Introduction

Rice is one of the most stable dietary staples for people all over the world. It plays a critical role in the country's food security. Nonetheless, rice is subjected to a variety of adverse conditions. Water deficit is the main threat for rice cultivation causes tremendous reduction in its yield productivity1,2,3,4. Limited water resources and the overgrowing population hinder the efforts done for rice crop genetic improvement5. However, the approach of developing and breeding new high yielding resilient rice genotypes considered the most efficient way to compensate yield losses under harsh environments1,2,3,6.

Water deficit environmental stress limits crop productivity and distribution worldwide. Availability and quality of water resources together with the increased demand exacerbate the effects of water scarcity events7. Incidences of aggressive water deficit are expected to occur during the near future as the global climate change is accelerating causing alterations in the rainfall patterns and distribution8,9. Water deficit impairs various physiological, biochemical and molecular processes in plants. Cells osmotic pressure plays an important role in conserving the plant water potential during water deficiency to maintain continuous growth10,11. However, extreme water deficiency decreases the turgor pressure which impairs cell growth and decreases photosynthesis rate12.

In response to water stress, rice plants accumulate several osmo-protectant that lower the maintain cell turgor. These osmo-protectants such as proline, glycine betaine and soluble sugar act to achieve osmotic adaption13,14. Reactive oxygen species (ROS) are by-products of many metabolic pathways such as respiration, photosynthesis, and photorespiration, that are occur in different organelles15. Under normal conditions there is equilibrium between the production and scavenging of the ROS. Water deficit stress can disrupt this balance and lead to excessive production of ROS, plant cells are protected against the harmful effects of ROS through its antioxidant defence system to maintain ROS levels and homeostasis. Tolerance to drought stress in higher plants is associated with a strong levels of antioxidant systems thus relatively higher ability to scavenge the ROS resulted from water stress-induced oxidative stress. Superoxide dismutase (SOD) and catalase (CAT) are major antioxidant enzymes that play important role in ROS homeostasis16,17. Furthermore, the resistance to water deficit includes signal transduction pathways to activate the expression of stress-responsive genes (REF)18. Among those genes, the genes that are involved in the production of osmo-protectants, antioxidant enzymes, water channel proteins, transcriptional factors and other genes involved in water stress adaption. Transcriptional factors play a crucial role in regulating of stress responsive genes19.

Efforts have been done to increase rice productivity by developing resilient high yielding rice varieties adapted to water stress20. Recent advancements in molecular analysis facilitated the identification of quantitative trait loci (QTLs), that have major effects, contributing to various agronomic traits related to water stress tolerance in rice21,22. Most of these QTLs are linked to grain yield under water shortage conditions. Expressional profiling of the candidate genes is a powerful tool to investigate the level that the genes are contributed to stress elevation. This could be done by conducting deferential gene expression for the tolerant and susceptible genotypes23. The characterization of the candidate genes responsive to water stress tolerance could provide insights about water stress resistance mechanisms in the pre-breeding lines. These candidate genes could be used to develop MAS systems for the incorporation of these genes form the pre-breeding lines in molecular breeding programs24.The SSR markers linked to these QTLs have been successfully used in mapping and introgression of such QTLs into elite rice genotypes via marker assisted selection approaches1,21,25,26,27,28. Recently, we have developed high yielding novel rice genotypes harboring QTLs linked to grain yield derived from a rTGMS line1. Water stress pre-breeding lines are intermediate germplasm that minimize the barrier in introgressing novel genetic variation into breeding programs. Breeders can use those lines to improve cultivated varieties for water deficit resistance. The advantage behind the pre-breeding lines is to strictly capture the desirable genetic background from the genetic stock by obtaining the minimum linkage drag. Accordingly, several water deficit pre-breeding lines were developed using different breeding approaches. Seven pre-breeding lines were selected in the current investigation to conduct analysis of their yield under field condition. We further estimated their physiological, biochemical, and molecular response at seedling stage in the screen house to compare the performance of these lines as compared with their parents.

Materials and methods

Plant material

The current investigation involved twelve rice genotypes; seven pre-breeding lines and their 5 parents. The seven pre-breeding lines are derived from the cross between the upland genotypes IR60080-46A, IRAT170 and the drought tolerant; WAB881SG9 with the Egyptian elite lowland cultivars Giza177 and Giza178, Supplementary Table 1.

