Identification of bacterial blight resistance genes in rice landraces from Yunnan Province, China

Bacterial blight (BB), a serious bacterial disease caused by pathogen Xanthomonas oryzae pv. oryzae (Xoo) affects rice growth and yield. Yunnan Province is regarded as a center of rice diversity in China and indeed the world, and has abundant rice landrace resources, which may offer prospective candidate donors in rice improvement and breeding. In this study, a set of 200 rice landraces were evaluated to determine their resistance to 10 pathogenic Xoo strains resistance by the leaf-clipping method. The results indicated that the tested rice landraces had different resistance levels against different Xoo strains. Multiple comparisons showed that the Xoo strain PXO99 was virulent to the tested rice landraces. Sixty-six rice landraces conferred resistance against at least one Xoo strain. These resistant rice landraces screened were then performed the presence of 14 cloned BB resistance genes by closely linked molecular markers and designed specific primers. The results showed that none of these resistant accessions contained xa13, Xa21, Xa27, and Xa45(t) homologous fragments, while 9, 24, 4, 7, 9, 15, 1, 5, 4 and 27 accessions contained Xa1, Xa2/Xa31(t), Xa14, Xa3/Xa26, Xa4, xa5, Xa7, Xa10, Xa23 and xa25 homologous fragments, respectively. Sequence analysis further revealed that nucleotide variations around functional nucleotide polymorphisms region were observed within these accessions containing the Xa1, Xa2/Xa31(t), Xa14, Xa3/Xa26, Xa4, xa5, Xa10, Xa23 and xa25 homologous fragments. These results along with phenotypic resistance spectrum supported that these accessions carried nine resistance homologous genes. Only one accession (Qishanggu_Wenshan) carried the Xa7 resistance gene. We also found that some resistant rice landraces, especially Xilandigu_Baoshan, and Laoyaling_Lincang without the above resistance genes, which mediated broad spectrum resistance to multiple Xoo strains, were identified as potential sources for breeding rice lines resistance to BB.


Introduction
Rice (Oryza sativa L) is one of the most important domesticated food crops in the worldwide. Bacterial blight (BB) caused by pathogen Xanthomonas oryzae pv. oryzae (Xoo), is considered to be one of the most destructive disease of rice crop (Niño-Liu et al. 2006;Savary et al. 2019). Xoo races often make the breakdown of rice resistance by highly pathogenic variability in a short period, which attacks rice by principally depending on diverse effectors secreted through a type III secretion system (T3SS) (Gu et al. 2005). Transcription activator-like effectors (TALEs) as major virulence effectors of Xoo strains, primarily activate host gene expression by combining to specific promoter region (effector binding elements, EBEs) of host susceptibility genes or resistance genes, thereby triggering susceptibility or resistance symptoms, respectively (Römer et al. 2009;Ji et al. 2016). Rice, meanwhile, also constantly employs its own resistance genes to defend the fierce attacks from variable Xoo races (Ji et al. 2018). Nowadays, plant host resistance, chemical measures and nitrogen management have been employed to control BB caused by Xoo races in rice breeding practice (Sombunjitt et al. 2017;Chukwu et al. 2019). However, breeding new resistant rice cultivars would be an effective, economic and environmental approach to better control BB disease (Deng et al. 2018). Consequently, more researchers are interested in discovering available resistance genes for controlling the disease.
