Genes, mechanisms and novel EST-SSR markers associated with metribuzin tolerance in wheat (Triticum aestivum L.): targets for improving photosynthetic capacity and yield

Weed infestation is one of the major yield-reducing factors in wheat in dry-land farming. Metribuzin is a broad-spectrum herbicide which allow effective weed management in wheat but the narrow safety margin results in crop damage decreasing grain yield. Improving our understanding of the genetic and genomic basis for metribuzin tolerance opens the potential to enhance herbicide tolerance and better productivity in wheat. The present investigation examines the genes involved in regulation of metribuzin tolerance including genetic/signalling pathways, transcription factors, phytohormones, and gene based EST-SSR markers related to photosynthesis and metabolic detoxication. Transcriptome sequencing of most diverse genotypes using high throughput NovaSeq 6000 RNA-Seq platform identied a total of 77,443 genes, of which 59,915 were known genes and 17,528 were novel genes. The integrative analyses of the expression proles of genes and pathways at 0 h, 24 h and 60 h herbicide exposure indicated that modulation of reactive oxygen species (ROS) homeostasis and endogenous increase of light-harvesting chlorophyll (Lhc) a/b-binding proteins, PSII stability factor HCF136, metabolic detoxication enzymes (peroxidase, cytochrome P450, glycosyltransferase, glutathione transferase, oxidoreductase), and glucose metabolism conferred metribuzin tolerance in wheat. The validation of DEGs related to photosynthesis (Lhc a/b-binding proteins and PSII stability factor HCF136) and metabolic enzymes (cytochrome P450, peroxidase) using RT-qPCR conrmed their responsiveness to metribuzin. Over-expression of transcription factors MYB, AP2-EREBP, ABI3VP1, bHLH, and NAC played a signicant roles in regulating photosynthetic and ROS scavenging activities during metribuzin stress. Transcripts with signicant enrichments (q-value < 0.05), related to photosynthesis and metabolic detoxication revealed 114 EST-SSRs which may be used as bio-markers. of the data suggests that high amount of sugars, modulation of ROS homeostasis and enhanced photosynthetic activity play a signicant role in regulating metribuzin tolerance in wheat. Our data identied master regulators controlling metribuzin tolerance that provide promising avenues for wheat industry.


Background
Wheat (Triticum aestivum L.) is a major global cereal crop in terms of production and area (FAO 2018) and is a staple food for 40% of the global population due to its wide adaptability and easy harvestability [1]. Wheat grows well between the latitudes of 30° and 60° N and 27° and 40° S [2]. The Australian Wheat Promotion Program (2011) proved an important milestone for systematic wheat research and real breakthroughs in productivity but many constraints remain related to dryland farming yields.
Weed infestation is a global problem, more so in Mediterranean-type climatic regions where wheat and weeds actively grow throughout the cropping season. More than 60% of the 20 million ha of arable lands in Australia has a typical Mediterranean-type climate that is characterised by cool, wet winters and hot dry summers [3,4]. High weed pressure affects tillering in wheat and thereby reduce grain yield. There are instances where weed infestations have reduced wheat yields up to 50% [5]. Weed infestation is estimated to cost $1.3 billion, equivalent to around 20% of the gross value of the Australian wheat crop.
Metribuzin (C 8 H 14 N 4 OS), a triazine herbicide (group C), is a broad-spectrum herbicide, registered for controlling a range of monocot (grass) and dicot (broad-leafed) weeds, including the most problematic annual ryegrass. Metribuzin is widely used in dryland farming systems in Australia and elsewhere. Preemergent application controls weeds during the early stages of growth, between radicle emergence from the seed and seedling leaf emergence through the soil. In particular, pre-emergent application of metribuzin reduced barley and brome grass by over 80% with good wheat yields in Western Australia [6].
Also, the residual activity of metribuzin controls the rst few ushes of germinating weeds when the crop is too small to compete and protects the crop from early weed competition. However, the narrow safety margin of metribuzin and lack of selectivity in wheat results in yield losses, limiting its wider use. Apart from its weed control e cacy, tolerance of wheat crops to this herbicide is equally important for maximum crop production.
Herbicide tolerance is an important agronomic trait that allows effective weed management in dryland farming. It is increasingly di cult to discover a new herbicide with a novel mode of action. Expanding the utility of existing broad-spectrum herbicides with a good environmental pro le through genetically enhanced herbicide tolerance is a useful strategy for effective weed management. There is unprecedented scope for developing metribuzin-tolerant wheat cultivars through molecular breeding due to the wide genetic variability for metribuzin tolerance in wheat [7]. Key genes, mechanisms and functional markers involved in herbicide tolerance can be explored using transcriptome analyses for marker-assisted selection (MAS) to develop herbicide-tolerant crops.
The present study aimed to identify (1) the most contrasting genotypes to metribuzin tolerance by assessing the dose-response relationships of nine potential genotypes in a detailed metribuzin doseresponse experiment and con rming these genotypes in eld conditions, (2) key genes, pathways, and mechanism associated with metribuzin tolerance in the most-tolerant and most-susceptible wheat genotypes using transcriptomic approach (Illumina NovaSeq 6000 platform), and (3) EST-SSR markers for MAS and breeding.