Field evaluation of the pre-breeding lines under well and water deficit irrigation condition

Phenotypic performance of the pre-breeding lines under water stress field conditions

The pre-breeding rice genotypes and their parents were subjected to well-watered and water shortage stress conditions, during the two successive rice summer growing seasons of 2017 and 2018. The plant material was laid out following the transplanting cultivation method in randomized complete block design (RCBD) with three replications for each condition for the two seasons under study. Each plot was five one-meter-long rows, 20 × 20 cm spacing for each genotype at both well- and stressed- watered conditions. In the well-irrigated condition, a sufficient submerged depth (around 5 cm) was maintained during all the growth stages after transplanting. While, the water stress condition was induced through applying one flush irrigation every 12 days without keeping standing water after irrigation. Other cultural practices were followed in the experimental field as recommended by the national rice research program. According to rice standard evaluation system29 phenotypic performance of the plant material was evaluated under both conditions by measuring; days to heading (days), plant height (cm), panicle length (cm), number of tillers per plant (tiller), relative amount of Chlorophyll content (SPAD), number of panicles per plant (panicle), sterility (%), 1000-seed weight (g), grain yield (g plant−1) and harvest index. The drought susceptibility index (DSI) was determined as DSI = (NS − S)/NS for each genotype where NS is the grain yield under normal conditions and S is the grain yield under drought conditions.

Climate and soil properties of the experimental field conditions

The field investigation was conducted at the experimental farm of Rice Research and Training Center, Sakha, Kafr El-Sheikh, Egypt (30° 57′ 12″ North latitude, 31° 07′ 19″ East longitude). Rice crop is cultivated during the summer where the high temperature with no rain are the main climatic features. The minimum temperature is 12–19 °C, the maximum temperature is 23–41 °C, relative humidity is 40–60%, solar radiation is 40–65% and wind speed are 1.8–6 ms−1. The experimental sites were located in the old Nile delta with clayey soil (18.6% sand, 22.4 silt and 59% clay).

Genotyping via SSR marker linked to QTLs for traits contributing to water shortage resistance

Genomic DNA were extracted from the fresh leaves of the new developed seven lines and their corresponding parents using the cetyltrimethylammonium Bromide (CTAB) method30. The quality and the quantity of the isolated DNA were tested using NanoDrop (Thermo scientific, Wilmington, DE), and the DNA of all samples was approximately adjusted to concentration of 20 ng µl−1. Thirteen simple sequence repeats primers (SSR) linked to QTLs contributing to water stress tolerance in rice were selected and used in the present study for molecular characterization and genetic diversity. All the SSR primers were provided by Sigma Aldrich Company, Germany. The nucleotide sequence, annealing temperature and the chromosome number of the 13 SSR primers used in this investigation are presented in Supplementary Table 2. The PCR amplification was carried out in a volume of 15 µl reaction mixture using iNtRON's Master mix Solution, i-TaqTM (Intron Biotechnology, South Korea). The mixture contained 2 µl of 15 ng µl−1 genomic DNA, 1 µl of each of forward and reverse primers (10 pmol), 7.5 µl of 2X Master mix solution and the final volume was made up to 15 µl with double distilled water. The components were mixed thoroughly and then placed in a thermal cycler (Biometra, Göttingen, Germany) following this profile: initial denaturation step at 94 °C for 3 min, followed by 35 cycles of the following steps, denaturation at 94 °C for 30 s., annealing (according to the primer) for 20 s. and extension at 72 °C for 30 s., a final extension step of 7 min at 72 °C was given. The amplified PCR products were separated by electrophoresis in 3% high resolution agarose gel (Sigma Aldrich, USA), stained with ethidium bromide. The gels were visualized and photographed using a gel documentation system (BioDocAnalyze, Biometra). The molecular size of the amplified bands for the different SSR markers was calculated based on the migration of band relative to the standard molecular size of DNA marker (Bio-Helix, 100 bp DNA ladder).

