Rice pyramided line IRBB67 (Xa4/Xa7) homeostasis under combined stress of high temperature and bacterial blight

Rice bacterial blight (BB) caused by Xanthomonas oryzae pv. oryzae (Xoo) implies substantial yield loss to rice. In times of climate change, increasing temperatures are observed and further acceleration is expected worldwide. Increasing temperature often turns into inhibition of host plant defense to pathogens. Recently, a reduced resistance in rice IRBB4 carrying Xa4, but an increase in resistance in IRBB7 carrying Xa7 resistance by increasing temperature has been reported. Influence of high temperature on both R genes (Xa4+Xa7) combined in IRBB67 was analyzed under growth chamber conditions and transcriptomic analysis performed. The pyramided line IRBB67 showed no differences in lesion length between both temperature regimes, demonstrating that non-effectiveness of Xa4 at high temperature did not affect IRBB67 resistance. Moreover, Xa4 complements Xa7 resistance with no Xoo spread in planta beyond the symptomatic area under both temperature regimes in IRBB67. Time course transcriptomic analysis revealed that temperature enhanced IRBB67 resistance to combined heat and Xoo. Our findings highlight altered cellular compartments and point at a role of the cell wall involved in Xoo resistance and heat stress tolerance in both susceptible (IR24) and the resistant (IRBB67) NILs. Interestingly, up-regulation of trehalose-6-phosphatase gene and low affinity cation transporter in IRBB67 suggest that IRBB67 maintained a certain homeostasis under high temperature which may have enhanced its resistance. The interplay of both heat stress and Xoo responses as determined by up-regulated and down-regulated genes demonstrates how resistant plants cope with combined biotic and abiotic stresses.

The world population is projected to reach 9.7 billion by 2050 and half the world population growth is expected in developing countries 1 . This world population growth coupled with the impact of climate change on agricultural production in those countries demands a rapid growth in the food supply and animal feed, while conserving forest resources for climate mitigation 2,3 . Previous reports have shown that an increase of seasonal temperature by 1 °C results in a decline of major grains yield in the range of 2.5 to 16% in the tropics and subtropics 4,5 . Unfortunately, due to their sessile lifestyle, plants have no chance to escape this environment (biotic and abiotic stresses) and must respond and adapt [6][7][8] .
Abiotic stress may imply positive or negative effects on plant defense responses [7][8][9] . According to these authors, the outcome of the interaction depends on the timing, nature and the severity of the stress. Temperature, water, relative humidity, light and circadian rhythm significantly influence plant defense and pathogen invasion 6 . A small variation in temperature can affect plant growth, but also plants' responses to pests and pathogens 10,11 . Most studies on plant responses to environmental changes were carried out under single stress and are therefore unsuccessful in explaining plant responses to more than one stress factor 9 . High temperature affecting host resistance to pathogens has been reported in tobacco infected with Tobacco mosaic virus 12 . Increased disease resistance to stripe rust (Puccinia striiformis f.sp tritici) was observed in wheat under high temperature (25-35 °C) and is likely caused by the significant expression of resistance gene Yr36, which is not effective under low temperature (15 °C) 13,14 . Similar reactions were reported in Arabidopsis, which shows resistance to virulent Pseudomonas syringae pv. tomato (Pst) strain DC3000 at 22 °C 14 and susceptible to the same strain at moderately elevated temperature between 27-30 °C 14,15 .
In rice, Onaga et al. 16,17 showed that the reaction of resistance genes in rice to Magnaporthe grisea under high temperature depends on the genetic background of the rice genotype, where high temperature may either compromise resistance genes' expression, or enhance the resistance response of gene Pi54. Dossa et al. 18 reported that abiotic stress (drought) resulted in decrease of Xa4 R-gene mediated resistance to bacterial blight. Studying the effect of high temperature on the rice Xanthomonas oryzae pv. oryzae (Xoo) interaction, Webb et al. 19 reported high temperature reducing the resistance of rice IRBB NILs carrying the Xa4 resistance gene to Xoo. According to these authors, an inverse response was observed in IRBB7, a NIL with Xa7 resistance gene. This inverse response raised the question about the R gene durability under climate change conditions. However, the pyramided lines such as IRBB67 (Xa4+Xa7) may be an alternative as pyramided lines are more durable and have broad-spectrum of resistance than monogenic lines 20 . Additionally, a recent study revealed that pyramided line IRBB61 carrying three R-genes (Xa4, xa5 and Xa7) showed more resistance to the combined high temperature and Xoo stress 21 . Therefore, there is a need to study the molecular mechanisms underlying the IRBB67 carrying two R-genes (Xa4 and Xa7) response to the pathogen under high-temperature conditions. The inhibition of plant resistance to pathogens under high temperature is often associated to an enhanced activity of RNA-silencing mediated resistance and an inhibition of effector-triggered immunity under which the pathogen effector is normally recognized by the host R gene 22,23 . Recently, a study from Huot et al. 24 reveals that high temperature enhances bacterial type III secretion and disease susceptibility. The resistance induced by the NB-LRR class of R gene is reduced by temperature increase due to less nuclear accumulation of SCN1 25 .
