULTS AND DISCUSSION Phenotypic responses of wheat to single and multiple environmental stresses Triticum aestivum cv. Stettler, a Canada Western Red Spring wheat variety with several desirable agronomic traits including high yield, higher grain protein concentra- tion, reduced lodging and resistance to several common wheat pathogens (DePauw et al., 2011), was selected for this study. It is also the common parent of the Canadian Nested Association Mapping panel. The productivity of high-yielding wheat cultivars, such as Stettler, is especially vulnerable to stresses during grain filling, which causes premature flag leaf senescence and reduced grain quality (Spiertz et al., 2006; Zhao et al., 2007). Flag leaf characteris- tics are associated with yield, as flag leaves provide the majority of total grain carbon (Biswal & Kohli, 2013; Blake et al., 2007). To best observe plant adjustments to sub- lethal stress and stress combinations, physiological, meta- bolic, hormonal and transcriptomic changes in flag leaves during grain filling were targeted for assessment. From spike initiation (Feekes scale stage 8) onwards, greenhouse potted Stettler plants were subjected to one of eight stress treatments, as described in the Materials and Methods sec- tion: control (C), heat (H), drought (D), salinity (S), heat and drought (HD), salinity and heat (SH), salinity and drought (SD) and, lastly, salinity, heat and drought (SHD) treatment. Forty-seven phenotypes were measured for each treat- ment, including 12 agronomic, seven physiological, and 22 metabolic traits along with measurements of six plant hor- mones. Of the seven stress treatments to which plants were subjected, the heat-treated plants showed the fewest agronomic, physiological and metabolic differences com- pared to the control (Table S1). Drought, salinity and all of the four stress combinations were significantly different from the control for most measured agronomic and physi- ological traits (Figures 1 and 2). This is apparent in visual comparisons and the relative reductions observed in above- and belowground biomass (Figure 2a). Typical stress-induced growth modifications included reduced plant height, smaller leaves, shorter spikes, senescence- induced yellowing, lower seed yield and reductions in 1000-kernel weight (TKW; Figure 1). An up to 80% decrease in seed yield was observed in double and triple stress treatments. This differs from stress combination studies in Arabidopsis thaliana L., where impacts of single stresses were considered negligible with only combinatorial stres- ses affecting plant fitness (Zandalinas et al., 2021), and the model plant Brachypodium distachyon (L.) P.Beauv., which was able to maintain 100-grain weight under HD and SD conditions (Shaar-Moshe et al., 2019). The single trait on which heat had the largest effect was flag leaf temperature, which increased by 27, 36, 40 and 47% for H, HD, SH and SHD treatments, respectively (Figure 2b). The majority of the 22 wheat metabolic traits varied considerably within and across treatments. Cysteine, proline and serine had the most notable trends, where the three amino acid con- centrations in flag leaves were significantly proportional to the presence of abiotic stress or the increasing stress inten- sities (Figure 1). In general, amino acid concentrations trended upwards, while organic acid concentrations had no discernable trends. Sucrose levels were highly variable but showed upward trends in osmotic stress treatments D, S and SD. Sucrose concentrations in the SHD treatment were similar to those in the control, presumably due to a significantly reduced photosynthetic assimilation rate when compared to all other treatments. Plants exposed to 80 mM NaCl accumulated sodium, with at least a seven- fold decrease in Ca/Na and K/Na ratios in all treatments involving salinity stress (Figure 1). Diagnostic traits, such as photosynthesis and biomass, trended downward from control values based on the number of stresses applied, with lowest trait values in the SHD treatment. Intrinsic water-use efficiency (WUEi) showed the opposite trend with values increasing relative to the control, and the high- est WUEi was found in the triple stress treatment (Figure 1). Of the plant hormones measured, salicylic acid (SA) showed the most marked difference across treatments with increasing concentrations across the single and combina- torial stress treatments. In particular, SA concentrations are high in the flag leaves of plants subjected to combina- tional stresses. Based on the agronomic, physiological, metabolic and hormonal modifications observed, treat- ments were ranked from lowest to highest impact as fol- lows: H, D, S, HD, SD, SH and SHD (Table S2). Transcriptomic comparisons of single and multiple environmental stresses A range of comparisons were made to assess differentially expressed transcripts (DETs) that were both commonly and uniquely expressed across treatments. Thresholds of at least a 2-fold change relative to the control and a false discovery rate (FDR) of 5% were used to determine DETs. When examining the single stress treatments, D and S have roughly three times as many common DETs (78) as H shared with either of the other two single treatments (29 and 23 DETs, respectively) (Figure 3a). The three single stress treatments, H, D and S, had fewer commonalities (seven DETs) than the double stress treatments (34 DETs) (Figure 3a,b). Overall, there were only 31 common and 4539 unique DETs across all seven stress treatments (Figure 3c). Approximately half of the 31 � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1120 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense common DETs were downregulated compared to the con- trol (Figure 3d). Single stresses had a similar number of uniquely expressed transcripts. HD and SD had 1.329 as many unique DETs compared to single stresses, while SH and SHD had 3.49 and 8.79 as many, respectively (Figure 3c). The groupings of DETs, according to which treatments elicited a response, can be found in Table S3. For annotation purposes, wheat gene lists were cre- ated for the following categories: common differentially expressed genes (DEGs) to all salt stress treatments (170 DEGs), common DEGs to all heat stress treatments (186 DEGs), common DEGs to all drought stress treatments (sex DEGs), unique DEGs to HD (290 DEGs), unique DEGs to SD (299 DEGs), unique DEGs to SH (734 DEGs) and unique DEGs to SHD (1852 DEGs). These gene lists were used to find the corresponding Arabidopsis homologs for improved annotation during the subsequent analyses. DEGs common to drought treatments were unable to be assigned to Gene Ontology (GO) categories due to the small number of genes. The seven Arabidopsis homologs identified were annotated as a NAC domain transcription factor, three HXXXD-type acyl-transferase family proteins, Figure 1. Boxplots demonstrating changes in wheat agronomic, physiological and metabolic traits in addition to flag leaf hormone concentrations as observed under the individual abiotic stresses and their combinations. All agronomic traits were measured at plant maturity. Physiological measurements and metabolite samples were taken 9 days after the start of stress treatments. Total organic acids represent a sum total of malic, oxalic and oxaloacetic acids. Asterisks denote differing levels of significance as follows when the treatment is compared to the control, P < 0.10 (*), P < 0.05 (**), P < 0.01 (***). � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1121 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense a high affinity K+ transporter and two guanylate kinases. No discernable links between these DEGs were identified. For the remaining categories, GO term enrichment analysis was conducted to isolate the top 100 categories into which the common and unique DEGs fell (Figure 4). Heat treat- ments all had DEGs involved in heat response and acclima- tion, chaperone-mediated protein folding and, to a lesser extent, sugar metabolism. Salinity treatments shared genes annotated to redox homeostasis, carbon utilization, nitrogen assimilation, cofactor catabolism and, to a lesser extent, amino acid metabolism. Genes unique to double stressors varied widely by stress and covered a wide array of metabolic processes, including nucleosome assembly, ABA biosynthesis and alternative respiration in HD; fatty acid metabolism, redox homeostasis and nitrogen/sulfur cycles in SD; and response to toxicity, the phenylpropa- noid pathway and fatty acid metabolism in SH. Uniquely expressed transcripts, particularly the large number associ- ated with SHD treatment (8.79 more DETs than for single stress treatments) were found to be in GO categories related to photosynthesis, glycolysis, hexose metabolism and a range of basic cell processes (Figure 4). Commonalities in differential gene expression, based on groupings into annotated metabolic pathways between Figure 2. Images of wheat plants grown under various stress treatments and their combinations. (a) Treatments include the well-watered control (C), heat (H), drought (D), salt (S), heat and drought (HD), salt and drought (SD), salt and heat (SH) and a combi- nation of salt, heat and drought (SHD). Each row of images from top to bottom is as follows: standard imaging depicting the differences in plant health and bio- mass, infrared images taken by a FLIR T-530 camera with average flag leaf temperature indicated in the bottom right, and standard imaging showing the effect of different stresses on root development. (b) Boxplots showing changes in the phenotypic measurements