A list of all ribosomal protein family (both large and small subunit) was obtained using a keyword search, “ribosomal” in the RGAP-DB¹ and Phytozome v11² databases. From the list, a total of 56 ribosomal small subunit genes (40S) were shortlisted by their names starting with the letter “S” as a prefix (S for Small subunit). Nucleotide sequences of all the 56 genes were retrieved from RGAP-DB. These 56 sequences were further confirmed for their ribosomal origin through nucleotide and protein BLAST³ search in the NCBI⁴ and Hidden Markov Model (HMM) of Pfam⁵ databases, respectively. The availability of only 34 ribosomal small subunit proteins, which are coded by 56 genes indicated the existence of paralogs of the genes coding for more than one RPS protein. For the gene expression analysis, we selected 34 RPS genes, one from each group representing their corresponding paralogous groups. Identical gene copies were, thus, excluded from further analyses. We have listed the 34 genes along with their paralogs in the Supplementary Table 9. The sequences of RPS genes were retrieved from RGAP-DB (a Nipponbare database). Our analysis of gene expression is performed in an indica rice variety (Samba Mahsuri). Hence, we cross-checked the sequence similarity of the coding as well as the 5′- upstream regions of the genes selected for analysis in the present investigation in both indica and japonica sequence databases by doing a detailed BLAST analysis of the japonica sequences on indica genome databases, OryGeneDB⁶ and EnsemblPlants⁷. Further, we have looked for the similarity of rice (40S) RPS genes with RPS genes of Arabidopsis and performed a BLAST search of the Rice sequences in TAIR⁸ database and listed the similarity percentage in Table 1.

To obtain the genome-wide chromosomal distribution of all RPS genes on 12 chromosomes of rice, we had submitted the locus numbers (obtained from RGAP-DB) of 56 RPS genes as an input to the OrygenesDB. Based on the OrygenesDB outputs, all RPS gene locations were indicated on the chromosome maps, and the number of genes on each chromosome has also been shown. Genes, which were selected for expression analysis, have been shown in bold in Figure 1.

To predict the gene structure features, such as intron and exon number, size, etc., we had submitted the genomic and corresponding cDNA sequences of the selected 34 genes to the Gene Structure Display Server (GSDSv2)⁹.

Since all the 34 genes exhibited differential expression under multiple stress treatments, we checked for the presence of putative cis-elements in the putative promoter regions of the corresponding genes. To verify this, we have used an in silico approach, in which the nucleotide sequence of about ≤ 1 kb upstream of each of the 34 RPS genes was retrieved from RGAP-DB, and the sequences were cross-checked for their identity/ deviation with the indica genome in OrygeneDB and EnsemblPlants databases. Further, the sequences that exhibited the highest similarity with indica genome in the databases were submitted to PlantCARE (Cis-Acting Regulatory Elements) Database¹⁰ (Rombauts et al., 1999; Dhadi et al., 2009; Ding et al., 2011) for identifying the regulatory elements. We have compared the putative promoter regions of some of the significant genes in two other databases, such as PlantPAN 2.0¹¹ and the JASPAR¹² and found the functional similarity of the elements with respect to PlantCare database confirming the authenticity of our observations.

Predicted protein sequences available from the RGAP-DB were used for predicting the protein structure, molecular weight, isoelectric point (pI) prediction in silico by the online tool PSORT¹³ and ExPASyProtParam¹⁴. By submitting the protein sequences to SMART¹⁵ (Simple Modular Architecture Research Tool), we have identified various ribosomal domains and motifs. Sequence submission to ExPASyProtParam helped us in obtaining the GRAVY indices of RPSs, which indicated the probable hydrophobicity of the proteins by their negative values for the indices that are relevant for the hydrophobic proteins.