At present, more than 40 resistance genes have already been found to mediate rice resistance against Xoo racespecific (Neelam et al. 2020). However, many BB resistance genes are not cloned and characterized types of proteins due to their unknown sequences. A recent study reported that 5 additional cloned Xa1 allelic resistance genes, Xa2, Xa31(t), Xa14, CGS-Xo1 11 , and Xa45(t) from different rice varieties had highly conserved structure, and the variable number of 93 amino acid residues (279 bp) ranging from 4 in XA14 to 7 in XA45(t) at their C-terminal tandem repeats (LRR region) can distinguish the type of Xa1 alleles (Ji et al. 2020). Meanwhile, sequence analysis confirmed that Xa2 gene was the same as Xa31(t) gene (Ji et al. 2020;Zhang et al. 2020); hence, we refer to this gene as Xa2/Xa31(t) in this paper. The CGS-Xo1 11 gene is considered to be Xo1, which confers resistance to Xanthomonas oryzae pv. oryzicola (Xoc), but it is unclear whether they are the same gene (Ji et al. 2020). The other Xa1 allelic genes conferred specific BB resistance against Xoo strains. Additionally, 12 cloned BB resistance genes, including Xa1 (Yoshimura et al. 1998), Xa3/Xa26 (Sun et al. 2004), Xa4 (Hu et al. 2017), xa5 (Iyer and McCouch. 2004), Xa7 , Xa10 (Tian et al. 2014), xa13 (Chu et al. 2006), Xa21 (Song et al. 1995), Xa23 (Wang et al. 2015), xa25 (Liu et al. 2011), Xa27 (Gu et al. 2005) and xa41(t) (Hutin et al. 2015) have been reported for encoding different types of proteins. Out of these genes, Xa1 and its allelic genes (Xa2/Xa31(t), Xa14, Xa45(t)), xa5,Xa7,Xa10,xa13,Xa23,xa25,Xa27 and xa41(t) conferring race-specific resistance to certain Xoo are tightly interrelated with TALEs from the pathogen, but the three other resistance genes Xa3/Xa26, Xa4, and Xa21, encoding kinase proteins, are not induced by TALEs (Jiang et al. 2020). These resistance genes are important resources for rice breeding, such as pyramiding rice cultivars with different resistance genes have been applied to increase the resistance against Xoo (Sombunjitt et al. 2017;Sutrisno et al. 2018). Thus, it is essential to identify the present of the resistance genes in rice germplasm. However, the selection of rice varieties comprising resistant genes is difficult and time-consuming through the traditional approach alone. Molecular markers within closely linked to the target genes can be used to identify germplasm with one or more resistance genes (Perumalsamy et al. 2010). The application of gene-linked/specific molecular markers should be complemented by gene homologous cloning method and sequences analysis to effectively analyze the type of resistance genes .
Yunnan Province, known as one of the centers of rice genetic diversity, is alternately distributed with different mountains, plains and basins, and the diverse geographic and climatic conditions have contributed to the remarkably rich diversity of its rice landraces (Zeng et al. 2007). In comparison to modern rice cultivars, Yunnan rice landraces have many desirable agronomic traits, higher genetic diversity, various stress tolerance, as well as excellent resistance characteristics (Cui et al. 2016. However, only limited studies have been carried out to identify BB resistant varieties among the Yunnan rice landraces, such as Zhachanglong, a rice landrace from Yunnan Province, harbors Xa3/Xa26, Xa22(t), and Xa31(t) resistance genes, which confers resistance to multiple Chinese Xoo strains (Sun et al. 2004;Wang et al. 2003Wang et al. , 2009. However, there is still insufficient information on the responses of Yunnan rice landraces against Xoo strains, and lack of information on the identification of BB cloned genes. In this work, we selected 200 representative rice landraces by previously collected from different regions in Yunnan. Evaluation of these rice landraces resistance to multiple Xoo strains was performed by using a scissors-based inoculation method under the same planting conditions. We then screened resistant rice landraces to identify the presence and type of previously cloned resistance genes, including Xa1 and its allelic genes (Xa2/ Xa31(t), Xa14, Xa45(t)), Xa3/Xa26,Xa4,xa5,Xa7,Xa10,xa13,Xa21,Xa23,xa25, and Xa27, except for xa41(t) due to its nucleotide sequence unpublished. Our studies could provide a reference and valuable information for further utilizing Yunnan rice landraces in resistance breeding.

