Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Pathotype and Genetic Diversity amongst Indian Isolates of Xanthomonas oryzae pv. oryzae

Pathotype and Genetic Diversity amongst Indian Isolates of Xanthomonas oryzae pv. oryzae

  • Deo Mishra, 
  • Manne Ramachander Vishnupriya, 
  • Madhusoodana Girija Anil, 
  • Kotilingam Konda, 
  • Yog Raj, 
  • Ramesh V. Sonti
PLOS
x

Abstract

A number of rice resistance genes, called Xa genes, have been identified that confer resistance against various strains of Xanthomonas oryzae pv. oryzae (Xoo), the causal agent of bacterial blight. An understanding of pathotype diversity within the target pathogen population is required for identifying the Xa genes that are to be deployed for development of resistant rice cultivars. Among 1024 isolates of Xoo collected from 20 different states of India, 11 major pathotypes were distinguished based on their reaction towards ten Xa genes (Xa1, Xa3, Xa4, xa5, Xa7, xa8, Xa10, Xa11, xa13, Xa21). Isolates belonging to pathotype III showing incompatible interaction towards xa8, xa13 and Xa21 and compatible interaction towards the rest of Xa genes formed the most frequent (41%) and widely distributed pathotype. The vast majority of the assayed Xoo isolates were incompatible with one or more Xa genes. Exceptionally, the isolates of pathotype XI were virulent on all Xa genes, but have restricted distribution. Considering the individual R-genes, Xa21 appeared as the most broadly effective, conferring resistance against 88 % of the isolates, followed in decreasing order by xa13 (84 %), xa8 (64 %), xa5 (30 %), Xa7 (17 %) and Xa4 (14 %). Fifty isolates representing all the eleven pathotypes were analyzed by southern hybridization to determine their genetic relatedness using the IS1112 repeat element of Xoo. Isolates belonging to pathotype XI were the most divergent. The results suggest that one RFLP haplotype that is widely distributed all over India and is represented in strains from five different pathotypes might be an ancestral haplotype. A rice line with xa5, xa13 and Xa21 resistance genes is resistant to all strains, including those belonging to pathotype XI. This three gene combination appears to be the most suitable Xa gene combination to be deployed in Indian rice cultivars.

Introduction

Xanthomonas oryzae pv. oryzae (Xoo) is the causal agent of bacterial blight, a serious disease of rice. Bacterial blight is endemic and causes serious yield losses for the rice crop grown in irrigated, low land areas across Asia. Host plant resistance is the most effective way of managing yield losses due to the disease as chemical control is not effective [1]. Almost thirty different rice genes (called Xa genes) that confer resistance against various races and pathotypes of Xoo have been identified [2]. Many of these resistance genes have been tagged with closely linked molecular markers and are being used in marker assisted selection [36]. These resistance genes display specificity with regard to their effectiveness against different pathogen races. Therefore knowledge of the pathotype diversity in the target pathogen population is essential for making an informed choice of resistance genes that are to be used in a breeding program.

In India, bacterial blight occurs in a large number of states with yield losses going up to 60-80% in severe infections. Pathotype studies conducted in India under the All India Coordinated Rice Improvement Project (AICRIP) during the 1970s and 1980s indicated that pathotypes Ia and Ib were distributed all over the country [7,8]. These two pathotypes exhibit similar reactions on differential cultivars and are characterized by incompatibility with rice varieties BJ1 (xa5 and xa13) and DV85 (xa5, Xa7 and xa24). They are distinguished from each other on the basis of reaction pattern on rice cultivar IR20 which carries the Xa4 resistance gene; pathotype Ia strains are incompatible on IR20 while pathotype Ib strains are compatible. IS1112, a repeat element native to Xoo, is a good probe for genotyping different strains of the pathogen as it is present in multiple copies and reveals substantial inter-strain variability [912]. Yashitola et al. (1997) performed pathotype analysis and DNA fingerprinting studies with IS1112 on a set of Indian isolates of Xoo that were collected between 1991 to 1995 [13]. They observed that the vast majority of strains (60/67) analyzed were incompatible with BJ1 and DV85 but were compatible with IR20. This suggested that all of these strains belonged to pathotype Ib and none were like pathotype Ia. RFLP haplotyping demonstrated that strains belonging to this pathotype clustered together in a dendrogram as a lineage comprised of closely related strains. The BXO1 Xoo strain was considered as the type strain for this lineage. The remaining strains (7/67) were found to be compatible with BJ1 and DV85 (therefore they were neither pathotype Ia nor pathotype Ib) and they had very diverse haplotypes. The BXO8 strain was considered as an example of this group of strains. Shanti et al. (2001) inoculated a collection of Xoo strains from Eastern India on a set of near-isogenic rice lines containing any one of several Xa genes in the IR24 rice varietal background [14]. They identified several Xoo strains that are compatible with the xa13 disease resistance gene and suggested that the combination of Xa4, xa5 and Xa21, when pyramided, would be effective against strains from Eastern India. Lore et al. (2011) inoculated 224 Xoo strains collected between 1999-2006 from the North western Indian state of Punjab on near-isogenic rice lines containing individual Xa genes and found that none of these strains were compatible with the xa13 disease resistance gene [15]. Approximately 93% of the isolates were incompatible with Xa21 while 95% of the isolates were found to be compatible with the xa5 resistance gene. In an earlier study of Xoo strains from the Punjab, several isolates were found to be compatible with xa13 [16].

All of these above mentioned studies were done with either a limited number of strains or were from one particular geographic area of India. In order to obtain a more comprehensive picture of the pathotype diversity of Xoo in India, we have collected 1024 isolates from twenty different states. These strains were inoculated on a set of near-isogenic lines containing any one of nine Xa genes which have been used in earlier studies to assess pathotype diversity of Indian strains of Xoo (14,15). Besides these lines, the IR8 rice variety which carries the Xa11 disease resistance gene was also included in the study. Thus, the effectiveness of ten different Xa genes was assessed and eleven different pathotypes were identified in this study. Representative strains from each of the pathotypes were subjected to DNA fingerprinting using the IS1112 probe and the relationship between RFLP haplotype and pathotype was analysed. The three gene combination of Xa21, xa13 and xa5 appears to be appropriate for development of bacterial blight resistant rice varieties in India.