across the stress treatments and their combinations for biomass, flag leaf temperature and root weight from left to right. � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1122 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense each of the three stresses regardless of the stress combina- tion applied, were examined to differentiate between plant responses to salt, heat and drought. DEGs across salt treat- ments (S, SH, SD, SHD) included genes involved in redox homeostasis and auxin signaling and genes encoding ARR-B transcription factors and peroxidases. Heat treat- ments (H, HD, SH, SHD) had upregulation of many genes associated with abiotic stress, heat shock factors and pro- teolysis (Figure S1). These responses have been widely documented for salt and heat stress (Abhinandan Figure 3. Visualizations of the common intersections of all differentially expressed transcripts (DETs) across the seven treatments. The three Venn diagrams depict the following: (a) compares the 1395 DETs from the single stress conditions (S, H and D), (b) compares the 2336 DETs between the double stress combi- nations (SH, HD and SD), (c) is a seven-way Venn diagram that compares the 8724 DETs from all treatments to identify 31 common DETs and (d) is a heatmap showing expression of the 31 common DETs across all stress treatments. � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1123 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense Figure 4. Top 100 results from GO term enrichment analysis using the Arabidopsis homologs of differentially expressed wheat genes that were in the following categories from left to right: common to all heat (H) treatments, common to all salt (S) treatments, unique to the heat drought treatment (HD), unique to the salt drought treatment (SD), unique to the salt heat treatment (SH) and unique to the triple salt, heat and drought treatment (SHD). � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1124 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense et al., 2018; Farooq et al., 2017; Kumar et al., 2020; Zandali- nas et al., 2020). Interestingly, drought treatments (D, HD, SD, SHD) only shared six DEGs. The lack of a unique drought response could be due to its overlap with aspects of salt and heat stress (Abhinandan et al., 2018; Shaar- Moshe et al., 2017). Instead, the stress response may be modulated when occurring in tandem with salt or heat stress, or certain pathways may be initiated when both are not present. The former possibility is further indicated in a comparison of the numbers of unique DETs between HD, SD and SH treatments. Both HD and SD had approximately 460 unique DETs while SH had over 29 that amount with 985 DETs (Figure 3b). SA may play a role in these tran- scriptional differences as average concentrations show similar trends. Concentrations of SA are lowest in the D treatment among the single stress treatments and highest in the SH treatment among all treatments. This would need to be explored more in depth to demonstrate possible cor- relations. This could explain why many genes improve plant performance under multiple stresses (He et al., 2016; Niu et al., 2020). The number of pathways affected increased with the number of stresses experienced, culminating in the highest effect being observed in the SHD treatment. Mounting changes were apparent in developmental genes, second- ary metabolites, proteolysis and transcription factor cate- gories. Double and triple stress treatments were observed to affect the ABA, jasmonic acid and auxin signaling path- ways. Expression of bZIP transcription factors appeared as uniquely expressed genes in treatments involving heat stress, while the regulation of MYB transcription factors was apparent throughout (Figure S2). Overall, salt and heat stress appeared to have the largest impacts on the path- ways involved in plant stress responses and the direction- ality of gene expression within a given pathway (Figure 4; Figures S1 and S2). This high variability in the directional- ity of gene expression, within transcription factor families and hormone signaling pathways, suggested shifting prior- ities and a partial shutdown of stress response pathways depending on the stress combination. Similar results have been observed in Brachypodium dystachion (Shaar-Moshe et al., 2017). The result is that expression patterns within a pathway do not appear to show trends across related stress treatments, but demonstrate an increasing mix of complex responses. This loss of predictability in transcrip- tional response highlights the complexity of identifying genes for improving stress tolerance. Unique DEGs can indicate adaptive mechanisms for plant stress or be associ- ated with regulation and the developmental fall-out caused by an inability to withstand the stress combination. Pheno- typic alterations supported these observations as values for diagnostic traits, such as photosynthesis and biomass, trended downward from control values based on the num- ber of stresses applied. Values reached a minimum in the SHD treatment (Figures 1 and 2b). The stress-related amino acids proline and cysteine along with WUE show the oppo- site trend, with values increasing relative to the control. The highest levels of these three traits were found in the triple stress treatment (Figures 1 and 2b). Wheat abiotic stress eFP browser To increase dataset accessibility and facilitate ongoing identification of stress-associated genes, an open access wheat abiotic stress eFP browser was created and inte- grated into the Wheat eFP Browser gene expression visual- ization tool (Ramirez-Gonzalez et al., 2018) hosted by the Bio-Analytic Resource (BAR) for Plant Biology at the Uni- versity of Toronto (https://bar.utoronto.ca/efp_wheat/cgi- bin/efpWeb.cgi?dataSource=Wheat_Abiotic_Stress) (Fucile et al., 2011; Waese et al., 2017). As a resource for the wheat research community, it allows for rapid visualization of flag leaf transcript levels across all stress treatments and their combinations without the need for bioinformatics soft- ware. An example can be seen in Figure S3. Given that Wheat eFP hosts a variety of datasets in wheat, possible non-target effects of gene stacking for stress tolerance dur- ing other growth stages and tissues could be explored. Visualization of ortholog expression in other major crop species, such as rice, barley (Hordeum vulgare) and maize (Zea mays), is also possible through the online BAR. Associations among phenotypic, transcriptomic and metabolomic data The associations among transcript expression and the measured phenotypes were detected using weighted gene co-expression network analysis (WGCNA) with the data from all treatments. The gene expression network modules were identified based on the means of hierarchical cluster- ing, yielding modules containing genes with high topologi- cal overlap. In total, transcripts were grouped into 16 modules ranging in size from 20 to 6866 transcripts (Table S4). Modules were assessed with all 47 external traits to identify module–trait correlations. Genes in mod- ules negatively correlated with agronomic traits vastly out- numbered those found in positively correlated modules. Correlations between the largest module (turquoise with 6866 genes) showed decreased expression relative to the control in H and D conditions, with greater relative expres- sion in S and all combinatorial stress conditions. This pat- tern resembles the biphasic J-shape characteristic of hormetic dose responses, in which low-dose stress can cause a response in the opposite direction of stronger stress (Calabrese & Mattson, 2017). Five modules, black, blue, turquoise, purple and red, had significant correlations to a wide range of phenotypic traits (Figure 5). Expression of transcripts in the black and red modules was positively correlated to most agronomic and physiological traits. The exceptions were canopy temperature and WUE, which � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1125 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://bar.utoronto.ca/efp_wheat/cgi-bin/efpWeb.cgi?dataSource=Wheat_Abiotic_Stress https://bar.utoronto.ca/efp_wheat/cgi-bin/efpWeb.cgi?dataSource=Wheat_Abiotic_Stress were negatively correlated to the gene expression in both of these modules. Expression of transcripts in these mod- ules was also negatively correlated to flag leaf concentra- tions of amino acids and SA (Figure 5). Purple, blue and turquoise modules had opposite results, with positive cor- relations between gene expression and amino acid and SA concentrations and negative correlations between the agronomic traits and several physiological traits such as photosynthetic rate, stomatal conductance and the ratio of intercellular to ambient CO2 concentrations (Figure 5). This demonstrates the tradeoff between plant productivity and stress resilience in commercial wheat cultivars as amino acid metabolism has been shown to be a key node in plant stress acclimation across multiple crop species (Cardoso et al., 2022). DETs from these modules and their annota- tions have been included in Table S5. These genes could further be tested for causal roles in stress responses and could serve as candidates for specific plant modifications required in high-intensity stress conditions. Ubiquitous indicators of stress Many abiotic stresses elicit similar phenotypic responses in wheat. These phenotypes can therefore be used to indi- cate sub-optimal growth conditions and identify general coping strategies. Unexpectedly, of the 40 agronomic, physiological and metabolic traits measured, only agro- nomic traits