Pseudomonas syringae Lytic Transglycosylases Coregulated with the Type III Secretion System Contribute to the Translocation of Effector Proteins into Plant Cells

Abstract

Pseudomonas syringae translocates virulence effector proteins into plant cells via a type III secretion system (T3SS) encoded by hrp (for hypersensitive response and pathogenicity) genes. Three genes coregulated with the Hrp T3SS system in P. syringae pv. tomato DC3000 have predicted lytic transglycosylase domains: PSPTO1378 (here designated hrpH), PSPTO2678 (hopP1), and PSPTO852 (hopAJ1). hrpH is located between hrpR and avrE1 in the Hrp pathogenicity island and is carried in the functional cluster of P. syringae pv. syringae 61 hrp genes cloned in cosmid pHIR11. Strong expression of DC3000 hrpH in Escherichia coli inhibits bacterial growth unless the predicted catalytic glutamate at position 148 is mutated. Translocation tests involving C-terminal fusions with a Cya (Bordetella pertussis adenylate cyclase) reporter indicate that HrpH and HopP1, but not HopAJ1, are T3SS substrates. Pseudomonas fluorescens carrying a pHIR11 derivative lacking hrpH is poorly able to translocate effector HopA1, and this deficiency can be restored by HopP1 and HopAJ1, but not by HrpH(E148A) or HrpH_1-241. DC3000 mutants lacking hrpH or hrpH, hopP1, and hopAJ1 combined are variously reduced in effector translocation, elicitation of the hypersensitive response, and virulence. However, the mutants are not reduced in secretion of T3SS substrates in culture. When produced in wild-type DC3000, the HrpH(E148A) and HrpH_1-241 variants have a dominant-negative effect on the ability of DC3000 to elicit the hypersensitive response in nonhost tobacco and to grow and cause disease in host tomato. The three Hrp-associated lytic transglycosylases in DC3000 appear to have overlapping functions in contributing to T3SS functions during infection.

The type III and type IV secretion systems (T3SS and T4SS, respectively) of gram-negative bacterial pathogens must penetrate the peptidoglycan layer in the bacterial periplasm to deliver virulence proteins into host cells. Specialized lytic transglycosylases (LTs) have been implicated in this process because of the perceived need to locally enlarge pores in the peptidoglycan mesh to accommodate the secretion machinery and because T3SS and T4SS gene clusters typically have an LT gene (by definition “specialized”) associated with them (18, 34). Specialized LTs are typically members of LT superfamily 1 and characteristically carry a motif with an invariant glutamate, which is the single catalytic residue responsible for the cleavage of the β-1,4 glycosidic bond between the N-acetylglucosamine and N-acetylmuramic acid units in peptidoglycan and the formation of the terminal 1,6-anhydromuramyl cleavage product (11). LT superfamily 1 is also referred to as the SLT superfamily.

Understanding the function of these LTs in virulence protein delivery and host interactions has proven challenging for several reasons. (i) With few exceptions, mutations in LT genes associated with T3SSs have little or no phenotype (6, 17, 47). (ii) Heterologous LTs may function in place of some specialized LTs (26, 34). (iii) The precedence in animal innate immune systems of intracellular pattern recognition receptors that recognize peptidoglycan fragments raises the possibility that specialized LT activity may have consequences beyond virulence protein delivery (23, 37). (iv) And finally, the VirB1 LT associated with the T4SS of Agrobacterium tumefaciens has a C-terminal extension that is cleaved, partially secreted, and independently contributes to virulence (8, 36), which suggests that specialized LTs may have multiple roles in host interactions.

Several LTs associated with the T3SS of plant pathogens have been partially characterized. The T3SSs of plant pathogens such as Ralstonia solanacearum, Xanthomonas spp., Erwinia amylovora, and Pseudomonas syringae are encoded by hrp genes, which are required for these bacteria to elicit the defense-associated hypersensitive response (HR) in nonhosts or to promote pathogenesis in hosts (nine of the T3SS genes are designated hrc genes because they are conserved in the T3SS of animal pathogens) (13). The T3SS is thought to function primarily to translocate effectors, which are variously known as Hop (for Hrp outer protein) or Avr (for avirulence) proteins, into host cells (4). Microarray analysis of R. solanacearum revealed three genes with similarity to LTs that are upregulated by HrpB, a positive activator of the hrp genes (40). In Xanthomonas, the hpa2 and hpaH genes of X. oryzae pv. oryzae and X. campestris pv. vesicatoria, respectively, are orthologous T3SS-associated LT genes, which have been mutated but show no virulence phenotype (38, 56). In E. amylovora a putative LT gene was found to be upregulated in planta and to make a small contribution to virulence (55).

Somewhat more is known about the T3SS-associated putative LTs of P. syringae. Microarray analysis of the kinetics of gene expression in P. syringae pv. tomato DC3000 after bacterial transfer into Hrp-inducing medium reveals three putative LT genes that are preceded by Hrp promoter sequences and are activated by the HrpL alternative sigma factor in coordination with genes encoding the T3SS and major effectors (21). These are PSPTO2678, PSPTO852, and PSPTO1378. There are fragmentary data on the question of whether these proteins are T3SS substrates themselves. PSPTO2678 (HopP1) behaves like a typical effector protein in a test for translocation into plant cells that involves the use of a native promoter and an AvrRpt2 avirulence domain translocation reporter (15). In contrast, the same test suggests that PSPTO852 (HopAJ1) and PSPTO1378 (HrpH) are not T3SS substrates (15).

PSPTO1378 is of particular interest because the gene is adjacent to the cluster encoding the T3SS machinery and has been provisionally referred to as conserved effector locus (CEL) open reading frame 1 (ORF1) because it is the first ORF in the CEL of the Hrp pathogenicity island (3). Importantly, this gene has also been provisionally designated ipx10 (for in planta expression) based on an IVET screen for DC3000 genes expressed in planta, and an ipx10 mutant is partially reduced in virulence in host tomato leaves (12). Here we provide evidence that the CEL ORF1 in two P. syringae pathovars makes important contributions to T3SS effector translocation, and the gene has consequently been given a hrp designation, specifically hrpH, which will be used hereafter in this report.

P. syringae pv. syringae 61 and P. syringae pv. tomato DC3000 provide useful models for studying the action of T3SS-specialized LTs. Strain 61 is a weak pathogen of bean whose hrp-hrc gene cluster (which includes hrpH), cloned in cosmid pHIR11, enables nonpathogens such as Pseudomonas fluorescens to elicit the HR in tobacco (Nicotiana tabacum) and to translocate heterologously expressed test effectors into plant cells (29). The P. syringae pv. syringae DNA cloned in pHIR11 also encodes the HopA1 effector, which acts as an HR-eliciting avirulence determinant in tobacco (5). The use of pHIR11 in nonpathogens enables the P. syringae T3SS to be studied in the absence of various, redundant, T3SS-related factors that have been identified in the completely sequenced genome of P. syringae pv. tomato DC3000 (35). DC3000 is a pathogen of tomato and the model plants Arabidopsis and wild tobacco Nicotiana benthamiana, and this strain is being intensively studied for its ability to both induce and suppress innate immune defenses in plants (1, 24, 39). The combined use of P. fluorescens(pHIR11) and P. syringae pv. tomato DC3000 permits both reductionist and comprehensive approaches to the function of plant pathogen T3SS LTs.

Here we report that (i) HrpH possesses the conserved glutamate characteristic of SLTs at position 148 but has a C-terminal extension that is atypical of T3SS-associated LTs, (ii) the glutamate at position 148 is essential for the toxicity of HrpH when expressed in E. coli, (iii) a HrpH-Cya hybrid can be translocated by the P. syringae T3SS, (iv) P. fluorescens carrying a pHIR11 derivative lacking hrpH is poorly able to translocate HopA1, and this deficiency can be restored by HopP1 and HopAJ1, but not by HrpH(E148A) or HrpH_1-241, (v) P. syringae pv. tomato DC3000 mutants lacking hrpH or hrpH, hopP1, and hopAJ1 combined are reduced in effector translocation, HR elicitation, and virulence in test plants, and (vi) the HrpH(E148A) and HrpH_1-241 variants have a dominant-negative phenotype when expressed in wild-type DC3000.