The hypothesis here is to determine the effects of Xa4 on Xa7 in the pyramided line IRBB67 carrying both R genes and to understand how this pyramided line responds to the combination of Xoo and heat stress. Combinations pathogen infection with high temperature or with drought stress alter gene expression by activation of specific programs, revealing that plant response to multiple stresses differs between reactions to one or more stresses 26 . Moreover, the study of Rasmussen et al. 27 showed that 61% of transcripts could not be predicted under single stress compared to double stress treatment. Understanding how resistant plants respond to Xoo and heat stress will provide information to be used for developing double stress tolerant rice varieties.

Results
High temperature affects plant morphology. Plant morphology change due to temperature rise was determined by measuring the plant height under both temperature regimes. All rice NILs were taller under high temperature compared to normal temperature (Fig. 1A). Under low temperature, the plant height ranged from 47.8 cm to 50.8 cm, and under high temperature from 55.5 cm to 63 cm. The highest plant height under low temperature was recorded on IRBB7 and under high temperature on IRBB4 (Fig. 1B).
Rice near isogenic line with R genes Xa4 and Xa7 combination confers strong resistance to BB under high temperature. Rice bacterial blight R genes Xa4 and Xa7 are among the major R genes mediating resistance to BB. In order to determine the high-temperature effect on the combination of the two R genes (Xa4 and Xa7) responses to BB, rice NIL IRBB67 carrying Xa4 and Xa7, along with IR24 (susceptible) and IRBB4 carrying Xa4 and IRBB7 carrying Xa7 were used. Plants were inoculated under both temperature regimes in the growth chamber with PXO145 (avrXa4+avrXa7). Disease evaluation after 11 days showed an inverse response between Xa4 and Xa7, with Xa7 resistance enhanced under high-temperature conditions. However, the gene combination in IRBB67 did not show a significant difference in disease lesion length between high and low temperatures. Under low temperature the average lesion length was 1.5 cm, under high temperature 1.3 cm on IRBB67 (Fig. 1C).
Evaluation of bacterial blight lesion length development recorded from 4 dpi to 11 dpi showed that IRBB67 resistance follows a similar pattern to that of IRBB7 under high-temperature conditions (Fig. 2). IRBB67 resistance was high compared to IRBB7 resistance under low-temperature with an average lesion length of 2.6 cm for IRBB7 at 11 dpi. Additionally, in planta bacterial counts showed evidence of complementation of Xa4 to Xa7 under high-temperature compared to low-temperature (Fig. 1D). The leaf segment plating beyond the visible lesion length showed no significant differences in the bacterial count in IR24 leaves under both temperature regimes. IRBB4 showed similar results at high-temperature while significant differences were observed at low-temperature. However, in IRBB7, the bacterial count between leaf segments was significantly different at high-temperature compared to low-temperature, suggesting that Xoo in planta spread less in IRBB7 under high-temperature compared to the spread in IRBB4. Unexpectedly, we did not detect bacterial growth beyond the symptomatic area under both temperature regimes in the pyramided NIL IRBB67.
Gene expression profiling of IR24 and IRBB67, Xoo and mock inoculated under normal and high temperature. To better understand the complementary effects of Xa4 on Xa7 response to both bacterial blight and high temperature, total RNA from leaves' samples of mock-inoculated and bacterial-inoculated plants grown in normal and high temperature conditions were single-end sequenced using Illumina Hiseq technology. The read length of 100 bp in the range of 25.2-82.1 million was generated (Table S2). Percentage of mapped reads of each sample ranged from 97% to 98.6%. The mapped reads were used for further analysis.
To identify differentially expressed genes (DEGs), we used the DESeq. 2 package 28 under R/bioconductor. The significant DEGs were identified based on the false discovery rate (FDR) of 0.05 and log-2 fold change. Using these criteria, we identified 4,683 transcripts differentially expressed comparing Xoo inoculated samples to mock inoculated samples from both IR24 and IRBB67 under both temperature regimes at 3, 72 and 120 hpi Plants under high temperature conditions were significantly higher compared to low temperature. The letters a and b represent the significant differences between low and high temperature on each NIL as determined by least significant differences means (LSD means) at α < 0.05. (B) Picture of rice NIL IR24 under both temperature regimes. Plants under high temperature conditions were significantly higher compared to low temperature. (C) Bacterial blight lesion length under low and high temperature regimes conditions. Bacterial blight lesion length was recorded at 32 days old (11dpi). The letters a and b represent the significant differences between low and high temperature on each NIL as determined by least significant differences means (LSD means) at α < 0.05. (D) PXO145 spread in planta in four rice NILs under two temperature regimes. Data were collected from leaf clipping inoculated leaves of 21 days old plants and collected 11 days after inoculation. A, B and C represent leaf segments of 5 cm length beyond bacterial blight symptomatic area. The letters a and b represent the significant differences between low and high temperature on each NIL as determined by least significant differences means (LSD means) at α < 0.05.