Materials and Methods

Bacterial collection and maintenance

Infected leaf samples were collected during 2004 - 2012, from different rice growing states of India representing various geographical locations (Figure 1). Although a majority of the states were sampled, because of constraints arising from the fact that a vast region was involved, certain states were sampled sparsely or were not sampled at all. All the samples were collected from private rice fields with owner’s permission. Samples were placed in auto seal plastic packets with silica gel and stored in a refrigerator until isolation of the pathogen. Single colonies of cultures isolated from disease samples were picked up from Peptone Sucrose Agar (PSA) plates, maintained in liquid medium at 4°C for routine work and in 20% glycerol at -80°C for long term storage.

thumbnail
Figure 1. Statewise distribution of X. oryzae pv. oryzae strains collected from India.

X. oryzae pv. oryzae strains were collected from different states in India. The numbers in parentheses indicate the number of strains isolated from each state. Overall, the states from which more isolates were collected are those that have been more extensively surveyed.

http://dx.doi.org/10.1371/journal.pone.0081996.g001

Pathotype analysis

The seeds of bacterial blight differential rice lines (IRBB-1, 3, 4, 5, 7, 8, 10, 13, 21, IR8), the susceptible check line (IR24) and 7 different gene pyramid lines (IRBB-52, 54, 55, 57, 58, 59, and 60) were provided by the International Rice Research Institute (IRRI), Philippines. The plants were grown in plastic trays (55 x 40 x 15 cm) in a greenhouse. The trays were filled with a mixture of soil and farmyard manure at a ratio of 3:1. N-P-K were supplied to the plants at the rate of 100-75-74 kg/ha as the basal dose in the form of urea, superphosphate, and muriate of potash. Sowing of the differentials was done at 20 days intervals to get 2 stages of plants (60 and 40 days old) at the time of inoculation. The trays were irrigated every day. Adequate plant protection measures were taken to ensure healthy and vigorous growth of the plants. Plants were clip-inoculated with bacterial suspensions of 109 cfu/ml [17]. Four leaves per plant were inoculated for each isolate-cultivar combination for 60 and 40 days old plants.

Disease observations were taken 14 days after inoculation by measuring lesion length. Lesion lengths <5 cm were considered as resistant, 5-10 cm were considered as moderately resistant and >10 cm were considered as susceptible. Pathotype grouping was done based on the reaction pattern onto the differentials.

Genotype analysis

A total of 50 Xoo isolates representing all pathotypes were used for genomic DNA isolation and fingerprinting. Two previously genotyped strains from the study of Yashitola et al. (1997), BXO1 and BXO8, were also included in the present analysis for comparison.

DNA isolation and Southern hybridization

Xoo cultures were grown for 24 h in peptone sucrose medium on a rotary shaker at 28°C. Genomic DNA was isolated using the phenol-chloroform method. Genomic DNA was digested using restriction endonucleases EcoRI (New England Biolabs Inc., Ipswich, MA, USA) for 1 h at 37°C. One microgram of completely digested DNA from each strain was separated by electrophoresis on 0.7% agarose gels, denatured, neutralized, and vacuum-transferred to Hybond N+ (GE Healthcare Bio-Sciences, Uppsala, Sweden) membranes according to the procedure given by Sambrook et al [18]. Blots were pre-hybridized in a solution of 0.5 M sodium phosphate (pH 7.2), 7% sodium dodecyl sulfate (SDS), 1% bovine serum albumin, and 1 mM EDTA for 3 h at 65°C. Probes were labelled with α32P-dATP and hybridized for 18 h at 65°C with constant shaking. Blots were washed three times (20 min/wash) at 65°C first with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate, pH 7.0), 0.1% SDS, and 5 mM sodium phosphate (pH 7.0) and subsequently with 0.5× SSC, 0.1% SDS, and 3 mM sodium phosphate buffer (pH 7.0). Autoradiography was done with X-ray film. A 1 kilobase DNA ladder (New England Biolabs Inc.) was added to all the gels as a marker. After transfer, the membrane was cut, and the lane carrying the marker was hybridized separately.

RFLP analysis

Twenty nine different fragments present between 1 kb and 5 kb were used for genotyping. The haplotype (fragmenting pattern) of each strain obtained by Southern hybridization was compared with all the other strains, and presence or absence of a particular fragment was recorded as 1 and 0, respectively. Similarity matrix was prepared using the dice coefficient option and dendrograms were prepared using the UPGMA (unweighted pair group method of averages) option of the software FREETREE [19]. The confidence limits of dendrograms were determined by bootstrap analysis with 2,000 replications using the same program. The bootstrap values are expressed as a percentage of these 2,000 replications. The dendrogram was drawn using the software TreeView [20].

Using the same data, a minimum spanning tree (MST) was also constructed for phylogenetic analysis using the MST gold software programme available at www.bellinghamresearchinstitute.com /software [21]. The pair wise distances between strains were calculated using the equidistant method option in the software programme [21]. A 1000 bootstrap iterations were done on 500 unique MSTs and the MST with highest average bootstrap percentage was taken as the representative MST. The MST was visualized using the software GVEdit for Graphviz version 1.01 [22]. Using the same software, a minimum spanning consensus network was also made, using the pathotyping data in Table 1, for the eleven pathotypes identified in this study. A binary score was generated by taking either a resistant or a moderately resistant interaction as 1 and a susceptible interaction as 0.

Host DifferentialbX. oryzae pv. oryzae pathotypes a
IIIIIIIVVVIVIIVIIIIXXXI
IRBB1 (Xa1)S S S S S S S S S R, MS
IRBB3 (Xa3)SSSSSSSSSRS
IRBB4 (Xa4)SSSSR, MMR, MR, MSRS
IRBB5 (xa5)SSSSR, MRRRRR, MS
IRBB7 (Xa7)SSSSSSSRRR, MS
IRBB8 (xa8)SSR, MR, MR, MRR, MSSRS
IRBB10 (Xa10)SSSSSSSSSR, MS
IRBB13 (xa13)RRRRRSSSSR, MS
IRBB21 (Xa21)R, MSR, MSR MR, MRR, MRS
IR8 (Xa11)SSSSSRSSSR, MS
IR24SSSSSSSSSR, MS

Table 1. Pathotyping of Indian isolates of X. oryzae pv. oryzae.