Xoo strains inoculation and phenotypic data analysis
Ten Xoo strains including one highly pathogenic strain from Philippines (PXO99), one representative race from Japan (T7147), two representative strains from China (C5 and C9), six pathogenic strains from Yunnan Province (Y8, YM1, YM187, HZHJ19, YJWS2 and YJDP2) were used to evaluate the resistance of rice landraces in Yunnan. All strains were cultured on Nutrient Agar medium by 28℃ for 48 h, then resuspended in double distilled water, and diluted to OD 600 nm = 0.8 for inoculation by using the NanoDrop2000. Each rice accession at booting stage (panicle development) was respectively inoculated with Xoo strains by the leafclipping method at 15:00 ~ 17:00 . Three replicates (three individual plants) for each accession were designed for each Xoo strain with five leaves inoculated per plant. Mock treatments with double distilled water and untreated plants served as the controls. Lesion lengths (cm) and leaf lengths (cm) were respectively measured across three leaves from each accession about 3 weeks after inoculation adopting the Standard Evaluation System (SES) for rice (IRRI, 2002). The leaf with lesion area 0-5% was scored 0, 6-10% was scored 1, 11-25% was scored 3, 26-50% was scored 5, 51-75% was scored 7, and 76-100% was scored 9. The lines with the score of 0, 1, 3, 5, 7 and 9 were considered as high resistance (HR), resistance (R), middle resistance (MR), middle susceptible (MS), susceptible (S), and high susceptible (HS), respectively. Analysis of variance (ANOVA) of lesion length rate (LLR) data was conducted by R programming language software (https:// www.r-proje ct. org). The R software package multcomp was used to conduct Tukey's multiple comparisons of the LLR between Xoo strains and varieties. The R software package ggplot2 was also performed to show the differences in the LLR values and disease levels among Xoo strains and varieties.

Linked molecular markers and designed specific primer
Total genomic DNA from the resistant rice landraces in Yunnan was isolated by using cetyltrimethylammonium bromide (CTAB) method, and was analyzed to determine the presence of 14 cloned BB genes Xa1 and its allelic genes (Xa2/Xa31(t), Xa14 and Xa45(t)), Xa3/Xa26,Xa4,xa5,Xa7,Xa10,xa13,Xa21,Xa23,xa25, and Xa27 by using genelinked markers or gene-specific primers reported, respectively (Table S2). The tested rice landraces were initially considered to contain disease-resistant allele markers of these genes, and further used to identify the homologous fragments of these genes by designing specific primers based on their nucleotide sequences in ORF region and polymorphic locus region related to resistance function (Table S2; Fig. S2).

PCR amplification and detection
For each PCR, 25 µL reaction mixture consisted of 12.5 µL 2 × Phanta® Max Master Mix (Vazyme, China), 1 µL of each forward and reverse primer (10 μmol L −1 ), 1 μL DNA (25 ng μL −1 ), 9.5 µL ddH 2 O. The PCR amplification program was as follows: 95℃ for 3 min; followed by 35 cycles at 95℃ for 15 s, 50-60℃ for 15 s, 72℃ for 30-60 s/kb, and finally extension at 72℃ for 5-10 min. The PCR amplified products were separated on 1%-3% (variable depending on product sizes) agarose gel electrophoresis. The presence of resistance genes in rice landraces were judged by comparing the fragments in corresponding control lines. If no or weak fragments on agarose gel, the above steps were repeated twice to ensure the accuracy of the results.

DNA sequencing analysis of selected rice landraces
The selected rice landraces carrying homologous fragments of resistance gene(s) around functional nucleotide polymorphisms region were then sequenced. The sequenced data were spliced into complete sequences by DNAMAN. The resistant and susceptible sequences of cloned genes were downloaded from NCBI (https:// www. ncbi. nlm. nih. gov/), except for Xa10 recessive allele, for which no published sequence was available. The spliced sequences were compared with the cloned gene sequences by using the DNA-MAN for multiple alignments, and their similarity/identity analysis were performed by using BLASTN programs (https:// blast. ncbi. nlm. nih. gov/ Blast. cgi).

