e-Xtra* A New Pathovar, Pseudomonas syringae pv. alisalensis pv. nov., Proposed for the Causal Agent of Bacterial Blight of Broccoli and Broccoli Raab

N. A. Cintas, USDA, ARS, PWA, 1636 E. Alisal Ave., Salinas, CA 93905; S. T. Koike, University of California, Cooperative Extension, Salinas 93905; and C. T. Bull, USDA, ARS, PWA, 1636 E. Alisal Ave., Salinas, CA 93905

year. The second crucifer crop can become ABSTRACT infected when planted in residue of a pre- Cintas, N. A., Koike, S. T., and Bull, C. T. 2002. A new pathovar, Pseudomonas syringae pv. vious crop that had one of the bacterial alisalensis pv. nov., proposed for the causal agent of bacterial blight of broccoli and broccoli blights. The broccoli raab bacterial blight raab. Plant Dis 86:992-998. pathogen caused disease in a second plant- ing of broccoli raab up to 4 months after The etiology of three foliar bacterial diseases of crucifers and the relationships between their incorporating the initial infected crop (7). causal agents were evaluated. Data from LOPAT, carbon utilization tests, and fatty acid analysis Because the pathogens of broccoli, broc- indicated that bacterial blights of broccoli and broccoli raab, and leaf spot of broccolini, were coli raab, and broccolini have not been caused by strains of Pseudomonas syringae. Data from phage sensitivity, ice nucleation, single carbon source utilization, Polymerase chain reaction using BOXA1R primer (BOX-PCR), and completely characterized, it is not known if host range analyses were identical for the pathogen causing leaf spot of broccolini and P. syrin- they have the same or similar host ranges. gae pv. maculicola. The broccoli raab and broccoli pathogens infected broccoli raab, all cruci- We are, therefore, unable to predict if the fers tested, tomato, and three monocots (California brome, oat, and common timothy). None of pathogen from one crop has the potential to the other pathogens tested (P. syringae pv. maculicola, P. syringae pv.tomato, or P. syringae pv. infect a second planting of a different host. coronafaciens) caused disease on broccoli raab or on both crucifers and monocots. Data from Having a better understanding of the host phage sensitivity, ice nucleation, single carbon source utilization, BOX-PCR, and host range range of these pathogens would be useful analyses were identical for the pathogens from broccoli raab and broccoli, but were different to make crop rotation recommendations for from other pathovars tested, and supported the hypothesis that a new pathovar of P. syringae pv. second plantings. Our goal was to charac- alisalensis pv. nov. caused a leaf blight on broccoli and broccoli raab. terize the pathogens to aid in making these management choices. The specific objec- Additional keywords: BOX-PCR, coronatine production, phage sensitivity tives of this work were: (i) to determine if these three new bacterial diseases are caused by the same pathogen; (ii) to clarify Broccoli (Brassica olearacea var. botry- over $11 million; 23). This leafy vegetable host ranges of these three bacterial patho- tis) is a valuable U.S. commodity both has a flavor between broccoli and mustard gens; and (iii) to clarify the taxonomy of nutritionally and economically. It is an greens and is widely grown in Europe (11). these pathogens. excellent source of vitamin C and provides Broccolini (Brassica oleracea var. botrytis dietary fiber, protein, iron, calcium, and × Brassica alboglabra), is another novel MATERIALS AND METHODS vitamin A. Broccoli also contains antican- crucifer grown in Monterey County. This Microorganisms and media. The bac- cer components (37). The financial benefit gourmet broccoli is a cross