a Based on mean lesion lengths, the nature of responses were classified into susceptible (S) (above 10 cm), moderately resistant (MR) (between 5-10 cm) and resistant (R) (up to 5 cm). The data presented are from experiments with 60 day old plants; similar results were obtained from inoculation of 40 day old plants. R, M is indicated when some isolates in the pathotype give a resistant reaction while other isolates give a moderately resistant reaction
b The Xa resistance gene in each host differential is given in parenthesis. IR8 is a cultivar with the Xa11 resistance gene and IR24 is the susceptible control.
CSV
Download CSV

Results

Pathotyping of X. oryzae pv. oryzae strains

A total of 1024 isolates of Xoo were collected from 20 different states in India (Figure 1) during 2004 to 2012. All of these isolates were pathotyped using 9 NILs harbouring Xoo resistance genes Xa1, Xa3, Xa4, xa5, Xa7, xa8, Xa10, xa13, Xa21 and one variety (IR8) that is reported to carry the Xa11 resistance gene. This analysis revealed 11 distinct pathotypes (Table 1; Figure 2). The pathotype III was the most frequent pathotype and accounted for 40.7 % of the isolates. Strains belonging to this pathotype were widely distributed throughout India with presence in 19/20 sampled states in the country (Table 2). Pathotype I was the second most prevalent pathotype comprising 20.5 % of the isolates that were analyzed. This pathotype was present in 14 of the states that were covered in this study. As compared to these two pathotypes, the remaining nine pathotypes were isolated less frequently with each of them being present in this study sample at a frequency that was <10%. Pathotype X exhibited an incompatible interaction with all differentials as well as susceptible control variety (IR24) indicating that this pathotype consists of less virulent strains. This pathotype was present in 11/20 sampled states and constituted 4 % of the total Xoo isolates. It is likely that pathotype X is a grouping of strains from other pathotypes that have, for unknown reasons, lost their virulence. Pathotype XI consisted of six isolates collected from Tripura in North-East India and showed a compatible interaction with all the differentials. Except for pathotype XI, all other pathotypes that are compatible with xa13 are incompatible with xa5 and those that are compatible with xa5 are incompatible with xa13. Pathotypes V and X are incompatible with both xa5 and xa13. Strains from all pathotypes, except for pathotype XI, are found in the Eastern Indian state of Odisha (Table 2). The pathotype III is the most frequent pathotype in 16/20 states covered in this study. Interestingly, pathotypes that are compatible with xa13 were not isolated in half of the states that have been surveyed. These include the states of Andhra Pradesh, Karnataka, Maharashtra, Gujarat, Himachal Pradesh, Uttarakhand, Jharkand, Kerala, Manipur and Madhya Pradesh. However, the sample size was limited in some of these states. Although a large number of isolates were collected from the states of Haryana and Punjab, only a few isolates (1 out of 72 isolates and 3 out of 130 isolates respectively) were compatible with xa13. In comparison, many of the isolates that are compatible with xa13 were found in the states of Eastern India, particularly in Odisha and Chhattisgarh (54 out of 152 and 25 out of 193, respectively).

thumbnail
Figure 2. Frequency of X. oryzae pv. oryzae pathotypes in India.

Eleven different pathotypes, designated as I to XI, were identified in this study. The x-axis indicates pathotype and the y-axis indicates % frequency amongst the 1024 isolates that were collected and analyzed.

http://dx.doi.org/10.1371/journal.pone.0081996.g002

StateaPathotypesTotal isolates
IIIIIIIVVVIVIIVIIIIXXXI
Andhra Pradesh14 (16.5)4 (4.7)56 (65.9)3 (3.5)5 (5.9)3 (3.5)85
Assam1 (14.3)4 (57.1)1 (14.3)1 (14.3)7
Bihar10 (18.2)9 (16.4)12 (21.8)1 (1.8)2 (3.6)21 (38.2)55
Chhattisgarh32 (16.6)1 (0.5)96 (49.7)4 (2.1)23 (11.9)17 (8.8)8 (4.1)12 (6.2)193
Gujarat12 (70.6)5 (29.4)17
Himachal Pradesh4 (57.1)1 (14.3)2 (28.6)7
Haryana12 (16.7)3 (4.2)32 (44.4)13 (18.1)8 (11.1)1 (1.4)3 (4.2)72
Jharkhand6 (100)6
Karnataka1 (6.7)10 (66.7)4 (26.7)15
Kerala1 (12.5)2 (25.0)4 (50.0)1 (12.5)8
Manipur2 (66.7)1 (33.3)3
Maharashtra17 (38.6)5 (11.4)18 (40.9)3 (6.8)1 (2.3)44
Madhya Pradesh1 (14.3)5 (71.4)1 (14.3)7
Odisha26 (17.1)20 (13.2)34 (22.4)10 (6.6)4 (2.6)5 (3.3)22 (14.5)10 (6.6)17 (11.2)4 (2.6)152
Punjab36 (27.7)9 (6.9)59 (45.4)3 (2.3)16 (12.3)1 (0.8)2 (1.5)4 (3.1)130
Tamil Nadu2 (25.0)4 (50.0)2 (25.0)8
Tripura1 (7.1)3 (21.4)4 (28.6)6 (42.9)14
Uttar Pradesh24 (25.0)3 (3.1)36 (37.5)4 (4.2)4 (4.2)14 (14.6)1 (1.0)10 (10.4)96
Uttarakhand27 (39.7)5 (7.4)16 (23.5)8 (11.8)1 (1.5)11 (16.2)68
West Bengal7 (18.9)1 (2.7)10 (27.0)3 (8.1)2 (5.4)14 (37.8)37
1024

Table 2. Distribution of X. oryzae pv. oryzae pathotypes in different states of India.

a The number of isolates of each pathotype is provided along with the total number of isolates collected from the state . The numbers in parenthesis indicate % frequency.
CSV
Download CSV

Considering the individual R-genes, Xa21 appeared as the most broadly effective, conferring resistance against 88 % of the isolates, followed in decreasing order by xa13 (84 %), xa8 (64 %), xa5 (30 %), Xa7 (17 %) and Xa4 (14 %) (Figure 3). The R-genes Xa1, Xa3 and Xa10 were not effective against any of the pathotypes, except for those strains belonging to pathotype X.

thumbnail
Figure 3. Effectiveness of Xa genes against Indian X. oryzae pv. oryzae isolates.