such as plant height, biomass, seed yield and harvest index were significantly lower in all stress treat- ments when compared to the control (Figures 1 and 2b). With the exception of heat stress (as a single stress), physi- ological and metabolic traits showed a response to all applied stresses. One interesting finding is the differential accumulation of SA in the flag leaves of plants subjected to various single and combined stresses. SA had a demonstrated response in H and all combinatorial stress treatments. The importance of SA accumulation to combi- natorial stress resilience has been demonstrated in maize; however, direct links to metabolic phenotypes have yet to be fully elucidated (Suraj et al., 2022; Yang et al., 2022). Khan et al. (2013) found evidence in wheat that the adverse effects of heat stress on photosynthesis can be alleviated by induced proline accumulation and its interaction with ethylene, as demonstrated by the application of SA. SA treatment increased the production of proline in plants subjected to heat stress, which in turn increased the osmotic potential of the cells, enabling the plants to take up more water. This had a positive effect on stomatal aper- ture and the photosynthetic machinery, resulting in improved efficiency of photosystem II and increased activ- ity of the enzyme RuBisCO. The cumulative effect of these changes was an increase in photosynthesis under heat stress conditions (Khan et al., 2013). Our finding suggests a potential function of this hormone in differential stress responses in the wheat flag leaves. Reduced height and biomass were rapid indicators of all applied stresses, but these adjustments were unable to maintain yields or TKW. Under D, S and HD treatments, wheat plants maintained a TKW similar to plants exposed to H stress, albeit with sig- nificant biomass reductions (Figure 1). This decoupling of the fitness-associated traits is most prominent between S and H stresses, where a significant reduction of biomass in S is not accompanied with reduced yield and TKW. This is possibly due to the accumulation of amino acids (Figure 1), as they have demonstrated roles in stomatal closure and ABA production (Batool et al., 2018; Liu et al., 2019). Amino acid concentration adjustments showed diminishing returns in the double and triple stress treatments, as yield and TKW reached minimum levels despite continued Figure 5. Correlations of module eigengenes (MEs) with 47 quantitative plant traits and 14 676 transcripts. Of the 14 676 transcripts, 618 were annotated as transcription factors. Each column represents one of the 47 traits with the corresponding trait–module correlation coefficients written in each cell. Cell color denotes the strength of the correlation, with red indicating a positive correlation and blue indicating a negative correlation. Bolded values represent significant correlations with a P-value of <0.05. � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1126 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense cysteine accumulation and spiking proline concentrations (Figure 1). The sudden proline accumulation, during expo- sure to stress combinations, demonstrated increased cell requirements for osmotic adjustment, membrane stabiliza- tion and scavenging of reactive oxygen species (Hayat et al., 2012). Studies testing the exogenous application of proline (Farooq et al., 2017) and transgenic approaches enhancing expression of key proline biosynthetic enzymes in wheat (Vendruscolo et al., 2007) have shown proline accumulation to be a general coping strategy in response to stress. Amplified stress-induced proline production can result in greater stress tolerance, but the possibility of free proline-induced growth inhibition should be acknowledged (Nanjo et al., 2003). Targeting of serine hydroxymethyl- transferases (e.g., SHMT1) (Liu et al., 2019) and pyrroline- 5-carboxylate synthetases (e.g., P5CS) (Vendruscolo et al., 2007) in wheat could have the potential to improve general stress tolerance in a pathway-specific manner. DEG candidates annotated as SHMT or P5CS were found within the black and blue modules (Table S5). They include TraesCS2B02G521700, TraesCS2D02G493600, TraesCS3D02G378700 and TraesCS3B02G395900. Direct application of SA in switchgrass (Panicum virgatum) has been shown to strongly promote accumulation of amino acids such as proline, serine, threonine and alanine (Li, Yu, et al., 2016). Significant positive correlations between the purple, blue and turquoise gene modules and amino acid and SA concentrations are promising for exploration of mechanisms regulating this interaction (Figure 5). Further pathway-specific gene candidates could be identified from the five gene expression network modules (black, blue, tur- quoise, purple and red) as they have the strongest correla- tions to the measured phenotypic traits (Figure 5). Gene candidates for general stress tolerance were additionally identified based on transcripts found to be dif- ferentially expressed in all stress treatments compared to the control. Thirty-one transcripts met these criteria, 30 of which were assigned to a module (Table S6). Genes included members of the cytochrome P450 family that could be implicated in ABA homeostasis and flavonoid bio- synthesis (Li & Wei, 2020), a glutathione S-transferase which is used to metabolize toxic substances in the cell (Liu et al., 2013), and a heat shock protein. A further three are annotated as protein kinases and an additional seven as transcription factors. Three of the transcription factors were BTB-TAZ domain proteins, whose roles in abiotic stress have yet to be elucidated, but which have been shown to increase drought tolerance through brassinoster- oid signaling and proline biosynthesis in transgenic Arabi- dopsis (Zhou et al., 2020). Two were bZIP transcription factors, proteins whose overexpression in Arabidopsis resulted in enhanced tolerance for a multitude of stresses, including heat, drought and salinity (Agarwal et al., 2019). The final two were ethylene response factors (ERFs), which modulate ethylene responses to stress (Xu et al., 2007), and a MYB-related protein. The MYB protein family includes members that improve plant performance under salt stress (Yu et al., 2017). The ERF (TraesCS4B02G299600) andMYB-related (TraesCS6D02G241900) transcription factor genes are considered top candidates, as they were also identified as hub genes due to their high module membership (MM) value and therefore their impact on the intra-modular gene network. The MM mea- sures transcript strength within the module, with higher MM values indicating a stronger influence on the intra- modular network. In the wheat cultivar ‘Stettler’, all but one of these common transcription factors were downre- gulated. These transcription factors present themselves as the main candidates for gene characterization studies, fol- lowed by gene stacking of the promising transcription fac- tors and pathway-specific genes for enhanced multi-stress tolerance. Hub genes and the prevalence of EAR motifs For a more general plant response to multiple stresses, transcripts annotated as transcription factors were mapped to the respective co-expression modules. This identified 618 transcripts in the network modules. Of the 618 tran- scripts, 155 transcripts (152 genes) were identified as hub genes by having an MM value greater than 0.9 (Table S7). Expression patterns across treatments and the module colors to which they belong can be found in Figure 6a. Hub genes included representatives of the WRKY, MYB, ERF, HSF, ARF, NAC, bHLH and NF-Y transcription factor gene families (Table S7). The top five represented families of transcription factors in the hub genes were WRKY, MYB, ERF, mitochondrial transcription termination factors and HSF, with 16, 13, 12, 11 and 11 members, respectively. Four of the WRKY and ERF transcription fac- tors are considered putative, three ERF transcription fac- tors are AP2/ERF proteins and MYB transcription factors included both MYB and MYB-like transcription factors. Among the 152 hub genes, each of the RNA polymerase sigma transcription factors, ARF, NAC and DOF zinc-finger protein transcription factor families comprised five genes. Important stress response families with at least four repre- sentatives included bHLH and NF-Y (Table S7). All these families have been previously implicated in wheat stress signal transduction pathways (Figure S4). Wheat transcrip- tion factor genes from the WRKY (He et al., 2016; Qin et al., 2015), MYB (Cai et al., 2011; Zhang et al., 2012), ERF (Xu et al., 2007), HSF (Kumar et al., 2018), NAC (Zhang et al., 2016) and bHLH (Yang et al., 2016) families showed distinct functions in multi-stress tolerance. Studies of the NF-Y family member TaNF-YA10-1 were more complex with the gene conferring drought tolerance while increas- ing plant sensitivity to salt (Ma et al., 2015). Characteriza- tion of members of the aforementioned transcription factor � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1127 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense gene families from wheat has been done using transgene expression techniques, predominantly as overexpression studies in Arabidopsis. Wheat ARF transcription factor gene family members still require characterization, but have been implicated in the abiotic stress response by affecting the ABA signaling pathway (Xu et al., 2020). Sequences for many of the characterized genes are pub- licly available through the NCBI GenBank (ncbi.nlm.nih. gov), and current IWGSC v1.1 gene annotations were found by cross-referencing