Results
Comprehensive screening to identify the most-tolerant and most-susceptible genotypes The dose-response parameters (ED 50 , LD 50 and GR 50 ) differed signi cantly in the tolerant and susceptible genotypes (Table 1). Chuan Mai 25 consistently had the highest values across ED 50 , LD 50 and GR 50 and highest fold tolerance (R:S ratio) while Ritchie consistently had the lowest values, amongst the nine genotypes (Table 1). Chuan Mai 25 (CM) survived 1717 g a.i. ha − 1 metribuzin dose whereas Ritchie (R) did not survive the eld-recommended rates of 200 g a.i. ha − 1 . The eld data con rmed these ndings ( Fig. 1). Chuan Mai 25 had the lowest senescence and highest survival rate, whereas, Ritchie had the highest senescence and lowest survival rate among all genotypes (Additional le 1: Table S1). This con rms that Chuan Mai 25 and Ritchie are the most-tolerant and most-susceptible genotypes, respectively.  Table 3. After complete transcriptome assembly, an average total of 77,443 genes was identi ed, of which 59,915 were known genes and 17,528 were novel genes. These genes were compared with the NCBI non-redundant protein for functional annotation.  and 60 h HE) revealed commonly and differentially expressed genes (HT-C-vs-HS-C, HT-24 h-vs-HS-24 h and HT-60 h-vs-HS-60 h) (Fig. 2). Most of the DEGs in the inter-groups are related to metribuzin tolerance; the distribution trends of differentially and non-differentially expressed genes, with up-regulated and down-regulated gene numbers in the control and treatments are presented in volcano plots (Additional le 2: Figure S1). The comparison revealed the sum total of commonly and differentially expressed genes: 24,108 in the control (Fig. 2, Additional le 1: Table S2), 24,783 at 24 h HE (Fig. 2, Additional le 1: Table  S3) and 23,664 at 60 h HE (Fig. 2, Additional le 1: Table S4). A comparison of HT-24 h-vs-HS-24 h and HT-60 h-vs-HS-60 h revealed a total of 7,736 and 7,460 unique DEGs (Fig. 2), respectively, related to herbicide tolerance in wheat. These DEGs were integrated to investigate the signi cantly enriched genes and pathways involved in the herbicide tolerance mechanism.