The x-axis indicates resistance gene and the y-axis indicates the frequency of isolates against which this gene confers resistance. Inoculation of rice leaves with isolates was done as described in Methods and lesion lengths were measured 14 days after inoculation. Lesion lengths <5 cm are considered as resistant, 5 to 10 cm are considered as moderately resistant and >10 cm are considered as susceptible.

http://dx.doi.org/10.1371/journal.pone.0081996.g003

The reaction of the isolates against a set of gene pyramid lines carrying several combinations of ‘Xa’ genes was assessed (Table 3). The IRBB52 line which harbours both Xa4 and Xa21 genes exhibited a moderate level of resistance against pathotype II and a resistant reaction against pathotype VI. The NILs that carry either one of these individual resistance genes exhibited a susceptible reaction with pathotype II and a moderate level of resistance against pathotype VI. In particular, the reaction with pathotype II indicates that the presence of both of the ‘defeated’ genes leads to an enhanced level of resistance, a phenomenon that has been previously described and termed as quantitative complementation (QC) [3,4,23]. Similarly QC was also observed in IRBB 52, IRBB54 (xa5 + Xa21), IRBB55 (xa13 + Xa21) and IRBB 58 (Xa4 + xa13 + Xa21) lines which exhibited a moderate level of resistance against pathotype XI as compared to the susceptible reaction in lines carrying the respective single resistance genes. In another example of QC, IRBB54 exhibited a moderate level of resistance towards pathotype IV as compared to the susceptible interaction exhibited by lines carrying either xa5 or Xa21. In an important observation, the IRBB59 line (xa5 + xa13 + Xa21) is found to exhibit resistance to pathotype XI even though the lines carrying each of the individual genes are susceptible. Interestingly, the IRBB55 line exhibits a susceptible response towards pathotype VI while the line carrying Xa21 alone exhibits a moderate level of resistance against this pathotype. This suggests that, in the presence of xa13, there is a reduction in the level of resistance that is provided by Xa21 against pathotype VI.

Gene Pyramidsb Xoo pathotypes a
I II III IV V VI VII VIII IX X XI
IRBB52 (Xa4 + Xa21) R Mc R S R Rc R R R R Mc
IRBB54 (xa5 + Xa21) R S R Mc R R R R R R Mc
IRBB55 (xa13 + Xa21) R R R R R Sd M R M R Mc
IRBB58 (Xa4 + xa13 + Xa21) R R R R R M R R M R Mc
IRBB59 (xa5 + xa13 + Xa21) R R R R R R R R R R Rc
IRBB60 (Xa4 + xa5 + xa13 + Xa21) R R R R R R R R R R Rc

Table 3. Reactions of X. oryzae pv. oryzae pathotypes on Xa gene pyramid lines.

a Based on mean lesion length, the nature of responses were classified into susceptible (S) (above 10 cm), moderately resistant (M) (between 5-10 cm) and resistant (R) (up to 5 cm).
b The Xa genes that are present in the gene pyramid line is indicated in parentheses.
c The gene pyramid lines exhibit more resistance than the lines having the single ‘R’ genes.
d The gene pyramid line exhibits less resistance than the lines having the single ‘R’ genes.
CSV
Download CSV

In order to determine the shortest path by which the eleven pathotypes could be derived from each other, we have performed a minimum spanning network analysis (Figure S1) using the pathotyping data from Table 1. We have added, to the different edges (connecting lines) in this network, a hypothetical directionality by assuming that in most cases the evolution of a pathotype would have involved acquisition of the ability to break down a host resistance gene. Pathotype V is incompatible with the xa5 and the xa13 resistance genes. Therefore, pathotype V could be an ancestral pathotype from which strains that are compatible with xa5 or xa13 could have evolved by acquisition of the ability to breakdown these resistance genes. The hypothesized change from pathotype V to pathotype III might have involved acquisition of the ability to break down the xa5 resistance gene as well as Xa4. Pathotype II might have arisen from pathotype I through acquisition of the ability to break down the Xa21 resistance gene or from pathotype IV through acquisition of the ability to break down the xa8 resistance gene. Pathotype XI may have arisen from pathotype II through acquisition of the ability to breakdown the xa13 resistance gene. The hypothesized change from pathotype V to pathotype VII might have involved acquisition of the ability to break down the xa13 resistance gene. The hypothesized change from pathotype VII to VI would have resulted in susceptibility to the Xa11 resistance gene and the change from pathotype VII to VIII would have required acquisition of the ability to break down the xa8 resistance gene. In comparison to pathotype VII, pathotype VIII is also avirulent on the Xa7 resistance gene. Pathotype IX may have arisen from pathotype VIII through acquisition of the ability to breakdown the Xa4 resistance gene.

RFLP analysis of Xanthomonas oryzae pv. oryzae strains

RFLP based genotyping of 50 isolates belonging to 11 different pathotypes was done by scoring 29 bands obtained by using IS1112, an Xoo insertion sequence element, as the DNA fingerprinting probe. Five strains each of pathotypes V, VIII, IX and X; four strains each of pathotypes I, III, IV and VI; six strains each of pathotypes II and VII and two strains of pathotype XI were analyzed (Table 4). Two other strains, namely BXO1 and BXO8, which were pathotyped and genotyped in an earlier study [13] were also included as reference strains. The RFLP and phylogenetic analysis using UPGMA revealed that 20 haplotypes were present amongst the 50+2 isolates that were studied (Figures 4 and Figure S2). The most frequent RFLP haplotype consisted of 13 isolates including the following: 4 strains belonging to pathotype III, each of which was isolated from four different states (Andhra Pradesh, Chhattisgarh, Punjab and West Bengal), 4 strains belonging to pathotype V isolated from different states (Andhra Pradesh, Bihar, Maharashtra and Punjab) as well as two strains each of pathotypes I and IV and one strain from pathotype X. This indicates that pathotypes III as well as V consisted of genetically closely related strains which had dispersed to several widely distributed locations in India. The BXO1 strain, which was included as a reference strain, is found to have the same RFLP haplotype as these 13 strains. The study of Yashitola et al. (1997) had indicated that strains with the same RFLP haplotype as BXO1 were widely distributed in India. The BXO1 strain belongs to pathotype III. Isolates of pathotype VI were isolated from a specific region in India (the state of Odisha) and all the four genotyped isolates of this pathotype have a single haplotype. Isolates of pathotype IX that were isolated from Odisha (East), Uttar Pradesh (North) and Tamil Nadu (South) clustered together in the dendrogram. In contrast isolates belonging to pathotypes I, II and X were found to be composed of genetically diverse strains as they grouped into different clusters. Genotyping of two strains belonging to pathotype XI indicated that they belong to a haplotype which is quite divergent from the rest of the strains. Also, the BXO8 strain which had been previously found to be a very diverse strain [13], is also found to be an outlier in this study.