the sequence to the Ensembl- Plants T. aestivum database (plants.ensembl.org/Triticum_ aestivum). Only one characterized gene, TaWRKY33 (TraesCS6B02G175100), was found on the list of identified hub genes. Expression of this gene was downregulated in D, S, SD and SHD treatments. The inclusion of heat stress in H, HD and SH treatments appeared to maintain expres- sion levels similar to those of the control (Table S7). Over- expression of TaWRKY33 in transgenic Arabidopsis lines resulted in lower water loss during drought and higher sur- vival rates after exposure to 45°C. The observation of TaWRKY33’s responsiveness to both ABA and methyl jas- monate suggests the possibility that this transcription fac- tor may integrate the two signaling pathways (He et al., 2016). These data justify identification of favorable alleles of TaWRKY33 and using them for stacking and field tests in wheat. When common sequence motifs were analyzed, 49 (32.2%) of the hub genes were found to contain the ethylene-responsive element binding factor-associated amphiphilic repression (EAR) motif (Figure 6a). The EAR motif, defined by the consensus sequence patterns of either LxLxL or DLNxxP (Kagale et al., 2010), is the most predominant transcriptional repression motif in plants (Kagale & Rozwadowski, 2010). A majority of the EAR motif-containing hub genes (42 of the 49 genes) had the conserved LxLxL repression motif (Figure 6b). Interest- ingly, the identified hub transcription factor gene TaWRKY33 also contains the EAR motif. Expression of transcription factors with EAR motifs varied across the seven treatments, as some were up- and others were downregulated in response to stress. Twenty EAR motif- containing hub genes showed a distinct pattern of increased upregulation upon introduction of stress combi- nations, with peak expression occurring in SH or SHD treatments (Figure 6a). Figure 6. The expression patterns of differentially regulated transcription factors during abiotic stress. (a) A heatmap with Euclidean distance cluster- ing based on normalized transcript expression values (z-scores) for the 155 differentially expressed transcription factors identified from any of the abi- otic stress treatments. Treatments include the well-watered control (C), heat (H), drought (D), salt (S), heat and drought (HD), salt and drought (SD), salt and heat (SH) and a combination of salt, heat and drought (SHD). Coloration on the right side of the heatmap displays the module eigengenes (MEs) linked to each corresponding transcript. Names of the 49 transcripts con- taining EAR motifs are denoted in red with adjacent asterisks. Of the 49 transcripts with EAR motifs, 42 contained the conserved LxLxL repression motif. (b) Conserved patterns in the protein sequence surrounding the LxLxL repression motif. � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1128 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense http://ncbi.nlm.nih.gov http://ncbi.nlm.nih.gov http://plants.ensembl.org/Triticum_aestivum http://plants.ensembl.org/Triticum_aestivum Proteins containing the EAR motif play diverse biologi- cal roles by negatively regulating genes involved in develop- mental, hormonal and stress signaling pathways. EAR repressors can also act as positive regulators of the abiotic stress response (Joo et al., 2013). These diverse functions are apparent from the varied expression patterns observed across the hub gene transcription factors containing the motif (Figure 6a). Annotations of EAR motif-containing hub genes included ERF and MYB transcription factors as previ- ously described, as well as WRKY, mTERF, HSF and MADS- box transcription factors, among many others (Table S7). A sustained response to environmental stress is metabolically demanding, necessitating the use of negative gene regula- tion mechanisms to manage the impacts on growth and development (Kazan, 2006). In particular, ERF repressors containing the EAR motif are thought to apply strong regu- latory control on stress-related genes to limit acute stress responses (Dong & Liu, 2010). Only a single ERF transcrip- tion factor with an EAR motif, TraesCS5D02G229400, had increased expression as stress severity increased. Four others (TraesCS1A02G218100, TraesCS1D02G059200, TraesCS2A02G514200, TraesCS7B02G028700) showed diverse expression patterns across stress treatments with similar or lower expression in SHD compared to the control (Figure 6a), supporting the apparent complexity of stress- dependent gene regulation even among EAR motif- containing transcription factors. As the literature surrounding combinatorial abiotic stress responses in wheat expands, careful selection of tar- get genes will be imperative for cultivar improvement. Chosen gene combinations and the affected pathways will need to activate, and act synergistically, under a wide array of conditions. To support the future of adaptive gene stack- ing in wheat, we provide a visual summary of our pro- posed promising gene families and the downstream effects as characterized in wheat-specific literature to date (Figure 7). Transcript IDs for the suggested gene candi- dates within these families, as identified in this study, have been highlighted in Figure 7 with gene lists included in the supplemental files to enable region-specific gene selection. These pathways and candidate genes were identified in the relatively stress-tolerant cultivar Stettler, but they pro- vide targets for the search for more favorable alleles in other wheat germplasm. Additionally, we provide our com- plete multi-omics dataset for cross-validation and for use in further gene discovery (Data S1). We hope this facilitates the implementation of multi-stress research and gene stacking into current wheat breeding programs for abiotic stress tolerance. The direct application of our results is currently dependent on co-transformation-mediated gene stacking of the identified candidate genes into commercial wheat cultivars. Incorporation of these findings into marker- assisted selection-based breeding programs would require further screening of multi-stress-resistant cultivars. Crosses of parents with opposing phenotypes and the sub- sequent generation of double haploid or recombinant inbred line mapping populations would further help to identify causal alleles for conventional breeding applica- tions. It is cultivar development through the comprehen- sive use of multi-omics knowledge bases, characterization studies and a mix of advanced molecular approaches that will give us the potential to meet future demands for wheat under a changing climate. CONCLUSIONS In summary, this study provides insights into the complex interplay between gene expression and metabolomic, physiological and agronomic traits in wheat under abiotic stresses. The genetic basis of stress responses is highly intricate, necessitating a hierarchical approach to compre- hend the sequence of events from stress perception to response development. Under sub-lethal conditions, the plant’s response is to reconfigure its physiology towards a new steady state. While this process can be partially pre- dictable, careful analysis is required to fully understand the mechanisms involved. Our results highlight the impact of different stress combinations on plant growth and devel- opment, with the triple combination of salt, heat and drought stresses having the largest effect on stress response pathways, resulting in reduced biomass and seed yield. We identified several key genes and pathways that could serve as targets for breeding more stress-tolerant wheat varieties, such as TaWRKY33 and other genes involved in proline synthesis and ABA homeostasis. In addition, the identification of hub genes that include tran- scription factors from various families, such as WRKY, MYB, ERF, HSF, ARF, NAC, bHLH and NF-Y, and the pres- ence of EAR motifs in 32% of the hub genes suggest that the integration of multiple stress signals in wheat is likely regulated by a complex transcriptional network that involves positive and negative regulation mechanisms. Therefore, careful selection of target genes is critical for cultivar improvement in wheat under abiotic stresses, and the proposed gene families and pathways identified in this study, along with the reference multi-omics dataset pro- vided, offer promising targets for adaptive gene stacking and marker-assisted selection-based breeding to enhance crop resiliency in the face of increasingly complex environ- mental conditions. MATERIALS AND METHODS Growth parameters and stress treatments The Canadian Spring wheat cultivar ‘Stettler’ developed and con- tributed by the Swift Current Research and Development Centre of Agriculture and Agri-Food Canada was chosen for study. Growth conditions and stress treatments were applied in a similar manner � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1129 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense to previous studies (Elferjani & Soolanayakanahally, 2018). Seeds were obtained from the Seed Increase Unit at the Agriculture and Agri-Food Canada Indian Head Research Farm in Saskatchewan. Before beginning the experiment, 11-L pots were filled with a mix of 80% peat moss and 20% sand, weighed and watered four times during 24 h to calculate water holding capacity. Four seeds were sown per pot and were later thinned to two after emergence. Seedlings were maintained under 20°C/16°C and 16 h/8 h day/ night temperature and photoperiod. After 2 weeks, temperatures were raised to 24°C/18°C, while the photoperiod remained con- stant. All pots were fertilized weekly with N:P:K (15:20:15) and