A draft genome sequence and functional screen reveals the repertoire of type III secreted proteins of Pseudomonas syringae pathovar tabaci 11528

Background Pseudomonas syringae is a widespread bacterial pathogen that causes disease on a broad range of economically important plant species. Pathogenicity of P. syringae strains is dependent on the type III secretion system, which secretes a suite of up to about thirty virulence 'effector' proteins into the host cytoplasm where they subvert the eukaryotic cell physiology and disrupt host defences. P. syringae pathovar tabaci naturally causes disease on wild tobacco, the model member of the Solanaceae, a family that includes many crop species as well as on soybean. Results We used the 'next-generation' Illumina sequencing platform and the Velvet short-read assembly program to generate a 145X deep 6,077,921 nucleotide draft genome sequence for P. syringae pathovar tabaci strain 11528. From our draft assembly, we predicted 5,300 potential genes encoding proteins of at least 100 amino acids long, of which 303 (5.72%) had no significant sequence similarity to those encoded by the three previously fully sequenced P. syringae genomes. Of the core set of Hrp Outer Proteins that are conserved in three previously fully sequenced P. syringae strains, most were also conserved in strain 11528, including AvrE1, HopAH2, HopAJ2, HopAK1, HopAN1, HopI, HopJ1, HopX1, HrpK1 and HrpW1. However, the hrpZ1 gene is partially deleted and hopAF1 is completely absent in 11528. The draft genome of strain 11528 also encodes close homologues of HopO1, HopT1, HopAH1, HopR1, HopV1, HopAG1, HopAS1, HopAE1, HopAR1, HopF1, and HopW1 and a degenerate HopM1'. Using a functional screen, we confirmed that hopO1, hopT1, hopAH1, hopM1', hopAE1, hopAR1, and hopAI1' are part of the virulence-associated HrpL regulon, though the hopAI1' and hopM1' sequences were degenerate with premature stop codons. We also discovered two additional HrpL-regulated effector candidates and an HrpL-regulated distant homologue of avrPto1. Conclusion The draft genome sequence facilitates the continued development of P. syringae pathovar tabaci on wild tobacco as an attractive model system for studying bacterial disease on plants. The catalogue of effectors sheds further light on the evolution of pathogenicity and host-specificity as well as providing a set of molecular tools for the study of plant defence mechanisms. We also discovered several large genomic regions in Pta 11528 that do not share detectable nucleotide sequence similarity with previously sequenced Pseudomonas genomes. These regions may include horizontally acquired islands that possibly contribute to pathogenicity or epiphytic fitness of Pta 11528.

As mentioned in the text of the article [1], Pta 11528 encodes a full-length homologue of HopAB2. Therefore, in Figure 1, HopAB2 should not have been marked with asterisks. A corrected summary of the Hop protein repertoire in Pta 11528 is shown in Figure 1 of this Correction. We confirmed the presence of a full-length hopAB2 using capillary sequencing. Unfortunately, in the draft assembly of Illumina sequence data presented in the paper [1], there was a mis-assembly error that resulted in deletion of 271 nucleotides from the 5' end of the hopAB2 gene. This type of error is, unfortunately, not uncommon in assemblies of short sequence reads, though recent versions of the Velvet assembly software seem to be less prone to such errors. We are currently generating 454 GS-FLX sequence data from Pta 11528 genomic DNA and hope to make public an improved genome assembly and annotation in due course.
Pta 11528 encodes a full-length homologue of T3SS helper protein HrpA2 (Locus tag C1E_5326 in our annotation; RefSeq: ZP_05641290.1). Therefore hrpA2 should have been shown in boldface and underlined in Figure 1 of the manuscript [1]. This has been remedied in Figure 1 of this Correction Contrary to the original manuscript [1], HopR1 is degenerate in Pta 11528. In the Pta 11528 draft assembly, the hopR1 gene was split into at least two open reading frames (RefSeq: ZP_05639788.1, ZP_05639787.1; locus tags C1E_3889, C1E_3890) suggesting that is a degenerate pseudogene. We confirmed the presence of an internal stop codon in hopR1 using capillary sequencing. This degeneracy should have been indicated by marking hopR1 with a double asterisk (**) in Figure 1. This has been remedied in Figure 1

of this Correction
Pta 11528 encodes a full-length HopM1 homologue (Ref-Seq: ZP_05641297.1; locus tag C1E_5336; GenBank: ACR46722.1). The fact that HopM1 is intact and not degenerate should have been indicated in Figure 1 (by highlighting hopM1 in boldface and underlined) in the original manuscript [1]. This has been remedied in Figure  1 of this Correction.
We regret any inconvenience caused by these errors and are grateful to Dr Magdalen Lindeberg for bringing them to our attention.
Comparison of the complement of Pta 11528 genes encoding candidate T3SS substrates with those of the three fully sequenced P. syringae genomes Figure 1 Comparison of the complement of Pta 11528 genes encoding candidate T3SS substrates with those of the three fully sequenced P. syringae genomes. In addition to validated hop genes, also are included are several genes for T3SS helper proteins (hrpA1, hrpA2, hrpZ1, hrpW1, hrpK1 and hopP1) and several former candidates that are probably not true hop genes (hopAH2, hopJ1, hopAJ2 and hopAN1) (the HopDB website, http://www.pseudomonas-syringae.org); these genes are indicated by parentheses. Those genes that are conserved in Pta 11528 are shown in boldface and underlined. Pta 11528 also contains three hop genes that do not have orthologues in the three sequenced genomes: hopAR1, hopF1 and hopW1. * No close homologue of avrPto1 was found in Pta 11528; however, there is a gene encoding a protein that shares 43% amino acid identity with AvrPto1 from Pto DC3000. ** In the Pta 11528 genome hrpZ1 and hopR1 appear to be degenerate pseudogenes.