Growth chamber early stage water deficit stress condition

Growth chamber conditions for the physio-biochemical and gene expression profiling

Seeds of the twelve rice genotypes were sterilized and germinated on a wet paper in Petri dishes and incubated at 26–27 °C for five days. The germinated seeds were transplanted to 13D pots (0.88 l), (East Riding Horticulture, York, UK) containing water-saturated soil consisting of 70% Kettering Loam soil (Supplied by Boughton, UK), 23% Vitax John Innes No. 3 (Leicester, UK), 5% silica sand and 2% Osmocote Extract Standard 5–6 month slow-release fertilizer (ICL, Ipswich, UK). Plants were grown and maintained in controlled-environment growth chambers (Conviron Controlled Environments Ltd, Winnipeg, MB, Canada) following RCBD experimental design. The growth conditions were: 13 h 30 °C: 11 h 24 °C light:dark cycle, PAR 1000 μmol m−2 s−1 and 60% relative humidity. The plants in the pots were constantly based in water, and water supply to the soil from the top once a week unless otherwise specified. Accordingly, the plants were grown in water-flooded trays for up to 29 days. The stress treatment was applied by draining the water and stopping watering the plants, while the control treated plants remained under flooded condition.

Estimation of water stress physio-biochemical responsive biomarkers

Leaves of the 8-days of water withholding plants were used to estimate the physio-biochemical biomarkers. This period was enough for mild water deficit symptoms to appear on the sensitive genotype, Giza177. The proline content was estimated according to Bates et al.31. Super oxidase activity (SOD) using the method described by Zhang32. Lipid peroxidation (MDA) was measured using the thiobarbituric acid (TBA) test as described by Velikova et al.33. Total soluble sugars (TSS) were quantified according to Zhang et al.34, and the relative water content was measured using the method described by Barrs and Weatherly35.

Expression profiling of water stress related genes in the pre-breeding lines

Based on the field performance under water deficit stress and its physio-biochmical response, the RBL112 pre-breeding line was selected to analyze the expression profile of the water stress related genes, Supplementary Table 3. Its parents, IR60080-46A and Giza 178, were added for comparison along with the high yielding IR64 rice variety (IRRI, Phillipine) which was used as a highly sensitive control. Total RNA was isolated from the leaves of the well-watered and water-stressed plants grown in the growth chamber after 5-days from the stress. The extraction was done using a Quick-RNA™ MiniPrep kit Zymo Research, Irvine, USA) according to the manufacturer’s instructions. After cDNA amplification, the 10 μL reaction was diluted with 60 μl of nuclease free water. A 5 μL of the cDNA was used in a 20 μL qRT-PCR reaction using the SYPR Green, JumpStart™ Taq Ready Mix (Sigma-Aldrich, Poole, UK, Cat.#S5193), assay using a CFX Connect Real-Time PCR Detection System (Bio-Rad, Watford, UK). Transcript sequences of the drought-related genes investigated were obtained from the Rice Genome Annotation Project database (http://rice.uga.edu/) then used to design gene-specific primers Supplementary Table 3.

Statistical analysis

The statistical analysis of the phenotypic characteristics was carried out using analysis of variance by means of ‘MSTAT’’ computer software package, and means were compared based on the least significant difference (LSD) test at the 5% and 1% probability levels. Heatmap of the genes relative expression levels was done using gplots package36 statistical software R (R Core Team37) version 4.1.1. Relative expression levels of the target genes in the different samples were calculated from UBC1 normalized target signals using the 2−ΔΔCT method38.

Research ethics

Experimental research and field studies on plants (either cultivated or wild), including the collection of plant material, complied with relevant institutional, national, and international guidelines and legislation. A prior approval was undertaken from Offices of Research, Innovation and Commercialization, Princess Nourah bint Abdulrahman University, Riyadh 11671, Saudi Arabia.

Results

Phenotypic performance of the pre-breeding lines and their corresponding parents under well and stress-irrigated conditions

Growth performance and characteristics of the pre-breeding lines and their parents

In both field experimental conditions, analysis of variance revealed the existence of highly significant differences among the pre-breeding lines and their parents, (Supplementary Tables 4 and 5 for well-watered and stress conditions, respectively). The plant height and number of tillers, in the genotypes records were reduced under water shortage stress, Fig. 1 and Supplementary Table 5). While for the days to heading, some genotypes have lower values under water stress such as Giza177 and Giza 178, IR60080-46A and RBL44 as compared to their values under well-irrigated condition. In contrary, there was a delay in the heading of IRAT170, WABSG9, RBL112, RBL35 and RBL133 under water stress, compared to the well irrigated conditions.