between broc- teria used in these studies are listed in Ta- of broccoli to the U.S. agricultural industry coli and chinese kale (gai lan). Acreage ble 1. For all experiments, P. syringae pv. is significant. The value of the 1999 U.S. and demand for this crop are expected to coronafaciens, P. syringae pv. syringae, P. broccoli crop was over $493 million (1). increase over the next few years as con- syringae pv. maculicola, and P. syringae Approximately 95% of U.S. broccoli pro- sumers become more familiar with it. pv. tomato were used as controls. The bac- duction occurs in California, with Mon- Since 1995, leaf lesions caused by bacte- teria were stored at –80°C in a solution of terey County in the Central Coast Region, ria have begun to appear on all three of 50% glycerol and 50% nutrient broth (NB; producing 50% of the state’s acreage. these hosts. Bacterial leaf spot of cauli- Difco Laboratories, Detroit, MI) and rou- Monterey County is the leading broccoli flower, caused by P. syringae pv. maculi- tinely cultured on King’s medium B agar producing county in the country (25). cola, was initially documented in 1911 (KMB; 16). All chemicals were purchased Broccoli production in Monterey County in (29). Although this pathogen is now known from Sigma Chemical Company (St. 1999 was worth over $359 million (23). to infect many crucifers (3,42), the symp- Louis, MO) unless otherwise indicated. Broccoli raab (Brassica rapa subsp. toms on broccoli raab and broccoli are not Isolation of the causal agent. Sympto- rapa), also known as rappini, has become similar to those caused by P. syringae pv. matic leaves were surface-sterilized with an important crucifer grown in coastal maculicola. Bacterial blight of broccoli (0.525%) sodium hypochlorite for 1 min California for east coast markets (worth raab was the first crucifer blight described followed by rinsing in sterile distilled wa- in the Salinas Valley of California (21). ter three times. Small (3 × 3 mm) sections &RUUHVSRQGLQJÃDXWKRUÃ&Ã7Ã%XOOÃ Initial symptoms consist of small water- of tissue were aseptically excised from leaf E-mail: FEXOO#SZDUVXVGDJRY soaked flecks on the lower foliage. These spot margins and macerated in 40 µl of flecks expand and become surrounded by sterile distilled water. The resulting sus- *The e-Xtra logo stands for “electronic extra” and bright yellow borders, which eventually may pensions or a dilution were streaked on indicates the HTML abstract available on-line coalesce and result in large necrotic areas. sucrose peptone agar (4) and incubated at contains supplemental material not included in the Later, a similar bacterial blight of broccoli 24 to 26°C. After 3 to 5 days, single colo- print edition. emerged (20). Most recently, a bacterial leaf nies were purified and stored. Accepted for publication 9 May 2002. spot of broccolini developed on transplants Isolations of fungi from diseased tissue (5). Although initial characterization of these were attempted because fungal diseases pathogens suggests that they are related cause lesions on the leaves of crucifers in Publication no. D-2002-0628-01R pathovars of P. syringae, specific characteri- the Salinas Valley (18,22). Small leaf sec- This article is in the public domain and not copy- zation was incomplete (6). tions were aseptically removed and placed rightable. It may be freely reprinted with custom- ary crediting of the source. The American Phyto- In the Salinas Valley, many fields are on acidified potato dextrose agar (2 ml of pathological Society, 2002. planted with two crops of crucifers in 1 lactic acid per liter). Plates were incubated

