HrpK and HopB1 (PSPTO1406) are secreted via the DC3000 TTSS. (A) Immunoblot analysis using anti-HopB1 antibodies showed that HopB1 is made in small amounts in cell (C) extracts of DC3000 grown under hrp-inducing conditions. HopB1 was not made in the DC3000 hopB1 mutant, UNL130, or the hrpK polar mutant, UNL132, suggesting that the protein band observed was due to the hopB1 gene and that hopB1 is in the same transcriptional unit as hrpK. (B) P. syringae pv. tomato DC3000 strains carrying the hopB1-containing plasmid pLN63 (phopB1) were grown under hrp-inducing conditions and separated into cell (C) and supernatant (S) fractions, concentrated 13.3 and 133 fold, respectively, and analyzed by SDS-PAGE and immunoblots with anti-HopB1 antibodies. The results clearly showed that HopB1 is secreted via the TTSS. (C) P. syringae pv. tomato DC3000 strains carrying a construct, pLN64 (phrpK-F), that made HrpK C-terminally fused to the FLAG epitope tag were grown under hrp-inducing conditions, and secretion experiments were performed as above. Anti-FLAG antibodies were used as the primary antibody in the immunoblot to detect HrpK-FLAG. HrpK-FLAG is abundantly found in the supernatant fractions of DC3000 and the hrpK mutant. (B and C) A DC3000 hrcC mutant (hrcC), which is defective in the TTSS, was used as a negative control. All strains also carried pCPP2318, which encodes mature β-lactamase that remains in the cytoplasm and acts as a lysis control.

Bioinformatics anlaysis of HopB1. (A) Schematic representation of P. syringae ORFs with similarity to a region within HopB1. Regions from P. syringae pv. tomato DC3000 PSPTO4996 and P. syringae pv. syringae B728a HolPtoAC_Psy that share similarity to HopB1 are depicted with striped boxes. Homologous regions between HolPtoAC_Psy and PSPTO4993 and 4996 are depicted with dark grey boxes. An insertion sequence (ISPsy5) interrupting PSPTO4993 and 4996 is depicted as a black box. Numbers above boxes indicate amino acid positions. (B) Protein alignment of the region within HopB1 that is similar to other proteins in the databases. Additional proteins that are not shown in panel A are shown in the alignment adjacent to their accession numbers. Their genus and species names, ORF designations, and database accession numbers are as follows: P. fluorescens PfO-1, ORF Pflu 0200366, ZP_00263996; P. luminescens supsp. laumondii TTO1, ORF Plu2400, NP_929645; and V. vulnificus YJ016, ORF VVA1030, NP_937086. The MegAlign program from DNAStar (Madison, Wis.) was used for protein alignment.

Pathogenicity assays and bacterial multiplication in planta indicate that a DC3000 hrpK mutant but not a DC3000 hopB1 mutant is significantly reduced in virulence on tomato and Arabidopsis. The following strains were used in pathogenicity and bacterial growth assays: wild-type DC3000; the hopB1 mutant, UNL130 (ΔhopB1); the DC3000 nonpolar hrpK mutant, UNL111 (ΔhrpK); UNL111 carrying pLN64 encoding HrpK-FLAG [ΔhrpK(phrpK)]; and a DC3000 hrcC mutant. Three- to 4-week-old L. esculentum cv. Moneymaker plants and A. thaliana Col-0 were dip inoculated into different bacterial suspensions at an OD₆₀₀ of 0.2 with 0.02% Silwet L-77. Disks measuring 0.64 cm² for tomato plants and 0.4 cm² for Arabidopsis were excised, ground in water, serially diluted, and plated on KB plates containing appropriate antibiotic markers. (A) Disease symptoms produced on tomato on day 4. Bacterial growth was tracked for a period of 4 days in tomato (B) and Arabidopsis (C). The results show that the hrpK mutant was reduced in symptom production and bacterial growth in planta, whereas the hopB1 mutant was similar to the wild type. The hrpK mutant phenotype was complemented by pLN64 carrying hrpK-flag.

HrpK requires a TM domain for proper function. (A) Schematic representation of HrpK truncations and deletions. Numbers indicate amino acid positions in HrpK. The CyaA column indicates detection of cAMP production in the CyaA translocation assay. All constructs were checked for protein stability by SDS-PAGE and immunoblotting with anti-CyaA antibodies. The constructs and their corresponding products are listed as follows: pLN377, full-length HrpK-CyaA; pLN917, HrpK_1-382-CyaA; pLN1170, HrpK_1-467-CyaA; pLN1171, HrpK_1-582-CyaA; pLN1172, HrpK_1-698-CyaA; pLN1173, HrpK_{1-698,750-780}-CyaA; pLN1174, HrpK_{1-382,750-780}-CyaA; and pLN1175, HrpK_{1-80, 750-780}-CyaA. (B) HrpK_{1-698/750-780} failed to complement the reduced HR produced by the DC3000 hrpK mutant UNL111. N. tabacum cv. Xanthi was infiltrated at an OD₆₀₀ of 0.2 (1 × 10⁸ cells/ml), and 10-fold serial dilutions were performed. The pictures show leaves that were infiltrated with bacteria at a cell density of 1 × 10⁷ cells/ml. The following strains were infiltrated: wild-type DC3000; the DC3000 hrpK mutant UNL111 (ΔhrpK); UNL111 carrying pLN774 expressing full-length HrpK-HA (phrpK); and UNL111 carrying pLN1167 expressing the TM domain deletion (phrpKΔTM). The construct pLN1167 carrying the TM domain deletion was unable to complement the reduced HR phenotype exhibited by UNL111.