Figure 1
figure 1

Growth characteristics of the pre-breeding lines and their parents under water shortage stress (orange columns) as compare with normal irrigated conditions (blue columns).

Full size image

Short-stature rice genotypes are preferable for rice breeders, Giza178 and RBL133 were the shortest under well irrigated conditions. However, under stress conditions RBL35 and RBL133 were the shortest, with no significant difference with Giza177. Interestingly, RBL 35 exhibited the highest reduction (24.42%) in the plant height due to water stress. At the same time, RBL 9 values for number of tillers were the best under both conditions under study. The number of tillers of pre-breeding line RBL10 was affected by water shortage with about 31.45%, while the line RBL112 was the least affected genotype (3.86%). On the days to heading, RBL9 was the best pre-breeding line under both conditions, as early maturing varieties have fewer water requirements.

The pre-breeding lines and their parents yield related characteristics response under water shortage stress

Panicle number per plant and panicle length significantly decreased due to water stress treatment as compared to the well-irrigated conditions (Supplementary Table 6, Fig. 2a,b). However, the reduction in the number of reproductive tillers in the genotype RBL112 was the lowest similarly to the genotype IR60080-46A. At the same time, RBL9 values under both conditions were the best compared to other genotypes under study. While Giza177 has the lowest number of reproductive tillers under water stress. Among the pre-breeding lines and parents, RBL44 has the longest panicles under both conditions. The Egyptian popular cultivars, Giza177 and Giza178 have the lowest values for panicle length under water-stress condition. Furthermore, water shortage stress increased the sterility percentage of all the genotypes. RBL10 sterility percentage value was the highest under normal irrigated condition, while RBL133 and Giza177 records were the highest under stress condition (Fig. 2a,b). RBL112 has the lowest sterility percentage values under water stress treatment.

Figure 2
figure 2

Yield related parameters of the pre-breeding lines and their parents under normal irrigated conditions (a) and water shortage stress (b).

Full size image

Grain yield performance and drought susceptibility of the pre-breeding lines under field conditions

The effect of water deficit stress on yield was evident, as there was a reduction in the overall GY mean under water stress of 32.9% as compared with the well-irrigated condition. Giza 177 records showed that the cultivar was the most sensitive to water stress conditions. Its GY was the lowest under stress (22.7 g plant−1) (Fig. 3, Supplementary Table 6) and the DSI value was the highest (65.9%). Under water deficit stress, the RBL112 GY value was significantly higher than others under water stress conditions, with the lowest DSI value (20.8%) among the pre-breeding lines. However, the lowest DSI value was recorded by the parent genotype IR6006480-A (10.3%) but the yield was not comparable to the RBL112. RBL35 showed the highest 1000-seed weight under both well- and stress-irrigated conditions (36 and 31 g, respectively), Supplementary Table 6. All the pre-breeding lines have DSI values lower than the popular Egyptian cultivars. Furthermore, the harvest index values of the pre-breeding line, RBL112 were the highest under both conditions (44.5 and 38.2%, respectively), Supplementary Table 6. While, Giza177, RBL35 and RBL133 have the lowest harvest index values under water shortage stress.

Figure 3
figure 3

Grain yield (g plant−1) performance of the tested genotypes under normal irrigated conditions (blue columns) and water shortage stress condition (orange columns) and DSI estimates (black dots).

Full size image

Physio-biochemical response of the pre-breeding lines under growth chamber conditions