Sl. No.  IsolateLocationRice cultivaraYear of isolationPathotype
1IXO221Andhra Pradesh MTU10102005I
2IXO278Uttar Pradesh Arize 64442005I
3IXO812Andhra Pradesh Arize 65162007I
4IXO675Punjab Pusa 442007I
5IXO89UttarakhandNA2004II
6IXO884Maharashtra MTU 70292008II
7IXO685Kerala NA2006II
8IXO725Maharashtra H930252007II
9IXO411Chhattisgarh Mahamaya2006II
10IXO651Punjab Pusa Basmati 12006II
11IXO639Punjab Pusa 442006III
12IXO281Andhra Pradesh MTU10012005III
13IXO74Chhattisgarh Arize 64442004III
14IXO189West Bengal H930242005III
15IXO92Punjab PR1182005IV
16IXO367UttarakhandNA2006IV
17IXO35Chhattisgarh IRBB 52004IV
18IXO98Himanchal Pradesh NA2005IV
19IXO99Andhra Pradesh BPT 52042005V
20IXO518Gujarat Arize 64442006V
21IXO414Bihar Arize 61292006V
22IXO134Maharashtra NA2005V
23IXO220Punjab Arize 61112005V
24IXO620Odisha MTU 70292005VI
25IXO644Odisha NA2005VI
26IXO621Odisha Surendra2005VI
27IXO630Odisha NA2005VI
28IXO222Odisha NA2005VII
29IXO599Uttar Pradesh Arize 64442005VII
30IXO597Uttar Pradesh Arize 64442005VII
31IXO631Odisha NA2005VII
32IXO390Odisha NA2005VII
33IXO792Chhattisgarh MTU 70292007VII
34IXO90Uttar Pradesh BPT 52042004VIII
35IXO493Odisha NA2006VIII
36IXO169Uttar Pradesh Arize 64442005VIII
37IXO141Odisha Surendra2005VIII
38IXO365Odisha MTU 70292005VIII
39IXO93Uttar Pradesh Indrasan2005IX
40IXO608Odisha MTU 70292007IX
41IXO603Odisha MTU 70292007IX
42IXO159Uttar Pradesh Roopali2005IX
43IXO97Tamil Nadu NA2005IX
44IXO53Karnataka NA2004X
45IXO151Haryana HKR 472005X
46IXO842Chhattisgarh Arize 64442007X
47IXO645Odisha Arize Tej2005X
48IXO704Punjab Rasi 70012007X
49IXO1088Tripura MTU 70292009XI
50IXO1104Tripura MTU 70292009XI

Table 4. List of representative X. oryzae pv. oryzae strains that were used for Restriction Fragment Length Polymorphism Analysis.

a Rice cultivar from which the strain was isolated; NA indicates that the information is not available
CSV
Download CSV
thumbnail
Figure 4. Restriction Fragment Length Polymorphism analysis of X. oryzae pv. oryzae strains.

Genomic DNA was isolated from 52 X. oryzae pv. oryzae strains and southern analysis was performed using the IS1112 probe as described in Methods. These 52 strains included 50 strains from eleven pathotypes identified in this study as well as X. oryzae pv. oryzae strains BXO1 and BXO8 from the study of Yashitola et al (13). Twenty different haplotypes were identified and at least one representative of each haplotype is shown. The IXO strain number is indicated on each lane along with pathotype. A one kb DNA ladder was added as a size marker. X. oryzae pv. oryzae strain BXO1 is loaded in the second lane of each blot for comparison. Each strain was analyzed at least three times, and 29 bands that were consistently visible in all the three replicates were used for scoring.

http://dx.doi.org/10.1371/journal.pone.0081996.g004

Additional phylogenetic analysis was performed, using the same RFLP fragment data, by constructing a minimum spanning tree which can help in hypothesizing on ancestral or emergent RFLP haplotypes (Figure 5). The RFLP haplotype representing 13 strains (14 including BXO1) occupies an approximately central position in the tree. As indicated above, this is a widely distributed RFLP haplotype in India and includes strains from five different pathotypes. These observations, taken together, suggest that this might be an ancestral RFLP haplotype. Interestingly, this RFLP haplotype includes 4 strains of pathotype V which is hypothesized to be an ancestral pathotype. An RFLP haplotype that is shared by 7 strains (four of pathotype VI and one each of pathotypes VII, VIII and X) is connected to this haplotype and could be derived from it. Also, the postulated ancestral RFLP haplotype is connected by an edge having a strong bootstrap value to an RFLP haplotype shared by two pathotype VII strains suggesting that the latter might be an emergent haplotype. A different RFLP haplotype (strain IXO93.9) is connected to five different haplotypes. Support for this grouping is provided by the observation that two nodes representing three strains belonging to pathotype IX are connected to IXO93.9. Interestingly, two pathotype II strains that share an RFLP haplotype are connected to IXO93.9 through an edge that has a high bootstrap value. As was observed in the tree made by UPGMA, the pathotype II strains appear at multiple places in the minimum spanning tree indicating that they are a diverse set of strains. Further studies on pathotype II strains using additional rice differentials might reveal significant pathological differences between them.

thumbnail
Figure 5. Phylogenetic relationships of X. oryzae pv. oryzae strains analyzed using a Minimum spanning tree.