MATERIALS AND METHODS

Bacterial strains and plasmids.

Bacterial strains and plasmids used in the present study are listed in Table 1. E. coli was grown in Luria-Bertani (LB) broth at 37°C. P. syringae and P. fluorescens cells were grown in King's B (KB) medium at 28°C (33). E. coli TOP10 cells were used for general cloning and Gateway manipulations. Antibiotics were used at the following concentrations (in μg/ml): kanamycin (Km), 50; gentamicin (Gm), 10; tetracycline (Tc), 40; streptomycin (Sp), 100; ampicillin (Amp), 100; rifampin (Rif), 50; and cycloheximide, 2. For marker exchange, Gm and Km were used at half concentrations.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Genotype or relative phenotype	Source or reference
Strains
E. coli
C2110	PolAts nalidixic acid resistant	32
TOP10	F−mcrA Δ(mrr-hsdRMS-mcrBC) φ80lacZΔM15 ΔlacX74 deoR recA1 araD139 Δ(ara-leu)7679galU galK rpsL endA1 nupG	Invitrogen
BL21(DE3)	F−ompT hsdSB(rB− mB−) gal dcm (DE3)	Invitrogen
S17-1	thi pro ΔhsdR hsdM+recA (chr::RP4-2-Tc::Mu-Km::T7)	46
P. fluorescens 55	Wild type, nalidixic acid resistant	29
P. syringae pv. tomato
DC3000	Wild type; Rifr	14
CUCPB5112	DC3000 with hrcC replaced with an nptII cassette lacking a transcription terminator; Kmr Rifr	52
CUCPB5469	hrpH deletion mutant containing FRT scar; Rifr	This study
CUCPB5477	hopP1 deletion mutant containing FRT scar; Rifr	B. H. Kvitko
CUCPB5470	hopAJ1 deletion mutant containing FRT scar; Rifr	This study
CUCPB5471	hrpH hopP1 deletion mutant containing FRT scar; Rifr	This study
CUCPB5472	hrpH hopAJ1 deletion mutant containing FRT scar; Rifr	This study
CUCPB5473	hrpH hopP1 hopAJ1 deletion mutant containing FRT scar; Rifr	This study
CUCPB5474	ΔhrpH242-495 deletion mutant containing FRT scar; Rifr	This study
CUCPB5460	hopQ1-1 deletion mutant containing FRT scar; Rifr	51
CUCPB5475	hrpH hopQ1-1 deletion mutant containing FRT scar; Rifr	This study
CUCPB5476	hrpH hopP1 hopAJ1 hopQ1-1 deletion mutant containing FRT scar; Rifr	This study
Plasmids
pHIR11	pLAFR3 containing the P. syringae pv. syringae 61 hrp-hrc gene region; Tcr
pLN18	pHIR11 derivative with shcA and hopA1 replaced by an nptII cassette; Tcr Kmr	31)04
pCPP3297	pLN18 containing an unmarked deletion in hrcC; Tcr Kmr	45
pCPP5271	pBluescript II SK(+) containing hopA1-cya-hrpK1	This study
pCPP5316	pHIR11 containing hopA1-cya translational fusion; Tcr	This study
pCPP5703	pCPP5316 ΔhrpH::FRTGmr; Tcr Gmr	This study
pET-DEST42	pET Gateway destination vector allowing for T7-regulated expression of a protein with a C-terminal His6-V5 tag; Ampr	Invitrogen
pENTR/SD-TOPO	Entry vector for Gateway cloning; Kmr Cmr	Invitrogen
pCR2.1-TOPO	PCR cloning vector; Kmr Ampr	Invitrogen
pRK2013	Helper plasmid; ColE1 replicon TraRK+Mob+; Kmr	22
pCPP3234	pVLT35 containing Gateway reading frame B cassette and codons 2 to 406 of cya; Sp/Smr Cmr	45
pCPP5295	pBBR1MCS containing Gateway reading frame B cassette and codons 2 to 406 of cya; Gmr Cmr	B. H. Kvitko
pCPP5371	pBBR1MCS containing avrPto1 promoter, Gateway reading frame B cassette, and codons 2 to 406 of cya; Gmr Cmr	B. H. Kvitko
pCPP5372	pBBR1MCS containing avrPto1 promoter, Gateway reading frame B cassette, and C-terminal HA tag; Gmr Cmr	B. H. Kvitko
pCPP5209	pKD4 FRT cassette plasmid with nptII replaced by aaCI; Gmr	51
pCPP5301	pRK415 containing Gateway reading frame B cassette and lacZ; Tcr	51
pCPP5264	pRK415 with C1 and flp; Tcr	51
pCPP5721	pENTR/SD-TOPO containing hrpH; Kmr	This study
pCPP5722	pENTR/SD-TOPO containing hrpH(E148A); Kmr	This study
pCPP5723	pCPP5301 containing the left and right flanking regions of hrpH with an FRT Gmr cassette; Gmr Tcr	This study
pCPP5724	pCPP5301 containing the left and right flanking regions of hopAJ1 with an FRT Gmr cassette; Gmr Tcr	This study
pCPP5728	pCPP5301 containing the left and right flanking regions of hrpH242-495 with an FRT Gmr cassette; Gmr Tcr	This study
pCPP5608	pK18mobsacB containing hopQ1-1 deletion fragment; Kmr	51
pCPP5725	pET-DEST42 expressing hrpH; Ampr	This study
pCPP5726	pET-DEST42 expressing hrpH(E148A); Ampr	This study
pCPP5704	pCPP5295 carrying hrpH expressed from its native promoter and fused to cya; Gmr	This study
pCPP5705	pCPP5295 carrying the hopP1 gene under its native promoter and fused to cya; Gmr	This study
pCPP5706	pCPP5371 expressing HopAJ1-Cya; Gmr	This study
pCPP5312	pCPPP5371 expressing AvrPto-Cya; Gmr	This study
pCPP5708	pCPP3234 expressing HrpH-Cya, Sp/Smr	This study
pCPP5709	pCPP3234 expressing HrpH17-495-Cya; Sp/Smr	This study
pCPP5710	pCPP3234 expressing HrpH(E148A)-Cya; Sp/Smr	This study
pCPP5711	pCPP3234 expressing HrpH1-241-Cya; Sp/Smr	This study
pCPP3256	pCPP3234 expressing HopP1-Cya; Sp/Smr	A. R. Ramos
pCPP3259	pCPP3234 expressing HopAJ1-Cya; Sp/Smr	A. R. Ramos
pCPP5712	pCPP3234 expressing HrpH-stop-Cya; Sp/Smr	This study
pCPP5713	pCPP3234 expressing HopP1-stop-Cya; Sp/Smr	This study
pCPP5714	pCPP3234 expressing HopAJ1-stop-Cya; Sp/Smr	This study
pCPP5715	pCPP3234 expressing MltDDC3000-stop-Cya; Sp/Smr	This study
pCPP5718	pCPP3234 expressing HrpH(E148A)-stop-Cya; Sp/Smr	This study
pCPP5727	pCPP3234 expressing HrpH1-241-stop-Cya; Sp/Smr	This study
pCPP5716	pCPP5371 expressing HopA1-Cya; Gmr	This study
pCPP5717	pCPP5371 expressing HrpZ1-Cya; Gmr	This study
pCPP5707	pCPP5372 expressing HrpH-HA; Gmr	This study
pCPP5720	pCPP5372 expressing HopAJ1-HA; Gmr	This study
pCPP5736	pCPP5372 expressing HrpH(E148A)-HA; Gmr	This study
pCPP5737	pCPP5372 expressing HrpH1-241-HA; Gmr	This study

DNA manipulations.