To assess the effect of water stress on the physiological and biochemical behavior of the pre-breeding lines, we assessed the proline content, SOD, TSS, MDA and RWC (Fig. 4 and Supplementary Table 7). ANOVA revealed that the differences among the tested genotypes under well and stress irrigation conditions were significant. Regarding the proline content, there were no great differences among the pre-breeding rice genotypes and the parents under the well-irrigated conditions. However, the proline content was markedly increased in all genotypes under the water deficit treatment in comparison with the well irrigated condition. Nevertheless, this increase was significantly higher in both the advanced pre-breeding lines, RBL112and RBL44. Similarly, water stress treatment caused great elevation in SOD content in all the rice genotypes. The pre-breeding lines RBL112, RBL44, RBL10 and the tolerant parents IRAT170, WAB881SG9 and IR60080-46A showed higher activity of SOD compared to other genotypes. Interestingly, there were positive correlation between proline content and MDA estimates of the genotypes under well irrigated conditions and their corresponding estimates under stress conditions (75.5 and 74.4%, respectively). The level of TSS in the tested genotypes under both flooded and drought stress conditions. The data revealed that the highest concentration of TSS was found in the advanced breeding line RBL44 (52.58 mg g−1 FW) followed by RBL112 (50.73 mg g−1 FW). MDA content was markedly increased in the leaves of the tested rice genotypes under water dificit stress conditions compared to control plants. The sensitive parent Giza177 displayed the highest MDA content, while the genotypes RBL112, RBl44 and IRAT170 were less affected by drought treatment. The genotypes RBL112, RBL44, IRAT170 and WAB881SG9 maintained higher relative water content in their leaves under drought condition. Moreover, the RWC values of all genotypes decreased in response to water stress treatment, while the pre-breeding line was the best able to maintain a high level of water in its leaf tissues.

Figure 4
figure 4

The biochemical assays of fresh tissues under water shortage stress (orange columns) as compared to normal irrigated conditions (blue columns).

Full size image

Marker assisted selection (MAS) of the pre-breeding lines as compared to their parents

In order to detect the similarity among the pre-breeding lines and their water stress tolerant parents. The pre-breeding lines and their parents were genotyped using 13 SSR markers associated with QTLs for water shortage related characteristics (Supplementary Table 2). Eleven out of the 13 SSR markers were found to be polymorphic; altogether, 28 alleles were detected, Supplementary Table 9. The microsatellite marker RM22 reproduced two different alleles (180 and 195 bp, respectively). RM22-180 bp was detected in the sensitive parent, Giza177. The RM22-195 bp allele was detected in the tolerant parents IR60080-46A, WAB881SG9, and the moderately tolerant parent, Giza178. The latter allele was detected in the lines RBL112, RBL44, RBL10, RBL35, and RBL133. The primer pairs RM525 have amplified three alleles, 120, 140, and 170 bp. The allele RM525-170 bp was detected only in the IR60080-46A parent, and was transmitted to the lines RBL112 and RBL35. RM472 generated two alleles; a unique one was generated for the sensitive parent Giza177 (RM472-295 bp). While the RM472-300 bp allele was common in all the other tested genotypes. At the same time, the RM228 generated two alleles: the RM228-120 bp was located in the moderately tolerant parent Giza178 and RBL112, and the other one, RM228-135 bp was detected in the sensitive, the tolerant parents and their derived lines. Furthermore, RM324 amplified three different fragments, of which the IR60080-46A (RM324-165 bp fragment) was the only tolerant parent that did not carry the allele located in the sensitive parent (RM324-150 bp). RBL112 was the only pre-breeding line that carried the allele RM324-165 bp. The highest polymorphism was detected using the microsatellite marker RM3805, generating four alleles. The tolerant parents IR60080-46A and IRAT170 were harboring the RM3805-130 bp allele. This fragment was also segregated into the pre-breeding lines, RBL112, RBL44, RBL7, and RBL9. The SSR marker RM260 generated three marker fragments. The tolerant parents and the pre-breeding lines RBL112, RBL44, RBL7, RBL9 and RBL10 are harboring the same allele size. While for the SSR marker RM175 the tolerant parents, RBL7, RBL9 and RBL10 generated the same fragment size. RM328, 228, RM201 and RM319 were not able to differentiate the tolerant parents from the sensitive ones.