Transcription Factors And Pathway Analysis Of Degs
A total of 59 transcription factors (TFs) associated with 7,227 DEGs from complete transcriptome were identi ed (Additional le 2: Figure S2, Additional le 1: signi cant enrichments occurred for carbon metabolism, fructose and mannose metabolism, homologous recombination, biosynthesis of amino acids, plant-pathogen interaction (R genes), pyrimidine metabolism, galactose metabolism and amino sugar and nucleotide sugar metabolism. The genes regulated around these pathways were early genes generated in response to metribuzin stress. Increased exposure to metribuzin, 60 h HE, caused over-expression of photosynthetic enzymes, ROS scavengers, and phytohormones (glutathione and ascorbic acid) (Fig. 3b).

Hub Genes Related To Metribuzin Tolerance In Wheat
A total of 107 photosynthesis-related genes were differentially regulated (logFC ≥ 5.0, q-value < 0.05) under metribuzin stress, including genes related to PSI, PSII, light harvesting chlorophyll protein complex (Lhc) a/b-binding proteins, PSII stability factor HCF136, PSII oxygen-evolving complex (Additional le 1: Table S6). There is an active participation of Lhc a/b-binding proteins in response to metribuzin stress (Fig. 4). Lhc a-binding protein and Lhc b-binding protein is embedded in thylakoid membrane of PSI and PSII in chloroplast, respectively, which primarily collect and transfers light energy to photosynthetic reaction centres. In this study, most of the Lhcs' were down-regulated in susceptible Ritchie but upregulated in tolerant Chuan Mai 25 under metribuzin stress with 60 h HE. The over-expression of Lhca1, Lhca4, Lhcb1-6 in the tolerant Chuan Mai 25 conferred metribuzin tolerance. Enzymatic and nonenzymatic components were synthesised in response to metribuzin stress. The enzymatic components comprise several antioxidant enzymes (Additional le 1: Table S7) involved in alleviating oxidative stress.

Validation of differential expressed genes by RT-qPCR
The fold-changes logFC of gene expression obtained from RNA-seq analysis and RT-qPCR largely corresponded (Table 4) and correlated (r = 0.84) with each other. The two genotypes, Chuan Mai 25 (tolerant) and Ritchie (susceptible), used in transcriptome sequencing responded differently to metribuzin. The RT-qPCR con rmed the involvement of genes related to photosynthesis and metabolic detoxi cation in metribuzin tolerance. For example, the DEG TraesCS5D01G323800 and TraesCS1D01G096400 involved in metabolic detoxi cation of herbicide had 3.8 fold increase in expression of cytochrome P450 and peroxidase, respectively, in Chuan Mai 25 than Ritchie. The DEG TraesCS7B01G486500 had 3.1 fold increase in expression of PSII stability/assembly factor HCF136 in Chuan Mai 25 than Ritchie. The DEG TraesCS7D01G276300 expressing Lhc a/b-binding proteins in chloroplast had 3.7 fold increased expression in Chuan Mai 25 than Ritchie (Table 4).

Discussion
We focused on unravelling gene networks, mechanisms and pathways associated with metribuzin tolerance in hexaploid wheat using a unique top-to-bottom three tiered strategy. In the rst tier, metribuzin effects were investigated in 946 wheat germplasms (Australian winter wheat collection) from different regions of the world [7]. Our metribuzin tolerance screening identi ed promising contrasting genotypes. Identi cation of the most contrasting genotypes is a pre-requisite for better resolution and deeper insight into genes and mechanisms involved in herbicide tolerance. Therefore, in the second tier, a detailed doseresponse experiment and eld screening were conducted using potential contrasting genotypes to identify the most contrasting genotypes. Chuan Mai 25 and Ritchie were the most contrasting genotypes for metribuzin tolerance when compared with the present known sources. Discovery of the most contrasting genotypes lays a strong foundation for genetic and genomic studies to assist in the development of herbicide-tolerant cultivars with a wide safety margin (Fig. 5). The third tier focused on transcriptome sequencing of Chuan Mai 25 and Ritchie using the Illumina NovaSeq6000 platform. The DEGs identi ed gene networks, pathways/metabolic enzymes and mechanism(s) contributing to metribuzin tolerance in wheat.