Genotypic detection of cloned BB resistance genes in disease-resistant rice landraces
The availability of tightly linked molecular markers or reported gene-specific primers were exploited to detect the type of resistance genes of interest. To estimate the frequency of BB cloned resistance genes in the 66 resistant rice landraces, we first screened these resistant rice landraces by using the reported gene-linked markers or gene-specific primers of 14 cloned BB resistance genes, Xa1 and its allelic genes (Xa2/Xa31(t), Xa14 and Xa45(t)), Xa3/Xa26,Xa4,xa5,Xa7,Xa10,xa13,Xa21,Xa23,xa25, and Xa27, respectively. The results indicated that Xa1 and its allelic genes (Xa2/Xa31(t), Xa14, and Xa45(t)), Xa3/Xa26,Xa4,xa5,Xa7,Xa10,Xa23,xa25, and Xa27 genes could be detected their corresponding resistant allele markers in 56% (37), 10.6% (7), 47% (31), 22.7% (15), 12.1% (8), 7.6% (5), 6.1% (4), 40.9% (27), and 53% (35) accessions, respectively, while xa13 and Xa21 genes were absent in these rice landraces (Fig. S1, Table S3). Interestingly, we found that 63 accessions were comprised two or more disease-resistant allele markers (Table S3), but most of them showed narrower spectrum resistance to the tested Xoo strains in this study, especially those carrying the xa5, Xa23, and Xa27 resistant allele markers (Table S1). On the contrary, 3 accessions (Lengshuigu_Yuxi, Xilandigu_Baoshan and Laoyaling_Lincang) were found not to contain the above disease resistance allelic markers, but 2 accessions (Xilandigu_Baoshan and Laoyaling_Lincang) of them had broad-spectrum resistance (Table S1; Table S3). In general, a rice cultivar with multiple resistances genes provides more resistance than a single resistance gene (Singh et al. 2001;Rajpuroit et al. 2010). On the basis of this result, it was possible that some of the resistant allele markers detected in our study were false positive fragments.  To avoid whether our results included false positive allele markers with the same sizes as resistance genes, we chose the rice landraces with the above resistant allele markers and 2 accessions (Xilandigu_Baoshan and Laoyal-ing_Lincang) without the above resistance allele markers to further evaluate the homologous fragments of resistance genes using the designed specific primers (Fig. S2; Table S1). Xa1 and its allelic genes can be distinguished mainly by the variable number of LRR region at their C-terminal tandem repeats (Ji et al. 2020;Zhang et al. 2020). We designed two pairs of primers to amplify the ORF and LRR regions of Xa1 allelic genes in 37 accessions, respectively. The results indicated that 9 accessions were detected to contain Xa1 expected homologous fragments, 24 accessions were detected to contain Xa2/Xa31(t) expected homologous fragments, 4 accessions were detected to contain Xa14 expected homologous fragments, no accessions were detected to contained Xa45(t) expected homologous fragments ( Fig. 2A; Table S3). Similarly, we also designed specific primers for Xa3/Xa26, Xa4,xa5,Xa10,Xa23,xa25, and Xa27 to amplify their homologous fragments depending on their nucleotide sequences in ORF region and polymorphic loci region related to resistance function. The results showed that all 7 accessions with Xa3/Xa26 resistant allele marker (Fig. 2B), 15 accessions with xa5 resistant allele marker (Fig. 2C), 5 accessions with Xa10 resistant allele marker (Fig. 2D), 4 accessions with Xa23 resistant allele marker (Fig. 2E), 27 accessions with xa25 resistant allele marker were amplified to have expected homologous fragments of their corresponding resistance genes (Fig. 2F), respectively. For Xa4, 9 of 31 accessions with resistant allele marker had the homologous fragments of Xa4 resistance gene, whereas 22 accessions were absent the same fragments as IRBB4 ( Fig. 2G; Table S3). For Xa7, 1 of 8 accessions with resistant allele marker had the homologous fragments of Xa7 resistance gene ( Fig. 2H; Table S3). For Xa27, all 35 accessions with resistant allele marker could not amplify the expected fragments as their resistant/susceptible lines (Table S3). We also found that two broadspectrum resistant accessions, Xilandigu_Baoshan and Laoyaling_Lincang without resistant allele markers and the susceptible control IR24 were detected to carry Xa3/Xa26, xa5, xa25 homologous fragments (Fig. 2). Although some false positive resistant allelic accessions were excluded by homologous gene cloning method, the phenotypic resistance of some accessions with resistant homologous fragments was different from that of the resistance control lines/ varieties in this study (Table S1). We suspected that the phenotypic resistance difference in rice might be affected by nucleotide sequence variations, especially functional nucleotide polymorphisms region related to resistant/susceptible gene (Fujino et al. 2011;Wu et al. 2013), besides its own genetic background (Cao et al. 2007;Zhou et al. 2009).