The cell wall plays several roles such as a physical barrier against insects and pathogens and providing shape and structure but is also involved in cell-cell communication and osmotic regulation. Cell wall biochemistry may play a decisive role in defense response to pathogens, as revealed specifically for bacterial pathogens [30][31][32][33] . In this study, the cell wall was generally affected by high-temperature combined with Xoo, especially in IRBB67. Similarly, the cell wall proteins such as AGPs (arabinogalactans-proteins) which link the cell wall with the plasma membrane and the cytoskeleton 34,35 and were described to be involved in defence against bacterial pathogens 30 were shown to be downregulated by high temperature and Xoo in both rice NILs (IR24 and IRBB67) at an early stage of inoculation (Table S4). Further up-and downregulations of cell wall related genes as temperature response were observed. Our results point at an important role of phytohormone regulation and cell wall modifications in response to combined stress of high temperature and bacterial blight, with particularly SA signaling and cell wall proteins likely being important in IRBB67 response to the combined stress.

IRBB67 regulates several biological processes and molecular functions under high-temperature conditions.
To determine the effects of high-temperature conditions on rice, DEGs from mock-inoculated plants at high and low temperature were compared at each time point and within NILs. A total of 332 DEGs were differentially expressed in both mock inoculated NILs. Several DEGs were down-regulated in both NILs in response to high temperature. Differences in rice transcript accumulation between high and low temperatures were observed at 3 hpi at which most of the DEGs were down-regulated (Table S5).  www.nature.com/scientificreports www.nature.com/scientificreports/ up-regulation in the resistant reaction at 120 hpi. The DEGs, 32 in total enriched in nucleus at 120 hpi in IRBB67 were significantly down-regulated at 3 hpi.
Besides, 05 DEGs (LOC_Os07g34520.2; LOCOs01g12490.1; LOC_Os08g30020.3; LOC_Os07g34520.3; LOC_ Os11g46850.1) were up-regulated in IR24 at 3 hpi and LOC_Os08g04500.2 encoding for terpene synthase was up-regulated at 72 hpi. In the resistant NIL IRBB67, 9 DEGs were up-regulated (Table S5) including MYB family transcription factor (LOC_Os02g53670.1) and MYB transcription factors, which are involved in plant development, but also in defense responses to hormone or stress treatments 36 . Additionally, flavin mono-oxygenase www.nature.com/scientificreports www.nature.com/scientificreports/ (LOC_Os01g12490.1), which plays an important role in n-tryptophan (Trp)-dependent indole-acetic acid (IAA) biosynthesis in plants and regulates plant growth and development, was among the nine up-regulated DEGs in IRBB67. Taken together, these findings suggest that IRBB67 responds to heat stress by inducing plant growth and development related genes, but also defense response genes.

High-temperature during pathogen infection affects rice cellular compartments.
To explore the effects of high-temperature on rice during pathogen infection, we compared transcript accumulation between high and low temperatures of Xoo inoculated samples from both IR24 and IRBB67. A total of 156 DEGs (Table S6) were induced from both IRBB67 and IR24 NILs, demonstrating the repression of several DEGs in response to Xoo inoculation between high and low temperatures compared to that of mock-inoculated (Table S5). Besides, only 19 DEGs were shared between mock and Xoo-inoculated samples (Fig. S1).
At 3 hpi, no significant DEG was found to be up or down-regulated, except LOC_Os08g39850.2 encoding for lipoxygenase, a chloroplast precursor involved in the programmed cell death pathway and biotic and abiotic stress response in plants 37 , which was down-regulated in IRBB67 (Table S6). At 72 hpi, 14 DEGs were expressed in IR24 with down-regulation of 9 DEGs and up-regulation of 5 DEGs. In the resistant NIL, the difference in the response to Xoo between high and low-temperature was generally significant at 72 hpi compared to the susceptible response at the same time point. Fifty four and 39 DEGs were down-regulated and up-regulated, respectively, in IRBB67 at 72 hpi. A late response was observed in the susceptible NIL at 120 hpi, where 63 DEGs were expressed with 7 down-regulated DEGs and 56 up-regulated. Only 5 significant DEGs were expressed in IRBB67 at 120 hpi and all were up-regulated. These 5 DEGs included genes encoding for terpene synthase, putative, expressed (LOC_ Os08g04500.1, LOC_Os08g04500.2), methyltransferase, putative, expressed (LOC_Os05g01140.1), Jacalin-like lectin domain containing protein, putative, expressed (LOC_Os12g14440.1), helix-loop-helix DNA-binding domain containing protein, expressed (LOC_Os01g01840.1) and WRKY65, expressed (LOC_Os12g02470.1) (Table S6).
Functional analysis of the 156 DEGs showed down-regulation of plasma membrane (GO:0005886) and membrane (GO:0016020) genes across the three time points in IR24 and at 72 and 120 hpi in IRBB67 (Fig. 4). Genes of the cell wall (GO:0005618), extracellular region (GO:0005576), and external encapsulated structure (GO:0030312) were significantly down-regulated at 72 hpi in both NILs while transferase activity (GO:0016740) was down-regulated in IR24 at 120 hpi, and vacuole (GO:0005773) was down-regulated in IRBB67 at 72 hpi. The late response observed in the susceptible NIL IR24 at 120 hpi was correlated with up-regulation of the following GO terms: transcription regulator activity (GO:0030528), transcription factor activity (GO:0003700), DNA binding (GO:0003677) and nucleic acid binding (GO:0003676) in IR24 at 120 hpi (Fig. 4).