Mechanism(s) for metribuzin tolerance in wheat
Metribuzin stress limits CO 2 xation and over-reduction of the electron transport chain resulting in ROS [8]. Herbicides generate an abiotic stress that produces ROS, such as O 2 · − , H 2 O 2 , 1 O 2 , OH·, which are extremely toxic and trigger membrane lipid peroxidation and rapid destruction of cellular constituents, resulting in oxidative stress and cell injury or death [9]. Metribuzin is a potent PSII inhibitor. It binds the target site D1 protein in PSII and inhibits electron ow between the primary electron acceptor and plastoquinone. This leads to selective and speci c cleavage of the D1 protein. The D1 protein turnover cause the breakdown of PSII, reducing photosynthetic electron transport chain, which produce superoxide radicals and singlet oxygen in the chloroplasts [10][11][12]. This limits the generation of the energy currencies of cells, ATP and NADPH, inhibiting CO 2 xation in the Calvin cycle.
The present study suggests that metribuzin tolerance in wheat is metabolism-based (Fig. 6). Two major metabolic pathways-glycolysis and pentose phosphate pathway-are over-regulated in response to early metribuzin stress in tolerant wheat. The co-ordinated interplay between these metabolic pathways increases-the in ux of energy (ATP), reducing powers [reduced nicotinamide adenine dinucleotide (NADH), NADPH and avin adenine dinucleotide (FADH 2 )], and intermediates for biosynthetic and metabolic detoxi cation processes [13,14] (Additional le 1: Table S7) are essential for supporting the antioxidant system and preventing oxidative damage to DNA, proteins and lipids [15].
Early genes regulated in response to metribuzin stress (24 h HE) (Fig. 6) belong to carbon metabolism, fructose and mannose metabolism, homologous recombination, amino acid biosynthesis, pyrimidine metabolism, galactose metabolism and amino sugar and nucleotide sugar metabolism. Metabolites such as fructose and mannose are synthesised to protect membranes and proteins from oxidative stress by ROS. Genes involved in homologous recombination are signi cantly enriched to repair harmful breaks in DNA and restore the essential molecular function in cells [16,17]. Galactose is involved in glucose synthesis, and pyrimidines serves the role of ATP for glucose synthesis (Zrenner et al., 2006) [18], promoting nutrient remobilisation and preventing senescence.
Increased exposure to metribuzin (60 h) caused over-expression of photosynthetic and metabolic enzymes, antenna proteins (Lhc a/b-binding proteins), PSII stability/assembly factor HCF136, and glutathione/ascorbic acid (Fig. 6). Photosynthetic enzymes and antenna proteins are involved in carbon xation and glucose synthesis catalysed by Rubisco (Additional le 1: Table S6). Glutathione metabolism removes free radicals and prevents oxidative damage to DNA, proteins and lipids. Ascorbic acid (antioxidant) functions as a cofactor for enzymes in photosynthesis, and the synthesis of plant hormones [19] and affects gene expression and transcription, cell division, and growth [20].

