Plant amphipathic proteins delay the hypersensitive response caused by harpinPssandPseudomonas syringaepv.syringae

Harpin_Pssfrom the plant pathogenPseudomonas syringaepv.syringaeis a proteinaceous elicitor that induces a hypersensitive response (HR) in non-host plants. The plant products which recognize harpin_Pssin the triggering of the HR are not yet known. According to the elicitor-receptor model, we hypothesize that an exogenous cell membrane receptor infiltrated into the intercellular space will interfere with the interaction between harpin_Pssand the putative receptor. We demonstrate a plant amphipathic protein (AP1) which can postpone the HR induced by harpin_Pssas well asP. syringaepv.syringae.AP1 was extracted by solubilizing proteins from healthy leaves in the non-polar n-octanol buffer followed by a polar Tris buffer. The amphipathic extracts were then further separated by gel filtration and anion exchange chromatography to obtain highly purified AP1. Similar proteins can be extracted from cotton, tomato, and sweet pepper. The N-terminal amino acid sequence of AP1 is conserved among cotton, tomato, and sweet pepper. The postponement of the harpin_Pss-mediated HR was characterized as a competitive dosage-dependent pattern of AP1. An analysis of the bacterial population development indicates that the effect of AP1 on the postponement of bacteria-mediated HR was attributed to the suppression of bacterial growth during the early stages of the HR. The time course analysis of the infiltration indicates that the postponement of HR resulted from the co-interaction between AP1 and the bacteria. Based on these results, we suggest that the postponement of bacteria-mediated HR is due to the interference of the interaction between harpin_Pssand the putative receptor in the plant. Our research provides a new approach to elucidating the role that plants may play in the nonhost response caused by pathogens.     

Harpinpsta from Pseudomonas syringae pv. tabaci Is Defective and Deficient in Its Expression and HR-inducing Activity

To elucidate the role of harpins produced by Pseudomonas syringae, the corresponding hrpZ gene was isolated from P. s. pv. tabaci. The sequence information revealed that this gene carries a serious mutation with 326 bp lacking in the central region and potentially encodes only 140 N-terminal amino acids because of a frame shift. The investigation of biological properties using recombinant harpin indicated harpin_psta was incapable of inducing HR in both host and nonhost plants. Based on an immunoblot analysis to detect harpin from P. s. pathovars in hrp-inducing medium, the truncated harpin_psta was neither expressed nor secreted into the culture medium. These results suggest that harpin is not the sole determinant of the host-parasite specificity in P. s. pv. tabaci. Received 10 August 2000/ Accepted in revised form 21 December 2000     

<Emphasis Type="Italic">N</Emphasis>,<Emphasis Type="Italic">N</Emphasis>-dimethylsphingosine,an inhibitor of sphingosine kinase,induces phytoalexin production and hypersensitive cell death of Solanaceae plants without generation of reactive oxygen species

Plant recognition of elicitors derived from pathogens induces various resistant reactions, including production of reactive oxygen species, hypersensitive cell death and accumulation of phytoalexins. Previously, we isolated a ceramide elicitor from Phytophthora infestans, which activates O₂ ⁻ production of potato suspension-cultured cells. In this study, we employed nine ceramide-related chemicals to test their elicitor activity. Although, none of the tested chemicals induced O₂ ⁻ production, N,N-dimethylsphingosine (DMS) induced accumulation of phytoalexin in potato tubers. In potato, tobacco and Nicotiana benthamiana, DMS also induced rapid cell death. DMS-treated potato cells stained with 4′,6-diamidino-2-phenylindole (DAPI) showed chromatin condensation, and isolated DNA from DMS-treated cells had ladder pattern, confirming that DMS-induced plant cell death is a hypersensitive reaction-like programmed cell death. Involvement of ceramide signaling in induction of plant defense reactions is discussed. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Genetic analyses of Pseudomonas syringae isolates from Belgian fruit orchards reveal genetic variability and isolate-host relationships within the pathovar syringae, and help identify both races of the pathovar morsprunorum

