Figure 1.
Symptoms of Pseudoperonospora cubensis infection on susceptible Cucumis sativus cv. ‘Vlaspik’.
Images were collected of the adaxial (left column) and abaxial (right column) leaf surfaces at 1, 2, 3, 4, 6, and 8 days post-inoculation (dpi). A., 1 dpi, B., 2 dpi, C., 3 dpi, D., 4 dpi, E., 6 dpi, F., 8 dpi.
Figure 2.
Experimental design and sample collection.
Cucumber cv. ‘Vlaspik’ leaves were inoculated on the abaxial leaf surface with 10–30 10 μl droplets of a 1×105 sporangia/ml solution. Samples were collected at 1, 2, 3, 4, 6, and 8 days post-inoculation (dpi) using a #3 cork borer to isolate tissue immediately around the infection point (black circles). Samples from cucumber leaves mock-inoculated with 10 μl droplets of dH2O were collected in the same manner at 1 dpi. Leaf disks from each time point were pooled for RNA extraction. mRNA-Seq libraries were made for each time point from two separate biological replicates. Within a biological replicate, libraries were barcoded and sequenced in multiple lanes with the exception of the mock-inoculated library, which was not barcoded.
Figure 3.
Comparison of total mRNA-Seq reads, reads mapped and number of genes expressed.
A. Total number of reads, total reads mapped, and number of genes expressed as determined from pooling of both biological replicates are shown. Reads were mapped to the C. sativus genome [6] using Bowtie version 0.12.5 [57] and TopHat version 1.2.0 [56]. Fragments per kilobase pair of exon model per million fragments mapped (FPKM) values for the annotated C. sativus genes were calculated using Cufflinks version 0.9.3 [58]. Genes were considered expressed if the FPKM value and 95% confidence interval lower boundary FPKM value was greater than 0.001 and zero, respectively. B. Effect of sampling depth on detection of expressed genes. For all time points 5, 10, 15, 20, 25, and 30 million reads were randomly selected from the total pool of reads. Read mapping and expression abundance estimates were performed as described above. Solid lines indicate number of genes expressed and the dashed lines indicate number of reads mapped at different time points. dpi, days post-inoculation.
Figure 4.
Correlation matrix of Cucumis sativus expression profiles during infection by Pseudoperonospora cubensis.
Tissue samples were collected from C. sativus at different time points of infection with Ps. cubensis. Normalized transcript abundances for 14,476 genes were calculated as fragments per kilobase pair of exon model per million fragments mapped (FPKM) with Cufflinks version 0.9.3 [58]. Pearson correlation coefficient of log2 FPKM values were calculated for all pair-wise comparisons using R (http://cran.r-project.org/). Hierarchical clustering was performed using Pearson correlation distance metric and average linkage with the Multiple Experiment Viewer Software version 4.5 [60]. The bootstrap support values shown on tree nodes were obtained from 1000 bootstrap replicates. dpi, days post-inoculation.
Figure 5.
Comparison of orthologous gene expression in Cucumis sativus and Arabidopsis thaliana in a compatible interaction.
Microarray expression profiles were obtained from time-course analyses of genes expressed in A. thaliana (Arth) during infection by H. arabidopsidis [41]. Single copy orthologous genes between C. sativus (Cusa) and A. thaliana were identified using OrthoMCL [63]. Log2 transformed expression values of single copy orthologous genes expressed in the C. sativus mRNA-Seq dataset (log2 fragments per kilobase pair of exon model per million fragments mapped [FPKM]) and A. thaliana microarray dataset (log2 intensity) are shown as scatter plots. SCC, Spearman correlation coefficient; dpi, days post-inoculation.
Table 1.
Number of genes differentially expressed between different time points.
Figure 6.
Heat map of eigengenes representing each gene module.
The mock control and post-infection time points are represented in columns and the eigengenes for each of the 11 identified coexpression modules [49] are presented in rows. The numbers of genes in each module are given in parentheses. The eigengene values, which range from 0 to 1, are a measure of centrality and indicate the relative expression levels of all genes in the module (see Materials and Methods). dpi, days post-inoculation.
Figure 7.
Trend plots of the normalized gene expression values from six identified gene co-expression modules.
Modules consisting of genes down-regulated at 2 dpi (Module B), genes up-regulated at a single time point (Modules G, H, I, K), and genes up-regulated at two time-points (Module J) are shown. The number of genes in each module is shown in parentheses.