IRBB67 mediated resistance to bacterial blight under low temperature conditions. To determine the difference in transcript accumulation after Xoo inoculation between the resistant (IRBB67) and the susceptible (IR24) NILs, selected DEGs in the IRBB67 and IR24 were compared at each time point. A total of 145 DEGs were differently induced in IRBB67 and IR24 after Xoo inoculation under low temperature (Table S7). Among them, 102 DEGs (51 up-regulated and 51 down-regulated) were induced at 3hpi, 40 (six up-regulated and 34 down-regulated) at 72hpi and 85 (37 up-regulated and 48 down-regulated) at 120hpi. DEG numbers decreased at 72hpi, however, a late response to Xoo was observed at 120hpi with induction of 85 DEGs. Additionally, 51 DEGs were specifically induced at 3hpi and there were 24 DEGs in common between the 3 time points.
GO enrichment analysis of the DEGs revealed no significant GO terms for p-value of 0.05. However, DEGs (LOC_OS02g40130 and LOC_Os02g40190) which belong to the group of protein kinases related genes, involved in cell death, response to biotic stimulus and response to stress were up-regulated in the resistance NIL (IRBB67). Protein kinases play important roles in activation of plant defense mechanisms and signal transduction. Among the 145 DEGs which are differentially induced in IRBB67 and IR24, 21 DEGs were related to protein kinases. Four(04) LRR type receptor like kinase genes were induced among which three (LOC_ Os11g29110, LOCOs_02g40130 and LOC_Os11g29090) were significantly up-regulated in IRBB67 at 3hpi and LOC_Os05g46090 was down-regulated at 120hpi. One DEG encoding for cysteine-rich receptor-like protein kinase (LOC_Os02g12130) was down-regulated in IRBB67 (Table S7). Other types of receptor like kinase such as LOC_Os02g40180, LOC_Os06g38760, LOC_Os06g16300, LOC_Os09g18594, LOC_Os11g07170 and LOC_ Os09g19500 were down-regulated in IRBB67.
Additionally, protein kinase genes (LOC_Os05g41950, LOC_Os11g44250) were specifically up-regulated in IRBB67 at 3hpi and a DEG (LOC_Os09g18159) which encodes for light repressible receptor protein kinase, putative expressed was only up-regulated at 120 hpi. Two DEGs encode for wall associated kinases (LOC_ Os11g47140 and LOC_Os1146860) and LOC_Os07g03920 which encodes for lectin-like receptor kinase, and LOC_Os06g38650 which encodes for RLKs were significantly up-regulated at 3 and 120hpi in IRBB67 in comparison to IR24 after Xoo inoculation under low temperature conditions. LOC_Os11g46850 which encodes for wall associated kinase was significantly up-regulated in IRBB67 only at 3hpi as well as receptor kinase (LOC_ Os02g40190) and cyclin-dependent kinase G-1 (LOC_Os02g39010) in IRBB67 (Table S7).
In addition to protein kinases related genes, LOC_Os07g05400 encoding for Ferredoxin-NADP reductase, chloroplast precursor, putative expressed, LOC_Os06g38120 which encodes for low-affinity cation transporter, and LOC_Os06g38110, encoding for expressed protein, were significantly up-regulated at 3hpi in IRBB67. Uncharacterized glycosyl hydrolase Rv2006/MT2062, putative expressed, encoded by LOC_Os09g20390, LOC_ Os11g44950 (glycosyl hydrolase family 3 protein) showed significant up-regulation in IRBB67 at 120 hpi, and LOC_Os07g46660, encoding ubiquitin carboxyl-terminal hydrolase domain containing protein expression was up-regulated in IRBB67 from 72-120 hpi, while glycine-rich cell wall protein (LOC_Os03g07270) showed up-regulation at 120 hpi (Table S7). www.nature.com/scientificreports www.nature.com/scientificreports/ High-temperature conditions enhance IRBB67 resistance to bacterial blight. In order to understand the mechanisms by which the resistant rice variety IRBB67 harboring Xa4 and Xa7 resistance genes, responds to the combined stress of high temperature and bacterial blight, DEGs from the resistant NIL IRBB67 and the susceptible IR24 after inoculation with Xoo under high temperature were compared. A total of 188 transcripts were differentially expressed between IRBB67 and IR24 at 3, 72 and 120 hpi under high-temperature conditions. At 3 hpi, 113 DEGs were expressed with 56 down-regulated and 57 up-regulated. At 72 hpi, 99 transcripts were differentially expressed (44 down-regulated and 55 up-regulated) and 145 DEGs at 120 hpi with 77 down-regulated and 68 up-regulated (Table S8).