To predict the three-dimensional structure and to check for the presence of ligand binding sites on each RPS, we have used the Phyre2¹⁶ (Protein Homology/AnalogY Recognition Engine v2; Kelley et al., 2015) and 3DLigandSite¹⁷ programs (Wass et al., 2010), respectively. Results from these tools predicted the protein properties like percentage of α-helix, β-sheet, disordered protein fraction and ligands of each protein, ligand binding residues on the proteins and 3D model of the ligand bound protein. Multiple sequence alignment using ClustalW¹⁸ also deciphered the phylogenetic relationships among the 34 RPS proteins in MEGA6¹⁹ (Molecular Evolutionary Genetic Analysis) platform, which constructed an unrooted phylogenetic tree showing evolutionary sequence similarity among the proteins with bootstrap values of 100.

Oryza sativa L. ssp. indica var. Samba Mahsuri (BPT-5204) seeds were surface sterilized using 70% ethanol for 50 s followed by washing with 4% sodium hypochlorite for 15–20 min and 5–6 washes with sterile double-distilled water. After this, the seeds were dried on sterile blotting paper and were placed on solid MS media at 28 ± 2°C and 16 h light/8 h dark photoperiod cycle (Moin et al., 2016b).

For native expression studies, the surface sterilized seeds were allowed to germinate on solid MS medium. Some of the seeds were soaked for 16 h to collect the embryo and endosperm tissues using a sterile surgical blade. Shoot and root apical meristems of 3 d old seedlings and roots and shoots of 7 d old seedlings were collected for the expression studies. After transferring the seedlings from MS medium to soil in the pots; roots, shoots, leaves, root-shoot transition region, flowers, grains and panicles were collected (Moin et al., 2016b).

For abiotic stress treatments, 7 d old seedlings were exposed to ABA (100 μM), PEG (10%), NaCl (250 μM), and H₂O₂ (10 μM) by dipping them in the respective solutions. Around five seedlings in replicates were collected after 15 min, 3, 6, 12, 24, 48, and 60 h duration. Roots and shoots were collected separately for RNA isolation and cDNA synthesis. The untreated seedlings in water were used as control samples for normalization in qRT-PCR (Moin et al., 2016b).

To analyse the expression of the RPS genes under biotic stress, rice samples infected with Xanthomonas oryzae pv. oryzae (Xoo) and Rhizoctonia solani were collected 20 and 25 d post infection, respectively. For the biotic stress experiments, 1-month-old rice plants were used for treatments. The untreated samples of the same age were used as corresponding controls. The suspension of Xoo bacterium was applied at the edge of the leaves using a sterile blade (Moin et al., 2016b) and agar blocks of Rhizoctonia solani were placed on the leaf sheaths (Park et al., 2008) as per the standard protocols (and maintained under controlled culture room conditions.

All the samples were collected as three biological and technical replicates, and qRT-PCR was performed with their respective control samples. Rice specific act1 and β-tub genes were used as internal reference genes for normalization and the mean values of relative fold change were calculated as per ΔΔC_T method (Livak and Schmittgen, 2001). For native expression studies, the normalization was performed with only rice specific act1 and β-tub genes (ΔC_T). Bar diagrams and heat maps were generated by Sigmaplotv11 and Morpheus²⁰ programs, respectively using the means of the fold change that were calculated from the qRT-PCR data.

Total RNA was isolated from the treated and untreated samples using Tri-Reagent (Takara Bio, UK) according to the manufacturer's protocol. The concentration of the isolated RNA was checked using a Nanodrop spectrophotometer, and 2 ug of RNA was used to synthesize first strand cDNA using reverse transcriptase (Takara Bio, UK). The synthesized cDNA was diluted ten times, and 2 μl of it was used to perform qRT-PCR and 10 μM of the primers for each gene was used per reaction. The primers specific to RPS genes were designed by using the on line tool Primer-3²¹ the sequences of forward and reverse primers are listed in Supplementary Table 5. The amplification protocol included an initial denaturation at 94°C for 2 min, with 40 cycles of 94°C for 30 s, primer specific annealing temperature for 25 s, an extension of 72°C for 30 s. This was followed by constructing a melt curve at the end to estimate the amplification efficiency of each gene. The fold change was calculated as per ΔΔC_T method, and the mean of these values was considered as the final fold change of the transcript levels (Moin et al., 2016b). Means of relative fold change were plotted in Sigmaplot v11 with standard error to produce the bar diagrams for all the genes in a specific stress.