Functional nucleotide sequences analysis of resistance genes in selected rice landraces
Xa1 and its allelic resistance genes can be distinguished on the basis of the substructure of variable 93-aa (279 bp) tandem repeats at the LRR domain, which are to be likely involving the strength of rice-Xoo resistance interactions (Ji et al. 2020;Zhang et al. 2020). We selected all 9 accessions with Xa1 (2010 bp), 24 accessions with Xa2/Xa31(t) (1731 bp) and 4 accessions with Xa14 (1572 bp) to compare the sequence identity of LRR region, respectively. These accessions sequenced were 99.8% identical to Xa1, 99.8% identical to Xa2, and 99.3% identical to Xa14 (Fig. S3). The results, along with phenotypic data, indicated that these accessions carried Xa1 homologous allelic resistance genes. A total of 9 accessions comprising Xa3/Xa26 homologous fragments including 2 accessions, Xilandigu_Baoshan and Laoyaling_Lincang without resistant allele marker around the resistant and susceptible function region (Xiang et al. 2006;Hur et al. 2013) were sequence alignment. Seven accessions had the same TGCA (+ 452-456 bp) in exon 1 region as that in the Xa3/Xa26 resistant allele, but 2 accessions with susceptible allele marker had AATC characteristic of the xa3/xa26 susceptible allele (Fig. S4). Xa4 and its recessive allele xa4 exists one base substitution (+ 456 C → G) in the exon 1 coding region, resulting in the replacement of aspartic acid (D) residue with glutamic acid (E). Residue D45 is speculated to be associated with resistance phenotype (Hu et al. 2017). Comparison sequences analysis revealed nine accessions were the same as xa4 (C), with a few base deletions and substitutions in the promoter region (Fig. S5).
xa5 is a natural allele of dominant susceptible gene Xa5. Sequence analysis of 15 accessions with resistant allele marker and two accessions, Xilandigu_Baoshan and Laoy-aling_Lincang with susceptible allele marker revealed that two nucleotide (TC) substitutions were identical with Xa5 in polymorphic loci related to resistance function, and single base mutation was appeared in partial accessions (Fig. S6). These results confirmed that these accessions contained Xa5 homologous gene, but not xa5 resistance gene. Xa7, Xa10 and Xa23 are three executor resistance genes. Xa7 is a recently cloned executor resistance gene, which has two types of EBE sequences in the promoter region, namely Type I (IRBB7, DV85, and AUS 308) and Type II (Lao Zao Gu, MOTIA, and DANGAR). And they have the same CDS sequences . One accession (Qishanggu_ Wenshan) amplified sequence was identical to that of Type I (IRBB7) (Fig. S7). Unlike other cloned BB resistance genes, the susceptible allele sequence of Xa10 has not been published. Sequence analysis indicated that five accessions were equal to Xa10 at coding region, but had a few base mutations and substitutions in the specific TALE AvrXa10 binding  (Fig. S8). Similarly, Xa23 and its susceptible allele xa23 share identical coding sequence, but xa23 lacks cognate TALE AvrXa23 binding element (EBE AvrXa23 ) due to 7 bp polymorphism bases insertion in the promoter region. All four accessions were found to have the same sequence as xa23 in the EBE AvrXa23 region (Fig. S9). These evidences confirmed that 4 accessions carried xa23 homologous allele gene, instead of Xa23 resistance gene. xa25 and its dominant susceptible allele Xa25 have four types of Xa25/xa25 alleles, including two recessive xa25 resistance alleles (Minghui63 and OsSWEET13 Nip /OsSWEET13 Kit ) and two dominant Xa25 susceptible alleles (OsSWEET13 ZS97 and OsSWEET13 IR24 ), are identified as dependent on EBE variability at promoter region correlated with gene function (Zhou et al. 2015;Cheng et al. 2017;Xu et al. 2019). All 27 accessions with xa25 resistance alleles shared cognate xa25 Nip (OsSWEET13 Nip ) EBE sites but also base changes in promoter region, and were considered to have xa25 Nip (OsSWEET13 Nip ) type homologous gene (Fig. S10), whereas Xilandigu_Baoshan and Laoyaling_Lincang with Xa25 susceptible allele marker had Xa25 IR24 and Xa25 ZS97 EBE sites in corresponding region, respectively.