Functional analysis of the differential DEGs between IRBB67 and IR24 after inoculation with Xoo under high temperature conditions revealed no functional enrichment at 3 hpi, which suggested no response to both high temperature and Xoo at an early stage of inoculation from both rice NILs. Examination of GO terms at 72 and 120 hpi suggested that under combined stress of Xoo and high temperature, the rice transcriptome is largely devoted to catalytic activity. At 72 hpi, 65 DEGs were functionally enriched in catalytic activity (Fig. 5). Catalytic activity (GO:003824), transferase activity (GO:0016740), kinase activity (GO:0016301) and transferase activity, transferring phosphorus-containing groups (GO:0016772) were the most significant functional groups at 120 hpi. Catalytic activity was shown to be a major regulator in the response to high temperature and Xoo in IRBB67, as well as kinase activity (24 DEGs), transferase activity (35 DEGs) and transferase activity and transferring phosphorus-containing groups (24 DEGs) which belong also to catalytic activity groups (Fig. 5). www.nature.com/scientificreports www.nature.com/scientificreports/ Besides these four functional groups, DEGs encoding wall associated kinases together with DEGs encoding for low-affinity cation transporter (LOC_Os06g38120.1) and expressed proteins (LOC_Os06g38110.1, LOC_Os06g38210.1, LOC_Os06g38210.2) were significantly up-regulated in IRBB67. Additionally, IRBB67 preferentially responds to the pathogen infection with up-regulation of NB-ARC/LRR disease resistance protein (LOC_Os11g29090.1) and NB-ARC domain containing protein (LOC_Os11g44990.1) and down-regulation of stress response genes such as DEGs encoding for NB-ARC domain containing protein (LOC_Os07g02570.1, involved in stress response, LOC_Os01g24820.1, and LOC_Os11g46210.1, involved in protein binding and plasma membrane). Receptor like kinases function like cell membrane sensors of stimuli and lectin-like receptor kinase 7 (LOC_Os07g03920.1 and LOC_Os07g03970.1) were up-regulated in IRBB67 compared to IR24 in response to high-temperature and Xoo (Table S7). DEG LOC_Os09g20390 encoding for uncharacterized glycosyl hydrolase Rv2006/MT2062, putative, expressed involved in trehalose 6 phosphate metabolism, also showed significant up-regulation to combined Xoo and high-temperature in IRBB67.

RNA-Seq validation by RT-PCR.
To validate gene expression pattern determined by RNA-Seq, qRT-PCR was performed on 10 candidate genes (Fig. 6). Expression pattern of the candidate genes determined by qRT-PCR data was in consistent with their expression pattern in RNA-Seq (Fig. 6), although some smaller variation can be seen. LOC_Os11g44250.1, LOC_Os06g38110.1 and LOC_Os06g38120.1 were not induced in IR24 under both temperature regimes confirming the RNA-seq data. The expression of OsSweet14 (LOC_Os11g31190.1) was much higher in IR24 as determined by qRT-PCR compared to RNA-seq data.

Discussion
Abiotic stresses are major crop production constraints and are being aggravated by climate change. Under field conditions, plants are exposed to combined abiotic and biotic stresses and the damage induced by several stress factors are more than that of single stress. Increasing temperature affects rice response to BB. Plants have evolved mechanisms to respond to external stimuli such as biotic and abiotic stresses 7,8 and interactions were described in the rice Xoo interaction 38 . Also, virulence factors of the pathogen may play a decisive role in the interaction of bacterial pathogens with host plant components 33,38 . Host plant immunity allows plants to counter-attack the invading pathogen, following pathogen recognition mediated by resistance (R) proteins 22 . Our results on the effects of high temperature on the response of rice NIL IRBB67, which pyramids bacterial blight resistance genes www.nature.com/scientificreports www.nature.com/scientificreports/ (Xa4+Xa7), to bacterial blight, revealed a complementation effect of Xa4 to Xa7. Xa4 appears to be among the most widely used R genes in rice breeding programs in Asia 39,40 . As previously reported, Xa4 resistance to Xoo decreases with temperature increase while the inverse trend is seen with Xa7 19 . Moreover, the pyramiding NIL IRBB67 harboring the two R genes showed resistance to Xoo strain PXO145 (avrXa4+avrXa7) with no significant differences between both temperature regimes. Also, disease progression recorded on IRBB67 under both temperature regimes is similar to that of IRBB7 under high temperature, suggesting that Xa4 may not loose completely the resistance and is complemented by Xa7 in IRBB67. High temperature altered Xa4 response in IRBB4 to Xoo. The mechanisms by which the two R genes respond to Xoo is still unclear, although it is known that Xa7 resistance is mediated by avrXa7 and gives pathogenic fitness cost to the pathogen 41 . Hence, we hypothesized that these two R genes use different resistance mechanisms to mediate resistance to Xoo. The mechanisms by which Xa4 triggers the response to the pathogen seems to be affected by temperature rise. In contrast, Xa7 resistance mechanisms may confer an abiotic tolerance component which enhances its reaction to Xoo under temperature rise, therefore, Xa4 resistance mechanism may not be affected by temperature when it is combined with Xa7 in IRBB67. This corroborates an early report that high temperature can negatively impact the temperature sensitive resistance to stem rust in oat cultivars harboring Pg3 and Pg4 genes 42 . Moreover, Xa7-mediated resistance to avrXa7 pathogen shows stronger and faster response at high temperature than at low temperature 21 .