Rice genome has 12 chromosomes carrying 56 RPS genes. Based upon the OrygeneDB output, the RPS genes were found to be distributed throughout the rice genome covering all 12 chromosomes. Both the arms of chromosome randomly carried the RPS genes. Each chromosome carried at least one member of the RPS gene family. Most of the genes are located on chromosomes 3, 1, 7, and 11, with chromosome 3 bearing the highest number of eleven RPS genes. Both the chromosomes 1 and 7 carried a total of eight and nine genes, respectively. Chromosome 9 is the only example with only one RPS gene present near the telomere region of its long arm. The chromosome 10 and 1 carried three and eight RPS genes, respectively (Figure 1).

RPS genes, like other eukaryotic genes, showed the organization of introns, exons and UTRs. On submitting the nucleotide sequences of the 34 RPS genes in the Gene Structure Display Server (GSDS), we have obtained the predicted gene structure showing all exons, introns, UTRs for each gene and the number and relative position of each of the components. The BLAST analysis of RGAP-DB sequences with indica variety database showed that the coding regions of the 34 genes in both the varieties were similar. The percent similarity of the coding and the 5′-upstream regions of the genes found from the BLAST analysis has been listed in Supplementary Table 1. The similarity of RPS genes in rice with the corresponding RPS gene in Arabidopsis has also been searched with the help of BLAST analysis in TAIR database. All the 34 RPS genes studied in this report exhibited very high similarity with their counterparts in Arabidopsis also. The minimum similarity percentage is 78 and the maximum is 95.

Results from the GSDS suggested that most of the RPS genes consisted of both introns and exons (Supplementary Figure 1). RPS25a has a long stretch of intronic sequence covering most of its length. It is also the longest gene among the selected genes with a total length of approximately 6 kb. RPS17 and RPS30 lacked introns and possessed a single and small CDS flanked by UTRs. RPS6, RPS19, RPS23a, and RPS7 lacked UTRs and are only composed of intronic and exonic sequences. RPS23a is the smallest gene with a total size of 0.5 kb. The number of exons also varied from a minimum of two in RPS23a and RPS29 to a maximum of three exons in RPS3a. The details of properties related to the gene structure have been provided in Supplementary Table 2.

In accordance with the observations regarding the stress responsive differential expression and tissue specific expression of several RPS genes, we tried to correlate the transcript levels with the presence of stress-responsive elements in the 5′ upstream regions of the individual genes. Among the 5′ regions of the RPS genes, we have shortlisted some of the common stress-responsive elements that were present in single or multiple copies, and we have listed the elements that might have significance in tissue specificity of expression in Supplementary Table 7. Most of the genes have elements for meristem and endosperm specificity in their promoter regions.

All the RPS genes have at least two stress responsive elements in their immediate upstream regions (up to 1 kb). RPS30 has one copy of ABRE and TGACG, and RPS18b has one copy of LTR and TC rich repeat as their cis-acting regulatory elements (CAREs). Other than these two genes, putative promoter regions of the rest of the genes have more than two elements. SARE, EIRE, motifIIb, DRE and ERE, which are responsive to SA, elicitor, ABA, dehydration and ethylene, respectively and are present in single copies in the putative promoter regions of some of the genes. RPS23 showed five repeats of ABREs, three copies of TGACG motif, three copies of CGTCA motif and one copy each of TGA, Box W1, W box, and GARE elements. RPS4a exhibited three copies of ABREs, two repeats each of TGA, TGACG, and CGTCA elements. RPS15 has two repeats of MBS. Similarly, the genes (RPS4, 10a, 6a, and 5) that have exhibited remarkable up-regulation under biotic stress conditions have multiple MeJa responsive elements and some of them contain box W1 and TCA elements that are related to responsiveness under pathogen attack. All the stress responsive cis-acting elements were pictorially represented by their physical mapping in Figure 2 and Table 2 mentioning the function of the corresponding motifs. The elements having roles in tissue specificity for each of the genes have been listed in Supplementary Table 7. The BLAST analysis for the 5′-upstream regions of the genes of both the varieties showed that 8 out of 34 genes had shown a slight variation in the sequences of upstream regions (97%) in the two varieties. The differences in regulatory elements due to the dissimilarity in their sequences have been compared and shown in the Supplementary Figure 7. The upstream regions could not be identified for three genes (S15a, S18a, and S18b) for which the promoter analysis has been performed using the japonica sequences only.