full- strength Hoagland’s micronutrient solution (Hoagland & Arnon, 1950). Five weeks after sowing, at spike initiation (Feekes scale stage 8), eight pots were assigned to each of the seven abi- otic stresses or stress combinations and the control for a total of 64. Single treatments included control (C), heat (H), drought (D) and salinity (S). Combined treatments were heat and drought (HD), salinity and heat (SH), salinity and drought (SD) and, lastly, salinity, heat and drought (SHD) treatment. Control plants were maintained at 100% field capacity using tap water. To apply heat stress, plants were placed in two greenhouses. The first main- tained day/night temperatures of 24°C/18°C and housed C, D, S and DS treatments. The second housed the treatments requiring heat stress (H, HD, SH, SHD). The day-time temperature was set to 30°C from 9 am to 3 pm before returning to the standard 24°C/ 18°C temperatures. Drought treatments (D, HD, SD, SHD) were imposed by maintaining pots at 40% field capacity. Salinity treat- ments were applied by increasing NaCl concentrations, starting at 20 mM, increasing to 40 mM for the first 2 weeks and lastly increasing to 80 mM, which was maintained for the remainder of the experiment. All treatments involving salinity (S, SH, SD, SHD) were given by watering with 250 mL salt water to either 100% or 40% field capacity. Treatments were applied from 5 weeks after sowing and continued until physiological maturity for yield and other agronomic measurements. Data collection and sample processing Half of all plants in the experiment were used for physiological and yield measurements, with the remaining half used for meta- bolomics and transcriptomics sampling (n = 32; four replicates per treatment). Flag leaves were used for all sampling due to their importance in grain filling. At the heading stage (Feekes scale 10.1), gas exchange measurements were performed using a Li- 6400XT portable photosynthesis system equipped with a 6400-08 chamber attached to a 6400-02B LED light source (LI-COR Inc., Lin- coln, NE, USA). This corresponded to approximately 9 to 10 days after imposing stress treatments. Measurements were made on the flag leaf between 9:30 am and 11:30 am to evaluate changes in net photosynthesis (A, lmol m�2 sec�1) and stomatal conduc- tance (gs, mol H2O m�2 sec�1). During measurements, the leaf chamber temperature was kept at 24°C and 30°C depending on the greenhouse conditions. The remaining parameters were set as follows: Ca = 400 lmol CO2 mol�1, photosynthetically active radiation = 1000 lmol m�2 sec�1, airflow = 500 lmol sec�1, a rel- ative humidity of 55–65% and a vapor pressure deficit of 1.2 � 0.1 kPa. The order of the measurements was randomized across treatments, days and the measurement period. WUEi was later calculated as WUEi = A/gs. Four leaf discs were sampled using a paper punch from the same leaf used for gas exchange and oven-dried at 50°C for 72 h for recording the leaf mass per unit area (LMA, mg mm�2). At the ripening stage (Feekes scale 11), thermal images were taken between 11:30 am and 12:30 pm using a FLIR T-530 cam- era. Leaf temperature of the flag leaves was then determined with FLIR tools software (FLIR Systems, Inc. Wilsonville OR, Figure 7. Summary of transcription factors and pathway-specific gene families containing promising gene stacking candidates for the improvement of multi- stress tolerance in wheat. Signaling pathways shown for the identified gene candidates are summarized from the literature. Groupings for effective gene stack- ing could be made by reducing overlap of affected signaling pathways (transcription factors) and enhancing expression of pathway-specific genes depending on the dominant stress in a given region. Expression of candidate genes was affected under combinatorial stress treatments of heat, drought and salt in wheat flag leaves. Transcript IDs of candidate genes identified in this study have been provided and additional genes of interest could be identified using the supple- mental tables. � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1130 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense USA). The final plant height was measured on the same day. From plants set aside for metabolomic and transcriptomic sam- pling, flag leaves were collected at ripening and immediately flash-frozen in liquid nitrogen prior to storage at �80°C. At crop maturity, plants designated for physiological and yield measure- ments were harvested individually, threshed to collect the seeds and weighed. The sampled leaf discs and a 2-mg subset of ground seed were analyzed at the Agriculture and Agri-Food Canada (AAFC) Stable Isotope Facility in Lethbridge, Alberta. The tissues were combusted and analyzed by an online continuous flow dual ana- lyzer coupled to an isotope ratio mass spectrometer (Europa Sci- entific Integra, Cheshire, England, UK). Leaves and grain were analyzed for carbon content, nitrogen content and stable isotope ratios (d13C and d15N). All d13C values were converted to D13C using the Farquhar et al. (1989) equation with an isotopic compo- sition of the air to PeeDee Belemnite of �8.3&. Leaf C-to-N ratio and photosynthetic nitrogen-use efficiency (lmol CO2 g�1 N sec�1) were calculated from these values. Powdered leaf samples were then further analyzed at AGVISE laboratories (Northwood, ND, USA) for elemental analysis using inductively coupled plasma mass spectrometry. Amino acids were extracted from 10 mg of powdered freeze- dried tissue following the protocol described in Inaba et al. (1994) with some modifications. Briefly, 1.5 mL of 80% (v/v) ethanol solu- tion (40°C) was added to each sample, samples were shaken for 30 min at 40°C and the supernatant was recovered by centrifuga- tion (1290 g for 10 min) at 4°C. Amino acids were derivatized from the extracted samples using the Waters AccQTag Reagent Kit (Waters, Milford, Massachusetts, USA). The derivatized samples were subjected to high-performance liquid chromatography (HPLC) as described in the Waters AccQTag chemistry package instruction manual, with an excitation wavelength of 285 nm and an emission wavelength of 320 nm on a Waters Amino Acid Col- umn (3.9 9 150 mm) and a 2475 scanning fluorescence detector (Waters, Milford, Massachusetts, USA). The column was set at 37°C with an injection volume of 5 lL. Waters AccQTag buffer (100 mL AccQTag Buffer concentrate + 1000 mL deionized water), acetonitrile and deionized water were used as mobile phase A, mobile phase B and mobile phase C, respectively. Concentrations of each amino acid (pmol lL�1) in the samples were calculated using chromatogram peak area values against a calibration curve (10, 25, 50, 100 and 150 pmol lL�1) of known amino acid calibra- tion standards (WAT 088122, Waters, Milford, Massachusetts, USA) with a-aminobutyric acid as internal standard. The values were then converted to lmol mg�1 using the extraction volume and weight of the initial sample. Sugars and organic acids were extracted from 10 mg pow- der of freeze-dried leaf tissue samples. One milliliter of 75% (v/ v) methanol solution containing 0.1% formic acid was added to each sample and samples were mixed by vortexing for 10 sec, followed by sonication in a water bath at room temperature for 15 min. The supernatant was obtained by centrifugation (20 000 g for 15 min) at room temperature. The resulting supernatant was filtered through a 0.2-lm PVDF filter syringe onto HPLC slit vials (Waters, Milford, Massachusetts, USA) and stored at �20°C until use. An evaporative light scattering detector and a UPLC photodiode array detector were used for processing sugars and organic acids, respectively. These metabolites were identified and quantified by comparing with peaks of known standards. Plant hormones were extracted, purified and analyzed by liq- uid chromatography–electron spray ionization–tandem mass spec- trometry as described elsewhere (Preston et al., 2009). RNA isolation and sequencing Total RNA was extracted from the ground flag leaf samples using the Qiagen RNeasy plant RNA extraction kit (Qiagen, Hilden, DE) following the manufacturer’s recommendations. RNA integrity and quality were assessed using an Agilent 2100 Bioanalyzer sys- tem (Agilent Technologies Inc., Santa Clara, California, USA). RNA concentrations were estimated using the Qubit RNA HS assay kit (Thermo Fisher Scientific, Waltham, Massachusetts, USA) and Illu- mina library preparation was done using the TruSeq mRNA library preparation kit (Illumina, San Diego, California, USA). The result- ing libraries were multiplexed (12 samples per lane of a flow cell) and sequenced (paired-end sequencing with 125 cycles) using an Illumina HiSeq 2500 system (Illumina, San Diego, California, USA), generating a total of 1.013 billion paired reads (253 Gb; Table S8). Expression profiling Sequencing reads were filtered using Trimmomatic (version 0.37) (Bolger et al., 2014) by (i) removing reads when the average qual- ity per base dropped below 15 within a four-base wide sliding window, (ii) removing adapter sequences and (iii) trimming lead- ing and trailing low-quality (Q < 15) bases. Clean reads were aligned to the wheat Chinese Spring reference genome (IWGSC v1.0, EnsemblPlants) with RSEM (version 1.3.3) (Li & Dewey, 2011). The transcripts per million (TPM) and expected read counts gener- ated by the RSEM algorithm were used in downstream analyses. The TPM values were transformed by adding one and taking the natural logarithm to allow the use of statistical tests that assume a normal data distribution. WGCNA and differential RNA expression analyses Prior to