A collection of Pseudomonas syringae and viridiflava isolates was established between 1993 and 2002 from diseased organs sampled from 36 pear, plum and cherry orchards in Belgium. Among the 356 isolates investigated in this study, phytotoxin, siderophore and classical microbiology tests, as well as the genetical methods REP-, ERIC- and BOX- (collectively, rep-) and IS50-PCR, enabled identification to be made of 280 isolates as P. syringae pv. syringae (Pss), 41 isolates as P. syringae pv. morsprunorum (Psm) race 1, 12 isolates as Psm race 2, three isolates as P. viridiflava and 20 isolates as unclassified P. syringae. The rep-PCR methods, particularly BOX-PCR, proved to be useful for identifying the Psm race 1 and Psm race 2 isolates. The latter race was frequent on sour cherry in Belgium. Combined genetic results confirmed homogeneities in the pvs avii, and morsprunorum race 1 and race 2 and high diversity in the pv. syringae. In the pv. syringae, homogeneous genetic groups consistently found on the same hosts (pear, cherry or plum) were observed. Pathogenicity on lilac was sometimes variable among Pss isolates from the same genetic group; also, some Psm race 2 and unclassified P. syringae isolates were pathogenic to lilac. In the BOX analyses, four patterns included 100% of the toxic lipodepsipeptide (TLP)-producing Pss isolates pathogenic to lilac. Many TLP-producing Pss isolates non-pathogenic to lilac and the TLP-non-producing Pss isolates were classified differently. Pseudomonas syringae isolates that differed from known fruit pathogens were observed in pear, sour cherry and plum orchards in Belgium.     

Hypersensitive response suppression by type III effectors of plant pathogenic bacteria

The suppressor activity of four representative avirulence (avr) genes from the Pseudomonas syringae group against elicitors of a general hypersensitive response (HR) was examined in tobacco leaves inoculated with double transformants of Pseudomonas fluorescens containing both a cosmid plasmid (pHIR11) carrying the hrp gene cluster and a plasmid carrying each avr gene. The double transformants Pf (pHIR11) containing avrB, avrRps4, or avrPto failed to induce an HR, but that carrying avrRpt2 did induce an HR as Pf (pHIR11 + empty vector) did. Thus, some Avr proteins seem to have suppressor activity against a general HR and should promote aggressiveness of the pathogens.     

Identification and Discrimination of Pseudomonas syringae Isolates from Wild Cherry in England

A survey of wild cherry (Prunus avium) woodland plantations and nurseries was carried out in 2000/01. Trees with symptoms of bacterial canker were found in 20 of the 24 plantations visited and in three of seven nurseries. Fifty-four Pseudomonas syringae isolates from wild cherry together with 22 representative isolates from sweet cherry and 13 isolates from other Prunus spp., pear and lilac were characterised by physiological, biochemical, serological and pathogenicity tests. Isolates from wild cherry were predominantly P. syringae pv. syringae (Pss), but P. syringae pv. morsprunorum (Psm) races 1 and 2 were also found. Physiological and biochemical tests discriminated Psm races 1 and 2 from other P. syringae isolates. Agglutination and indirect-enzyme-linked immunosorbent assay tests with three different antisera showed that Psm race 1 and race 2 were very uniform and indicated high variability amongst other P. syringae isolates. However, pathogenic Pss isolates could not be distinguished from non-pathogenic isolates of P. syringae on the basis of physiological, biochemical or serological tests. Pathogenicity tests on rooted lilac plants and on micropropagated plantlets of lilac and two wild cherry clones differentiated Pss and Psm isolates and demonstrated a range of aggressiveness amongst Pss isolates. Serological tests could be used as an alternative to the classical physiological and biochemical tests to increase the speed of detection and discrimination of isolates, but pathogenicity tests are still necessary to discriminate the pathogenic Pss isolates.     

Involvement of nitric oxide generation in hypersensitive cell death induced by elicitin in tobacco cell suspension culture

Recent studies suggest that nitric oxide (NO), an important signaling and defense molecule in mammals, plays a key role in activating disease resistance in plants. We characterized NO production by tobacco Bright Yellow-2 cells pharmacologically after treatment with INF1, the major elicitin secreted by the late blight pathogen Phytophthora infestans, prepared from Escherichia coli. NO production rapidly occurred within 1h and reached a maximum level 3–6h after the addition of INF1. Carboxy-PTIO, a NO-specific scavenger, abolished INF1-induced NO production in a dose-dependent manner. Pretreatment of protein synthesis inhibitor cycloheximide and protein kinase inhibitor K252a blocked NO production 3–12h after INF1 treatment, indicating that NO production requires de novo protein synthesis and protein phosphorylation. In an investigation of the relations between NO generation and several defense responses induced by INF1, carboxy-PTIO completely suppressed activation of a 41-kDa protein kinase and cell death by INF1. Carboxy-PTIO also suppressed the induction of hypersensitive-related (hsr) genes HSR515 and HSR203J, the expression of which is strongly correlated with the hypersensitive response in plants. The results suggest that NO plays a crucial role in the induction of hypersensitive cell death.     