Discussions
Some previous studies have reported phenotypic resistance to BB in rice by the leaf-clipping method (Fred et al. 2016;Jiang et al. 2019). Molecular markers can greatly facilitate the development of new disease resistant materials, and effectively detect the presence of BB resistance genes in rice (Ullah et al. 2012;Hur et al. 2013). Several studies have been reported to detect the genotype of resistance genes in rice by using gene-linked molecular markers (Singh et al. 2001;Hajira et al. 2016). For example, gene-linked markers were used to detect Xa4, xa5, and Xa21 resistance genes in Pakistan rice (Sabar et al. 2016), and xa5, xa13, Xa21 and Xa27 resistance genes in wild rice species (Xia et al. 2010). However, little information is known about the resistance phenotypes and specific resistance genes of Yunnan rice landraces to BB. In the present study, when we characterized the phenotypic resistance response of 200 rice landraces to ten different Xoo strains, 33% (66) showed specific resistance reactions against at least one strain. The PXO99 strain showed stronger pathogenicity than the other strains towards these tested rice landraces. Jiang et al. (2019) reported that PXO99 also exhibited high virulence in wild rice species. Out of the 66 resistant rice landraces identified in this study, 63 accessions contained the combination of two or more BB resistance allele genes by using gene specific molecular markers/primers reported. Rice cultivar comprising multiple resistant genes tend to show a broader spectrum compared with cultivars with a single resistance gene (Huang et al. 1997;Rajpurohit et al. 2010;Luo et al. 2012). In this study, however, most accessions carrying multiple resistance alleles, such as Erhuanggu_Kunming, Wuming_Dali, Lizihong_Yuxi, and Wulixiang_Wenshan only showed moderately resistance to a single Xoo strain. Two rice landraces, Xilandigu_Baoshan and Laoyaling_ Lincang, showed a broad spectrum resistance, despite the lack of resistant alleles. When considering phenotypic differences, the influence of rice genetic background should be taken into account, such as Xa3/Xa26, Xa4 had different levels of in different rice materials (Cao et al. 2007;Zhou et al. 2009). Sequence variations in resistant alleles can also influence the phenotypic differences (Konishi et al. 2008;Fujino et al. 2011;Wu et al. 2013).