Expression profiling of water stress responsive genes

To emphasize the response of the tolerant genotypes in comparison with the sensitive ones. We further evaluated the gene expression of the most tolerant pre-breeding line and compared it with their parents along with the sensitive international check IR64 to understand the expression response of these genes to water deficit condition. A set of 22 drought responsive genes were manipulated, among these genes 14 transcription factor (TF) genes, 4 antioxidant enzymes genes and 4 drought functional genes, Supplementary Table 3. The heatmap of the expression profiling of these genes is presented by generating the − log10 of the differences between their relative expression estimates under normal irrigated and water deficit, Fig. 5. Apparently, among the 22 genes OsLEA3 and OsAPX2 were strongly upregulated under water deficit stress in all the genotypes as compared to their expression under normal irrigation. Both genes were clustered were separated in a single cluster based on the heatmap cluster analysis. Furthermore, the genes OsCATA, OsZIP23, OsDREB1C, OSDREB2A, OsAHL1, OsP5CS and OsNAC1 were upregulated in all the genotypes and were subclustered together. In contrast, the genes OsDRE2B, OSCDPK7 were downregulated in all the genotypes due to water deficit stress. The most notable expression profiling was detected for the genes Os64660, OsHSP101, OsDREB1A and OsSKIPa as the responses of these genes to water deficit were different in the sensitive check from the moderately tolerant and the tolerant genotypes. The genes Os64660 and OsHSP101 were downregulated in the sensitive check IR64 while it was upregulated in the parents and the RBL112. In contrary, the gene OsDREB1A was upregulated in the sensitive genotype while it was downregulated in the moderately tolerant Giza178 and the tolerant genotypes IR60080-46A and RBL112. While there were no differences in the expression of OsSKIPa for the genotypes IR64 and Giza178, its expression was upregulated in the tolerant genotypes IR60080-46A and RBL112.

Figure 5
figure 5

The hierarchical cluster heat map of the differential relative expression analysis as compared to the estimates under normal irrigated conditions.

Full size image

Discussion

Water deficit stress is considered the main constraint reported as a consequence of climate change. Developing new rice genotypes tolerant to water deficit stress is a prerequisite for sustainable rice cultivation. Several investigations have reported the effect of water deficit stress on yield performance of cultivars12,39,40. In the current investigation we are reporting the phenotypic performance, physio-biochemical and gene expression response of different advanced pre-breeding lines derived from crossing drought tolerant genotypes with commercial Egyptian rice cultivars. The present investigation showed that there were wide range of variability for the phenotypic and biochemical responses of the pre-breeding lines and their parents under both normal and stress irrigated conditions. Water deficit stress significantly affected the growth characteristics, yield and its related parameters, and the internal physio-biochemical characteristics for all the genotypes.

Yield is the final product of all contributions made by all the plant characteristics that either directly or indirectly influence the grain yield1,41,42. Among the tested genotypes, the Egyptian cultivars were higher than the other drought tolerant genotypes under normal irrigated conditions. However, the reduction in their yield due to water deficit stress was higher than the reduction for the tolerant parents. These findings were confirmed further by estimating the DSI. The DSI considers both grain yield under no-stress and stressed conditions and provides values to measure the genotype tolerance to water deficit stress43,44. The DSI of the sensitive genotype Giza177 was the highest among all the genotypes, Fig. 3. Interestingly, the RBL112 pre-breeding line DSI record was the lowest, similar to its parent, the IR60080-46A, the upland tropical japonica cultivar45 indicating that RBL112 is tolerant to water deficit stress. Furthermore, its yield records were the highest under water deficit stress and normal irrigated conditions. Nevertheless, this line was developed by backcrossing with the recurrent parent Giza178. Giza178 was the highest under normal irrigated conditions as compared to the other tested genotypes. All the pre-breeding lines under study have DSI estimates less than those for their two Egyptian cultivars parents. The sterility percentage under water deficit conditions for the RBL112 line was the lowest, Fig. 2-C. A Significant negative correlation record (r = − 0.886) was estimated between the genotype’s records of grain yield and sterility estimates under water deficit condition. As a matter of fact, water deficit stress increased sterility percentages in all the genotypes. This performance and effect of water deficit stress on sterility was reported in other investigations12,46.