The submission of the protein sequences from RGAP-DB to the ExPasyProtParam server revealed many putative properties. Their molecular weights (MW) varied from approximately 7 to 30 kDa with RPS4 having the highest MW of 29.8 kDa and RPS30 with the lowest MW of 6.9 kDa. In accordance with the MW, the RPS4 is the largest protein with a sequence of 265 amino acids, while RPS30 is the smallest with 62 amino acids. In silico analysis using ExPasyProtParam server helped us to get the theoretical pI-values of each of the RPS. Most of them were in the range of 9 to 12 with very few exceptions out of this range. For example, RPS6 has a very low pI of 4.86. Most of the RPS proteins are rich in positively charged amino acids like Arg and Lys compared to negatively charged amino acids like Asp or Glu. The only exceptions are the RPS21, which has an equal number of both the types and RPS6 having more of acidic amino acids. Percentage composition of secondary protein structures (disordered protein, alpha helix and beta sheet) has variations within the RPS family. RPS4 has the minimum value of 14% while the RPS30 has the maximum value of 76%, respectively; S10a exhibited 10%, and S29 and S13 have 64% α- helical content, respectively. The range on the percentage of β-sheet content also varied with RPS13, 13a having 0% and S27 showing 49% of the β-sheet content.

The Grand Average of Hydrophobicity (GRAVY) indices were below the value of zero for all 34 RPS proteins suggesting the hydrophobic nature of RPS proteins. We have also identified some of the interacting ligands of these RPS proteins from the 3DLigandSite. The metal and non-metal ligands include IF-1, FMN, FAD, Mg²⁺, Zn²⁺ etc. and the specific residues involved in the ligand binding have been listed in the Supplementary Table 3. Most of the RPS proteins have Low Complexity Region (LCR) motifs, which are the stretches of biased amino acid sequences having a long repetition of single or few amino acids and they have been connected to many protein functions and interactions. LCRs that are located terminally in a protein have been considered to be translationally enriched and have stress responsive roles (Coletta et al., 2010). Three dimension protein structures from 3D ligand site have been included in Supplementary Figure 2. The ribosomal protein sequences were submitted to SMART database, and it was found that all of them exhibited ribosomal domains. Some of the proteins also contain KOW domain (for e.g., RPS4, RPS4a). RPS6 has a SPT2 domain; RPS5a has a HLH domain, a CARD domain and an ANTAR domain. RPS18a has a SANT domain, which is a MYB- related domain that binds to DNA and helps in transcriptional regulation (Feller et al., 2011) and it also has a homeobox domain HALZ. All the domains associated with each of the RPS subunits are listed in Supplementary Table 8.

Protein sequence data was exploited to construct an unrooted phylogenetic tree of 34 RPS proteins in the MEGA6 software. Alignment with ClustalW employing a neighbor joining algorithm resulted in the generation of a tree that showed the evolutionary relationship between the RPS proteins. The tree suggested that a considerable number of cladogenesis events occurred during the course of evolution in the RPS gene family that was reflected in their protein sequence similarity and diversity. The earliest divergence separated the proteins into two major clades, of which one contains five proteins (RPS13, 13a, 3a, 19, and 21) and the other major clade carried all the other proteins, which has further diverged several times. These five RPS genes were checked for similarity with their paralogs and depicted as a phylogenetic tree in Supplementary Figure 3B. While RPS 19 and 21 have no paralogous members, RPS13 and RPS3a have two paralogous members, and they have shown similarity with their corresponding group. Nodes, from where some pairs of genes were diversified showed the bootstrap value of 100. Such high value indicates the high significance of the clade. The phylogenetic tree showing the evolutionary relationships among the RPSs is shown in Supplementary Figure 3.