In planta Xoo spread beyond the symptomatic area suggested that Xa4 complements Xa7 at low temperature in IRBB67, thereby reducing Xoo spread in planta. The predominance of Xa7 over Xa4 in IRBB67 mediates strong resistance to PXO145 under high temperature with no Xoo spread in planta beyond the symptomatic area. High temperature promoting higher bacterial multiplication in the host was reported to possibly depend on the type of R gene mediating resistance to the pathogen 43 as we observed in the current study as inverse reaction between Xa4 and Xa7. Moreover, it is plausible that these two genes may belong to two different classes of genes coding R proteins and are modulated by temperature. Consistently, SNC1 gene, a NB-LRR type of R gene, does not confer resistance at high temperature when activated compared to low temperature in Arabidopsis 44 , suggesting that Xa4 R gene may belong to this type of R gene. Further study will have to investigate this hypothesis since both R genes (Xa4 and Xa7) are not cloned and the putative proteins are not known. Wang et al. 15 suggest that high temperature inhibiting plant defense to pathogens may be regulated by defense signalling components or by a combination of multiple factors. Bacteria spread in IRBB4 under high temperature suggests that recognition of corresponding avrXa4 by Xa4 R gene might be compromised by high temperature while effector recognition increased in IRBB7 leads to resistance increase. These results are in contrast to Chen et al. 43 , whose findings suggest that high temperature promotes vigorously bacteria multiplication compared to low temperature. Additionally, the rice bacterial blight pathogen during its interaction with the host plant injects effector proteins that bind to host susceptibility genes (OsSweet), a sugar transporter gene. Induction of OsSweet14 caused by TAL effectors PthXo3 45 , avrXa7 46 , Tal5 47 and TalC 48 results in increase in Xoo growth and lesion development. In the resistant plant IRBB67, OsSweet14 (LOC_Os11g31190.1) was down-regulated, and more significantly at 120 hpi (Table S7).
Examining the gene expression in IR24 and IRBB67 in response to PXO145 and two temperature (high and low) regimes, we found that the differentially expressed genes at early time triggered an efficient response to the pathogen under both temperature regimes in IRBB67. However, in IR24, the pathogen progression in planta, enabled by late expression of genes in IR24 under both temperature regimes with more differentially expressed genes under high temperature (Fig. 3). In response to environment changes, cells initiate a gene expression program to adjust its physiology and metabolism to the new environment, preventing it for damage or death 49 . In this study, it appears that at 3 hpi, high temperature and Xoo modulate more DEGs in both resistant and susceptible NILs compared to that of low temperature. Phytohormones play important roles in host immunity and response to environmental stimuli [50][51][52][53] . In this study, phytohormone related genes were induced, especially SA which is up-regulated in the resistant NIL from 72-120 hpi under high temperature. In contrary to this finding, a similar study revealed repression of SA genes under combined stress of high temperature and bacterial blight in IRBB61 (Xa4, xa5 and Xa7) at 3 and 24hpi 21 , however, in our study, the SA pathway is up-regulated at 72 and 120hpi, suggesting that SA is involved in downstream reactions to sustain the combined Xa4/Xa7-mediated resistance to high temperature and bacterial blight. Moreover, one of the major SA signaling regulators (OsNPR1, LOC_Os01g09800.1) was significantly up-regulated by the pathogen under high temperature. Expression of SA related genes in both rice NILs appears to be a general response to pathogens, which is enhanced by high temperature in the resistant plant. Salicylic acid has numerously been reported to play an important role in resistance against pathogens [54][55][56][57][58][59][60] .