Plasmid DNA was isolated and manipulated by using standard methods (42). PCR was performed with ExTaq polymerase (Takara Bio, Inc., Otsu, Shiga, Japan) used according to the manufacturer's instructions. Restriction enzymes were obtained from New England Biolabs (Ipswich, MA). All constructs were created by using the Gateway cloning technology (Invitrogen, Carlsbad, CA). For entry clones, inserts were created by PCR with the primers shown in Table 2 and pHIR11 or genomic DNA of P. syringae pv. tomato DC3000 as a template. The PCR products were cloned into a pENTR/SD-TOPO vector and sequenced at the Cornell Bioresource Center with an ABI 3700 DNA analyzer. Entry clones were transferred into appropriate destination vectors, as indicated in Table 1. Each resulting construct was conjugated into P. syringae or P. fluorescens strains by triparental mating, using pRK2013, and the resulting transconjugants were selected on KB medium with appropriate antibiotics (19).

TABLE 2.

Oligonucleotide primers used to clone, tag, or mutate genes

Primer	Sequence (5′-3′)
HrpH-F	CACCATGCCCGCCGTCGCCTTC
HrpH-R	GTATTGGCGTGGATCACTGGC
HopP1-F	CACCATGACCATGGGTGTTTCAC
HopP1-R	AGCGGGTAAATTGCCCTGC
HopAJ1-F	CACCATGCGTTCCAGGGTT
HopAJ1-R	TCGGCGCAGGCTCTC
MltDDC3000-F	CACCATGTCGTCATCTATCAGCAAGCC
MltDDC3000-stop	TCAGTGCGGCAGGTAAACGGTC
HrpH-stop	TTATTGGCGTGGATCACTGGC
HrpH1-241-stop	TCAGTAGTCCACAGGCAGACC
HopP1-stop	TCAAGCGGGTAAATTGCCC
HopAJ1-stop	CTATCGGCGCAGGCTCTCC
HrpH(E148A)-F	CTGCTGCCGATGATCGCAAGCTCTTATAACCCC
HrpH(E148A)-R	GGGGTTATAAGAGCTTGCGATCATCGGCAGCAG
Hrpbox HrpH-F	CACCCTTACGGCTGAAAGGATTCA
Hrpbox HopP1-F	CACCAAACGTGTGCGGTTG
HrpH17-495-F	CACCATGGCTGTTCAAATCGCAGTTC
HopAJ1-LF-F	CACCGTAATCCAGTTACAGCGC
HopAJ1-LF-R	CTCTGGTACCGCTTTTACTCTCCTGTAAAG
HopAJ1-RF-F	CTCTGGTACCGTGTACTCACTTCATAGG
HopAJ1-RF-R	GAACGACAGCGAATTTCCGCC
HrpH-LF-F	CACCTGCCGTCTTCACGCTCC
HrpH-LF-R	CTCTGGTACCTGAGCGTGTCCACCCTAC
HrpH-RF-F	CTCTGGTACCTATTGAAAACCCCTGAAAAG
HrpH-RF-R	CGTTCACTCTCATGGTGGGTGGC
HrpH1-241-LF-F	CACCGGATTTCTGGAACAATTTACCGAC
HrpH1-241-LF-R	CTCTGGTACCATTCCAGTAGTCCACAGGC
HrpHPsy61-LF-F	CACCCTTGGCGAGTATGTCGTCC
HrpHPsy61-LF-R	CTCTGGTACCCGGAATGTCCTGAACGTGTC
HrpHPsy61-RF-F	CTCTGGTACCTGAGCCCTTAAAGCACC
HrpHPsy61-RF-R	GCAAGTCCATATCCAGTAGCGCAC
KD3/4-KpnI-F	ATTAGGTACCGTGTAGGCTGGAGCTGCTTC
KD3/4-KpnI-R	ATTAGGTACCCATATGAATATCCTCCTTA
hrpK-F	ATTCCGGCTATGACAGCTAATGAAATCGGGGCTAAATAGC
Xba-hrpK-R	TAATTCTAGAGCTTCGCGTCAGTGTCTGAC
Cya-F	CGGCGCTCAGGGCGCGAAACGGGCAGCAATCGCATCAGGC
Cya-R	GCTATTTAGCCCCGATTTCATTAGCTGTCATAGCCGGAAT
Xba-hopA-F	TAATTCTAGAGTTACGAGATTCGTGCGGCC
HopA-R	GCCTGATGCGATTGCTGCCCGTTTCGCGCCCTGAGCGCCG

The underlining denotes the restriction site for KpnI.

Construction of pCPP5316 and ΔhrpH derivative pCPP5703.

To construct the hopA1-cya fusion in pHIR11, hopA1, hrpK1, and cya were PCR amplified from pHIR11 and pCPP3234 with the primer pairs Xba-hopA-F and HopA-R, HrpK-F and Xba-HrpK-R, and Cya-F and Cya-R, respectively. The three fragments were joined by consecutive SOEing PCRs (27) and cloned into the XbaI site of pBluescript II SK(+) using terminal primer-introduced restriction sites to create pCPP5271. pCPP5271 was transformed into E. coli C2110(pHIR11) and plated on KB agar supplemented with Tc and Amp at the ts-polA nonpermissive temperature of 42°C to obtain integrants. Integrant strains were cured of the plasmid at 30°C by daily 1:1,000 subinoculations over 4 days. Recovered colonies were screened for Amp^s and PCR screened to confirm the hopA1-cya fusion and the construction of pCPP5316.

To create the ΔhrpH_Psy61 derivative of pCPP5316, a 1.1-kb fragment carrying the left flanking region of hrpH and a 1.2-kb fragment carrying the right flanking region of hrpH was PCR amplified from pCPP5316 using primer pair HrpH_Psy61-LF-F and HrpH_Psy61-LF-R and the primer pair HrpH_Psy61-RF-F and HrpH_Psy61-RF-R. Each PCR product contains a restriction site for KpnI. An FRT Gm^r cassette was PCR amplified from pCPP5209 using the primers KD3/4-KpnI-F and KD3/4-KpnI-R, producing PCR products with the restriction site for KpnI at both ends. All three PCR products were digested with KpnI, ligated together, and purified. A 3.5-kb final product was cloned into the pCR2.1-TOPO vector (Invitrogen). The hrpH_Psy61 mutation was introduced into pCPP5316 as described previously (2). The construct carrying the FRT Gm^r cassette between the left flanking region and the right flanking region of hrpH was electroporated into E. coli C2110 carrying pCPP5316 and spread onto LB plates, selecting for Tc and Km at 42°C. The recombinant strains were cured of the plasmid at 30°C by daily 1:1,000 subinoculations over 4 days. Recovered colonies were screened for Tc^r and Km^s and PCR screened to confirm mutation. Triparental mating was used to move pCPP5703 from E. coli to P. fluorescens.

Construction of HrpH(E148A)_.

The mutation of hrpH glutamate 148 to alanine was produced with a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). For the reaction, pENTR/SD-TOPO::hrpH was used as a target plasmid with the mutated primers HrpH(E148A)-F and HrpH(E148A)-R. The resulting construct was confirmed by sequencing and transferred via Gateway cloning reactions into expression vectors pCPP3234 and pET-DEST42 (Invitrogen) to produce the plasmids listed in Table 1.

Construction of P. syringae pv. tomato DC3000 LT gene deletion mutants.