A minimum spanning tree was drawn using the RFLP data. The digits in the edges represent percentage boot strap values obtained after 1000 iterations. A solid bold edge indicates a bootstrap value over 90%; solid line indicates a value over 70% and a dotted line indicates a value below 70%.

http://dx.doi.org/10.1371/journal.pone.0081996.g005

Discussion

Based on the study sample of 1024 strains, and the differential rice lines used in this study, eleven different pathotypes of this bacterium have been identified. Although a number of pathotypes were identified, pathotype XI isolated from the Eastern Indian state of Tripura was very unique in that it was compatible with all the tested Xa genes. Fortunately, the gene pyramid line carrying Xa21, xa13 and xa5 is resistant to this strain. In view of the possible dispersal of this pathotype to other locations in India, efforts should be made to deploy these three resistance genes in the genetic background of important Indian rice cultivars.

Genotype analysis, using the IS1112 probe, carried out on a subset of 50 strains that are drawn from the 11 pathotypes (plus the reference strains BXO1 and BXO8) revealed 20 RFLP haplotypes indicating that the genetic base of the Indian Xoo population is fairly diverse. At the same time, the bootstrap values for many of the clusters were not significant. Multiple RFLP probes, PCR based finger printing methods and genotyping using single nucleotide polymorphisms (SNPs) might be needed to delineate the phylogenetic relationships between these strains in a more accurate manner. Using the same RFLP data, a phylogenetic tree was also constructed using minimum spanning tree analysis. One particular RFLP haplotype shared by thirteen different isolates plus BXO1 could be an ancestral haplotype as it is centrally located in the tree and five different pathotypes have this haplotype.

Although it might be considered as being too simplistic, we have tried to use the data in Table 1 to generate a minimum path by which these eleven pathotypes could have arisen. As pathotype V strains are incompatible with both of the xa5 and xa13 disease resistance genes, pathotypes that are compatible with either one or the other of these two resistance genes could have arisen by the acquisition of specific TAL (transcription activator-like) effectors by an ancestral pathotype V strain. The Xa5 gene encodes transcription factor TFIIAγ and is presumably needed by Xoo TAL effectors to upregulate specific host susceptibility genes [2426]. The xa5 disease resistance gene probably does not support this role of TAL effectors in up-regulating expression of host susceptibility factors. Xoo strains that can breakdown xa5 mediated resistance are postulated to produce TAL effector/s that can utilize an alternate TFIIAγ gene in rice [25]. Therefore a pathotype (such as pathotype III) that can breakdown xa5 mediated resistance could have arisen from an ancestral pathotype V like strain by acquisition of TAL effector/s that can utilize the alternate TFIIAγ. Pathotypes I, II, III and IV differ from pathotype V in their ability to overcome xa5 as well as Xa4 mediated resistance. We speculate that there might have been some bridging strains that were able to overcome xa5 mediated resistance but were unable to overcome Xa4 mediated resistance. The previously described pathotype Ia which is reported to be incompatible with Xa4 could be such a pathotype.

The xa13 disease resistance gene represents a mutation in the promoter of Os8N3, a host susceptibility gene that encodes a sugar transporter (SWEET protein) [2729]. Xoo strains that overcome xa13 mediated resistance are capable of producing a TAL effector that can upregulate the expression of an alternate SWEET gene of the host [30]. It is possible that pathotype VII could have arisen from pathotype V by acquisition of such a TAL effector. Although it is speculative, we postulate that the first step in the evolution of Pathotypes VI, VII, VIII and IX (that can all breakdown xa13 mediated resistance) would have been the acquisition of such a TAL effector by an ancestral pathotype V like strain.

Does the minimum spanning tree in Figure 5 provide any support for the notion that many of the current day Indian pathotypes of Xoo could have arisen from an ancestral pathotype V like strain? The RFLP haplotype that has been suggested above to be an ancestral haplotype includes four out of five pathotype V strains analyzed in this study. Moreover, several pathotype III, pathotype IV and pathotype I strains share the same RFLP haplotype. Therefore, it is possible that pathotypes III, IV and I could have evolved from an ancestral pathotype V strain. Figure 5 also suggests that RFLP haplotypes that represent pathotype VII could have emerged from this supposedly ancestral RFLP haplotype. Therefore, it is possible that pathotype VII strains could have arisen from an ancestral pathotype V like strain. Figure 5 also suggests that at least some pathotype VIII and IX strains could have arisen from pathotype VII strains.

The suggestion that pathotype V might be an ancestral pathotype from which other pathotypes arose through acquisition of compatibility against either xa5 or xa13 provides an explanation for the observation that the pathotypes that are compatible with xa5 are incompatible with xa13 and vice versa. The exceptions include pathotype X strains which are incompatible with all the rice lines tested in this study. We suggest that pathotype X strains might have lost virulence as growth ceased in dried lesions of rice leaves. Spontaneous loss of virulence has been previously reported from aging cultures of Xoo [31]. Pathotype XI strains are compatible with both xa5 and xa13 disease resistance genes; one possible way in which this could have occurred is through acquisition of the ability to overcome xa13 mediated resistance by a pathotype (such as pathotype II) that is already compatible with xa5. An alternative possibility that pathotype XI strains arose through acquisition of the ability to overcome xa5 mediated resistance by a strain that is already compatible with xa13 cannot be ruled out and indeed does have some support from the minimum spanning tree analysis (Figure 5).

A number of strains belonging to pathotypes III and V had the same haplotype even though they were isolated from widely separated locations. This suggests the possibility that they have been dispersed through seed. Pathotype IX strains cluster together in the dendrogram indicating that they are closely related strains. These strains were isolated from three geographically separated states in India; an observation that is again suggestive of the possibility of dispersal through seed. The BXO1 strain of Xoo has been previously described by Yashitola et al. (1996) as belonging to pathotype Ib which belongs to a widely distributed lineage of this pathogen in India. In that study, the same RFLP haplotype as that of the BXO1 strain was found in 15/67 strains that had been analyzed with the IS1112 probe and these strains were widely distributed in India. The results presented here indicate that the BXO1 strain has the same RFLP haplotype as the most prevalent RFLP haplotype found in this study. The current study confirms the results of Yashitola et al. (1996) and indicates that strains with the same RFLP haplotype as BXO1 continue to be widely distributed in India. The same RFLP haplotype had also been found to be present in a majority of Xoo isolates collected from a wild rice species (Oryza nivara) growing naturally in Southern India and it had been suggested that this haplotype might have transferred from wild rice to cultivated rice [32].