Enzymatic and non-enzymatic components for ROS detoxi cation
The DEG analysis suggested that metribuzin tolerance is wheat is metabolism-based involving overexpression of several ROS-scavenging enzymes such as superoxide dismutase, catalase, glutathione Stransferase (GSTs), glutathione peroxidase, cytochrome P450 (CYPs), cytochrome reductase, cytochrome peroxidase, oxidoreductase, ABC transporters, glycosyltransferase (GT), UDP-galactosyltransferase and ubiquitin transferase to prevent oxidative stress during herbicide stress in the tolerant wheat genotype, Chuan Mai 25. Some of the herbicide is detoxi ed before it reaches target site. CYPs add a reactive group such as hydroxyl, carboxyl, or an amino group through oxidation to herbicide molecule, making it a polar molecule (phase I detoxi cation) and transferases (phase II detoxi cation enzymes) conjugates the addition of water-soluble group to the reactive site of polar molecule. The identi ed gene superfamilies or domains are essentially xenobiotic detoxifying enzymes involved in vacuolar sequestration of conjugated herbicide metabolites. The non-enzymatic components/phytohormones such as ascorbic acid and glutathione (GSH) have ROS scavenging function and plays a protective role during metribuzin stress.
GSH function with GSTs to detoxify herbicides by tagging electrophilic compounds for removal during oxidative stress [21,22].
Overexpression of ROS-responsive regulatory genes (Additional le 1: Table S6, S7), which regulate a large set of genes involved in acclimation mechanisms, is a powerful strategy for enhancing herbicide tolerance in wheat. The ability of wheat genotypes to metabolize herbicides are largely dependent on the genetic expression of these enzymes. Difference in metribuzin tolerance expression is a result of genetic polymorphisms resulting in an altered expression. This is con rmed by SNP discovery in metribuzintolerant and -susceptible wheat groups using 90K iSelect SNP genotyping assay. The polymorphic SNP loci between the two groups detected genes on chromosomes (2A, 2D, 3B, 4A, 4B, 7A, 7B, 7D) encoding metabolic detoxi cation enzymes (cytochrome P450, glutathione S-transferase, glycosyltransferase, ATPbinding cassette transporters and glutathione peroxidase) [23]. We have mapped QTLs for metribuzin tolerance in wheat. The genes underlying the QTL support range on chromosomes-1AS (oxidoreductase), 2DS (glycosyltransferase), 4AL (transferase activity) are involved in metabolic detoxi cation. The integration of present transcriptomic analyses, previous metribuzin-tolerant QTL mapping [6], and SNP discovery using 90K iSelect SNP genotyping assay in metribuzin-tolerant and -susceptible wheat genotypes [7] suggests that enzymatic components play a signi cant role in modulating ROS homeostasis and the acclimation response of wheat to metribuzin tolerance.
Over-Expression of Lhc a/b-binding proteins and PSII stability/assembly factor HCF136 confers metribuzin tolerance in wheat PSII functions as a water-plastoquinone oxidoreductase in oxygenic photosynthesis. The redox components, required for PSII function are localised on the heterodimer of the Dl and D2 proteins of the PSII reaction centre (Fig. 4). Lhc a/b-binding proteins are typically complexed with chlorophyll and xanthophylls and serve as the antenna complex, which regulate the distribution of excitation energy between PSII and PSI [24]. Regulation of Lhc a/b-binding proteins is an important mechanism in plants to modulate chloroplast functions [25,26]. This study suggests that over-expression of Lhc a/b binding proteins in metribuzin tolerant wheat (Chuan Mai 25), promotes carbon xation and modulates ROS homeostasis during metribuzin stress.
HCF136-the thylakoid-embedded large pigment-protein complexes of the photosynthetic electron transfer chain-is involved in the assembly of PSII reaction centre complexes, de novo synthesis of the D1 protein and the selective replacement of damaged D1 protein during PSII repair [27]. Lower expression of HCF136 in susceptible Ritchie during metribuzin stress resulted in the accumulation of damaged PSII proteins, which increased oxidative stress. Photosynthesis cease when degradation and PSII repair do not balance under herbicide stress. This implies that in susceptible wheat, a reduction in fundamental processes such as photosynthesis produce oxidative stress in chloroplast, which extends beyond PSII to cause a down-regulation of total carbon gain and imbalance between the rate of photo-damage to PSII and the rate of the repair of damaged PSII, reducing plant yield in susceptible genotypes. 90K iSelect SNP genotyping assay in our previous investigation detected polymorphism between tolerant and susceptible wheat genotypes in the gene encoding PSII assembly factor involved in PSII repair [23]. This suggests that metribuzin-tolerant wheat genotypes have inherently high photosynthetic e ciency.  [28][29][30], AP2/EREBP [31], WRKY [32,33], and bHLH families play important roles in plant responses to abiotic and biotic stresses [34].

