Analyses of genome architecture and gene expression reveal novel candidate virulence factors in the secretome of Phytophthora infestans

Abstract

Background: Phytophthora infestans is the most devastating pathogen of potato and a model organism for the oomycetes. It exhibits high evolutionary potential and rapidly adapts to host plants. The P. infestans genome experienced a repeat-driven expansion relative to the genomes of Phytophthora sojae and Phytophthora ramorum and shows a discontinuous distribution of gene density. Effector genes, such as members of the RXLR and Crinkler (CRN) families, localize to expanded, repeat-rich and gene-sparse regions of the genome. This distinct genomic environment is thought to contribute to genome plasticity and host adaptation.

Results: We used in silico approaches to predict and describe the repertoire of P. infestans secreted proteins (the secretome). We defined the "plastic secretome" as a subset of the genome that (i) encodes predicted secreted proteins, (ii) is excluded from genome segments orthologous to the P. sojae and P. ramorum genomes and (iii) is encoded by genes residing in gene sparse regions of P. infestans genome. Although including only ~3% of P. infestans genes, the plastic secretome contains ~62% of known effector genes and shows >2 fold enrichment in genes induced in planta. We highlight 19 plastic secretome genes induced in planta but distinct from previously described effectors. This list includes a trypsin-like serine protease, secreted oxidoreductases, small cysteine-rich proteins and repeat containing proteins that we propose to be novel candidate virulence factors.

Conclusions: This work revealed a remarkably diverse plastic secretome. It illustrates the value of combining genome architecture with comparative genomics to identify novel candidate virulence factors from pathogen genomes.

Figure 1

Gene Ontologies and Pfam domains enriched in the Phytophthora infestans secretome. The graphs show the number of proteins annotated with GO biological process (A), GO molecular function (B) and Pfam domains (C) and their frequency (number of proteins with annotation/total number of proteins) in the P. infestans secretome (yellow bars) and non-secreted proteins (black bars). Only GO and Pfam domains significantly enriched or depleted in the secretome are shown (chi-square test with Bonferroni correction, p-value -p-val- indicated on the leftmost part of the panels: ***, p-value < 0.01; **, p-value < 0.05; *, p-value > 0.1). GO and Pfam domains were classified by decreasing enrichment in the secretome (Enr., see methods). Full bars indicate ontology or domain enriched in the secretome, empty bars indicate ontologies or domain depleted from secretome. Ontologies and domains were color-coded for easier reference. Enr., enrichment or depletion fold; p-val, p-value of chi-square test.

Figure 2

Delimitation and effector content of Phytophthora infestans gene sparse regions (GSRs). A) Simulation of core ortholog gene segregation. Genes with both flanking intergenic regions (FIRs) longer than a value 'L' were considered as gene-sparse region (GSR) genes, whereas genes with both FIRs below L were considered as gene-dense region (GDR) genes. To quantitatively define GSRs, the % core orthologs among total genes falling in GDRs (blue) and GSRs (red) was calculated for values of L ranging from 100 bp to 5 kb. Core ortholog segregation rate was defined as the difference between the core ortholog content of the GDRs vs. GSRs (green). The percentage of core orthologs assigned to GDRs is shown as a black line. The highest core ortholog genes segregation rate was obtained for L = 1.5 kb. B) Distribution of P. infestans genes according to the length of their FIRs. All P. infestans predicted genes were sorted into 2-variable bins according to their 3'FIR (Y-axis) and 5'FIR (X-axis). The number of genes in bins is shown as a contour graph with a color code. The 1.5 kb limit for GSRs genes (dotted lines) delimits three groups of genes: genes in GDRs, GSRs, and in between (corresponding genes features and numbers are indicated in labels). C) A sample window from the P. infestans genome browser illustrating typical examples of GDRs and GSR (red background). In this 80 kb region, core ortholog genes are exclusively found in GDRs, secretome genes (yellow) and genes excluded from orthologous segments (OS, red box) are excluded from GDRs. D) Distribution of gene groups into the GDRs and GSRs of P. infestans. The proportion of non-secreted, secretome, known effectors, RXLR effector genes and CRN effector genes that occur in GSRs (red, with % indicated), GDRs (blue with % indicated) and in between (yellow) is shown.