To create the ΔhrpH mutant CUCPB5469 and ΔhopP1 hrpH double mutant CUCPB5471, the flanking region of hrpH was PCR amplified from genomic DNA of P. syringae pv. tomato DC3000 using primer pair HrpH-LF-F and HrpH-LF-R and primer pair HrpH-RF-F and HrpH-RF-R. Each PCR product contained a restriction site for KpnI. A FRT Gm^r cassette was PCR amplified from pCPP5209 using the primers KD3/4-KpnI-F and KD3/4-KpnI-R, producing PCR products with the restriction site for KpnI at both ends. All three PCR products were digested with KpnI, ligated together, and purified. The final product was cloned into pENTR/SD-TOPO (Invitrogen). The resulting construct was transferred via Gateway cloning reactions into pCPP5301 to produce the plasmids listed in Table 1. Plasmid pCPP5723 was introduced into wild-type DC3000 or CUCPB5477 by triparental mating and selected on KB plates containing Gm, Tc, and Rif. The selected transformants were cured of the plasmid at 30°C by daily 1:1,000 subinoculations over 4 days. The final transfers were diluted and plated on KB agar containing Rif, Gm, and 40 μg of X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside)/ml for blue and white screening. To make unmarked deletion mutants, the FRT Gm^r was deleted by transformation with Flp-expressing pCPP5264.

To create the ΔhopAJ1 mutant CUCPB5470, ΔhrpH hopAJ1 double mutant CUCPB5472, and ΔhrpH hopP1 hopAJ1 triple mutant CUCPB5473, the flanking region of hopAJ1 was PCR amplified from genomic DNA of DC3000 using the primer pair HopAJ1-LF-F and HopAJ1-LF-R and the primer pair HopAJ1-RF-F and HopAJ1-RF-R. The subsequent steps were the same as those used to make the hrpH mutant. The final plasmid, pCPP5728, was introduced into DC3000, CUCPB5469, or CUCPB5471 by triparental mating to make mutants.

To create the C-terminus-truncated hrpH mutant (CUCPB5474), the flanking region of hrpH_242-495 was PCR amplified from genomic DNA of DC3000 using the primer pair HrpH_1-241-LF-F and HrpH_1-241-LF-R and the primer pair HrpH-RF-F and HrpH-RF-R. The final plasmid, pCPP5724, was introduced into wild-type DC3000 by triparental mating and used to construct a mutation as described for the hrpH mutants.

To add the hopQ1-1 mutation to CUCPB5475 and CUCPB5476, the previously constructed pCPP5608 was transferred from E. coli S17-1 into each mutant by conjugation and used for directed mutagenesis as described previously (51).

Bacterial lysis assay.

E. coli BL21(DE3) harboring native hrpH or hrpH mutant derivatives expressed from pET-DEST42 were grown in 5 ml of LB broth supplemented with Amp at 37°C overnight and inoculated to new media and allowed to grow until the optical density at 600 nm (OD₆₀₀) had reached 0.5. The T7/lacO promoter of the vector, driving expression of the test proteins, was activated by addition of 0.5 mM IPTG (isopropyl-β-d-thiogalactopyranoside), and the OD₆₀₀ was further monitored. Immunoblot analysis confirmed equivalent production of hrpH and hrpH mutant derivatives. The experiment was repeated three times with similar results.

Plant growth and bacterial inoculations.

Fully expanded leaves from 6-week-old tobacco (N. tabacum cv. Xanthi) were used for HR assays. Four-week-old N. benthamiana or three-week-old tomato (Solanum lycopersicum cv. Moneymaker) plants were used for virulence assays. Plants were grown under greenhouse conditions and transferred to the laboratory 1 day before inoculation. The tested plants were maintained in the laboratory with a light intensity of 40 μmol/m² s at 25°C. For virulence and HR assays, bacteria were grown on KB plates overnight at 28°C and were prepared for inoculation by suspension in 10 mM MgCl₂. For HR assays, bacterial infiltration was performed with a blunt syringe, with the levels of inoculum differing among experiments as noted in the figure legends. Bacterial virulence assays were done by dipping or by syringe infiltration, as indicated. For dipping inoculations, bacteria were diluted to 3 × 10⁶ CFU/ml in 10 mM MgCl₂ solution with 0.02% Silwet L-77 (Lehle Seeds, Round Rock, TX). Plants were submerged upside down in bacterial suspension for 1 min. For syringe infiltration, leaves were infiltrated with a bacterial suspension of 3 × 10⁴ CFU/ml in 10 mM MgCl₂ with a blunt syringe. Inoculated plants were maintained under high humidity conditions at 22°C with 16 h of light and 8 h of darkness. To aid visualization of lesions in N. benthamiana leaves, leaves were distained using Carnoy's fluid (10% acetic acid, 30% chloroform, and 60% ethanol). For bacterial population counts, three leaf disks from three tomato leaves were ground in 300 μl of 10 mM MgCl₂, and serial dilutions were spotted onto KB plates with Rif and cycloheximide every 2 days. All experiments involving plant responses and bacterial growth were repeated at least three times, and independent tests yielded similar results.

Cya translocation reporter assays.

Translocation assays were performed as described previously (45). P. syringae pv. tomato DC3000 or P. fluorescens strains were grown on KB plates with appropriate antibiotics at 30°C overnight, scraped off the agar, and resuspended in 5 mM morpholineethanesulfonic acid (pH 5.5). To induce expression of cya fusions in vector pCPP3234, 0.1 mM IPTG was added. During these experiments, inoculated plants were maintained in the lab at 25°C until samples were collected. Leaf disks were collected 6 h after infiltration with a 1.0-cm-diameter cork borer and ground in 250 μl of 0.1 M HCl by using a Dremel tool pestle in a microfuge tube. The cyclic AMP (cAMP) levels were determined by using a Correlate-EIA cAMP immunoassay kit according to the manufacturer's instructions (Assay Designs Inc, Ann Arbor, MI). All translocation tests were repeated at least three times with similar results.

Protein secretion immunoblot analysis.

Lawns of DC3000 and mutant derivatives were grown overnight at 30°C on KB plates with appropriate antibiotics. Bacteria were harvested from the plates by resuspension in 5 ml of Hrp-inducing medium (supplemented with 0.2% [wt/vol] fructose and 0.2% [wt/vol] mannitol) (30), followed by brief vortexing. Bacteria were directly added to 100 ml of Hrp-inducing medium to an OD₆₀₀ of 0.3 and grown with shaking at 22°C to an OD₆₀₀ of 0.5. Cultures were centrifuged at 5,200 × g for 15 min, and the bacterial pellet was collected and resuspended in protein sample buffer for the cell fraction. The top 40 ml of the supernatant was spun at 20,800 × g for 40 min. The top 25 ml of this supernatant was then added to 5 ml of trichloroacetic acid, shaken, and incubated at 4°C overnight to precipitate protein. Protein was sedimented by centrifugation at 20,800 × g for 40 min, and the pellet was resuspended in protein sample buffer for the supernatant fraction. Proteins in cell and supernatant fractions were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to a polyvinylidene difluoride membrane, and probed with anti-AvrPto or anti-HrpZ antibodies, followed by secondary anti-rabbit immunoglobulin G-alkaline phosphatase conjugate antibodies (Sigma) at a dilution of 1:30,000.

Nomenclature.

We have given PSPTO1378 an hrp designation because the mutants are significantly reduced in HR elicitation, virulence, and effector translocation. The letter “H” was chosen for hrpH because it is one of only a few letters lacking potentially confusing overlaps with other hrp genes in the Hrp1 T3SSs of phytopathogens in the Pseudomonadaceae and Enterobacteriaceae. However, it is important to note that HrpH was used for HrcC before 1996 and the development of a unified nomenclature for broadly conserved hrp genes of phytopathogenic bacteria (13, 28). Regarding HopAJ1, its status as a T3SS pathway substrate remains uncertain and no change in its designation is proposed.

RESULTS

The P. syringae pv. tomato DC3000 HrpL regulon includes three putative LTs with distinct domain structures.