To gain insights into the expression patterns of RPS genes in rice development and also to provide a comparative analysis between RPs of both the large and small subunits, we performed qRT PCR of 34 RPS genes in 13 different tissues as described in Moin et al. (2016a) for the RPL genes. The mean values were utilized to construct heat maps to demonstrate clearly the native expression pattern of the RPS genes (Figure 3). The tissues used to study the native expression pattern included embryo and endosperm from 16 h old germinating seed, plumule and radicle from 3 d old seedlings; shoot and root of 7 d old seedling; shoot, root, flag leaves, panicle, flower, grains and root-shoot transition of mature 20 d old plants (after transfer to green house).

The complete tissue-specific up-regulation pattern of the RPS genes in native conditions has been summarized in Supplementary Table 4 and also in the form of heat maps (Figure 3). The up-regulation of RPS genes was significant in mature leaves (25 genes), shoots (21 genes), 7 d shoots (26 genes), plumules (21 genes), and root-shoot transition (23 genes). Apart from these four, seven and 15 genes were up-regulated in seedling radicles, 7 d old root and mature roots, respectively. An analogous pattern of gene expression was also observed under different stress conditions as well, where more number of RPS genes was found to be up-regulated in shoot than in the root tissues showing that the majority of the RPS genes were up-regulated in shoots, leaves and plumules compared to the other organs.

About 17 RPS genes showed expression in rice grains and panicles indicating their involvement in inflorescence development and grain filling. None of the genes was expressed in florets implying that RPS proteins have probably no significant roles in the development of this meiotically active tissue. About 18 and 15 RPS genes were up-regulated in embryo and endosperm, respectively showing that these genes might have a role in seed development and maturity. RPS23a, RPS27a, RPS10a, and RPS7 have exhibited no significant up-regulation in any of the tissues and RPS23 is up-regulated only in root-shoot transition. A considerable up-regulation of RPS28 and RPS29 was observed in roots under different stresses. RPS4, RPS5a, RPS7a, RPS9, RPS9-2, RPS13, RPS6, RPS17, RPS18b, RPS24, and RPS29 were up-regulated in most of the tissues, while RPS13a, RPS18a, RPS4 were up-regulated in shoots throughout the duration of all the stress conditions. Similarly, these genes were also expressive in shoots, leaves, plumules and in native conditions as well.

Our findings on the screening of a large-pool of activation tagged mutant population for high water-use efficiency revealed the significant up-regulation of RPL genes (RPL6 and RPL23A) by the integrated enhancers in two of the mutant plants with sustained productivity under limited water conditions (Moin et al., 2016a). In the follow-up study, an extensive tissue specific and differential expression study of the entire RPL gene family was carried out, which showed significant up-regulation of certain RPL genes under abiotic and biotic stresses (Moin et al., 2016b). Results from these two studies prompted us to further investigate the expression of 40S Ribosomal Protein genes in response to abiotic and biotic stress conditions and to arrive at proper combinations of RPS and RPL genes for subsequent genetic manipulation. For studying their differential expression, we have selected 34 RPS candidate genes from the respective paralogous groups in response to four abiotic stresses (ABA, PEG, NaCl, and H₂O₂) at seven time points and two biotic stresses (BLB and SB). ABA, PEG, NaCl, and H₂O₂ induce a stress similar to osmotic, drought, salinity and oxidative stress, respectively (Gao et al., 2007; Hamayun et al., 2010; Hossain et al., 2015; Sah et al., 2016). Treatment of the plants with these stress-inducing agents would mimic the corresponding stress situations and consequently activate numerous stress responsive pathways. Also, we also analyzed their native expression pattern in different tissues.