Figure 3

Characterization of the Phytophthora infestans "plastic secretome". A) Frequency of P. infestans genes excluded from orthologous segments between P. infestans and either the P. sojae or P. ramorum genome. The proportion (% of gene group) of all, secretome, known effectors, RXLR effector and CRN effector genes is shown. B) Venn diagram illustrating the number of P. infestans genes (i) residing in GSRs and (ii) not in genome segments orthologous between the three Phytophthora species and (iii) belonging to the secretome. This set of three criteria defines the plastic secretome. The P. infestans plastic secretome consists of 561 genes: 398 known effector genes and 163 others. C) Percentage of various P. infestans gene groups found in the plastic secretome (as a % of the whole gene group). D) The plastic secretome is enriched in in planta-induced genes. The proportion of either plastic secretome (green) or non-plastic secretome (grey) genes induced in planta is shown. Genes induced at any of the time points tested are also shown ('Any'). Tom., infected tomato; Pot., infected Potato; dpi, days post-inoculation.

Figure 4

PITG_02700: Trypsin-like serine protease. A) Multiple sequence alignment showing the sequence similarity between PITG_02700 and its paralogs and well-characterized human and Aedes homologs. Regions spanning the catalytic triad (indicated by *) are shown. Proteins belonging to the P. infestans secretome are labeled with a signal peptide (SigP.) icon. GIP1, Glucanase Inhibitor Protein 1. B) Position of PITG_02700 and other P. infestans trypsin-like serine proteases on the FIR heat map (Figure 2B). C) in planta expression pattern of three in planta-induced GIP-like genes (left) and three other secreted serine protease genes (right), including PITG_02700. Expression of the effector gene Avr3a is given as a reference. Dpi, days post inoculation.

Figure 5

PITG_02930: Berberine bridge enzyme. A) Multiple sequence alignment showing the sequence similarity between PITG_02930 and its paralogs and well-characterized plant and fungal homologs. The FAD binding residues are indicated by *. Proteins belonging to the P. infestans secretome are labeled with a signal peptide (SigP.) icon. Aligned regions are numbered in the same way as in panel B to facilitate matching to the predicted protein structure. Regions indicated in blue show better conservation than regions in pink. B) Modeled protein structure of PITG_02930 with the regions shown in panel A highlighted. C) Position of PITG_02930 and other P. infestans BBEs on the FIR heat map of P. infestans (Figure 2B). D) in planta expression pattern of the five P. infestans BBEs. Expression of the effector gene Avr3a is given as a reference. Dpi, days post inoculation.

Figure 6

PITG_18284: Alpha-carbonic anhydrase. A) Multiple sequence alignment showing the sequence similarity between PITG_18284 protein from the plastic secretome and its paralogs and well-characterized plant and human homologs. The CO₂ binding residues are indicated by *. Proteins belonging to the P. infestans secretome are labeled with a signal peptide (SigP.) icon. Aligned regions are numbered in the same way in panel B to facilitate matching the sequence to the predicted protein structure. Regions indicated in blue show better conservation than regions in pink. B) Modeled protein structure of PITG_18284 with the regions shown in panel A highlighted. C) Position of PITG_18284 and other P. infestans α-CA on the FIR heat map of P. infestans (Figure 2B). D) in planta expression pattern of five P. infestans α-CAs. Non-secreted α-CAs are not induced in planta (PITG_17808 and PITG_17844), whereas secreted α-CAs show early (PITG_17842 and PITG_18284) or late induction (PITG_14412). Expression of the effector gene Avr3a is given as a reference. Dpi, days post inoculation.

Figure 7

PITG_04202: Small cysteine rich proteins (SCR). A) Pairwise sequence alignment of SCR PITG_04202 and its closest paralog. B) Position of PITG_04202 and known SCRs genes on the FIR heat map of P. infestans. C) in planta expression pattern of known SCR genes (SCR58, SCR91 and SCR50) and PITG_04202. Expression of the effector gene Avr3a is given as a reference. Dpi, days post inoculation.

Figure 8

PITG_17477, PITG_06957, and PIG_06212: Repeat containing proteins (RCPs). A) Sequence identity dot plots showing internal amino-acid sequence repeats found in PITG_17477, PITG_06957, PIG_06212 (in green) and their closest paralogs (except for PITG_06957, which lack paralogs). Numbers refer to MEME amino-acid motifs found within the repeats as indicated. B) Position of RCP genes on the FIR heat map of P. infestans. C) in planta expression pattern of the RCP genes. Expression of the effector gene Avr3a is given as a reference. Dpi, days post inoculation.