Herbicide-tolerant wheats
The EST-based SSR markers identi ed in signi cantly enriched genes relating to photosynthetic and metabolic detoxi cation enzymes with present-absent variation (PAV), with signi cant differential expression will be a great resource for metribuzin tolerance breeding. The PAV is a sequence in one genome, but entirely missing in another genome. This is an important source of genetic diversity in plants [35,36]. We propose the use of functional speci c markers for a desired traits which reduces genotypephenotype gaps in crop plants to maximize genetic gains in breeding. High-throughput identi cation of PAV on a whole-genome level has become possible with the advent of next-generation sequencing (NGS) technologies, at affordable prices [37]. There is a rapidly rising trends in the application of genome editing based crop improvement using CRISPR/Cas genome engineering system [38,39]. The improved understanding of genetic and genomic knowledge of herbicide tolerance will open up the utilities for inducing multiple cleavage events, controlling gene expression, and site speci c transgene insertion.
In conclusion, the use of improved metribuzin-tolerant wheats will help farmers to (1) minimise the early cohorts of problematic weeds, removing early wheat and weed competition and increasing wheat productivity, and (2) promote crop rotations with other herbicide-tolerant crops, such as narrow-leafed lupin (Lupinus angustifolius L.) and canola (Brassica napus L.) to assist in sustainable farming systems.

Methods
Herbicide Metribuzin (C 8 H 14 N 4 OS), a triazinone herbicide was purchased from Syngenta Crop Protection.
Metribuzin binds its target site D1 protein in PSII and inhibits electron ow between the primary electron acceptor to plastoquinone, arresting photosynthesis. The metribuzin dose of 400 g a.i. ha -1 was used to create stimulus/stress in tolerant and susceptible genotype.

Plant material
Contrasting wheat genotypes, six herbicide-tolerant (HT) and three herbicide-susceptible (HS) ( Table 1) identi ed amongst 946 wheat genotypes from six continents (Australian winter cereals collection) [7] and Western Australian local cultivars [40] -were selected for the detailed dose-response experiment and eld screening. The two most-contrasting genotypes, Chuan Mai 25 (HT) (origin: China; type: spring) and Ritchie (HS) (origin: England; type: spring) were used for transcriptome sequencing. Chuan Mai 25 is an advanced cultivar, with a very good disease resistance, which was released in 1995.
Detailed dose-response and eld screening A detailed dose-response experiment with pre-emergent application was conducted in glasshouse to determine the most contrasting genotypes and dose-response parameters such as ED 50 value (rate of application for 50% reduction in visual senescence/chlorosis), LD 50 value (rate of application to kill 50% of plants) and GR 50 value (rate of application for 50% growth reduction). Seeds were sown in 10 cm pots and sprayed with eight metribuzin rates (0, 100, 200, 400, 800, 1600, 3200 and 6400 g a.i. ha -1 ) via a twin at-fan nozzle perpendicular to the direction of sowing in two passes at 200 kPa in a cabinet spray chamber calibrated to deliver 118 L water ha -1 . The trial was carried out in 2018 in a UWA glasshouse, Australia. The trial comprised nine rows by three columns for each metribuzin rate. Each genotype × herbicide treatment was replicated three times with seven plants per replicate. Plants were watered every 48 h to ensure that moisture was non-limiting. At 21 days after treatment (DAT), senescence/chlorosis was rated using a scale of 0 (no senescence/phytotoxicity) to 10 (100% senescence/dead). Percentage survival was determined by scoring as 'dead' or 'alive' for all the rates. The above-ground biomass was harvested and expressed as a percentage of the mean untreated control. Dose-response analyses were carried out using R extension package 'drc'. Senescence was tted to the three-parameter log-logistic function 'LL.3' where the lower limit is equal to 0 to determine ER 50 . Survival and biomass were tted to the four-parameter log-logistic function 'LL.4' to determine LD 50  and survival rate were recorded as described above. Dot plots for eld screening data were generated using the 'ggplot2' library in R.