RT-PCR analysis of the P. syringae pv. tomato DC3000 EEL confirms that hrpK and PSPTO1406 are polycistronic and that both predicted hrp promoters in the EEL region are active. (A) Organization of the EEL in DC3000. White boxes depict the ORFs of the EEL along with hrpK, which is part of the hrp-hrc cluster. Other hrp-hrc cluster genes, hrpL and hrpJ, which are not dealt with here are depicted as grey boxes. Arrows above ORFs indicate the predicted direction of transcription, and the black boxes indicate putative hrp promoters. The numbered lines beneath the ORFs represent PCR products produced by specific primer sets used below. (B to D) The numbers above each figure correspond to the PCR primer sets shown in panel A. +, experimental PCRs containing RT; −, control PCRs lacking RT; g, control PCRs without RT and DC3000 total DNA. Primer set 8 was used as a negative control because its PCR product spans two genes transcribed in opposite orientations, which would not produce RNA. (B) RT-PCR analysis of RNA isolated from DC3000 grown in KB, a rich medium. Primer sets 1, 2, and 4 produced an RT-dependent PCR product, indicating that the ORFs spanning these regions are transcribed. Notably, primer sets 3 and 5 to 7, corresponding to PSPTO1408, PSPTO1409, PSPTO1406, and hrpK, did not produce an RT-dependent PCR product. (C) RT-PCR analysis of RNA isolated from DC3000 grown under hrp-inducing conditions. Under these conditions, all primer sets produced an RT-dependent PCR product, indicating that the ORFs corresponding to primer sets 3 and 5 to 7 were only transcribed under hrp-inducing conditions. Primer set 6 produced an RT-dependent PCR product, which confirmed that hrpK and PSPTO1406 were in an operon. (D) High-stringency Southern blot analysis showed that the EEL region hybridized to all of the PCR products, confirming that the PCR products corresponded to the EEL.

HopB1 and HrpK are translocated into plant cells based on reporter assays; however, HrpK requires its C-terminal half to translocate. (A) AvrRpt2 translocation assays with HopB1 and HrpK AvrRpt2 fusions. P. syringae pv. phaseolicola NPS3121 carrying constructs that encoded either full-length AvrRpt2, an 80-amino-acid N-terminal deletion of AvrRpt2 (ΔAvrRpt2), full-length HopB1 fused to ΔAvrRpt2 (HopB1-ΔAvrRpt2), full-length HrpK fused to ΔAvrRpt2 (HrpK-ΔAvrRpt2), and the first 399 amino acids of HrpK fused to ΔAvrRpt2 (HrpK_1-399-ΔAvrRpt2) were infiltrated into A. thaliana Col-0 at an OD₆₀₀ of 0.5 and scored after 12 to 19 h for elicitation of the HR. N, no visible HR. (B) CyaA translocation assays with HopB1 and HrpK CyaA fusions. DC3000 and a DC3000 hrcC mutant defective in TTSS carrying constructs that encoded either full-length HopB1 (HopB1-CyaA), HrpK (HrpK-CyaA), or the first 382 amino acids of HrpK (HrpK_1-382-CyaA) fused to CyaA were infiltrated into N. benthamiana and assayed for cAMP production 7 h after infiltration as described in Materials and Methods. cAMP levels are reported in picomoles of cAMP per micrograms of protein with standard errors. The levels of cAMP indicated that the full-length HopB1-CyaA fusion and the full-length HrpK-CyaA fusion were translocated. However, the N-terminal HrpK-CyaA fusion had background levels of cAMP, indicating that it was not translocated.

PSI-BLAST searches show that HrpK is similar to HrpF, NolX, and PopF1. (A) Regions that were identified in PSI-BLAST searches were aligned with DNAStar MegAlign. Shown are S. fredii NolX, accession no. AAB17674; X. campestris pv. vesicatoria HrpF, accession no. AAB86527; R. solanacearum PopF1, accession no. NP_523114; and DC3000 HrpK, accession no. AE016860. (B) TM regions in HrpK and other putative type III translocators from plant-associated bacteria. TM regions were predicted with the TMpred program (http://www.ch.embnet.org/software/TMPRED_form.html). Only scores of > 1,000 are shown. See the text for additional details.

hrpK mutants are reduced in their ability to translocate Avr proteins. (A) N. tabacum cv. Xanthi was infiltrated at an OD₆₀₀ of 0.2 (2 × 10⁸ cells/ml) and 10-fold serial dilutions of DC3000, the DC3000 hrpK mutant UNL111 (ΔhrpK), and UNL111 (pLN64), which carried hrpK in trans [ΔhrpK(phrpK)]. The hrpK mutant showed a visible repeatable reduction in its ability to elicit the HR when compared to the wild type. (B) N. tabacum cv. Xanthi or transgenic N. tabacum cv. Xanthi lines expressing HrpK-HA were infiltrated with P. fluorescens 55 (Pfl) carrying cosmid pHIR11, a pHIR11 derivative (pLN468) lacking hrpK (ΔhrpK), or Pfl (pLN468) with pLN174, which carried hrpK (phrpK), each at an OD₆₀₀ of 0.4 (4 × 10⁸ cells/ml). Production of the HR was documented at 18 h postinfiltration.