Protective Effects Against Erwinia amylovora Induced by hrp Mutants in the Whole Plant are Only Partially Mimicked in Cultivated Cells

A virulent strain of Erwinia amylovora, the causal agent of fire blight of Maloideae, and two of its non-virulent hrp mutants (a secretory and a regulatory mutant) were inoculated into apple cell suspensions either alone or in mixed inoculations. In single inoculations, death of 4- to 5-day-old apple cells occurred only when the concentration of the virulent strain of E. amylovora reached a threshold inoculum concentration of 10⁴CFUml^–1, while high concentrations of the hrp mutants were unable to kill apple cells. When mixed inoculated with the virulent parental strain, both hrp mutants protected apple cells from death caused by the virulent strain. The protective effect was associated with a decrease in the population level of the virulent strain and it was dependent on the non-virulent to virulent inoculum concentration suggesting a competition between the non-virulent mutant and the virulent strain. However, no differential protective ability between the two types of mutants could be noticed, contrary to previous results obtained with apple seedlings or apple flowers in which the regulatory mutant was significantly more effective than the secretory mutant. In conclusion, inoculation of apple cell cultures with E. amylovora does not seem to be a model suitable for investigating mechanisms leading to protection.     

Gene for pathogenicity and ability to cause the hypersensitive reaction cloned from Erwinia amylovora

A genomic library of Erwinia amylovora isolate T was constructed in the cosmid pLAFR3 and maintained in Escherichia coli. Clones were transferred individually by conjugation into the non-pathogenic isolate P66 of E. amylovora. Transconjugants were screened for restoration of pathogenicity to pear by stab inoculation into sections of immature pear fruits. Three clones complemented P66 restoring pathogenicity and ability to cause the hypersensitive reaction (HR) in Phaseolus vulgaris. Restriction mapping and hybridization experiments showed that the three clones had a common 3·7 kb fragment of E. amylovora DNA. Sub-cloning and insertion mutagenesis with Tn5-lac confirmed that a determinant of pathogenicity and ability to cause the HR (hrp gene) was located on a 2·1 kb HindIII/BamHI fragment within the common DNA. Hybridization experiments using the 2·1 kb HindIII/BamHI fragment as a probe demonstrated that the hrp gene was located in the chromosome of isolate T and that homologous sequences were present in the non-pathogenic isolates P66 and S. Clones which restored hrp function did not affect the growth of isolate P66 in minimal or nutrient-rich media. Transconjugants of Pseudomonas syringae pv. phaseolicola race 1 harbouring the hrp gene(s) cloned from E. amylovora did not cause the HR in susceptible cultivars of bean but symptoms developed more slowly than in the absence of the clones or with pLAFR3 alone.     

Identification of dspEF, hrpW, and hrpN loci and characterization of the hrpN Ep gene in Erwinia pyrifoliae

Erwinia pyrifoliae, the causal pathogen of shoot blight in the Asian pear tree (Pyrus pyrifolia cv. Singo), is host-specific and endemic to Korea. To identify the genes associated with the hypersensitive response (HR) and pathogenicity, a genomic library of E. pyrifoliae WT3 was constructed, and the cosmid clone Escherichia coli (pCEP33) was selected. Sequence analysis of 19.7-kb pCEP33 determined disease-specific (dsp) region homolog and approximately 40% of the hrp genes, which included hrpW, hrpN_Ep, hrpV, hrpT, hrcC, hrpG, hrpF, and partial hrpE homologs, with respect to the cluster of Erwinia amylovora. Additionally, two open reading frames, ORFD and ORFE, were found downstream of the dspEF region. The results of the sequence analysis showed that the pCEP33 did not contain any hrp regulatory genes or most of the genes encoding components of the Hrp protein secretion system. The hrpN_Ep gene of E. pyrifoliae contained five intergenic nucleotide fragment insertions (INFIs) and produced the HR elicitor protein harpin_Ep, with a molecular mass of approximately 44kDa. The purified HrpN_Ep protein elicited faster and stronger HR when infiltrated into tobacco leaves than did HrpN_Ea from E. amylovora. To observe the role of the hrpL gene in the expression of HrpN_Ep, the pEL2 containing hrpL was used to transform E. coli (pCEP33). Expression of HrpN_Ep in E. coli (pCEP33 + pEPL2) was detected with an immunoblot using antiserum raised against HrpN_Ep, indicating a role of hrpL gene in enhancing the expression of HrpN_Ep.     