The domain structures of the HrpL-activated LT proteins of DC3000 are compared in Fig. 1 with the structures of several other LTs associated with type III and IV secretion systems. Targeting signals are not depicted in the figure, but HrpH and HopP1 each have an N-terminal region of 50 amino acids that possesses the three patterns associated with P. syringae T3SS substrates (35). Although HopAJ1 possesses two of these features—a high serine content (18%) and no acidic amino acids in the first 12 residues—it lacks a hydrophobic residue in position 3 or 4. Furthermore, unlike the other two LTs, HopAJ1 is predicted to have a signal peptide recognized by the Sec system for export to the periplasm based on SignalP 3.0 (10, 50). The three DC3000 proteins have distinct LT domains. HrpH has the MLTD_N and SLT domains of the MltD family 1D LTs, whereas HopP1 has just the SLT domain of the Slt family 1A LTs (11). HopAJ1 is a member of the MltB family 3 of LTs and is the only one of the three HrpL-regulated LTs to carry a peptidoglycan-binding domain.

FIG. 1.

Predicted domain architecture of the three P. syringae pv. tomato DC3000 HrpL-regulated LTs and mutant derivatives and comparison with other representative bacterial LTs. (A) Architecture of HrpH and mutant derivatives used in the present study. The Pfam domains are MltD_N (PF06471, the N-terminal domain of membrane-bound LT), SLT (PF01464, transglycosylase), and a Pfam class B domain shared with PSPTO3714. The location of the E148A mutation is indicated with a triangle. (B) Domain architecture of other relevant LT proteins: MltD, P. syringae pv. tomato DC3000 PSPTO3714; IagB, Salmonella enterica (NP_457271); IpgF, Shigella flexneri (NP_085298); Hpa2, Xanthomonas oryzae pv. oryzae (ABJ80895); and VirB1, Agrobacterium tumefaciens (NP_536285). The LysM domains in MltD are Pfam PF01476 (lysin motif). (C) Domain architecture of the other two HrpL-regulated LTs of DC3000. The HopP1 HrpW domain, determined by alignment with HrpW, includes the N-terminal T3SS targeting region. HopAJ1 has an MLTB domain (COG2951, membrane-bound lytic murein transglycosylase B) and a PGB domain (PF01471, peptidoglycan binding domain 1).

An unusual feature of HrpH is a C-terminal extension, comprised of residues 242 to 495, which is lacking from the T3SS-associated LTs of other plant and animal pathogens (Fig. 1A). The HrpH C-terminal region contains a Pfam-B domain, which is also found in the DC3000 PSPTO3714 MltD protein, but unlike PSPTO3714, HrpH lacks any LysM peptidoglycan-binding motifs. Instead, HrpH possesses, between residues 352 and 495, a proline-rich (19%) region that is hypervariable among the HrpH proteins in the three sequenced pathovars of P. syringae: tomato strain DC3000, phaseolicola 1448A, and syringae B728a (Table 3). HopP1 has the same two-domain structure of harpins like HrpW (Fig. 1C), and the N-terminal region of HopP1 is similar to the N-terminal region of HrpW (34a). However, whereas the C-terminal region of HrpW is similar to pectate lyases, the C-terminal region of HopP1 has the SLT domain.

TABLE 3.

Comparison of the predicted amino acid sequences of the P. syringae pv. tomato DC3000 HrpH with the HrpH of P. syringae pv. phaseolicola 1448A, P. syringae pv. syringae B728A, P. syringae pv. syringae 61, and P. viridiflava

Protein	DC3000 HrpH regions compared (% identity) at amino acids:
	1 to 241	242 to 350	351 to 495
P. syringae pv. phaseolicola 1448A HrpH	91.7	86.1	39.5
P. syringae pv. syringae B728A HrpH	85.5	84.3	36.1
P. syringae pv. syringae 61 HrpH	85.1	84.3	36.1
P. viridiflava HrpH	73.9	80.6	41.5
DC3000 MltD	50.6	52.8	14.3

The predicted catalytic glutamate 148 in the P. syringae pv. tomato DC3000 HrpH is required for toxicity when expressed in E. coli.

A multiple sequence alignment of HrpH, HopP1, and several representative LTs is presented in Fig. 2. Three of the specialized LTs (IagB, IpgF, and VirB1) have been shown to have peptidoglycanase activity in vitro, based on zymogram analyses (54). Figure 2 reveals that both of the P. syringae pv. tomato DC3000 proteins have the three consensus motifs and the glutamate in motif I that is predicted to be required for LT activity. Furthermore, differences in motif II are consistent with the membership of HrpH and HopP1 in different subfamilies of LT superfamily 1 (11). For HrpH the catalytic glutamate in motif I is at position 148. HopAJ1 does not have this motif, which is consistent with its membership in the MltB family (data not shown). To evaluate the functions of the different domains in HrpH, we have constructed derivatives (i) lacking the first 16 amino acids, (ii) with an alanine replacing glutamate 148, and (iii) lacking the C-terminal region carrying the Pfam-B motif and the proline-rich hypervariable region (Fig. 1A).

FIG. 2.

Multiple sequence alignment of the SLT domain from HrpH and other selected proteins. The proteins are as follows: Slt70, E. coli K-12 (POAGC3); IagB, S. enterica (NP_457271); IpgF, S. flexneri (NP_085298); Hpa2, X. oryzae pv. oryzae (ABJ80895); VirB1, A. tumefaciens (NP_536285); and HopP1, MltD, and HrpH from P. syringae pv. tomato DC3000 (AAO56180, AAO57184, and NP_791205, respectively). Boxes indicate three conserved sequence motifs typical for the goose egg white lysozyme-like domain (LT_GEWL). Overlines mark the conserved two α-helices (I and III) and the beta sheet (II) of the LT-GEWL family. The asterisk indicates the catalytic glutamate.

Overexpression of active specialized LTs can cause lysis of E. coli in liquid culture (9), and we accordingly examined the effects of inducing hrpH expression from a vector T7/lacO promoter with IPTG. Induced production of the native HrpH inhibited the growth of E. coli BL21(DE3) in culture, but it did not cause obvious cell lysis, as would be indicated by a decline in culture turbidity (Fig. 3). Induced production of HrpH(E148A) had no effect on E. coli growth (Fig. 3). Induced production of the HrpH_1-241 mutant also had no inhibitory effect on E. coli (data not shown), suggesting that the Pfam-B motif is also important for LT activity. Importantly, these observations suggest that HrpH is a typical family-1 LT in that its biological activity is dependent on the predicted glutamate.

FIG. 3.

Inhibition of E. coli growth in culture by induced production of HrpH. The data show growth in LB medium of E. coli BL21(DE3) cells harboring pET-DEST42 expressing hrpH or hrpH(E148A) from the vector T7lac promoter. Expression of the indicated LT genes was induced by the addition of 0.5 mM IPTG at time zero.

HrpH-Cya and HopP1-Cya can be translocated into plant cells by the P. syringae T3SS.

To compare the ability of the DC3000 HrpH, HopP1, HopAJ1, and HrpH derivatives to travel the P. syringae T3SS during bacterial interactions with plants, we constructed plasmids that generate fusions of the Cya reporter (Bordetella pertussis adenylate cyclase) to the C terminus of the test proteins and then performed translocation assays based on the calmodulin-dependent production of cAMP by Cya inside N. benthamiana leaf cells (45). Use of the semiquantitative Cya translocation assay permitted a direct comparison of the three proteins and the effects of promoter strength and mutations on translocation.

When expressed from an IPTG-induced vector tac promoter in P. syringae pv. tomato DC3000, HrpH-Cya and HopP1-Cya were translocated almost as well as AvrPto1-Cya (Fig. 4A). When expressed from its native promoter, HrpH-Cya was translocated at a lower, but significant level, whereas HopP1-Cya translocation was still equivalent to that of AvrPto1-Cya. HopAJ1-Cya was not translocated regardless of the promoter (Fig. 4A). None of the test proteins was translocated by a T3SS-deficient DC3000 ΔhrcC mutant.