To invade the host organism, pathogens need to encounter the host cell wall, the first physical defense barrier. In case of Xoo, it enters the host through hydathodes or wounds 61,62 . The plant cell wall is a battle ground between the pathogens and its host, where highly specific interactions between the pectic polysaccharides homogalacturonan, rhamnogalacturonan I and proteins such as arabinogalactan protein (AGP) and polygalacturonase-inhibiting protein (PGIP) of the cell wall and cell wall degrading enzymes of the pathogen interact, contributing to either an incompatible or a compatible interaction, underlining a decisive role of the plant cell wall in plant defense [30][31][32][33] . Further molecular resistance responses 63 follow, as the information of the foreign invasion is triggering reactions in the nucleus. In this study, the infiltration allowed the bacteria to gain access directly to the host xylem allowing the pathogen to interact directly with the host parenchyma cells 62,64 . It appeared that high temperature increased host colonization by the pathogen as the host cell wall is affected allowing the pathogen to move easily in planta when no R gene is present. Moreover, the cell wall is a source of nutrients to the pathogen deriving from cellulose, hemicellulose and pectic polysaccharide layers which the pathogen degrades during the infection process 33,65 . The general down-regulation of genes of cell wall metabolism, specifically of genes involved in AGP production and in cell wall degradation in the resistant plants as shown by Pageman analysis (Table S4) points at a role of the cell wall in resistance mediated by this rice NIL under high temperature. However, the comparison between low temperature and high temperature after mock inoculation showed that high temperature generally affects the membrane enclosure with down-regulation of cell wall related genes, external encapsulating structure and plastid Scientific RepoRtS | (2020) 10:683 | https://doi.org/10.1038/s41598-020-57499-5 www.nature.com/scientificreports www.nature.com/scientificreports/ related genes in the susceptible NIL IR24, while high temperature in combination Xoo repressed the entire cellular compartment in both rice NILs (Fig. 4). Dahal et al. 66,67 also reported the regulation of pathogenesis and stress related and metabolic proteins in interaction with a bacterial pathogen, resulting in suppression of cell wall metabolic proteins in susceptible tomato plants inoculated with Ralstonia solanacearum. With the mock inoculation, the nucleus is down-regulated in both IR24 and IRBB67 at 3 and 72 hpi, while a significant up-regulation was seen in IRBB67 at 120 hpi. However, with pathogen stress combined to high temperature, the up-regulation nucleus pathway across the three time points suggests that the host plant stays under permanent alert after pathogen invasion. To sustain the response to this combined stress, lipid metabolism was significantly up-regulated in IRBB67 from 72 to 120 hpi, while the significant expression was observed only at 72 hpi in IR24.
A change caused by high temperature stress on cell wall metabolism is an important physiological mechanism for heat stress tolerance 68 suggesting that the plant responds by regulating several biological processes and molecular functions under abiotic stress, such as high temperature. This hypothesis could explain the fact that no bacterial spread was detected beyond the symptomatic area in IRBB67 and corroborates the findings from IRBB61 under similar conditions 21 .
High temperature modulated resistance to the pathogen in IRBB67 suggests the existence of a shared pathway between biotic and abiotic stresses. Environmental changes induce plant cells to trigger several events that start with perception of the stimuli at the membrane level 69 . According to the same authors, receptor sensors located in the cell membrane activate several signal transductions that triggered calcium mobilization and other secondary signals to induce stress responsive genes. Calcium acts as second messenger in various stresses [70][71][72][73] and calcium signaling was seen to be up and down regulated in both NILs. However, calcium transport genes (LOC_ Os05g02940, LOC_Os03g27960) were up-regulated in IRBB67 at 3 hpi. Additionally, wall associated kinase genes LOC_Os07g03920 and LOC_Os07g03970 were significantly up-regulated in the resistant reaction. Cell walls' sensing pathogen invasion and high temperature triggering the plant response with up-regulation of receptor kinase gene LOC_Os07g03920 was also reported by Narsai et al. 74 who observed a similar response of the resistant cultivar to bacterial blight. Response to combined high temperature and Xoo showed that the resistant NIL devoted large parts of its transcriptome to catalytic activity (Fig. 5). Physiological mechanisms may contribute to IRBB67 resistance to Xoo under high temperature. Up-regulation of trehalose phosphate phosphatase gene (LOC_Os09g20390) which dephosphorylates trehalose phosphate synthase to release free trehalose 75 , recently reported to be involved in anaerobic germination in rice 76 may also play an important role in IRBB67 tolerance or adaptation to high temperature to trigger resistance to Xoo. Interestingly, up-regulation of the trehalase gene LOC_Os10g37660 was seen in the resistant NIL, suggesting that conversion of trehalose to glucose for carbohydrate metabolism may contribute to high temperature tolerance in IRBB67. Similarly, a low-affinity cation transporter gene (LOC_Os06g38120) was also up-regulated in the resistant NIL compared to IR24 in response to high temperature and Xoo, suggesting a possible role of this gene in maintaining a certain homeostasis of the resistant NIL cell membrane under high temperature. Considering these observations the resistance increase in IRBB7 under high temperature could possibly relate to the genes which are involved in alleviating the high temperature effects on the rice cell membrane by maintaining homeostasis during the stress. Further studies are required to prove the possible role of trehalose-6-phosphate and low-affinity cation transporter genes' functions in IRBB7 and IRBB67. Stress tolerance accompanied with up-regulation of resistance genes such as NB-ARC/LRR (LOC_ Os11g44990) and NR-ARC domain containing protein (LOC_Os11g44990) may contribute to enhance resistance to Xoo under high temperature in IRBB67 and possibly in IRBB7, while their activities might be reduced in IRBB4 under high temperature. Hu et al. 77 showed that Xa4 encodes a cell-wall associated kinase, and strengthening of the cell wall by Xa4 enhances resistance to bacterial infection. Moreover, genes involved in cell wall modification are down-regulated under high temperature and high temperature combined with Xoo inoculation, suggesting that Xa4 resistance is reduced under heat stress condition. Total resistance activity below the threshold under high temperature results in no defense 25 .