Histochemical studies on the accumulation of reactive oxygen species (O2 and H2O2) in the incompatible and compatible interaction of wheat—Puccinia striiformis f. sp. tritici

The generation and accumulation of reactive oxygen species (ROS), superoxide anion (O₂⁻) and hydrogen peroxide (H₂O₂), were studied in the interaction between wheat cv. ‘Suwon 11’ and two races of Puccinia striiformis f. sp. tritici (avirulent and virulent). Generation of O₂⁻ and H₂O₂ was analyzed histochemically using nitroblue tetrazolium (NBT) and 3,3-diamino-benzidine (DAB), respectively. At the pre-penetration stage during appressorium formation both stripe rust races induced H₂O₂ accumulation in guard cells. In the incompatible interaction, a rapid increase of O₂⁻ and H₂O₂ generation at infection sites was detected. The percentage of infection sites showing NBT and DAB staining was 36.1% and 40.0%, respectively, 12 h after inoculation (hai). At extended incubation time until 24 hai, percentage of infection sites showing H₂O₂ accumulation further increased, whereas those exhibiting O₂⁻ accumulation declined. The early infection stage from 12 to 24 hai coincided with primary haustoria formation in mesophyll cells. In contrast, in the compatible interaction, O₂⁻ and H₂O₂ generation could not be detected in most of the infection sites. In the incompatible interaction, intensive DAB staining was also determined in mesophyll cells, especially in cell walls, surrounding the infected cells 16–24 hai; thereafter, these cells contained fluorescing compounds and underwent hypersensitive response (HR). The number of necrotic host cells surrounding the infection sites increased continuously from 20 to 96 hai. It might be concluded that H₂O₂ accumulation during the early infection stage is associated with the occurrence of hypersensitive cell death and that resistance response is leading to arrest the avirulent race of the obligate stripe rust pathogen. In the compatible interaction at 96 hai, H₂O₂ accumulation was observed in mesophyll cells surrounding the rust lesion.     

Hydroperoxides of fatty acids induce programmed cell death in tomato protoplasts

Arachidonic acid (AA) is a fatty acid elicitor abundant in the glycerolipids of the late blight pathogen Phytophthora infestans and related Oomycete species. Lipoxygenases (LOX), which catalyze the addition of molecular oxygen to the 1 or 5 position of a cis, cis -1,4- Z, Z -pentadiene system in polyunsaturated fatty acids, is induced in host plants such as tomato and potato during infection by P. infestans. Here it is reported that AA, the LOX metabolites of AA, 5- and 15-hydroperoxyeicosatetraenioc acid (5- and 15-HpETE), and the LOX metabolite of linoleic acid, 9-hydroperoxyoctadecadienoic acid (9-HpODE), are potent inducers of programmed cell death (PCD) in tomato protoplasts. 5- and 15-HpETE increased DNA fragmentation as detected by t erminal deoxynucleotidyl transferase-mediated d U TP-X n ick e nd l abeling (TUNEL) and increased DNA laddering as visualized by ligation-mediated PCR. Background levels of DNA laddering were decreased in intensity by Zn²⁺ and increased by Ca²⁺, effects that are consistent with the reported action of these cations on PCD-associated endonucleases in other systems. H₂O₂, methyl jasmonate, and linoleic and linolenic acids were not toxic to tomato protoplasts at concentrations up to 350 μ , and lipid peroxides (LD₁₀₀ = 80 μ ) were far more potent inducers of death than free AA within this same concentration range. These results indicate the potential of fatty acid peroxides and LOX-related metabolism to engage an apoptotic type of PCD in higher plant cells.     