FIG. 4.

Evidence that HrpH and HopP1, but not HopAJ1, travel the P. syringae T3SS. (A) Translocation tests using P. syringae pv. tomato DC3000. N. benthamiana leaves were infiltrated via a blunt syringe with DC3000 at 3 × 10⁸ CFU/ml carrying plasmids expressing the indicated protein with a C-terminal Cya reporter fusion. The three vectors used different promoters: pCPP3234 (tac promoter), pCPP5295 (native promoters), and pCPP5371 (avrPto1 promoter). Regardless of the promoter, none of the test proteins was translocated by a T3SS-deficient DC3000 ΔhrcC mutant, as indicated by cAMP levels below 2.0 pmol/μg of total bacterial protein (data not shown). (B) Assays using P. fluorescens heterologously expressing a P. syringae T3SS to test for translocation of HrpH and derivatives HrpH_17-495, HrpH(E148A), and HrpH_1-241. N. benthamiana leaves were infiltrated with P. fluorescens at 3 × 10⁸ CFU/ml carrying pLN18 or ΔhrcC derivative pCPP3297 and pCPP3234 derivatives expressing the indicated test proteins as Cya fusions. In both panels, samples were collected from N. benthamiana leaves 6 h after inoculation, and data represent the means and standard deviations of populations measured from two leaf disks from each of two different leaves.

To test the effects of hrpH mutations on the translocation of HrpH-Cya itself, we expressed the test Cya fusions in P. fluorescens(pLN18). Cosmid pLN18 carries the P. syringae pv. syringae 61 hrp gene cluster from hrpK1 to hrpH and is less affected by dominant-negative effects of hrpH mutations than is DC3000 (discussed later). Deleting the first 16 amino acids of HrpH abolished translocation (Fig. 4B), but an E148A mutation and a truncated HrpH with the first 241 amino acids fused to Cya only partially reduced translocation. Interestingly, translocation of HrpH_1-241-Cya and HrpH(E148A)-Cya was more variable in repeated tests than translocation of HrpH-Cya. A ΔhrcC derivative of pLN18, pCPP3297, did not translocate HrpH-Cya. Thus, HrpH is typical of other P. syringae T3SS substrates examined in requiring the first 16 amino acids to travel the pathway (7, 45). Also, the ability of HrpH_1-241-Cya to be translocated suggests that the entire HrpH protein travels the T3SS rather than a C-terminal cleavage product.

The P. syringae pv. syringae 61 T3SS expressed in P. fluorescens is strongly impaired in HopA1-Cya translocation if hrpH is deleted.

Given the apparent redundancy in HrpL-activated LTs in the DC3000 genome, we decided to first assess the role of HrpH in effector translocation using P. fluorescens carrying pCPP5316, which is a derivative of pHIR11 with a translational fusion of cya to the native hopA1 gene (Fig. 5A). The complete sequence of the P. syringae pv. syringae 61 DNA insert in cosmid pHIR11 (GenBank accession number EF514224) was reported by Ramos et al. (41). The insert has a portion of the conserved effector locus (3), including hrpH and a truncated avrE1 gene. P. fluorescens(pCPP5316) elicits the HR in tobacco and translocates HopA1-Cya (Fig. 5B). The ability of pHIR11 and its derivatives to elicit the HR in tobacco is dependent on the translocation of HopA1 (2, 5); thus, HR elicitation provides an alternative indicator of HopA1-Cya translocation in this system. Deletion of hrpH from pCPP5316 resulted in a strong reduction in HopA1-Cya translocation and abolished HR elicitation (Fig. 5B). However, it is important to note that although the residual level of HopA1-Cya was low (70 pmol/μg protein), it was repeatedly higher than the level of translocation that we observe with various effector-Cya fusions using T3SS-deficient hrcC mutants of pHIR11 in P. fluorescens (<14 pmol/μg protein).

FIG. 5.

Contribution of HrpH and functionally equivalent LTs to the ability of T3SS-proficient P. fluorescens(pCPP5316) to translocate HopA1-Cya and elicit a HopA1-dependent HR in tobacco. (A) The P. syringae pv. syringae 61 DNA insert in cosmid pHIR11 contains a complete hrp-hrc gene cluster, including genes encoding positive regulators (diagonal lines) and T3SS substrates (shaded). pCPP5316 is pHIR11 with a hopA1-cya translational fusion. (B) Translocation of HopA1-Cya by P. fluorescens carrying pCPP5316 and pCPP5316 ΔhrpH, with or without test LT genes expressed in trans, as indicated by HR elicitation in tobacco leaves and by Cya reporter assays. Leaf tissue was photographed 24 h after inoculation with 6 × 10⁸ CFU/ml; a positive response (+) is indicated by confluent collapse and bleaching of the tissue. For Cya translocation reporter tests, samples were collected 6 h after syringe infiltration of tobacco leaf tissue with 6 × 10⁸ CFU/ml and assayed for cAMP levels as described in the text. The data show the means and standard deviations for two leaf disks from each of two different leaves. The LT genes were cloned with a stop codon and expressed by inducing the vector pCPP3234 tac promoter with 0.1 mM IPTG. The HR assay panels are aligned with the corresponding translocation bar graph results and labels below.

We also examined the ability HrpH, HrpH mutants, and other putative LTs—all derived from P. syringae pv. tomato DC3000—to restore HopA1-Cya translocation to P. fluorescens carrying the pCPP5316 ΔhrpH mutant when each LT gene was expressed from a vector tac promoter. HrpH, HopP1, and HopAJ1 all at least partially restored HopA1-Cya translocation and HR elicitation, whereas the PSPTO3714 MltD protein, HrpH(E148A) and HrpH_1-241 did not (Fig. 5). Thus, the three Hrp-regulated DC3000 LTs are interchangeable in promoting effector translocation, and the LT activity of HrpH appears to be essential for promoting effector translocation in this system. Furthermore, despite the divergence in the hypervariable C-terminal regions (Table 3), HrpH_PtoDC3000 can substantially complement the loss of HrpH_Psy61.

P. syringae pv. tomato DC3000 mutants lacking combinations of HrpH, HopP1, and HopAJ1 are altered in their ability to elicit the nonhost HR in tobacco and to translocate effectors into host tomato.

To assess the role in DC3000 of the three HrpL-regulated LTs, we constructed deletions affecting each of them individually and in various combinations and then tested the mutants for T3SS-dependent activities. When inoculated into nonhost tobacco leaves at 6 × 10⁶ CFU/ml, the ΔhrpH mutant was substantially reduced in HR elicitation activity, whereas the ΔhopP1 mutant was indistinguishable from the wild type and the ΔhopAJ1 mutant showed an intermediate reaction (Fig. 6). The differing contributions of these three LTs was more evident when tobacco was inoculated at 3 × 10⁶ CFU/ml. At this level, the ΔhrpH and ΔhopAJ1 mutants no longer elicited visible tissue collapse. HR eliciting activity was restored to each of these two mutants by expressing the corresponding LT gene from a vector tac promoter (Fig. 6). DC3000 double mutants that were ΔhrpH hopP1 or ΔhrpH hopAJ1 had the same HR-deficient phenotype of ΔhrpH or ΔhopAJ1 single mutants. However, the ΔhrpH hopP1 hopAJ1 triple mutant showed a wild-type ability to elicit the HR in repeated experiments (Fig. 6). Possible explanations for this unexpected result are considered in the discussion section.

FIG. 6.