A tremendous variation in the responses of the genotypes was recorded. Some genotypes started flowering under stressed conditions prior to well-irrigated conditions, while other genotypes were in contrary. Early maturing genotypes are favorable as they consume less water during the growing season compared to long duration genotypes5,47. These genotypes have more opportunities to escape the terminal drought event48,49. Accordingly, we found that RBL7 and RBL9 were the best performing pre-breeding lines. Water deficit stress reduces cell turgor, which inhibits both cell expansion and division. This phenomenon is very important for plant height and the number of tillers during plant growth50. Furthermore, water deficit-tolerant genotypes require a particular canopy structure for the plant shape51. This is obvious from the characteristics of the tolerant parents as compared to the Egyptian cultivars Giza177 and Giza178. Tolerant parents are significantly taller than the Egyptian cultivars, with more vigorous growth under water deficit stress (Fig. 1, Supplementary Table 6). However, short stature plants are preferable for rice breeders as they are not vulnerable to lodging52. RBL112 was the best pre-breeding line as its plant height value was the lowest and it was the least affected genotype sue to water deficit stress in this regard. Similarly, the reduction of the number of tillers due to stress in the pre-breeding line RBL112 was the lowest among the tested genotypes. These findings are consistent with those obtained by other groups53,54.

Intriguingly, significant correlations were detected between yield and the physio-biochemical biomarkers under water stress conditions. Proline is considered one of the most utilized indicators for screening drought-responsive plant material. Proline is one of the glutamate amino acids that is strongly activated and accumulated under stress55,56. In the current investigation, accumulation of proline was reported in the tested genotypes due to water stress conditions as compared to the normal irrigated one. The accumulation in the pre-breeding line RBL112 was the highest. Antioxidant enzymes, such as SOD, play an important role in plant tolerance to harsh environmental stress. Generally, all the genotypes exhibited low SOD enzymatic activity under normal irrigated conditions. While its activity increased markedly due to water deficit stress in the tested genotypes. These findings are similar to those obtained by Wang et al.57. In the current investigation, all the newly developed lines exhibited higher activity of the SOD as compared to the Egyptian parents under water shortage stress. RBL112 and RBL44 had the highest accumulated SOD activity in response to water shortage stress. Furthermore, increased accumulation of TSS in the leaves was reported to increase tolerance to water deficiency in plants58,59,60. The drought treatment significantly increased the TSS content in all the genotypes. The accumulation of soluble sugars in different plant species subjected to water deficit stress is well documented, and the accumulation rate was positively corelated with drought tolerance. Soluble sugars act as a key messenger to activate gene expression and enzymatic activities that are implicated with various development and metabolism events61,62,63.

As in other biochemical assays in the current investigation, water deficit stress markedly elevated the leaf MDA content of the twelve rice genotypes as compared with normal irrigated plants. However, the highest response with the highest MDA record was detected in the susceptible parent Giza177, in contrast, the pre-breeding genotype and the upland drought tolerant ones showed lower records. Furthermore, the MDA records showed a positive correlation estimate with the DSI (r = 0.896**). Together, these findings indicate that the pre-breeding lines RBL112 and RBL44 are drought tolerant. MDA content is largely adapted as a physiological index to indicate the damage that occurred as a consequence of stress. It explains the degree of lipid peroxidation, oxidative damage, and stress tolerance in plants. Similar records of MDA were found in salt, water shortage and heat resistant plants when exposed to stress64,65,66. Interestingly, MDA was significantly correlated with RWC under water shortage stress. Water deficit stress markedly decreased the RWC in all the genotypes under study. However, the pre-breeding lines RBL112 and RBL44 and the tolerant genotypes IRAT170 and WAB881SG9 exhibited a higher value of RWC explaining a considerable ability to maintain a high level of water in their leaf tissues. This performance confirms their tolerance to water stress to afford high yielding ability compared with the susceptible variety Giza177 and the moderated one Giza 178. Several reports have indicated that higher RWC values under water deficit stress are indicative of stress tolerance in many plant species67,68,69.

SSR markers are the most widely utilized molecular markers to detect allelic diversity among genotypes. In our study, SSR markers linked to water shortage stress related QTLs were used to conduct allelic diversity to support the findings of the phenotypic evaluation. The results of investigating allelic diversity using SSR markers revealed that the high yielding, with lowest DSI score, pre-breeding line RBL112 was found to be the most line that inherited several alleles from its upland parent IR60080-46A. RBL112 carries the alleles RM22-195 bp, RM525-170 bp, RM324-165 bp, and RM3805-130 bp from IR60080-46A, while only RM228-120 bp inherited from the Egyptian parent, Giza178. The previously mentioned alleles did not exist in the sensitive genotype Giza177. RM22 was reported to be linked to QTLs for proline content70 and grain yield71 where the phenotypic variance was high. While the microsatellite marker RM525 was reported to be linked to index of drought resistance72. Furthermore, QTLs that affect grain yield under water shortage stress were found to be linked to the SSR markers RM324 and RM3805 on chromosomes 2 and 6, respectively, which were reported by Venurprasad et al.,73,74,75.