Programmed cell death pathways induced by early plant‐virus infection are determined by isolate virulence and stage of infection

Programmed cell death (PCD) pathways caused by Turnip mosaic virus (TuMV) infection before symptom appearance were studied by light microscopy and electrolyte leakage following sap inoculation of Brassica carinata (Ethiopian mustard) TZ‐SMN‐44‐6 plants. Leaf responses to inoculation with avirulent (TuMV‐avir) and virulent (TuMV‐vir) isolates, and mock‐inoculation, were compared at 2, 20 and 52 h after inoculation (hai). The phenotypes induced were localized resistance (TuMV‐avir) and systemic susceptibility (TuMV‐vir). No visible TuMV symptoms were recorded in any inoculated plants during the 2–52 hai sampling period, but appeared as chlorotic spots in inoculated leaves at 5 days after inoculation. With TuMV‐vir alone, they were followed by systemic infection (mosaic). Dead cell number, deformation, percentage area and percentage integrated intensity, and conductivity of electrolyte leakage data, were analysed to examine their possible roles in stimulating cell death pathways. At 2 hai, dead cell number and percentage area were significantly greater for TuMV‐avir than TuMV‐vir infection or mock‐inoculation. Overall, isolate TuMV‐vir caused significantly greater cell deformation than TuMV‐avir, whereas wounding by mock‐inoculation had negligible effects. By 52 hai, isolate TuMV‐avir caused significantly greater electrolyte leakage than isolate TuMV‐vir or mock‐inoculation. This suggests both isolates triggered morphological changes consistent with apoptotic‐like PCD and necrosis‐like PCD that depended upon isolate virulence and stage of infection, respectively. These findings highlight how quantification of dead cell deformation and electrolyte leakage offer a new understanding of compatible and incompatible plant responses to early virus infection in plants.     

Expression of <Emphasis Type="Italic">Xanthomonas oryzae </Emphasis>pv. <Emphasis Type="Italic">oryzae hrp</Emphasis> Genes in XOM2, a Novel Synthetic Medium

To analyze the regulation of hrp expression and to detect and identify hrp-dependent secretion proteins of plant-pathogenic bacteria, an appropriate hrp-inducing medium is indispensable. In this study, two efficient hrp-inducing media for Xanthomonas oryzae pv. oryzae were designed by assaying the expression of a hrcU (the first gene of the hrpC operon) and a gus (β-glucuronidase) fusion gene. We modified XVM2, which is a hrp-inducing medium for X. campestris pv. vesicatoria, by adding 0.01% xylose in place of fructose and sucrose (0.18 and 0.34%, respectively) as a sugar source. The resulting medium induced approximately 15-fold more GUS activity from transformants containing a hrcU::gus gene than did XVM2. Moreover, a methionine-containing synthetic medium with 0.18% xylose as a sugar source was able to induce much stronger expression of HrcU::GUS, with GUS activity approximately 100-fold greater than that in XVM2. Induction depended on a regulator, HrpXo, and the PIP (plant-inducible-promoter) box, suggesting that HrcU::GUS was expressed in a hrp-dependent manner. The induction of operons hrpA to hrpF in XOM2 was also confirmed. These results suggest that both media, especially XOM2, are highly efficient hrp-inducing media for X. oryzae pv. oryzae. Received 7 October 2002/ Accepted in revised form 22 November 2002     

Flagellin from Pseudomonas syringae pv. tabaci induced hrp-independent HR in tomato

We have previously shown that flagellin of Pseudomonas syringae pv. tabaci is an elicitor that induces a hypersensitive reaction (HR) in nonhost tomato cells. Flagellin is the major HR elicitor produced by this pathogen, as shown by the inability of a flagellin-defective mutant, ΔfliC, to induce HR. Also, a ΔfliD mutant that secretes large amounts of monomer flagellins induces a strong HR in tomato. In this study, the possible involvement of an Hrp type III secretion system (TTSS) in flagellin-induced HR was investigated using flagella-defective mutants or Hrp TTSS-defective mutants. The hrcC gene encodes HrcC protein, which is required for Hrp pilus formation in the outer membrane. An hrcC mutation, introduced into the wild-type, ΔfliC, and ΔfliD mutants of P. syringae pv. tabaci did not affect swimming motility or flagellin secretion, whereas all ΔhrcC, ΔfliC, and ΔfliD mutants lost the ability to cause disease on host tobacco leaves. However, the ΔhrcC mutant and the ΔfliD/ΔhrcC double mutant were still able to induce HR cell death, expression of one of the defense-related genes hsr203J, and the generation of hydrogen peroxide in nonhost tomato cells. Thus, flagellin is required for both pathogenicity in host tobacco and HR in nonhost tomato. On the other hand, hrp TTSS is necessary for pathogenicity on host tobacco but is not indispensable to induce HR in nonhost tomato. These results clearly show that flagellin-induced HR is hrp-independent in tomato.The nucleotide sequence data reported are available in the DDBJ/EMBL/GenBank databases under the accession number AB049570     