Effect of deleting the three P. syringae pv. tomato DC3000 HrpL-regulated LT genes on elicitation of the HR in tobacco leaves. (A) HR threshold assays for LT mutants. The tested strains were DC3000 wild type (WT), three mutants with single deletions (ΔhrpH, ΔhopP1, and ΔhopAJ1, respectively), two double mutants (ΔhrpH hopP1 and ΔhrpH hopAJ1), one triple mutant (ΔhrpH hopP1 hopAJ1), ΔhrpH_242-496 producing a HrpH derivative lacking the C-terminal region, and a T3SS-deficient mutant (ΔhrcC) as a negative control. Strains were inoculated with a blunt syringae at 6 × 10⁶ and 3 × 10⁶ CFU/ml, as indicated, and leaves were photographed 48 h later. (B) HR threshold assays for complemented LT mutants. Two of the single-deletion mutants, ΔhopAJ1 and ΔhrpH, were complemented by expressing HopAJ1 and HrpH, respectively, from plasmids using a vector pCPP5372 avrPto1 promoter. For the plants used in this assay, the threshold bacterial population for HR elicitation was 6 × 10⁶ CFU/ml.

We also tested the ability of the DC3000 ΔhrpH and ΔhrpH hopP1 hopAJ1 triple mutants to translocate into N. benthamiana Cya fusions of the DC3000 proteins HopA1, AvrPto1, and HrpZ1 expressed using a vector AvrPto1 promoter. The avrPto1 promoter was chosen because microarray analysis revealed it to be typical of promoters that are strongly upregulated by HrpL, whereas the hopA1 promoter is weaker and the hrpZ1 is located downstream of hrpA1 in a polycistronic operon (21). For each test protein, the level of translocation was substantially reduced in the mutants (Fig. 7A). Furthermore, in three independent tests we observed that the ΔhrpH hopP1 hopAJ1 triple mutant was more impaired in the delivery of HopA1 than AvrPto1. We conclude that HrpH makes an important contribution to effector translocation, but deleting all three HrpL-regulated LT genes does not completely abolish effector translocation by DC3000.

FIG. 7.

Ability of P. syringae pv. tomato DC3000 wild-type, ΔhrpH mutant, and ΔhrpH hopP1 hopAJ1 triple mutant to translocate representative T3SS substrates into host tomato cells and secrete them in culture. (A) Translocation assay. Tomato leaves were infiltrated with DC3000 (WT), ΔhrpH mutant (ΔH), or ΔhrpH hopP1 hopAJ1 triple mutant (ΔH,P,AJ) cells that also carried pCPP5371 expressing the indicated T3SS substrates with a C-terminal Cya fusion. Bacteria were infiltrated via a blunt syringe at 6 × 10⁶ CFU/ml, and samples were collected from leaves 6 h later. The results present the apparent translocation (based on cAMP concentration) by the mutants relative to the WT for each test substrate, with the WT levels for each substrate normalized to 100. The data show the means and standard deviations for two leaf disks from each of two different leaves. The actual mean values of the wild type (pmol of cAMP/μg of total protein) were 127 for HopA1, 93 for AvrPto1, and 30 for HrpZ1 for the experiment shown. Note that the use of a moderate level of inoculum and the avrPto1 promoter in vector pCPP5371 for all three test substrates results in a relatively low level of translocation. None of the test proteins was translocated by a T3SS-deficient DC3000 ΔhrcC mutant, as indicated by cAMP levels below 2.0 pmol/μg of protein (data not shown). The experiment was repeated three times with similar results. (B) Assay for secretion of AvrPto1 and HrpZ1 produced from their native genes in the DC3000 chromosome under Hrp-inducing conditions in culture. The indicated DC3000 wild-type (WT) and mutant strains were grown in Hrp-inducing minimal broth to an OD₆₀₀ of 0.5 and separated by centrifugation into cell-bound “C” and supernatant “S” fractions. Proteins in each fraction were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and analyzed by immunoblotting with polyclonal antibodies to AvrPto1 and HrpZ1.

We also tested the ability of the DC3000 ΔhrpH and ΔhrpH hopP1 hopAJ1 triple mutants to secrete AvrPto1 and HrpZ1 in Hrp-inducing minimal medium. The use of antibodies against AvrPto1 and HrpZ1 enabled us to monitor secretion of the products of the native DC3000 genes. Unexpectedly, the LT mutations did not reduce the ability of either AvrPto1 or HrpZ1 to be secreted, although a ΔhrcQ-U mutation completely abolished secretion of both proteins (Fig. 7B). Thus, the HrpL-regulated LTs contribute to translocation in planta but not to secretion in culture.

DC3000 mutants lacking combinations of HrpH, HopP1, and HopAJ1 show reduced virulence in hosts tomato and N. benthamiana.

An hrpH mutant (ipx10) was previously reported to be partially reduced in lesion formation and growth in host tomato (12). Here, we extended that observation by also testing the ΔhrpH hopP1 hopAJ1 triple mutant in assays for tomato lesion formation and bacterial growth and also by examining the effect of the mutations on the ability of a DC3000 ΔhopQ1-1 derivative to produce disease lesions in N. benthamiana. The ΔhopQ1-1 derivative was used because HopQ1-1 acts as an avirulence determinant that prevents DC3000 from causing disease in N. benthamiana (51). Our results replicate the observations of Boch et al. (12) regarding the hrpH mutant (Fig. 8). Importantly, they show that the mutant lacking all three HrpL-regulated LTs is further reduced, but not completely abolished, in virulence.

FIG. 8.

Effect of ΔhrpH mutation and ΔhrpH hopP1 hopAJ1 triple mutation on the virulence of P. syringae pv. tomato DC3000 in tomato and N. benthamiana. (A) Bacterial speck symptoms in tomato leaves. Plants were inoculated by dipping in inoculum containing the indicated strains at 3 × 10⁶ CFU/ml and photographed 10 days later. (B) Bacterial growth in tomato leaves. Inoculum at 3 × 10⁴ CFU/ml was infiltrated with a blunt syringe into leaves and populations were measured at 0, 2, and 4 days after inoculation. The data show the means and standard deviations of populations measured from three leaf disks from each of three different leaves. (C) Effect on virulence in N. benthamiana of deleting hrpH, hopAJ1, and hopP1 from a DC3000 ΔhopQ1-1 mutant. The indicated strains were infiltrated at 3 × 10⁴ CFU/ml into leaf panels with a blunt syringe, and the leaf was photographed 7 days later. Circles mark the approximate inoculation area. To aid visualization of necrotic lesions, leaves were distained with Carnoy's fluid.

HrpH(E148A) and HrpH_1-241 display a dominant-negative phenotype in nonhost and host plants when expressed in wild-type DC3000.

While studying the ability of hrpH and its derivatives to complement the DC3000 ΔhrpH mutation, we observed that hrpH mutants had a dominant-negative phenotype. We explored this observation in more detail by expressing hrpH derivatives encoding HrpH(E148A) and HrpH_1-241 from a vector avrPto1 promoter in DC3000 and then assaying for altered HR elicitation in tobacco and for lesion formation and growth in tomato. Expression of the two mutants substantially reduced all three phenotypes (Fig. 9). However, unlike the DC3000 ΔhrcC negative control, production of HrpH(E148A) and HrpH_1-241 did not completely abolish any of these phenotypes. These observations further highlight the importance of HrpH and suggest that HrpH mutants lacking LT activity can unproductively compete with native HrpH.

FIG. 9.

Dominant-negative effect of HrpH(E148A) and HrpH_1-241 on the ability of wild-type P. syringae pv. tomato DC3000 to elicit the HR in nonhost tobacco and produce disease lesions and bacterial growth in host tomato. (A) HR elicitation in tobacco leaf. Bacteria were infiltrated with a blunt syringe at the indicated levels into a tobacco leaf, and the leaf was photographed 48 h later. The indicated plasmid-produced HrpH derivatives were expressed from the avrPto1 promoter in pCPP5372. (B) Lesion formation on tomato plants. Tomato plants were inoculated by dipping in the indicated strains at 3 × 10⁶ CFU/ml and photographed 10 days later. (C) Bacterial growth in tomato leaves. Inoculum at 3 × 10⁴ CFU/ml was infiltrated with a blunt syringe into different leaves, and populations were measured at 0, 2, and 4 days after inoculation. The data show the means and standard deviations of populations measured from three leaf disks from each of three different leaves.

