Processing and analysis of microarray data: Gene expression in Capsicum species challenged with Phytophthora capsici

The assessment of global gene expression under particular conditions is facilitated by microarrays. A wide range of statistical approaches are available to analyze the massive amount of data generated by microarrays. Some of these methods include dimension reduction tools such as Principal Component Analysis (PCA), supervised and unsupervised clustering methods like Self-Organizing Maps (SOM), Hierarchical Clustering (HC) or K-means. The appropriate choice of these methods depends on the characteristics of the data. A microarray approach was applied in this study to characterize gene transcription patterns under different times of root infection by Phytophthora capsici in two different Capsicum varieties. P. capsici is the causal agent of a serious fungal disease in C. annuum L. (chile). Criollo de Morelos-334 (CM334), a chile landrace with high levels of resistance to P. capsici and the susceptible variety, New Mexico 6-4 (NM6-4), were included in this study to gain a better understanding of the molecular basis of resistance to P. capsici. The processing steps selected for microarray data were background subtraction by negative controls, computation of ratios, natural log transformation, Lowess and Cross-Gene Error model. The data analysis methods selected were PCA, HC, and K-means applied to the twofold altered genes. In CM334, 87 genes were identified to be twofold altered in their transcription levels after 1, 4 or 24 h of P. capsici infection compared with time 0 (no infection). The same analysis in NM6-4 resulted in 207 genes altered across time points, while 100 genes were identified in the comparative analysis of CM334 vs. NM6-4 at time points 0, 1, 4 and 24. PR proteins were not detected in CM334 while 3 different kind were detected in NM6-4. NM6-4 clearly showed an hypersensitive response in infected roots. Few proteins were regulated in CM334 including a disease response protein, cell wall proteins, transcription factors and antimicrobial compounds. The twofold regulated genes showing similar transcription patterns after different times of infection were clustered together by the different algorithms. Genes in the same cluster are probably functionally related or commonly regulated. Further characterization of these clusters may provide more clues toward resistance mechanisms and pathways in the pathosystem Capsicum-Phytophthora.