A functional gene-for-gene interaction is required for the production of an oxidative burst in response to infection with avirulent Pseudomonas syringae pv. tomato in Arabidopsis thaliana

An early event correlated with the gene-for-gene hypersensitive response (HR) is the accumulation of active oxygen species (AOS), also known as the oxidative burst. We present data that genetically demonstrates that the oxidative burst is a downstream component of the RPS2- avrRpt2gene-for-gene signal cascade. An in planta AOS assay using the fluorescent probe 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) was modified for use with the Arabidopsis thaliana / Pseudomonas syringae pv.tomato (P. syringae pv. tomato) model system. An oxidative burst occurred between 8 and 15 hpi with avirulent P. syringae pv. tomato(avrRpt2), but not with virulent P. syringae pv. tomato. This burst preceded cell death and was not observed in the RPS2 Arabidopsis mutantsrps2-101C and rps2-201 inoculated with avirulent P. syringae pv. tomato. An HR-like response has been observed when plants undergoing a systemic acquired resistance (SAR) response are challenged with a normally virulent pathogen (manifestation stage of SAR), however an HR-like oxidative burst was not detected by the in planta AOS assay during this stage of SAR.     

SGT1 contributes to maintaining protein levels of MEK2DD to facilitate hypersensitive response-like cell death in Nicotiana benthamiana

Gene silencing revealed that the mitogen-activated protein kinase (MAPK) cascade in Solanaceae consisted with MEK2-WIPK/SIPK, is required for R protein-induced hypersensitive response (HR) cell death and/or resistance. Overexpression of MEK2^DD results in HR-like cell death. MEK2^DD is a phospho-mimic and constitutive active form harboring mutations at putative phosphorylation sites of upstream MAPKKK. The molecular mechanism that induces HR-like cell death is unknown. Here we report SGT1 is required for the accumulation of MEK2^DD protein, not MEK2^WT. Virus-induced gene silencing of SGT1 resulted in low protein accumulation of MEK2^DD. This result suggests that SGT1 has a positive role in the accumulation of the MEK2 active form protein to facilitate signal transduction.     

Harpin-elicited hypersensitive cell death and pathogen resistance require the NDR1 and EDS1 genes

Plants sprayed with harpin, a bacterial protein that induces hypersensitive cell death (HCD), develop systemic acquired resistance (SAR) without macroscopic necrosis. HCD sometimes accompanies the development of resistance conferred by resistance (R) genes. In Arabidopsis, some R genes require one or both of the signalling components NDR1 and EDS1 for function. This study addresses whether HCD, NDR1 and EDS1 are required for induction of SAR by harpin. When Arabidopsis and tobacco leaves were sprayed with harpin, microscopic hypersensitive response (micro-HR) lesions developed. Systemic expression of PR genes and the development of resistance were accompanied by micro-HR, except in the ndr1-1 mutant, in which harpin induced micro-HR without the development of resistance or expression of the PR-1 gene. Cell death and resistance did not occur following treatment with harpin in plants that could not accumulate salicylic acid. Harpin also failed to induce resistance in Arabidopsis eds1-1 mutants. Therefore, harpin-induced resistance seems to develop concomitantly with cell death and resistance requires NDR1 and EDS1.     

Hpa1 secretion via type III secretion system in <Emphasis Type="Italic">Xanthomonas oryzae</Emphasis> pv. <Emphasis Type="Italic">oryzae</Emphasis>

In many Gram-negative plant pathogenic bacteria the type III secretion system (TTSS), encoded by hrp genes, is essential for pathogenicity in the host and induction of a hypersensitive reaction (HR) in nonhost plants. The expression of hrp genes has been suggested to be repressed in complex media, whereas it is induced in planta and under certain in vitro conditions. We recently reported that XOM2 medium allows efficient hrp expression by Xanthomonas oryzae pv. oryzae. In this study, we investigated hrp-dependent secretion of proteins by the bacteria in vitro. Using modified XOM2, in which bovine serum albumin was added and the pH was lowered to 6.0, we detected at least 10 secreted proteins and identified one as Hpa1. This is the first evidence of protein secretion via TTSS in X. oryzae pv. oryzae.