DISCUSSION

Our findings indicate that HrpH is the primary specialized LT associated with the Hrp T3SS of P. syringae and that HopP1 and HopAJ1 may be secondary specialized LTs whose functions overlap with HrpH. The complex structure of HrpH, which includes a C-terminal proline-rich hypervariable region, raises the possibility that the protein may have multiple functions. HrpH is a T3SS substrate that is capable of being translocated into plant cells, and it contributes strongly to the translocation of type III effector proteins in planta but not to their secretion in culture. The strong phenotype of a site-directed mutation affecting the catalytic glutamate in HrpH indicates the importance of HrpH LT activity in the overall function of the P. syringae T3SS. Below, we will compare the properties of HrpH, HopP1, and HopAJ1 and discuss the contrasting effects of mutations involving these proteins on various phenotypes associated with the P. syringae T3SS.

Although the hrpH, hopP1, and hopAJ1 genes share the properties of possessing predicted LT domains and being upregulated by HrpL, they differ in their distribution and structure. The three sequenced strains of P. syringae—DC3000, 1448A, and B728a—represent divergent pathovars in the three major clades that comprise the species (43, 44). The hrpH gene is found in all three strains, which suggests that hrpH is universally present in P. syringae, but an HrpL-regulated hopAJ1 homolog is missing from B728a, and hopP1 is from missing from both 1448A and B728a (35). Structurally, HopAJ1 looks like a typical MltB LT with a Sec-dependent signal peptide, but the DC3000 hopAJ1 gene is distinguished by its activation by HrpL. In contrast, HopP1 is constructed like a two-domain harpin, with an N-terminal region that is similar to HrpW1. HopP1 is also like harpins in lacking any cysteine and in possessing an ability to elicit programmed cell death when infiltrated into the intercellular spaces of tobacco leaves (34a).

In considering the functions of phytopathogen T3SS components, particularly specialized LTs, it is worth noting that the successful delivery of type III effectors involves (i) traversing the bacterial envelope, including the peptidoglycan layer, (ii) traversing the plant cell wall and plasma membrane, and (iii) evading or defeating plant basal defenses that may be triggered by pathogen-associated molecular patterns, which reportedly include peptidoglycan (20). The expected primary target for a T3SS LT would be the bacterial peptidoglycan, but LT substrates with plant signaling roles may also be in the plant cell wall (48, 49), and LTs could also have a role in sequestering or processing peptidoglycan fragments that may be presented to plant cells as a by-product of T3SS assembly.

The localization patterns of the three DC3000 Hrp LTs may provide further clues to their function(s) in the T3SS assembly process. HopP1 is translocated into plant cells as efficiently as AvrPto1, and it does so when expressed from its native promoter and assayed for translocation using either AvrRpt2 or Cya translocation reporters (15, 50). In short, its translocation behavior is indistinguishable from that of effectors that are thought to function primarily within plant cells. However, as noted above, tobacco is hypersensitive to purified HopP1 delivered to the apoplast, which suggests that the protein has biological activity outside of plant cells.

HrpH expressed from its own promoter was reported not to be translocated when an AvrRpt2 translocation reporter was used with Arabidopsis test plants (15). However, we used the same low-copy-number vector (pBBR1MCS) and native promoter and observed a moderate level of translocation using a Cya translocation reporter in N. benthamiana. Although the levels of translocated HrpH would likely be lower with a single copy of hrpH, our data indicate that HrpH travels the T3SS and that its ability to do so, as with other P. syringae T3SS substrates, is dependent on the first 16 amino acids of the protein. Furthermore, the hypervariability of the proline-rich C-terminal region of HrpH suggests the possibility of selection for divergence as a result of interaction with host factors. The observation that HrpH is a T3SS substrate may also explain the relatively weak deleterious effect of HrpH expression in E. coli. Unlike P19, a plasmid R1 specialized LT (9), strongly expressed HrpH appears to inhibit bacterial growth without causing massive lysis, as would be indicated by a sharp decline in culture turbidity. One explanation for this is that HrpH inefficiently enters the E. coli periplasm via the T3SS-like flagellar biogenesis pathway (16, 53).

In contrast to HopP1 and HrpH, we saw no compelling evidence that HopAJ1 is a T3SS substrate. A 14-amino-acid N-terminal fragment of a P. syringae pv. maculicola ES4326 homolog (previously designated HopPmaG), fused to an AvrRpt2_81-255 reporter was found to be translocated into Arabidopsis in a functional screen for type III effectors (25). However, a full-length fusion of the same protein to AvrRpt2_101-255 was neither translocated nor secreted (50). Here, we saw no evidence for translocation of the full-length HopAJ1-Cya protein despite using the moderately strong, IPTG-induced tac promoter and a similarly strong Hrp promoter (P_avrPto1). Furthermore, we observed that the same tac-expressed HopAJ1-Cya protein that failed to be translocated could nevertheless enhance the translocation of test effectors in LT-deficient P. fluorescens(pCPP5316) and LT-deficient DC3000. Thus, it is likely that HopAJ1 functions in the bacterial periplasm to promote T3SS penetration of the peptidoglycan layer, but it does so following export by the Sec system. Interestingly, the strong dominant-negative phenotype exerted by HrpH(E148A) and HrpH_1-241 raises the possibility that LT-deficient HrpH derivatives may disrupt the activity of multiple LTs.

The foregoing observations point to a primary role for HrpH and HopAJ1 in enabling the nascent T3SS to penetrate the peptidoglycan layer. Thus, it was unexpected that a DC3000 ΔhrpH mutant would be deficient in effector translocation rather than secretion and that a ΔhrpH hopP1 hopAJ1 triple mutant would have a stronger ability to elicit the HR in tobacco leaves than a ΔhrpH or ΔhopAJ1 mutant. One key to these puzzling observations is probably the partial nature of the T3SS defect caused by the mutations: the mutants retain a significant residual ability to elicit the HR and to translocate effectors. P. syringae effector repertoires are now thought to contain a mixture of effectors with potential avirulence activity and suppressors of such avirulence activity. Indeed, several DC3000 effector mutants have been shown to produce a stronger HR in tobacco than the wild type (31), and our data suggest that effectors can differ in the degree to which they are impaired in their translocation by the ΔhrpH hopP1 hopAJ1 triple mutant. Thus, it is possible that a T3SS impaired by the lack of HrpH, HopP1, and HopAJ1 could deliver fewer suppressors and thereby produce an unexpectedly strong HR. However, such an incomplete or unbalanced effector repertoire would be less likely to function well in promoting virulence, which is consistent with our observations.

The residual T3SS activity of the DC3000 LT mutants may be explained by a residual ability of the T3SS machinery to penetrate the peptidoglycan layer without specialized LTs (albeit less efficiently) and/or by the contribution of other LTs. Whatever the explanation, the residual capacity to penetrate the peptidoglycan layer is apparently sufficient for wild-type levels of secretion in culture. It is possible that demands on wild-type levels of T3SS performance are simply higher for translocation in planta, but it is also important to consider the possibility that HrpH has additional activities that contribute to translocation per se. Importantly, HrpH is also a T3SS substrate, and its proposed primary activity could make it the first T3SS protein to enter the T3SS channel that is not itself a channel structural component. Therefore, HrpH could be the first T3SS protein to be delivered to plant cells. In that regard, it is interesting that HopA1-Cya translocation could be partially restored to P. fluorescens(pCPP5316 ΔhrpH) by HopP1, whose translocation behavior mimics that of true effectors. Future investigations will probe the activities of P. syringae Hrp LTs in planta.