The genes that exhibited ≥ 3 fold transcript level on the log₂ scale were considered as up-regulated. The pattern of expression of the genes depended mostly on the type of stress signal and also the tissue in which the expression was studied. We have categorized the genes in terms of their timing and the intensity of response. Some of the genes were up-regulated as early as 15 min-3 h after the onset of the stress, while a few of them responded between 3 and 12 h after exposure while others were up-regulated 12 h after the application of the stress. We have differentiated them as Immediate Early (IE), Early (E), and Late (L) expressing genes, respectively.

Since the level of expression varied from 3 to several 100 folds, we have further categorized the up-regulated genes as highly expressive genes (≥30 fold), moderately expressive (between 10 and 30 fold) and low-expressive (between 3 and 10 fold). The expression of the RPS genes in the shoot samples has been observed to be higher compared to the corresponding root samples under almost all the abiotic stresses at different time intervals. Under some conditions, the up-regulation of the genes in the shoot samples has been detected to be several 100 folds, which is not the case for the corresponding root samples. In the case of shoot samples, all the genes showed significant up-regulation for all the stress conditions for at least one time point, and they were predominantly IE type in their response, except under H₂O₂ treatment.

Although all the genes were up-regulated under all the stress conditions in shoots, their mode of response varied with the nature of the stress. For example, ABA induced the up-regulation of 26% of the genes (RPS7, 15a, 18a, 25a, 4, 7a, 10, 10a, and 13a) and these maintained a significant expression (≥3-fold) until the last time point used (60 h). Of these, 50% were highly expressive (RPS7, 15a, S8a, and 25a) with >30-fold transcript level throughout the stress treatment up to 60 h time point. RPS7 and RPS18a showed up-regulation up to several 100 folds. PEG and NaCl also induced the up-regulation of a majority of genes in shoots until 60 h time point, except RPS7a and RPS23 and RPS5, RPS10a, and RPS23a under PEG and NaCl treatments, respectively, which became down-regulated.

Some of the RPS genes maintained remarkably high transcript levels (≥100 fold). These include RPS4a, RPS9, RPS9-2, RPS13, RPS17, RPS19, RPS27, and RPS30 under PEG treatment and RPS4a, RPS7, RPS9-2, RPS18a, and RPS23a under NaCl treatment. The up-regulation of RPS4, RPS7, RPS9, RPS10, RPS19, and RPS26 under salt treatment is in agreement with the earlier studies (Kawasaki et al., 2001). Unlike in ABA, PEG, and NaCl treatments, several RPS genes were late responding in shoots under H₂O₂ treatment. Moreover, the level of up-regulation was also not as high as in other treatments. About 60% of the genes were up-regulated 15 min post treatment. About 30 genes have been found to maintain a considerable level of expression, and among these, RPS4a, RPS9-2, RPS15a, RPS18, and RPS28 have exhibited ≥100 fold up-regulation. In roots, RPS19 was down-regulated persistently in response to all the abiotic stresses at all the time intervals. RPS10a, RPS18, RPS18a, RPS18b, RPS24, RPS27a, and RPS6a were also observed to maintain a very low transcript level. Some of the genes that were initially up-regulated, subsequently, became down-regulated with the progression of the stress treatment. The expression pattern of all the genes under different stress treatments in the shoot and root tissues has been represented in the form of heat maps (Figures 4A,B).

There is an overlap in the expression of the genes that are up or down-regulated in response to the stresses at several time points, and this has been demonstrated in the form of Venn diagrams (Supplementary Figures 4, 5). Table 3 presents a detailed list of genes that were up/down-regulated in different stress conditions in the shoot and root tissues separately. The percentage of genes over-expressed at certain time point under the stress conditions in shoot and root has been represented in the form of Pie charts (Figures 5A,B), and the genes are tabulated in the tables that accompanied the figures. The genes that were commonly up-regulated under all the treatments, particularly during the early response may have significance in stress related signaling (Kilian et al., 2007) as the early responsive genes play a crucial role in defense against the environmental stresses (Mahajan and Tuteja, 2005; Ouyang et al., 2007).