network analysis, transcripts that were not expressed in three out of four replicates or had an expression value below 0.5 TPM were filtered from the dataset. Both high-and low-confidence transcripts were included. To further reduce the number of tran- scripts and to use transcripts with varying expression, we used analysis of variance (ANOVA) and an FDR of 5%. A final set of 14 676 DETs were used to create the co-expression networks using the R program-based package WGCNA (version 1.69) (Lang- felder & Horvath, 2008). Gene co-expression networks were gener- ated by defining the co-expression similarity using absolute values of Pearson correlation coefficients. A default b value of 12, as suggested by the authors of the WGCNA package, was used to penalize weak correlations and create the adjacency matrix. The default method of hierarchical clustering was then used to detect the clusters of interconnected transcripts known as modules, and transcripts within the modules were annotated using data from IWGSC v1.1 annotations. A threshold of 0.9 was applied to the MM value of annotated transcription factors to identify those which strongly influenced the intramodular network (hub genes). Module trait correlations were then used to identify modules of interest. In tandem, the list of 73 901 genes was used to identify tran- scripts that were differentially expressed in one or more of the stress treatments compared to the control. The package DESeq2 (Love et al., 2014) was employed and DETs were selected if they demonstrated a greater than 2-fold change in expression com- pared to the control and had an FDR below 5%. Only DETs were used for subsequent formation of gene lists, which were searched against the Ensembl Plants database (http://plants.ensembl.org/) for homologs in A. thaliana. The Arabidopsis homologs were used in the GO term analysis using Metascape (https://metascape.org/) � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1131 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense http://plants.ensembl.org/ https://metascape.org/ (Zhou et al., 2019) and pathway mapping using MapMan v 3.6.0 (https://mapman.gabipd.org/) (Usadel et al., 2009). Hub genes and transcripts in modules of interest lists were further filtered by removing those that were not considered to be differentially expressed. Trait data analysis R version 3.6.0 (R Core Team, 2019) along with packages emmeans (Lenth et al., 2020) and multcomp (Hothorn et al., 2008) were used to test for differences in the 40 measured phenotypes across stress treatments. ANOVAs were performed on linear models containing treatment as a fixed effect for each trait. The Tukey multiple comparison test was used to adjust P-values and deter- mine the compact letter displays for statistical groupings. AUTHOR CONTRIBUTIONS RS and SK conceived the study and managed the project. VB, RE, RS and SK performed the greenhouse study. SP and RS performed the metabolite study. EN performed hormone analysis. VB and SK performed the transcriptome study. LD, VB, SK and RS analyzed the data. NV, AP, EE and NP devel- oped the eFP Browser. LD and MK wrote the first draft of the paper. SK, RS and LD edited and finalized the manuscript. RS, SK and NP contributed the materials, reagents and web tools. All authors read and approved the manuscript. ACKNOWLEDGMENTS The authors thank Dr. Ron Knox (Swift Current, AAFC) and Janice Schmidt (NRC) for critical reading of the manuscript, Dr. Hamid Naeem, Seed Increase Unit (Indian Head, AAFC) for providing the Stettler seed, Dr. Branimir Gjetvaj (Saskatoon, AAFC) for taking photos, Krista Thompson (Saskatoon, AAFC) for assistance with metabolite profiling, Ayako Nambara (University of Toronto) for technical support on plant hormone analysis and Debbie Maizels (Zoobotanica) for the artwork in Figure 7. This work was sup- ported by Agriculture and Agri-Food Canada to RS and the National Research Council Canada through the Canadian Wheat Improvement Flagship Program to SK. Towards eFP Browser web support NP received funding from NSERC. COMPETING INTERESTS The authors declare no competing interests. SUPPORTING INFORMATION Additional Supporting Information may be found in the online ver- sion of this article. Data S1. Phenotypic (physiological, agronomic, hormonal and metabolic) changes in wheat under combinatorial abiotic stresses. Data S2. Expression levels (transcripts per kilobase million values) of wheat genes across different abiotic stresses and their combinations. Data S3. Expression levels (read count values) of wheat genes across different abiotic stresses and their combinations. Figure S1. Pathway mapping of the Arabidopsis orthologs to the 170, 186 and six differentially expressed wheat genes common to all salt, all heat and all drought treatments, respectively. Salt treat- ments include salt (S), heat and drought (HD), salt and drought (SD), salt and heat (SH) and salt, heat and drought (SHD). Heat treatments include heat (H), SH and SHD. Drought treatments include drought (D), SD, HD and SHD. All values are average expression across treatments relative to the control (C). Threshold expression values were a 2-fold change and a false discovery rate of 5%. Color scale represents relative expression in the units of log2(fold change) with positive values (red) indicating upregula- tion and negative values (blue) indicating downregulation. Graphs (a), (b) and (c) demonstrate the phenotypic effects of abiotic stress on wheat development with letters denoting significance group- ings. Graphs (d), (e) and (f) show salt accumulation and amino acid responses to the various stressors with letters denoting sig- nificance groupings. Asterisks denote differing levels of signifi- cance as follows when the treatment is compared to the control, P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). Figure S2. Pathway mapping of the Arabidopsis orthologs to the 290, 299, 734 and 1852 differentially expressed wheat genes unique to the heat and drought (HD), salt and drought (SD), salt and heat (SH) and salt, heat and drought (SHD) treatments, respectively. Not all genes were found to have orthologs in Arabi- dopsis; therefore, 112, 123, 259 and 720 wheat genes were unable to be mapped to a pathway in HD, SD, SH and SHD treatments, respectively. All values are average expression across treatments relative to the control. Threshold expression values were a 2-fold change and a false discovery rate of 5%. Color scale represents expression in the units of log2(fold change) with positive values (red) indicating upregulation and negative values (blue) indicating downregulation. Figure S3. Screenshot of the wheat abiotic stress eFP browser hosted by the University of Toronto. The figure above shows the transcriptional response of an ERF transcription factor in wheat flag leaves to single and combinatorial stresses. The ERF tran- scription factor was differentially expressed in all stress treat- ments compared to the control and was found to be a hub gene. Figure S4. Integrated signaling mechanisms at the cellular level under multiple abiotic stresses. Table S1. Wheat agronomic, physiological, hormonal and meta- bolic traits with hormone concentrations found in flag leaves as observed under the individual abiotic stresses and their combinations. Table S2. Rankings of stress treatments according to severity of adjustments required relative to the control. Table S3. Lists of DETs categorized according to the stress treat- ments in which they are differentially expressed. Table S4. Summary of network modules as determined by weighted gene co-expression analysis. Table S5. List of DETs from the five modules with significant cor- relations to a range of agronomic, physiological, hormonal and metabolic traits. Table S6. Transcript IDs, expression (in TPM) across stress treat- ments and annotations for the 30 genes assigned to a gene expression network module and found to be common across all stress treatments. Table S7. Transcript IDs, annotations, expression (TPM) and mod- ule color of the 155 transcripts identified as hub genes based on having a module membership (MM) value above 0.9. Table S8. Summary statistics for transcriptome sequencing of plants under various abiotic stresses and combinations thereof. OPEN RESEARCH BADGES This article has earned an Open Data badge for making publicly available the digitally-shareable data necessary to reproduce the � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1132 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://mapman.gabipd.org/ reported results. The data is available at https://bar.utoronto.ca/ efp_wheat/cgi-bin/efpWeb.cgi?dataSource=Wheat_Abiotic_Stress and https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE183007. DATA AVAILABILITY STATEMENT The raw phenotypic data (physiological, agronomic and metabolic traits) generated in this study are presented in Data S1 in the Supporting Information. The raw sequenc- ing data and processed transcriptome (TPM) data have been deposited into the National Center for Biotechnology Information Gene Expression Omnibus under accession number GSE183007 (https://www.ncbi.nlm.nih.gov/geo/ query/acc.cgi?acc=GSE183007). The TPM and read count data are also provided in Supporting Information (Data S2 and S3). The samples and/or replicates were paired between the different multi-omics experiments, including phenomics, metabolomics and transcriptomics. Expression data as TPM were also databased at the Bio-Analytic Resource for Plant Biology (Toufighi et al., 2005). An input image for the Wheat eFP Browser (Ramirez-Gonzalez et al., 