Our results confirmed that many of rice landraces possessed one or more Xa1, Xa2/Xa31(t), Xa3/Xa26, Xa4, xa5, Xa10, Xa14, Xa23, and xa25 homologous genes, but did not necessarily show strong resistance because of nucleotide changes in their coding sequences and/ or promoter region. These results have provided the important information for the identification of BB resistance genes in Yunnan rice landraces. Xa1 allelic genes belong to nucleotide-binding and leucine-rich repeat (NLR) class of typical resistance genes, and confer similar resistance to BB (Ji et al. 2020). Some recent studies revealed that Xa1 could recognize multiple intact TALEs to mediate broad-spectrum resistance against Xoo (Ji et al. 2016;Tang et al. 2019). However, most Asian Xoo races harbor a set of truncated TALEs (also termed as interfering TALEs, iTALEs), which interfere with the recognition of total TALEs by XA1, and suppresses resistance mediated by Xa1 (Ji et al. 2016). The resistant reference line IRBB1 (Xa1) showed narrow-spectrum resistance in this study. Although all 37 rice landraces with Xa1 allelic genes shared higher sequence identity at the Xa1 allele, there were some base mutations in LRR region that may contribute to phenotypic differences in resistance. The three dominant resistance genes Xa3/Xa26, Xa4, and Xa21, encoding kinase proteins, are not induced by TALEs, of which Xa3/ Xa26 and Xa21 involve receptor-like kinases (RLK) proteins that confer race-specific resistance to multiple Xoo strains, while Xa4, encoding wall-associated kinase (WAK) protein, is an unusual resistance gene of rice. In this study, Xa3/Xa26 and Xa4 homologs were found in 7, and 9 accessions, while Xa21 was absent in all resistant rice landraces. Other studies also reported that Xa21 resistance gene was not widely distributed in rice (Xia et al. 2010;Jiang et al. 2019). Xa3 and Xa26 have been confirmed as actually the same gene and renamed Xa3/Xa26 (Sun et al. 2004;Xiang et al. 2006). Xa3/Xa26-mediated resistance to Xoo strain is affected by rice genetic background, and IRBB3 show higher resistance against Xoo strains than Minghui 63 (Cao et al. 2007;Zhou et al. 2009). Consistent with this, we observed that a weaker resistance spectrum in Minghui63 than IRBB3 in the study (Table S1). Moreover, compared with the control line IRBB3 and Minghui63, rice landraces with Xa3/Xa26 homologs showed different resistance spectrum. We speculated that they might be related to their own genetic backgrounds. Xa4 locates closely to the Xa3/Xa26 gene locus and confers a race-specific resistance against Xoo in rice (Ma et al. 1999;Sun et al. 2003;Zhang et al. 2009), which has been widely utilized in rice breeding project since 1970s. However, some rice resistant varieties with only Xa4 have become susceptible to Xoo in Asia (Krattinger et al. 2017). In the present study, the reference line IRBB4 (Xa4) was susceptible to the PXO99, T7147, C5, C9, HZHJ19, YJWS2 and YJDP2 strains, and only showed moderate resistant to YM1 and YM187 strains in this study.
xa5 encoding transcription factor IIA gamma subunit (TFIIAγ) protein, is distinct from its susceptible gene Xa5 by two nucleotide (TC) substitutions, resulting in an amino acid variation from valine acid (V) to glutamic acid (E) at 39th position (V39E) related to gene expression in rice (Chen et al. 1997;Iyer and McCouch 2004). The xa5 resistance gene confers BB broad resistance, which depends on TALEs-mediated resistance against Xoo race, but differs from the SWEET gene in rice (Jiang et al. 2006;Huang et al. 2016). In our studies, the tested rice landraces with the xa5 resistant allele had a narrower resistance spectrum than the control line IRBB5 (xa5), and existed two nucleotide substitutions (TC) with the same as Xa5. Previous studies suggested that Pakistani rice varieties possessed xa5 gene (Naveed et al. 2010;Sabar et al. 2016), but neither of those studies sequenced the allele. Two recessive resistance genes xa13 and xa25, belonging to nodulin MtN3/SWEET family, are promoter variant alleles of the dominant susceptible genes Xa13 (also known as Os8N3/SWEET11) and Xa25 (also known as Os12N3/SWEET13) in the EBE regions, respectively (Yang et al. 2006;Yuan et al. 2013;Zhou et al. 2015;Zaka et al. 2018). xa13 was absent in any of the resistant rice landraces tested in this study. Other studies reported that other rice varieties were rarely found the existence of xa13 (Hajira et al. 2016;Jiang et al. 2019). xa25 was first identified in indica variety Minghui63 (Chen et al. 2002;Liu et al. 2011). Previous studies have reported that two types of recessive xa25 alleles are Minghui63 and OsSWEET13 Nip /OsSWEET13 Kit from Nipponbare, Kitaake as resistance genes, while two dominant Xa25 alleles are OsSWEET13 ZS97 from Zhenshan97 and OsSWEET13 IR24 from IR24 as susceptibility genes (Zhou et al. 2015;Cheng et al. 2017). In this study, the rice landraces with xa25 resistance alleles contained xa25 Nip (OsSWEET13 Nip ) type homologous gene.