Among the Xoo resistance genes studied, Xa21 was found to be most effective towards Indian Xoo strains followed by xa13 and xa8. All three genes appear to be good candidates to be deployed in Indian rice cultivars. The current study also throws light on the suitability of different Xa gene combinations for deployment in India. Since the most effective ‘R’ genes are Xa21 and xa13, a combination of these two genes will be a natural choice. However, as per this study, a number of Indian Xoo strains that are compatible with xa13 are incompatible with xa5. Therefore, a gene combination of Xa21, xa13 and xa5 might be an added advantage. This gene combination is also effective against pathotype XI. Two different research groups in India have pyramided the Xa21 and xa13 resistance genes into the popular rice varieties Pusa Basmati-1 and Samba Mahsuri [5,6]. In the latter work, the xa5 resistance gene was also incorporated along with Xa21 and xa13 to create a three gene pyramid line.

The present study also shows examples of a phenomenon called quantitative complementation (QC) wherein the presence of two different ‘Xa’ genes provides an increased level of resistance as compared to either of the single resistance genes. In this study, as described in Table 3, several such examples of QC could be observed. Interestingly, we have also observed an example of antagonistic interaction between Xa genes, wherein the pyramid of Xa21 and xa13 was less resistant than the line having Xa21 alone. This antagonistic interaction was observed when this gene pyramid was inoculated with isolates belonging to pathotype VI. A similar example of pathotype specific antagonistic interaction between Xa21 and xa13 has been previously described [33]. A negative interaction between Xa21 and xa5 only in the genetic background of a specific rice cultivar has also been previously described [34]. These examples highlight the need to take into consideration the possible effects of genetic background of the cultivar and pathotype prevalence in the region while making decisions on deployment of Xa gene pyramid lines in bacterial blight affected rice growing regions.

In summary, 1024 Xoo strains were collected from 20 different states in India and subjected to pathotyping and RFLP analysis. This has provided interesting insights into the genetic and pathotypic diversity of Indian strains of Xoo. However, several states were either not sampled or were sampled rather sparsely. Also, very little information is available about the TAL effectors and the host susceptibility genes that are used by Indian strains of Xoo to break down rice resistance genes such as xa5, xa13, etc. Future studies should be aimed at addressing these issues.

Supporting Information

Figure S1.

A Minimum spanning network of eleven different X. oryzae pv. oryzae pathotypes. The network was developed using data from Table 1 as indicated in methods. The direction of the hypothetical change is indicated by the arrow. The resistance gene against which compatibility is gained (+) or lost (-) during the change from one pathotype to another is indicated. Dotted lines are given when alternate edges are possible. The digits given on the dotted edges indicate the percentage of minimum spanning trees having that particular edge.

doi:10.1371/journal.pone.0081996.s001

(TIF)

Figure S2.

Dendrogram of 52 Indian strains of X. oryzae pv. oryzae derived from restriction fragment length polymorphism analysis using the IS1112 repeat element. The dendrogram was constructed and bootstrap values calculated as described in methods. The digits in the nodes represent percent boot strap values after 2000 iterations. The IXO number along with the pathotype is indicated for each strain. The scale bar represents genetic divergence. The BXO1 and BXO8 strains previously described by Yashitola et al. 1997 (13) were included for comparison.

doi:10.1371/journal.pone.0081996.s002

(TIF)

Author Contributions

Conceived and designed the experiments: DM MGA RVS. Performed the experiments: DM MRV MGA KK YR. Analyzed the data: DM MRV MGA RVS. Contributed reagents/materials/analysis tools: DM MGA RVS. Wrote the manuscript: MGA DM RVS.