2018) was generated and integrated into the tool. As part of the effort to create a Wheat Abiotic Stress view, the ability to query with both v1.0 and v1.1 wheat genome identifiers was enabled for the Wheat eFP Browser. REFERENCES Abhinandan, K., Skori, L., Stanic, M., Hickerson, N.M.N., Jamshed, M. & Samuel, M.A. (2018) Abiotic stress signaling in wheat – an inclusive over- view of hormonal interactions during abiotic stress responses in wheat. Frontiers in Plant Science, 9, 734. Available from: https://doi.org/10.3389/ fpls.2018.00734 Acevedo, E., Silva, P. & Silva, H. (2002) Wheat growth and physiology. In: Bread wheat: Improvement and production. Rome, Italy: Food and Agri- culture Organization of the United Nations. Plant Production and protec- tion series No. 30, p. 567. Available from: http://www.fao.org/3/y4011e06. htm#bm06 Agarwal, P., Baranwal, V.K. & Khurana, P. (2019) Genome-wide analysis of bZIP transcription factors in wheat and functional characterization of a TabZIP under abiotic stress. Scientific Reports, 9, 4608. Available from: https://doi.org/10.1038/s41598-019-40659-7 Ashe, P., Shaterian, H., Akhov, L., Kulkarni, M. & Selvaraj, G. (2017) Con- trasting root and photosynthesis traits in a large-acreage Canadian durum variety and its distant parent of algerian origin for assembling drought/heat tolerance attributes. Frontiers in Chemistry, 5, 121. Avail- able from: https://doi.org/10.3389/fchem.2017.00121 Ashraf, M. (2010) Inducing drought tolerance in plants: recent advances. Biotechnology Advances, 28(1), 169–183. Available from: https://doi.org/ 10.1016/j.biotechadv.2009.11.005 Asseng, S., Ewert, F., Martre, P., R€otter, R.P., Lobell, D.B., Cammarano, D. et al. (2015) Rising temperatures reduce global wheat production. Nature Climate Change, 5, 143–147. Available from: https://doi.org/10.1038/ nclimate2470 Batool, S., Uslu, V.V., Rajab, H., Ahmad, N., Waadt, R., Geiger, D. et al. (2018) Sulfate is incorporated into cysteine to trigger ABA production and stomatal closure. The Plant Cell, 30(12), 2973–2987. Available from: https://doi.org/10.1105/tpc.18.00612 Bheemanahalli, R., Sunoj, V.S.J., Saripalli, G., Prasad, P.V.V., Balyan, H.S., Gupta, P.K. et al. (2019) Quantifying the impact of heat stress on pollen germination, seed set, and grain filling in spring wheat. Crop Science, 59 (2), 684–696. Available from: https://doi.org/10.2135/cropsci2018.05.0292 Biswal, A.K. & Kohli, A. (2013) Cereal flag leaf adaptations for grain yield under drought: knowledge status and gaps. Molecular Breeding, 31, 749– 766. Available from: https://doi.org/10.1007/s11032-013-9847-7 Blake, N.K., Lanning, S.P., Martin, J.M., Sherman, J.D. & Talbert, L.E. (2007) Relationship of flag leaf characteristics to economically important traits in two spring wheat crosses. Crop Science, 47, 491–494. Available from: https://doi.org/10.2135/cropsci2006.05.0286 Bolger, A.M., Lohse, M. & Usadel, B. (2014) Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, 30(15), 2114–2120. Available from: https://doi.org/10.1093/bioinformatics/btu170 Cai, H., Tian, S., Liu, C. & Dong, H. (2011) Identification of a MYB3R gene involved in drought, salt and cold stress in wheat (Triticum aestivum L.). Gene, 485(2), 146–152. Available from: https://doi.org/10.1016/j.gene. 2011.06.026 Calabrese, E.J. & Mattson, M.P. (2017) How does hormesis impact biology, toxicology, and medicine? NPJ Aging and Mechanisms of Disease, 3, 1– 8. Available from: https://doi.org/10.1038/s41514-017-0013-z Cardoso, L., Freire, F. & Daloso, D. (2022) Plant metabolic networks under stress: a multi-species/stress condition meta-analysis. Journal of Soil Sci- ence and Plant Nutrition, 23, 4–21. Available from: https://doi.org/10. 1007/s42729-022-01032-2 Davenport, R.J. & Tester, M. (2000) A weakly voltage-dependent, nonselec- tive cation channel mediates toxic sodium influx in wheat. Plant Physiol- ogy, 122, 823–834. Available from: https://doi.org/10.1104/pp.122.3.823 DePauw, R.M., Knox, R.E., Clarke, F.R., Clarke, J.M. & McCaig, T.N. (2011) Stettler hard red spring wheat. Canadian Journal of Plant Science, 89, 945–951. Available from: https://doi.org/10.4141/CJPS08227 Dong, C.-J. & Liu, J.-Y. (2010) The Arabidopsis EAR-motif-containing protein RAP2.1 functions as an active transcriptional repressor to keep stress responses under tight control. BMC Plant Biology, 10, 47. Available from: https://doi.org/10.1186/1471-2229-10-47 Dong, W., Wang, M., Xu, F., Quan, T., Peng, K., Xiao, L. et al. (2013) Wheat oxophytodienoate reductase gene TaOPR1 confers salinity tolerance via enhancement of abscisic acid signaling and reactive oxygen species scavenging. Plant Physiology, 161, 1217–1228. Available from: https:// doi.org/10.1104/pp.112.211854 Elferjani, R. & Soolanayakanahally, S. (2018) Canola responses to drought, heat and combined stress: shared and specific effects on carbon assimi- lation, seed yield, and oil composition. Frontiers in Plant Science, 9, 1124. Available from: https://doi.org/10.3389/fpls.2018.01224 Farooq, M., Bramley, H., Palta, J.A. & Siddique, K.H.M. (2011) Heat stress in wheat during reproductive and grain-filling phases. Critical Reviews in Plant Sciences, 30, 491–507. Available from: https://doi.org/10.1080/ 07352689.2011.615687 Farooq, M., Nawaz, A., Chaudhry, M.M., Indrasti, R. & Rehman, A. (2017) Improving resistance against terminal drought in bread wheat by exoge- nous application of proline and gamma-aminobutyric acid. Journal of Agronomy and Crop Science, 203, 464–472. Available from: https://doi. org/10.1111/jac.12222 Farquhar, G., Ehleringer, J. & Hubick, K. (1989) Carbon isotope discrimina- tion and photosynthesis. Annual Review of Plant Biology, 40, 503–537. Fucile, G., Di Biase, D., Nahal, H. et al. (2011) ePlant and the 3D data display initiative: integrative systems biology on the world wide web. PLoS One, 6, e15237. Available from: https://doi.org/10.1371/journal.pone.0015237 Gao, Z., He, X., Zhao, B., Zhou, C., Liang, Y., Ge, R. et al. (2010) Overexpres- sing a putative aquaporin gene from wheat, TaNIP, enhances salt toler- ance in transgenic Arabidopsis. Plant and Cell Physiology, 51(5), 767– 775. Available from: https://doi.org/10.1093/pcp/pcq036 Genc, Y., Mcdonald, G.K. & Tester, M. (2007) Reassessment of tissue Na+ concentration as a criterion for salinity tolerance in bread wheat. Plant, Cell & Environment, 30(11), 1486–1498. Available from: https://doi.org/10. 1111/j.1365-3040.2007.01726.x Hayat, S., Hayat, Q., Alyemeni, M.N., Wani, A.S., Pichtel, J. & Ahmad, A. (2012) Role of proline under changing environments. Plant Signaling & Behavior, 7(11), 1456–1466. Available from: https://doi.org/10.4161/psb. 21949 He, G.H., Xu, J.Y., Wang, Y.X., Liu, J.M., Li, P.S., Chen, M. et al. (2016) Drought-responsive WRKY transcription factor genes TaWRKY1 and TaWRKY33 from wheat confer drought and/or heat resistance in Arabi- dopsis. BMC Plant Biology, 16, 116. Available from: https://doi.org/10. 1186/s12870-016-0806-4 Hoagland, D.R. & Arnon, D.J. (1950) The water culture method for growing plants without soil. California Agriculture Experimental Station Circular No., 347, 1–32. � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 Multi-omics atlas of abiotic stresses in wheat 1133 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://bar.utoronto.ca/efp_wheat/cgi-bin/efpWeb.cgi?dataSource=Wheat_Abiotic_Stress https://bar.utoronto.ca/efp_wheat/cgi-bin/efpWeb.cgi?dataSource=Wheat_Abiotic_Stress https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE183007 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE183007 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE183007 https://doi.org/10.3389/fpls.2018.00734 https://doi.org/10.3389/fpls.2018.00734 http://www.fao.org/3/y4011e06.htm#bm06 http://www.fao.org/3/y4011e06.htm#bm06 https://doi.org/10.1038/s41598-019-40659-7 https://doi.org/10.3389/fchem.2017.00121 https://doi.org/10.1016/j.biotechadv.2009.11.005 https://doi.org/10.1016/j.biotechadv.2009.11.005 https://doi.org/10.1038/nclimate2470 https://doi.org/10.1038/nclimate2470 https://doi.org/10.1105/tpc.18.00612 https://doi.org/10.2135/cropsci2018.05.0292 https://doi.org/10.1007/s11032-013-9847-7 https://doi.org/10.2135/cropsci2006.05.0286 https://doi.org/10.1093/bioinformatics/btu170 https://doi.org/10.1016/j.gene.2011.06.026 https://doi.org/10.1016/j.gene.2011.06.026 https://doi.org/10.1038/s41514-017-0013-z https://doi.org/10.1007/s42729-022-01032-2 https://doi.org/10.1007/s42729-022-01032-2 https://doi.org/10.1104/pp.122.3.823 https://doi.org/10.4141/CJPS08227 https://doi.org/10.1186/1471-2229-10-47 https://doi.org/10.1104/pp.112.211854 https://doi.org/10.1104/pp.112.211854 https://doi.org/10.3389/fpls.2018.01224 https://doi.org/10.1080/07352689.2011.615687 https://doi.org/10.1080/07352689.2011.615687 https://doi.org/10.1111/jac.12222 https://doi.org/10.1111/jac.12222 https://doi.org/10.1371/journal.pone.0015237 https://doi.org/10.1093/pcp/pcq036 https://doi.org/10.1111/j.1365-3040.2007.01726.x https://doi.org/10.1111/j.1365-3040.2007.01726.x https://doi.org/10.4161/psb.21949 https://doi.org/10.4161/psb.21949 https://doi.org/10.1186/s12870-016-0806-4 https://doi.org/10.1186/s12870-016-0806-4 Hothorn, T., Bretz, F. & Westfall, P. (2008) Simultaneous inference in gen- eral parametric models. Biometrical Journal, 50(3), 346–363. Available from: https://doi.org/10.1002/bimj.200810425 Inaba, K., Fujiwara, T., Hayashi, H., Chino, M., Komeda, Y. & Naito, S. (1994) Isolation of an Arabidopsis thaliana mutant, mto1 that overaccumulates soluble methionine (temporal and spatial patterns of soluble methionine accumulation). Plant Physiology, 104(3), 881–887. Available from: https:// doi.org/10.1104/pp.104.3.881 IWGSC, Appels, R., Eversole, K., Stein, N., Feuillet, C., Keller, B. et al. (2018) Shifting the limits in wheat research and breeding using a fully anno- tated reference genome. Science, 361, eaar7191. Available from: https:// doi.org/10.1126/science.aar7191 Joo, J., Choi, H.J., Lee, Y.H., Kim, Y.