Identification of Genetic Variation between Obligate Plant Pathogens Pseudoperonospora cubensis and P. humuli Using RNA Sequencing and Genotyping-By-Sequencing

Abstract

RNA sequencing (RNA-seq) and genotyping-by-sequencing (GBS) were used for single nucleotide polymorphism (SNP) identification from two economically important obligate plant pathogens, Pseudoperonospora cubensis and P. humuli. Twenty isolates of P. cubensis and 19 isolates of P. humuli were genotyped using RNA-seq and GBS. Principle components analysis (PCA) of each data set showed genetic separation between the two species. Additionally, results supported previous findings that P. cubensis isolates from squash are genetically distinct from cucumber and cantaloupe isolates. A PCA-based procedure was used to identify SNPs correlated with the separation of the two species, with 994 and 4,231 PCA-correlated SNPs found within the RNA-seq and GBS data, respectively. The corresponding unigenes (n = 800) containing these potential species-specific SNPs were then annotated and 135 putative pathogenicity genes, including 3 effectors, were identified. The characterization of genes containing SNPs differentiating these two closely related downy mildew species may contribute to the development of improved detection and diagnosis strategies and improve our understanding of host specificity pathways.

Fig 1. Principal components analysis of RNA-seq and GBS SNP data for P. cubensis and P. humuli isolates, maximizing for SNPs or isolates retained

a. Includes 19 isolates and 1,290 biallelic SNPs. Blue circles (8) represent P. humuli isolates. P. cubensis isolates are represented by red diamonds (9; cucumber and cantaloupe hosts) and green asterisks (2; squash host). b. Includes 30 isolates and 135 biallelic SNPs. Blue circles (16) represent P. humuli isolates. P. cubensis isolates are represented by red diamonds (11; cucumber and cantaloupe hosts) and green asterisks (3; squash host). Please note, two of the green asterisks overlap. c. Includes 31 isolates and 11,922 biallelic SNPs. Blue circles (14) represent P. humuli isolates. P. cubensis isolates are represented by red diamonds (14; cucumber and cantaloupe hosts) and green asterisks (3; squash host).d. Includes 38 isolates and 5,044 biallelic SNPs. Blue circles (18) represent P. humuli isolates. P. cubensis isolates are represented by red diamonds (15; cucumber and cantaloupe hosts) and green asterisks (5; squash and pumpkin hosts).

Fig 2. Principal components analysis of RNA-seq and GBS SNP data for P. cubensis isolates, maximizing for SNPs or isolates retained.

a. PCA using 1,290 biallelic SNPs for P. cubensis isolates only. Red diamonds (9) represent cucumber and cantaloupe isolates and green asterisks (2) represent squash isolates. b. PCA using 135 biallelic SNPs for P. cubensis isolates only. Red diamonds (11) represent cucumber and cantaloupe isolates and green asterisks (3) represent squash isolates. c. PCA using 11,922 biallelic SNPs for P. cubensis isolates only. Red diamonds (11) represent cucumber and cantaloupe isolates from NY, and green asterisks (2) represent squash isolates from NY. The black asterisk represents CDM-SQ, a squash isolate from SC. The cyan diamond represents CAcuc2008, a cucumber isolate from CA. The two blue diamonds represent cucumber isolates from NC. d. PCA using 5,044 biallelic SNPs for P. cubensis isolates only. Red diamonds (12) represent cucumber and cantaloupe isolates from NY, and green asterisks (3) represent squash isolates from NY. The black asterisk represents a squash isolate from SC (CDM-SQ). The black diamond represents a pumpkin isolate from NC (CDM-PM). The cyan diamond represents CDM-CA, a cucumber isolate from CA. The blue diamonds (2) represent cucumber isolates from NC.

Fig 3. Principal components analysis of RNA-seq and GBS SNP data for P. humuli isolates, maximizing for SNPs or isolates retained.

a. PCA using 1,290 biallelic SNPs for P. humuli isolates only. The green circle represents the 490–5 isolate from Japan. The blue circles (3) represent isolates from NY. The magenta circles (4) represent isolates from Oregon. b. PCA using 135 biallelic SNPs for P. humuli isolates only. The blue circles (4) represent isolates from NY. The magenta circles (6) represent isolates from Oregon and Washington. The black circles (3) represent isolates from Vermont. The green circles (2) represent isolates from Japan. The cyan circle represents an isolate from Wisconsin. c. PCA using 11,922 biallelic SNPs for P. humuli isolates only. The magenta circles (3) represent isolates from OR and WA. The blue circles (6) represent isolates from NY. The black circles (2) represent isolates from VT. The cyan circle represents an isolate from WI. The green circles (2) represent isolates from Japan. d. PCA using 5,044 biallelic SNPs for P. humuli isolates only. The magenta circles (6) represent isolates from OR and WA. The blue circles (6) represent isolates from NY. The black circles (2) represent isolates from VT. The green circles (3) represent isolates from Japan. The cyan circle represents an isolate from WI.

Fig 4. Characterization of unigenes containing PCA-correlated SNPs between Pseudoperonospora cubensis and P. humuli isolates sequenced using RNA-seq and GBS.

Using the P. cubensis reference genome [24,25], unigenes were identified that contained PCA-correlated SNPs from GBS and RNA-Seq data. The number of unigenes identified in each technique, as well as the number of overlapping unigenes are shown in (A). (B) Unigenes classified as putative pathogenicity genes. Gene ontology (GO) of unigenes from (C) RNA-Seq (n = 179) and (D) GBS (n = 89) assigned in terms of the associated biological processes, cellular components and molecular functions.