The other four dominant resistance genes Xa7, Xa10, Xa23 and Xa27 are known as executor dominant resistance genes, which differ from typical resistance gene, encoding relatively small proteins in rice (Porter et al. 2003;Zhang et al. 2015). Our studies indicated that Executor resistance genes were not widely distributed in Yunnan rice landraces. Similarly, Yang et al. (2019) reported that O. officinalis populations in Yunnan Province with strong resistance to Xoo strains were absent the Executor resistance genes. Executor genes confer broad and durable resistance by cognate TALE-dependent binding to their respective EBE in the promoter elements, leading to hypersensitive response in rice plants . The resistant and susceptible alleles of executor genes encode the identical protein, but they differ from each other in the promoter region correlated with nucleotide sequence polymorphisms. Xa10 gene expression depends on directly interaction between its promoter and AvrXa10, but its recessive allele has not been reported in rice (Gu et al. 2008;Wang et al. 2017). We found that rice landraces with Xa10 homologous gene had sequence variations in their cognate EBE AvrXa10 positions. In addition, they contributed narrower phenotypic resistance than that of the control line IRBB10. Xa23 is from O. rufipogon with an extremely broad spectrum resistance to BB . Xa23 and its recessive gene xa23 have the same sequence in their coding regions, but xa23 lacks the binding elements EBE AvrXa23 in its promoter region, which is closely related to disease resistance in rice (Wang et al. 2015). Our findings indicated that the EBE AvrXa23 were absent from the tested rice landraces, and their resistance phenotype was narrower than O. rufipogon. A previous study revealed that the EBE AvrXa23 is present only in the promoter region of the dominant resistance gene Xa23 (Cui et al. 2013). Xa27 is first reported and cloned executor resistance gene, and its specific expression is related to sequences in its promoter region like other executor genes (Gu et al. 2004;Bimolata et al. 2013). In the present study, 35 accessions were detected to lack of Xa27 homologs, although they possessed Xa27 resistance allele markers. By comparison, Xia et al. (2010) reported that O. rufipogon was found to harbor Xa27 resistance gene by using its molecular maker, whereas no further analyzed that O. rufipogon carried the type of Xa27 gene. Xa7 is a newly cloned executor resistance gene by three different research groups, respectively Luo et al. 2021;Wang et al. 2021). Just like Xa10, Xa23 and Xa27, Xa7 BB resistance is lost by a variation consisting of an 11-bp insertion and a base substitution (G to T) in EBE AvrXa7 region. Most cultivars, landraces, and wild accessions are lack of the Xa7 gene (Wang et al. 2021). Our results also indicated that the Xa7 gene was absent in most of the tested Yunnan rice landraces, only Qishanggu_Wenshan contained this gene and had a similar resistance spectrum to IRBB7 (Table S1). A notable finding in our studies is that Xilandigu_Baoshan and Laoyaling_Lincang with none of the above cloned resistance genes still have stable broad-spectrum in the later resistance identification (data not shown).

Conclusions
We have used a combination of phenotypic and genotypic identification methods to analyze the BB resistance levels and the BB cloned resistance genes of Yunnan rice landraces. A total of 200 representative rice landraces from different origins had been described as the distinct phenotypic resistance reactions towards 10 Xoo strains, of which 66 accessions were screened to confer specific resistance against at least one strain. The 66 resistant accessions were performed to analyze the distribution of 14 cloned BB resistance genes by linked molecular markers and homologous gene sequence method. These resistant accessions lacked of xa13, Xa21, Xa27 and Xa45(t) resistance genes. There were 9, 24,4,7,9,15,5,4,27 resistant accessions containing Xa1,Xa2/Xa31(t),Xa3/Xa26,xa4,Xa5,Xa10,Xa14,xa23 and xa25 Nip homologous genes, respectively. Only Qishanggu_Wenshan had Xa7 resistance gene. Remarkably, Xilandigu_Baoshan, and Laoyaling_Lincang conferring high and broad resistance to BB have potential applications as donor varieties for rice breeding. This work provides the important information about the BB resistance levels and the distribution of BB resistance homologous genes in rice landraces from Yunnan Province, China.