Tissue collection and RNA isolation
The two most-contrasting wheat genotypes, Chuan Mai 25 (HT) and Ritchie (HS) were analysed for transcriptome sequencing. Seeds were surface sterilised in 3% NaClO, for 10 min and washed three times followed by 24 h imbibition in double-distilled water. The seedlings were propagated in 1 L pots containing river sand in a growth chamber at 25°C/15°C (day/night) with a 16 h photoperiod at a light intensity of 800±200 µE m -2 s -1 and 8 h dark and relative humidity (55%). Twelve days after sowing fully-grown seedling (3-5 leaf stage) in three replicates were uniformly sprayed with metribuzin at 400 g ai ha -1 using a twin at-fan nozzle as described above. Young leaf tissue was harvested aseptically from the control and treatments after 24 h and 60 h herbicide exposure (HE) and frozen immediately in liquid nitrogen and stored at −80°C for RNA isolation. Total RNA was isolated using an RNA plant mini kit  [43] and the coding ability of novel transcripts was predicted by using CPC (v0.9-r2 available at http://cpc.cbi.pku.edu.cn) [44]. The novel coding transcripts were merged with reference transcripts to obtain the complete reference before mapping the clean reads to the reference genome using Bowtie2 [45]. Subsequently, novel genes, SNP and INDELs were detected.

Gene expression analysis
The gene expression level for each sample was calculated using a RSEM software (v1.2.12 available at http://deweylab.biostat.wisc.edu/RSEM) [46]. The relative transcript abundance in the different treatment groups was obtained using the FPKM method (fragments per kilobase of transcript per million mapped reads) [47]. FPKM = (1000000*C)/(N*L/1000), C represents the amount of fragment mapped to the speci c transcripts, N represents the amount of fragment mapped to any transcripts and L represents the base amount of the speci c transcripts. Differentially expressed genes (DEGs) are detected based on the poisson distribution of DEGseq [48]. To delineate herbicide resistance mechanism(s) in the wheat transcriptome, DEGs were grouped into three combinations, namely, HT-C-vs-HS-C, HT-24h-vs-HS-24h and New EST-SSR markers for metribuzin tolerance in wheat A universal web-tool, PolyMorphPredict (http://webtom.cabgrid.res.in/polypred/) [51] was used to identify SSRs contributing to metribuzin tolerance in wheat. Gene transcripts related to photosynthesis (PSI and PSII), ROS scavengers, phytohormones and metabolic detoxi cation, with PAV or higher-fold expression (logFC ≥ 5.0, q-value < 0.05) (Additional le 1: Table S6, S7), were used to survey potential SSRs. The mono-, di-, tri-, tetra-, penta-, and hexa-nucleotides were designed with minimum repeat numbers of 10, 6, 5, 5, 5, and 5 for the SSRs, respectively.
Validation of RNA-seq analysis and DEGs related to photosynthesis/antenna proteins, and metabolic enzymes using RT-qPCR To con rm the RNA-seq results, 20 DEGs were randomly selected and assessed using RT-qPCR. RT-qPCR primers were designed using Primer 3 software [52,53]. First-strand cDNA synthesis was done using  Table 4.
The expression pattern of signi cantly enriched DEGs under metribuzin treatment (60 h HE), were assessed. We mainly focused on DEGs related to photosynthesis and metabolic enzymes: TraesCS7B01G486500 (PSII stability factor HCF136), TraesCS7D01G276300 (Lhc a/b-binding proteins), TraesCS5D01G238300 (Lhc a/b-binding proteins), TraesCS5D01G323800 (cytochrome P450), TraesCS7B01G441100 (cytochrome P450) and TraesCS1D01G096400 (peroxidase). manuscript; and KHMS, PS and GY revised the paper. All authors discussed the results, read and approved the nal manuscript for publication.