References

  1. 1. Singh R, Das B, Ahmed K, Pal V (1980) Chemical control of bacterial leaf blight of rice. International Journal of Pest Management 26: 21-25. doi:10.1080/09670878009414278.
  2. 2. Liu-Nino DO, Ronald PC, Bogdanove AJ (2006) Xanthomonas oryzae pathovars: model pathogens of a model crop. Mol Plant Pathol 7: 303-324. doi:10.1111/j.1364-3703.2006.00344.x. PubMed: 20507449.
  3. 3. Yoshimura S, Yoshimura A, Iwata N, McCouch SR, Abenes ML et al. (1995) Tagging and combining bacterial blight resistance genes in rice using RAPD and RFLP markers. Molecular Breeding 1: 375-387. doi:10.1007/BF01248415.
  4. 4. Huang N, Angeles E, Domingo J, Magpantay G, Singh S et al. (1997) Pyramiding of bacterial blight resistance genes in rice: marker-assisted selection using RFLP and PCR. Theoretical and Applied Genetics 95: 313-320. doi:10.1007/s001220050565.
  5. 5. Joseph M, Gopalakrishnan S, Sharma R, Singh V, Singh A et al. (2004) Combining bacterial blight resistance and Basmati quality characteristics by phenotypic and molecular marker-assisted selection in rice. Molecular Breeding 13: 377-387. doi:10.1023/B:MOLB.0000034093.63593.4c.
  6. 6. Sundaram RM, Vishnupriya MR, Biradar SK, Laha GS, Reddy GA et al. (2008) Marker assisted introgression of bacterial blight resistance in Samba Mahsuri, an elite indica rice variety. Euphytica 160: 411-422. doi:10.1007/s10681-007-9564-6.
  7. 7. Reddy A, Srinivasan T, John V (1980). Breeding for bacterial leaf blight resistance in rice. Coimbatore, India: Tamil Nadu Agricultural University. pp. 5-13.
  8. 8. Seshu DV (1989). Salient findings from the multilocation evaluation of the international rice bacterial blight nursery. Manila, Philippines: International Rice Research Institute. pp. 167 - 176.
  9. 9. Leach JE, Rhoads ML, Cruz CV, White FF, Mew TW et al. (1992) Assessment of genetic diversity and population structure of Xanthomonas oryzae pv. oryzae with a repetitive DNA element. Appl Environ Microbiol 58: 2188-2195. PubMed: 1353345.
  10. 10. Nelson RJ, Baraoidan MR, Cruz CMV, Yap IV, Leach JE et al. (1994) Relationship between phylogeny and pathotype for the bacterial blight pathogen of rice. Appl Environ Microbiol 60: 3275-3283. PubMed: 16349380.
  11. 11. Leach JE, Leung H, Nelson RJ, Mew T-wW (1995) Population biology of Xanthomonas oryzae pv. oryzae and approaches to its control. Current Opinion in Biotechnology 6: 298-304. doi:10.1016/0958-1669(95)80051-4.
  12. 12. Adhikari TB, Cruz C, Zhang Q, Nelson RJ, Skinner DZ et al. (1995) Genetic Diversity of Xanthomonas oryzae pv. oryzae in Asia. Appl Environ Microbiol 61: 966-971. PubMed: 16534980.
  13. 13. Yashitola J, Krishnaveni D, Reddy AP, Sonti RV (1997) Genetic diversity within the population of Xanthomonas oryzae pv. oryzae in India. Phytopathology 87: 760-765. doi:10.1094/PHYTO.1997.87.7.760. PubMed: 18945099.
  14. 14. Shanti ML, George M, Cruz CV, Bernardo M, Nelson R et al. (2001) Identification of resistance genes effective against rice bacterial blight pathogen in eastern India. Plant Disease 85: 506-512. doi:10.1094/PDIS.2001.85.5.506.
  15. 15. Lore JS, Vikal Y, Hunjan MS, Goel RK, Bharaj TS et al. (2011) Genotypic and pathotypic diversity of Xanthomonas oryzae pv. oryzae, the cause of bacterial blight of rice in Punjab State of India. Journal of Phytopathology 159: 479-487. doi:10.1111/j.1439-0434.2011.01789.x.
  16. 16. Singh S, Sidhu J, Huang N, Vikal Y, Li Z et al. (2001) Pyramiding three bacterial blight resistance genes (xa5, xa13 and Xa21) using marker-assisted selection into indica rice cultivar PR106. Theoretical and Applied Genetics 102: 1011–1015. doi:10.1007/s001220000495.
  17. 17. Kauffman H, Reddy A, Hsieh S, Merca S (1973) An improved technique for evaluating resistance of rice varieties to Xanthomonas oryzae. Plant Dis Rep 57: 537-541.
  18. 18. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory.
  19. 19. Hampl V, Pavlícek A, Flegr J (2001) Construction and bootstrap analysis of DNA fingerprinting-based phylogenetic trees with the freeware program FreeTree: application to trichomonad parasites. Int J Syst Evol Microbiol 51: 731-735. doi:10.1099/00207713-51-3-731. PubMed: 11411692.
  20. 20. Page RD (1996) Tree View: An application to display phylogenetic trees on personal computers. Comput Appl Biosci 12: 357-358. PubMed: 8902363.
  21. 21. Salipante SJ, Hall BG (2011) Inadequacies of minimum spanning trees in molecular epidemiology. J Clin Microbiol 49: 3568-3575. doi:10.1128/JCM.00919-11. PubMed: 21849692.
  22. 22. Ellson J, Gansner ER, Koutsofios E, North SC, Woodhull G (2004) Graphviz and dynagraph—static and dynamic graph drawing tools. Graph drawing software. Springer. pp. 127-148.
  23. 23. Li Z-K, Sanchez A, Angeles E, Singh S, Domingo J et al. (2001) Are the dominant and recessive plant disease resistance genes similar?: A case study of rice R genes and Xanthomonas oryzae pv. oryzae races. Genetics 159: 757-765. PubMed: 11606550.
  24. 24. Iyer AS, McCouch SR (2004) The rice bacterial blight resistance gene xa5 encodes a novel form of disease resistance. Mol Plant Microbe Interact 17: 1348-1354. doi:10.1094/MPMI.2004.17.12.1348. PubMed: 15597740.
  25. 25. Sugio A, Yang B, Zhu T, White FF (2007) Two type III effector genes of Xanthomonas oryzae pv. oryzae control the induction of the host genes OsTFIIAγ1 and OsTFX1 during bacterial blight of rice. Proc Natl Acad Sci U S A 104: 10720-10725. doi:10.1073/pnas.0701742104.
  26. 26. Boch J, Bonas U (2010) Xanthomonas AvrBs3 family-type III effectors: discovery and function. Annu Rev Phytopathol 48: 419-436. doi:10.1146/annurev-phyto-080508-081936. PubMed: 19400638.
  27. 27. Chu Z, Yuan M, Yao J, Ge X, Yuan B et al. (2006) Promoter mutations of an essential gene for pollen development result in disease resistance in rice. Genes Dev 20: 1250-1255. doi:10.1101/gad.1416306. PubMed: 16648463.
  28. 28. Yang B, Sugio A, White FF (2006) Os8N3 is a host disease-susceptibility gene for bacterial blight of rice. Proc Natl Acad Sci U S A 103: 10503-10508. doi:10.1073/pnas.0604088103. PubMed: 16798873.
  29. 29. Chen L-Q, Qu X-Q, Hou B-H, Sosso D, Osorio S et al. (2012) Sucrose efflux mediated by SWEET proteins as a key step for phloem transport. Science 335: 207-211. doi:10.1126/science.1213351. PubMed: 22157085.
  30. 30. Antony G, Zhou J, Huang S, Li T, Liu B, et al. (2010) Rice xa13 recessive resistance to bacterial blight is defeated by induction of the disease susceptibility gene Os-11N3. The Plant Cell Online 22: 3864-3876.
  31. 31. Rajeshwari R, Sonti RV (2000) Stationary-phase variation due to transposition of novel insertion elements in Xanthomonas oryzae pv. oryzae. J Bacteriol 182: 4797-4802. doi:10.1128/JB.182.17.4797-4802.2000. PubMed: 10940020.
  32. 32. Yashitola J, Reddy A, Sonti RV (2000) A widely distributed lineage of Xanthomonas oryzae pv. oryzae in India may have come from native wild rice. Plant Disease 84: 465-469. doi:10.1094/PDIS.2000.84.4.465.
  33. 33. Jeung JU, Heu SG, Shin MS, Vera Cruz CM, Jena KK (2006) Dynamics of Xanthomonas oryzae pv. oryzae populations in Korea and their relationship to known bacterial blight resistance genes. Phytopathology 96: 867-875. doi:10.1094/PHYTO-96-0867. PubMed: 18943752.
  34. 34. Sundaram RM, Vishnupriya MR, Laha GS, Rani NS, Rao PS et al. (2009) Introduction of bacterial blight resistance into Triguna, a high yielding, mid-early duration rice variety. Biotechnol J 4: 400-407. doi:10.1002/biot.200800310. PubMed: 19253322.