Apart from the four abiotic stresses, the expression level of 34 RPS genes was also studied in response to the treatments with Xoo and R. solani pathogens that cause very serious Bacterial Leaf Blight (BLB) and Sheath Blight (SB) diseases, respectively in rice. In the BLB infected samples, 14 (40%) genes were down-regulated, and the rest of them were up-regulated (60%). RPS6a (26 fold), RPS9 (18 fold), RPS10a (29 fold), and RPS4 (13 fold) showed a high level of transcript upregulation when compared with the other up-regulated genes.

The qRT-PCR analysis of 34 RPS genes was performed from both the shoot and the leaf samples infected with R. solani. The up-regulation of genes was observed particularly in the shoot tissues for a majority of the RPS genes. Thirteen (37%) genes were down-regulated in response to this pathogen in the shoot region whereas the 22 RPS (63%) genes were up-regulated. Among the genes that were expressive, RPS4 and RPS5 have exhibited 22 and 14 fold up-regulation, respectively, which is the highest compared with the other genes. The expression pattern of all the genes under two biotic stresses has been depicted graphically in Figures 6A,B.

Previously available information about the RPL genes (Moin et al., 2016b) has been utilized and subsequently combined with the information obtained from the current study on the expression patterns of RPS genes. The details regarding their stress responsiveness and tissue-specific expression pattern under native condition have been summarized in tabular form to select suitable combinations of RPS and RPL genes for stress alleviation in transgenic plants. All the RPL and RPS genes up-regulated in response to four individual stress treatments were depicted in Supplementary Table 6. Highly up-regulated RPL and RPS genes in each of the 13 different tissues under native conditions have been represented in Figure 7. Four heat maps combining the expression of RPL and RPS genes under four abiotic stresses have been represented separately (Figures 8 A–D) and Supplementary Figure 6 shows the comparative expression analysis of both RPL and RPS genes under 13 different tissues in the form of a heat map.

One of the intriguing observations from this comparative analysis is that majority of the RPL genes were up-regulated in both shoots and roots of the plant under native conditions. But under all abiotic stress conditions, the maximum up-regulation was observed primarily in roots than in shoots. On the contrary, under the native conditions, the expression of RPS genes was mostly confined to shoots and the leaves compared to roots. Taken together, the up-regulation of the RPS genes and RPL genes was predominantly noticed in shoots and roots, respectively at all the time intervals under all the stress conditions.

According to the previous studies, RPL23A, L19.3, and L38 have shown to exhibit a high level of expression under ABA, PEG and NaCl treatments. RPL6, 12, 23A, 18A, and 13A have maintained high up-regulation under H₂O₂ treatment (Moin et al., 2016b). In the present study, RPS13a, 18a, 4, and 4a have shown significant up-regulation in almost all the stresses in shoots, while RPS29 and RPS28 were highly expressive in roots. In the case of RPL genes, NaCl caused down-regulation of a majority of genes when compared to ABA and PEG. On the contrary, almost all the RPS genes showed up-regulation under salt stress in shoots. The H₂O₂ treatment caused down-regulation of 60 and 80% of RPL genes in shoots and roots, respectively when compared with MeJA, SA and cold treatments, which induced up-regulation of many of them. Most of the RPS genes were up-regulated under all the stresses in shoots, whereas in roots, 8, 22, 5, and 28% of the RPS genes were down-regulated under ABA, PEG, NaCl, and H₂O₂ treatments, respectively. The H₂O₂ treatment induced the down-regulation of many of them. RPL30, 44, 22, 14, 29, 13A, and 15 were down-regulated under ABA, PEG, and NaCl treatments and RPS27a, 10, 20, 6, and 9 were down-regulated in response to all the treatments (ABA, PEG, NaCl, and H₂O₂). RPL10 was significantly activated upon infection with Xoo pathogen, whereas among the RPS genes, RPS4 became commonly up-regulated under both Xoo and R. solani treatments.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.