-K. & Song, S.I. (2013) A transcrip- tional repressor of the ERF family confers drought tolerance to rice and regulates genes preferentially located on chromosome 11. Planta, 238(1), 155–170. Available from: https://doi.org/10.1007/s00425-013- 1880-6 Kagale, S., Links, M.G. & Rozwadowski, K. (2010) Genome-wide analysis of ethylene-responsive element binding factor-associated amphiphilic repression motif-containing transcriptional regulators in Arabidopsis. Plant Physiology, 152(3), 1109–1134. Available from: https://doi.org/10. 1104/pp.109.151704 Kagale, S. & Rozwadowski, K. (2010) Small yet effective: the ethylene responsive element binding factor-associated amphiphilic repression (EAR) motif. Plant Signaling & Behavior, 5(6), 691–694. Available from: https://doi.org/10.4161/psb.5.6.11576 Kazan, K. (2006) Negative regulation of defence and stress genes by EAR- motif-containing repressors. Trends in Plant Science, 11(3), 109–112. Available from: https://doi.org/10.1016/j.tplants.2006.01.004 Khan, M.I., Iqbal, N., Masood, A., Per, T.S. & Khan, N.A. (2013) Salicylic acid alleviates adverse effects of heat stress on photosynthesis through changes in proline production and ethylene formation. Plant Signaling & Behavior, 8(11), e26374. Available from: https://doi.org/10.4161/psb. 26374 Khatkar, D. & Kuhad, M.S. (2000) Short-term salinity induced changes in two wheat cultivars at different growth stages. Biologia Plantarum, 43(4), 629–632. Available from: https://doi.org/10.1023/A:1002868519779 Kumar, R., Goswami, S., Singh, K., Dubey, K. et al. (2018) Characterization of novel heat-responsive transcription factor (TaHSFA6e) gene involved in regulation of heat shock proteins (HSPs) - a key member of heat stress-tolerance network of wheat. Journal of Biotechnology, 279, 1–12. Available from: https://doi.org/10.1016/j.jbiotec.2018.05.008 Kumar, R., Masthigowda, M.H., Kaur, A., Bhusal, N., Pandey, A., Kumar, S. et al. (2020) Identification and characterization of multiple abiotic stress tolerance genes in wheat. Molecular Biology Reports, 47, 8629–8643. Available from: https://doi.org/10.1007/s11033-020-05906-5 Kumar, S., Beena, A.S., Awana, M. & Singh, A. (2017) Physiological, bio- chemical, epigenetic and molecular analyses of wheat (Triticum aesti- vum) genotypes with contrasting salt tolerance. Frontiers in Plant Science, 8, 1151. Available from: https://doi.org/10.3389/fpls.2017.01151 Langfelder, P. & Horvath, S. (2008) WGCNA: an R package for weighted cor- relation network analysis. BMC Bioinformatics, 9(1), 559. Available from: https://doi.org/10.1186/1471-2105-9-559 Lenth, R., Singmann, H., Love, J., Buerkner, P. & Herve, M. (2020) Emmeans: estimated marginal means. R package version 1.4. 4. The American Statistician, 34(4), 216–221. Li, B. & Dewey, C.N. (2011) RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics, 12(1), 323. Available from: https://doi.org/10.1186/1471-2105-12-323 Li, B., Liu, D., Li, Q., Mao, X., Li, A., Wang, J. et al. (2016) Overexpression of wheat gene TaMOR improves root system architecture and grain yield in Oryza sativa. Journal of Experimental Botany, 67(14), 4155–4167. Avail- able from: https://doi.org/10.1093/jxb/erw193 Li, G.Z., Li, H.X., Xu, M.J., Wang, P.F., Xiao, X.H. & Kang, G.Z. (2020) Func- tional characterization and regulatory mechanism of wheat CPK34 kinase in response to drought stress. BMC Genomics, 21, 577. Available from: https://doi.org/10.1186/s12864-020-06985-1 Li, Q., Wang, W., Wang, W., Zhang, G., Liu, Y., Wang, Y. et al. (2018) Wheat F-box protein gene TaFBA1 is involved in plant tolerance to heat stress. Frontiers in Plant Science, 9, 521. Available from: https://doi.org/10.3389/ fpls.2018.00521 Li, Y. & Wei, K. (2020) Comparative functional genomics analysis of cyto- chrome P450 gene superfamily in wheat and maize. BMC Plant Biology, 20(1), 93. Available from: https://doi.org/10.1186/s12870-020-2288-7 Li, Z., Yu, J., Peng, Y. & Huang, B. (2016) Metabolic pathways regulated by abscisic acid, salicylic acid and c-aminobutyric acid in association with improved drought tolerance in creeping bentgrass (Agrostis stolonifera). Physiologia Plantarum, 159(1), 42–58. Available from: https://doi.org/10. 1111/ppl.12483 Liu, D., Liu, Y., Rao, J., Wang, G., Li, H., Ge, F. et al. (2013) Overexpression of the glutathione S-transferase gene from Pyrus pyrifolia fruit improves tolerance to abiotic stress in transgenic tobacco plants. Molecular Biol- ogy, 47(4), 515–523. Available from: https://doi.org/10.1134/S0026893313040109 Liu, Y., Mauve, C., Lamothe-Sibold, M., Gu�erard, F., Glab, N., Hodges, M. et al. (2019) Photorespiratory serine hydroxymethyltransferase 1 activity impacts abiotic stress tolerance and stomatal closure. Plant, Cell & Envi- ronment, 42(9), 2567–2583. Available from: https://doi.org/10.1111/pce. 13595 Love, M.I., Huber, W. & Anders, S. (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology, 15(12), 550. Available from: https://doi.org/10.1186/s13059-014-0550-8 Ma, X., Zhu, X., Li, C., Song, Y., Zhang, W., Xia, G. et al. (2015) Overexpres- sion of wheat NF-YA10 gene regulates the salinity stress response in Arabidopsis thaliana. Plant Physiology and Biochemistry, 86, 34–43. Available from: https://doi.org/10.1016/j.plaphy.2014.11.011 Mustafa, G., Akhtar, M.S. & Abdullah, R. (2019) Global Concern for Salinity on Various Agro-Ecosystems. In: Akhtar, M.S. (Ed.) Salt Stress, Microbes, and Plant Interactions: Causes and Solution, Vol. 1. Singapore: Springer, pp. 1–19. Available from: https://doi.org/10.1007/978-981-13-8801-9_1 Nanjo, T., Fujita, M., Seki, M., Kato, T., Tabata, S. & Shinozaki, K. (2003) Toxicity of free proline revealed in an Arabidopsis T-DNA-tagged mutant deficient in proline dehydrogenase. Plant and Cell Physiology 15, 44, 541–548. Niu, X., Luo, T., Zhao, H., Su, Y., Ji, W. & Li, H. (2020) Identification of wheat DREB genes and functional characterization of TaDREB3 in response to abiotic stresses. Gene, 740, 144514. Available from: https://doi.org/10. 1016/j.gene.2020.144514 Preston, J., Tatematsu, K., Kanno, Y., Hobo, T., Kimura, M., Jikumaru, Y. et al. (2009) Temporal expression patterns of hormone metabolism genes during imbibition of Arabidopsis thaliana seeds: a comparative study on dormant and non-dormant accessions. Plant Cell & Physiology, 50, 1786–1800. Qin, Y., Tian, Y. & Liu, X. (2015) A wheat salinity-induced WRKY transcrip- tion factor TaWRKY93 confers multiple abiotic stress tolerance in Arabi- dopsis thaliana. Biochemical and Biophysical Research Communications, 464(2), 428–433. Available from: https://doi.org/10.1016/j.bbrc.2015.06.128 R Core Team. (2019). R: A language and environment for statistical comput- ing. Vienna, Austria: R Foundation for Statistical Computing. https:// www.R-project.org/ Ramirez-Gonzalez, R.H., Borrill, P., Lang, D., Harrington, S.A., Brinton, J., Venturini, L. et al. (2018) The transcriptional landscape of polyploid wheat. Science, 361, eaar6089. Available from: https://doi.org/10.1126/ science.aar6089 Shaar-Moshe, L., Blumwald, E. & Peleg, Z. (2017) Unique physiological and transcriptional shifts under combinations of salinity, drought, and heat. Plant Physiology, 174(1), 421–434. Available from: https://doi.org/10.1104/ pp.17.00030 Shaar-Moshe, L., Hayouka, R., Roessner, U. & Peleg, Z. (2019) Phenotypic and metabolic plasticity shapes life-history strategies under combina- tions of abiotic stresses. Plant Direct, 3(1), e00113. Available from: https://doi.org/10.1002/pld3.113 Shailani, A., Joshi, R., Singla-Pareek, S.L. & Pareek, A. (2020) Stacking for future: pyramiding genes to improve drought and salinity tolerance in rice. Physiologia Plantarum, 172, 1352–1362. Available from: https://doi. org/10.1111/ppl.13270 Spiertz, J.H.J., Hamer, R.J., Xu, H., Primo-Martin, C., Don, C. & van der Put- ten, P.E.L. (2006) Heat stress in wheat (Triticum aestivum L.): effects on grain growth and quality traits. European Journal of Agronomy, 25, 89– 95. Available from: https://doi.org/10.1016/j.eja.2006.04.012 Suraj, H., SharathKumar, M. & van Kan, J. (2022) Too hot to defend: a tale of salicylic acid. Trends in Plant Science, 28, 4–6. Available from: https:// doi.org/10.1016/j.tplants.2022.10.001 � 2023 His Majesty the King in Right of Canada and The Authors. The Plant Journal published by Society for Experimental Biology and John Wiley & Sons Ltd. Reproduced with the permission of the Minister of Agriculture and Agri Food Canada., The Plant Journal, (2023), 116, 1118–1135 1134 Letitia Da Ros et al. 1365313x, 2023, 4, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/tpj.16332 by C anadian A griculture L ibrary, W iley O nline L ibrary on [22/04/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://doi.org/10.1002/bimj.200810425 https://doi.org/10.1104/pp.104.3.881 https://doi.org/10.1104/pp.104.3.881 https://doi.org/10.1126/science.aar7191 https://doi.org/10.1126/science.aar7191 https://doi.org/10.1007/s00425-013-1880-6 https://doi.org/10.1007/s00425-013-1880-6 https://doi.org/10.1104/pp.109.151704 https://doi.org/10.1104/pp.109.151704 https://doi.org/10.4161/psb.5.6.11576 https://doi.org/10.1016/j.tplants.2006.01.004 https://doi.org/10.4161/psb.26374 https://doi.org/10.4161