992 Plant Disease / Vol. 86 No. 9 under lights at 24°C and examined after 3 ters using a standard method (30). Fatty strains present on each plate. Plates were to 7 days for fungal growth. acids were analyzed with the Sherlock incubated at 27°C and results were re- Pathogenicity tests. The causal agents Microbial Identification System Version corded after 3, 7, and 14 days. Three or of the new diseases and strains of the con- 2.11 (MIDI Inc., Newark, DE) that used an four replications were made for each or- trol pathovars were tested on plants from automated GC 6890 Hewlett-Packard gas ganism and the experiment was conducted the host range of P. syringae pv. tomato, P. chromatograph fitted with a 25 × 0.2 mm three times. syringae pv. maculicola, P. syringae pv. phenyl methyl silicone-fused silica capil- Isolation of phage and test of patho- coronafaciens and the hosts of the new lary column, an HP 7673 automatic sam- gen sensitivity. A bacteriophage was re- diseases. Hosts tested included: broccoli pler, and HP Chem Station Software (30). covered from standing water in a commer- (cv. Greenbelt), cauliflower (Brassica ol- The analyses included assessing the degree cial broccoli raab field in the Salinas eracea subsp. botrytis cv. White Rock), of similarity of fatty acid composition. Valley in 1998. Enrichment flasks were broccoli raab (cv. Spring), broccolini (cv. Analyses were conducted three times for prepared by adding 50 ml of field water to Aspabrock), California brome (Bromus each strain. Strains were also compared 2 g of CaCO3 and 3 ml of overnight NB carinatus), corn (Zea mays cvs. Kandy three times by their carbon source utiliza- cultures of the broccoli raab pathogen. The Korn and Silver Queen), oat (Avena sativa tion profiles on Biolog GN microplates flasks were incubated at 27°C overnight. cv. Montezuma), rye (Secale cereale cv. (Biolog, Inc., Hayward, CA). Standard methods were used to recover and Merced), common timothy (Phleum prat- Carbon utilization. Because the use of isolate phage as well as determine sensitiv- ense), and tomato (Lycopersicon esculen- four carbon sources, erythritol, trigonel- ity of strains to the isolated phage (17). A tum cv. Early Girl). line, ascorbic acid, and D-tartrate can dis- routine test dilution giving confluent lysis To test pathogenicity, NB was inocu- tinguish P. syringae pv. coronafaciens, P. of the propagating strain was calculated lated with individual strains and cultures syringae pv. syringae, and P. syringae pv. from a prepared phage stock. A 300-µl were shaken at 200 rpm for 24 h. Cultures maculicola, the ability of the uncharacter- aliquot of an overnight bacterial culture were centrifuged at 7,000 rpm for 10 min, ized strains to use these carbon sources was added to 10 ml of melted soft agar. the broth decanted, and the bacterial pellet was tested under more rigorous conditions. The suspension was poured over nutrient resuspended in sterile, distilled H2O. The Standard mineral base agar (SMB; agar (Difco Laboratories) in a 100 × 15 resulting bacterial suspensions were ad- Na2HPO4·H2O, 4.5 g; KH2PO4, 4.5 g; mm plate. Ten µl drops of 1× and 10× con- justed to an optical density of 0.10 at 600 NH4Cl, 1.0 g; MgSO4·7H2O, 0.5 g; 5% centrations of the routine test dilution were nm with a spectrophotometer (Shimadzu ferric ammonium citrate, 1 ml; 0.5 % placed on the solidified soft agar plates. UV-1601, Kyoto, Japan). This concentra- CaCl2 solution, 1.0 ml; agar, 16 g; H2O, 1 The plates were incubated at 27°C for 48 h tion corresponded to approximately 1 × 107 liter; 10) was used in these experiments. before recording lysis. CFU/ml as determined by dilution plating. After washing, all glassware was rinsed six Ice nucleation. Ice nucleating