Interestingly, the cluster analysis showed that the new pre-breeding line was clustered with its upland parent, IR60080-46A. While the genotype Giza178 was clustered with the sensitive genotype IR64. Taken together, these results explain why the response of the pre-breeding line is most likely to be similar to its upland parent based on the 22 drought responsive genes. These 22 genes were transcription factor genes, antioxidant enzymes genes and drought functional genes. Most of the evaluated genes are either upregulated or repressed due to water deficit stress. Notably, there was a great upregulation in the relative expression profile of late embryogenesis abundant 3 (OsLEA3) in the leaves of all the genotypes in comparison to control conditions. Which indicates the crucial role of the LEA3 gene in drought tolerance in rice. Similar response was detected for the relative expression pattern of the cytosolic ascorbate peroxidase 2 enzyme, OsAPX2, which was reported to be a critical key enzyme in the growth and reproduction of rice under water deficit stress67,76.

In our study, we found that the transcription factor genes, OsNAC1, OSDREB2A, OsDREB1C, and OsZIP23 were clustered together with the genes OsP5CS, OsAHL1, and OsCATA. Similarly, OsDREB2A, OsP5CS OsNAC1, and OsCATA were reported to be highly expressed with OsLEA3 in OsMYB6 transgenic plants than in wild-type plants under drought stress68,77. Unlike our study, OsMYB6 was downregulated in all genotypes except for Giza178. Furthermore, the high yielding pre-breeding RBL112 showed the highest reduction in the expression level of this gene. This downregulation of the MYB6 gene might be due to the upregulation of the OsWRK13 gene, which was found to downregulate MYB genes78. Moreover, the genes OsDREB1A, OSDREB2E, OsCPK4, OsHSP101, and OsSKIPa showed different expression profiles among the tested genotypes. Dehydration-responsive element-binding (DREBs) are major plant TFs that control the expression of several stress-related genes79. Unlike OsDREB2A and OsDREB1C, which were upregulated for all the genotypes, OsDREB1A and OsDREB2E upregulate responses under water stress in IR64 only. This repression in those genes in the genotypes IR60080-46A, RBL112, and Giza178 might be due to specific methylation in these genes80 due to water stress as compared to normal irrigation. Furthermore, heat shock proteins (HSPs) have a key role in plant responses to abiotic stress as they act as molecular chaperons to prevent protein folding81,82. Our findings revealed the upregulation of OsHSP101 in the upland genotype and RBL112 due to water deficit stress, unlike its response in the sensitive check, IR64. The ski-interacting protein-a gene, OsSKIPa encodes a TF that controls the expression of several stress-related genes. In our investigation OsSKIPa was found to be induced by stress as compared to normal irrigation in the upland parent and the high yielding pre-breeding line under water deficit. This gene was reported to maintain cell viability and stress tolerance in rice83.

Conclusions

Understanding the performance of the pre-breeding lines under stress is of great importance for breeders to selecting the best lines to be utilized for successful breeding program. The pre-breeding genotypes under study showed significant differences in their phenotypic, physiochemical and molecular response to water deficit stress. Under water deficit stress, grain yield showed highly significant correlation with the physiological characteristics under water stress. Furthermore, high negative correlation was found between MDA and RWC. The new high yielding pre-breeding lines was higher in proline content, SOD, TSS and RWC while having low MDA content in their leaves. Expression profiling of the high yielding pre-breeding line compared to its upland and Egyptian lowland parents revealed that its expression pattern was more similar to the upland parent. Water deficit stress upregulated the genes OsLEA3, OsAPX2, OsNAC1, OSDREB2A, OsDREB1C, OsZIP23, OsP5CS, OsAHL1 and OsCATA. The results of the Cluster analysis and heatmap of the relative expression pattern of the 22 genes. The genes OsDREB1A, OSDREB2E, OsCPK4, OsHSP101 and OsSKIPa showed different expression profile among the tested genotypes.