Under low temperature, although no significant enrichment was seen between both rice NILs, up-regulation of defense related genes in the resistance NIL suggests an early recognition of PXO145 in IRBB67 that triggered the defense reaction. Moreover, the up-regulation of the defense genes in IRBB67 in comparison to the susceptible NIL IR24 may suppress pathogen growth and spread in planta 78 . Given that trehalose is a universal stress molecule, the up-regulation in the resistant NIL IRBB67 of LOC_Os09g20390 compared to the susceptible suggests that it may be involved in the defense response to Xoo.

Conclusion
In the light of our overall results, high temperature affects the host response to Xoo. Time course transcriptome profiles revealed evidence that the resistance of IRBB67 is enhanced under high temperature as several physiological changes were observed in comparison to the susceptible NIL IR24. The response to high temperature with regulation of cell membrane homeostasis might confer high temperature stress tolerance in IRBB67. Complementation effects of Xa4 affecting Xoo spread in planta as observed in IRBB67 under both temperature regimes suggest that this R gene may not be completely lost under high temperature, and that stress (high temperature) tolerance failure could possibly explain the decrease in Xa4 effectiveness compared to that of Xa7 under temperature increase. Our study showed that R gene Xa7 can complement the partly failure of Xa4 under high temperature, and a combination of both genes is therefore suggested to stabilize resistance of rice to Xoo under climate change. Thus, further investigations of how IRBB67 mediates resistance to Xoo under high temperature will reveal insight into the crosstalk between the abiotic and biotic stress regulatory pathways.

Plant inoculation.
The inoculum of the Philippines' Xoo race 7, strain PXO145 (avrXa4+avrXa7), was prepared from a 3 day-old culture. Twenty-one day-old seedlings of IR24, IRBB4, IRBB7 and IRBB67 were inoculated by the leaf clipping method 78 while IR24 and IRBB67 plants were inoculated at 3 points of infiltration using syringe needless with PXO145 and with sterilized demineralized water (mock) for RNA-Seq gene expression analysis 79 . The leaf at the second position of the main tiller was inoculated and disease assessment was performed daily from 4 days post inoculation to 11 dpi. Needless syringe inoculated leaves were sampled at 3, 72 and 120 hours post inoculation (hpi). One leaf from 5 plants each for each rice NIL and treatment were taken at each time point and immediately frozen in liquid nitrogen, and stored at −80 °C.
Total RNA extraction, library construction, and RNA sequencing. Total RNA was extracted from needless syringe inoculated leaves using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's protocol. Total RNA was then treated with DNase (Promega) and quantified using NanoDrop. RNA integrity was checked using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). For each time point and condition, two biological replicates were prepared. A single-end fragment library of 100 bp length was generated from cleaned total RNA following instructions of the TruSeq RNA Sample Preparation kit. Cluster generation of the produced libraries was performed using Illumina_ TruSeq SR Cluster Kit v3 -cBot -HS, and sequenced on a HiSeq. 2000 platform (Illumina) with single-end 100-bp reads and submitted to the Microarray and Deep-Sequencing Core Facility of the University Medical Center Göttingen (Germany). Sequence images were transformed with Illumina software BaseCaller to bcl files, which were demultiplexed to fastq files with CASAVA v1.8.2.
Transcriptome data analysis. The reads of each sample were mapped to the rice genome version 7 of the Rice Genome Annotation Project (RGAP) at MSU using clcbiogenomics workbench v7.0.4 with the following parameters: 2 for mismatch cost, 3 for insertion cost and deletion cost, 0.5 for length fraction and 0.8 for similarity fraction. The expression data for each sample was exported from clcbiogenomics workbench and analyzed in the R/Bioconductor environment loading DESeq. 2 26 , gplot, ggplot2 packages. The differentially expressed genes were analyzed based on a generalized linear model likelihood ratio test assuming negative binomial data distribution via DESeq. 2. Candidate genes were filtered to a minimum of 4X fold change and FDR-corrected p-value <0.05. Functional association enrichment analysis was conducted following the methodology of Du et al. 80 . RNA-Seq data's (raw read and mapped files) are available through the NCBI Gene Expression Omnibus with accession number GSE79011.
Time course quantitative real-time PCR validation. Quantitative RT-PCR using SYBR Green detection reagents in Step One Plus (Applied Biosystems, USA) was used to validate the expression of ten candidate genes shown to be differentially regulated from RNA-Seq data. Total RNA from the third biological replicate was used according to each time point. The sequence of each gene was obtained from the Rice Genome Annotation Project database, RGAP7 (http://rice.plantbiology.msu.edu/), and the sequences of genes were used for primers design using qPCR Assay Design tool of Integrated DNA Technology (IDT, http://sg.idtdna.com/site). Rice actin gene was selected as a reference gene in qRT-PCR. 2 −∆∆Ct method as described by Schmittgen and Livak 81 and used to determine the relative expression, and the log2 transformation of the relative expression was used to compare with the RNA-Seq data. All samples were studied in triplicate PCR. The primers and corresponding sequences are listed in Table S1.

Data availability
The RNA-Seq data (raw read and mapped files) are available through the NCBI Gene Expression Omnibus with accession number GSE79011.