ability Tween 20 (0.05%) was added to the sus- times with distilled water. Filter-sterilized was tested as previously described (10). pensions, and each suspension was sprayed active carbon radical (0.3%) was added to Cultures grown for 4 days at 21°C on glu- until runoff using a hand mister. After in- molten agar after autoclaving. Bacteria cose yeast extract peptone agar (glucose, 5 oculation, plants were placed in a humid were initially cultured on SMB agar with g; yeast extract, 5 g; peptone, 5 g; agar 5 g; chamber for 48 h, then maintained in a 0.3% glucose as the sole carbon source H2O, 1 liter) were tested for their ability to greenhouse at 20 to 25°C and evaluated for before transferring single colonies to the nucleate ice at –4°C. Cells were removed symptoms after 14 days. For negative con- agar with the test carbon sources. Negative from the plates and suspended in 5 ml of trols, plants were treated with sterile, dis- and positive controls were among the eight water. Suspensions (>1 × 109 CFU/ml) tilled H2O plus Tween 20 (0.05%). An experimental unit was six plants in a six- pack container. Each treatment was repli- Table 1. Bacterial strains used in this study cated three times for each host. This ex- Strain Original host Source periment was conducted twice, once with plants produced and incubated at the Pseudomonas syringae pv. alisalensis USDA/ARS greenhouse and once at the BS91z Brassica rapa subsp. rapa (broccoli raab) ATCC – This study University of California Cooperative Ex- accession BAA - 566 BS92 B. rapa subsp. rapa (broccoli raab) This study tension (UCCE) facility. For tests evaluat- 1097-11 B. rapa subsp. rapa (broccoli raab) Koike (28) ing reactions on tomato, the experiments BS228 B. rapa subsp. rapa (broccoli raab) This study were conducted four times. BS130 B. oleracea var. botrytis (broccoli) This study Characterization of pathogens. From BS132 B. oleracea var. botrytis (broccoli) This study each host, three to five strains were tested BS136 B. oleracea var. botrytis (broccoli) This study for morphological, biochemical, and P. syringae pv. maculicola physiological characteristics. When possi- 0649-18 B. oleracea var. botrytis (cauliflower) H. Bouzar ble, the strains came from separate disease 0788-18 B. oleracea var. botrytis (cauliflower) Koike (28) BS224 B. oleracea var. botrytis (cauliflower) Koike (28) incidences. BS220 B. oleracea var. botrytis × B. alboglabra (broccolini) H. Bouzar Morphology was determined by examin- BS221 B. oleracea var. botrytis × B. alboglabra (broccolini) H. Bouzar ing strains grown on mannitol glutamate BS235 B. oleracea var. botrytis × B. alboglabra (broccolini) This study yeast extract agar plates (MGY; 15). The BS236 B. oleracea var. botrytis × B. alboglabra (broccolini) This study potassium hydroxide solubility test (KOH BS244 B. oleracea var. botrytis × B. alboglabra (broccolini) This study test) to determine gram character and levan P. syringae pv. tomato formation, oxidase test, potato soft rot, 0482-1 Lycopersicon esculentum (tomato) Koike (28) 0183-4 L. esculentum (tomato) Koike (28) arginine dihydrolase, tobacco hypersensi- 0683-23 L. esculentum (tomato) Koike (28) tivity (LOPAT) test were conducted as P. syringae pv. coronafaciens described previously (34). NCPPB600z Avena sativa (oat) Koike (28) For fatty acid analyses, strains were cul- 345 Secale cereale (rye) Cupples (8) tured on tryptic soy broth agar (30 g tryptic ICMP 3114 A. sativa (oat) Koike (28) P. syringae pv. syringae soy broth, 15 g agar, 1 liter H2O; Difco Laboratories, Detroit, MI) at 28°C for 24 B301Dr Pyrus communis (pear) Xu et al. (41) h, then extracted for fatty acid methyl es- z Type strain.

