In planta gene expression analysis of Xanthomonas oryzae pathovar oryzae, African strain MAI1

Abstract

Background: Bacterial leaf blight causes significant yield losses in rice crops throughout Asia and Africa. Although both the Asian and African strains of the pathogen, Xanthomonas oryzae pv. oryzae (Xoo), induce similar symptoms, they are nevertheless genetically different, with the African strains being more closely related to the Asian X. oryzae pv. oryzicola (Xoc).

Results: Changes in gene expression of the African Xoo strain MAI1 in the susceptible rice cultivar Nipponbare were profiled, using an SSH Xoo DNA microarray. Microarray hybridization was performed comparing bacteria recovered from plant tissues at 1, 3, and 6 days after inoculation (dai) with bacteria grown in vitro. A total of 710 bacterial genes were found to be differentially expressed, with 407 up-regulated and 303 down-regulated. Expression profiling indicated that less than 20% of the 710 bacterial transcripts were induced in the first 24 h after inoculation, whereas 63% were differentially expressed at 6 dai. The 710 differentially expressed genes were one-end sequenced. 535 sequences were obtained from which 147 non-redundant sequences were identified. Differentially expressed genes were related to metabolism, secretion and transport, pathogen adherence to plant tissues, plant cell-wall degradation, IS elements, and virulence. In addition, various other genes encoding proteins with unknown function or showing no similarity to other proteins were also induced. The Xoo MAI1 non-redundant set of sequences was compared against several X. oryzae genomes, revealing a specific group of genes that was present only in MAI1. Numerous IS elements were also found to be differentially expressed. Quantitative real-time PCR confirmed 86% of the identified profile on a set of 14 genes selected according to the microarray analysis.

Conclusions: This is the first report to compare the expression of Xoo genes in planta across different time points during infection. This work shows that as-yet-unidentified and potentially new virulence factors are appearing in an emerging African pathogen. It also confirms that African Xoo strains do differ from their Asian counterparts, even at the transcriptional level.

Figure 1

In planta quantification of bacteria. Bacterial growth in 8-week old rice variety Nipponbare, in sections A, B, C, D, and E of the leaf at 0 and 12 h, and 1, 3, 6, 10, and 15 days after inoculation. The experiment was repeated three times with three leaves per time point. Error bars indicate standard errors.

Figure 2

Functional categorization of diferentially expressed genes. Genes of Xoo strain MAI1 found as differentially expressed in planta were grouped into nine categories: biological process unknown; hypothetical protein; protein synthesis; cell envelope and motility; phage-related and IS elements; metabolism; signal transduction; secretion, transport, and binding proteins; and virulence-related sequence. The proportion of each category of the total number of genes is given as a percentage.

Figure 3

Clusters of transcripts based on patterns of differential expression. Differentially expressed transcripts were clustered, using the k-means method. The mean expression levels of genes in each cluster are shown as a centroid graph. Error bars represent standard deviations of expression within the cluster. Seven clusters were created, with clusters 1, 2, 3, and 4 comprising up-regulated genes and clusters 5, 6, and 7 comprising down-regulated genes at 1, 3, and 6 dai, respectively. The x axis represents time-points during infection (1, 3, and 6 dai) and the y axis the expression level.

Figure 4

Comparing expression of genes through microarray and QRT-PCR assays. We used real-time PCR analysis to confirm the differential expression of 14 genes of the African strain MAI1 of Xanthomonas oryzae pv. oryzae. The genes represented various biological functional classes of interest. Although fold change in gene expression was generally higher for QRT-PCR than for the microarray, good correlation existed between the two data sets. A subset of 5 genes corresponding to xopX (ACD57163), HrpF protein (FI978263), hypothetical proteins (FI978252 and FI978328), and the avrXa7 gene (AF275267) showing a fold-change higher than 10 in QRT-PCR is illustrated in the figure. Three replicates were analysed in both microarray and QRT-PCR experiments. Vertical bars represent standard deviations.