Plant Disease / September 2002 993 were held on ice until tested. After adjust- Genotypic characterization using rep- Pathogenicity of P. syringae patho- ing an ice-ethanol water bath to –4°C, the PCR. The polymerase chain reaction of vars. The broccoli raab and broccoli bacterial suspensions were immersed in the repetitive bacterial sequences, (rep-PCR), pathogens caused leaf blight symptoms on bath for 5 min. Samples that froze were was used to determine the relationship be- all crucifers tested, a leaf speck on tomato, considered ice nucleation positive. This tween P. syringae strains causing bacterial and leaf lesions on several monocots: experiment was conducted twice. blights on broccoli raab, broccoli, and broc- brome, oat, and common timothy (Table Characterization of coronatine pro- colini. PCR amplification of repetitive se- 2). Corn and rye plants remained asymp- duction genes by Southern hybridiza- quences was accomplished with the tomatic. In contrast, P. syringae pv. macu- tion. DNA used as probes in Southern BOXA1R primer (5‡-CTA-CGG-CAA- licola strains caused leaf spot symptoms on hybridization experiments was isolated GGC-GAC-GCT-GAC-G-3‡), synthesized all crucifers except broccoli raab, had vari- using Quiagen Tip 100 (Diagen GmbH, by Operon Technologies, Inc. (Alameda, able reactions on tomato, and did not cause Hilden, Germany). Genomic DNA was CA), that is designed to prime DNA synthe- symptoms on any monocot host. P. syrin- isolated using CTAB as described in pub- sis from the BOXA subunit of the BOX gae pv. tomato strains caused bacterial lished methods (2,33). Restriction endonu- element (31,38). The PCR was carried out speck symptoms on tomato, but all other cleases were used as recommended by the according to published methods for BOX- hosts remained asymptomatic. P. syringae supplier (GIBCO-BRL Life Technologies, PCR using purified genomic DNA (33). pv. coronafaciens strains caused symptoms Gaithersburg, MD or Boehinger Mann- DNA samples were prepared by large-scale on oat, rye, and brome, but did not induce heim, Indianapolis, IN). DNA was sepa- chromosomal DNA isolations (2). The PCR symptoms on any cruciferous host, corn, rated by gel electrophoresis at 50 V for 3 to was carried out in an MJ Research DNA- timothy, or tomato. The P. syringae pv. 12 h in 0.7% agarose gels (SeaKem LE; Engine thermo-cycler with a heated lid in syringae strain did not cause symptoms on FMC BioProducts, Rockland, ME). the ‘block’ mode (MJ Research, Watertown, any host tested. None of the control plants Southern hybridizations were performed MA) with the use of Taq polymerase (Perkin inoculated with water and Tween 20 devel- on Hybond N membranes (Amersham, Elmer, Foster City, CA). Amplified DNA oped symptoms. Pathogenecity test results Arlington Heights, IL) using DNA probes fragments were examined by agarose gel were consistent for all strains within a that were random-prime labeled with di- electrophoresis in 1.5% agarose gels in 0.5× pathovar group, except for P. syringae pv. goxigenin-11-dUTP. Labeling, hybridiza- Tris acetate EDTA buffer, TAE (31). Gels maculicola on tomato. In general, results tion, and detection were conducted accord- were stained with ethidium bromide and from individual experiments were identical ing to manufacturer’s recommendations DNA was photographed on a UV transillu- for all strains. However, variable reactions (Boehinger Manheim). The probe was the minator with Polaroid type 55 film. Finger- occurred on tomato with P. syringae pv. coronatine production genes present on prints generated from different strains were maculicola and the pathogen from broc- pTPR1 cloned from P. syringae pv. tomato compared visually. colini (Table 2). DC3000 (8). Fragments used as probes Phenotypic characterization of bacte- were purified from low melting point aga- RESULTS rial pathogens. LOPAT reactions con- rose gels (SeaKem GTG; FMC BioPro- Isolation of the causal agent. No fungi firmed that strains were P. syringae. All ducts). Adsorption and elution to glass were isolated from broccoli raab, broccoli, strains were positive for levan production, milk was used to remove solutes as de- and broccolini lesions. Large numbers of produced a hypersensitive response on scribed in the manufacturer’s recommenda- fluorescent pseudomonads were consis- tobacco, and were KOH positive, i.e., gram tions (Geneclean Kit; Bio 101 Inc., Vista, tently isolated from these lesions on negative (Table 3). The strains were oxi- CA). Labeled probes were used at 0.25 blighted broccoli, broccoli raab, and broc- dase negative, arginine dihydrolase nega- µg/ml. DNA to be probed was transferred colini plants. Isolated suspect bacteria were tive, and did not rot potato slices. from agarose gels to nylon membranes by used to inoculate suspect host plants. Dis- The broccoli raab and broccoli strains, standard methods (33). Hybridization ease resulted, bacteria were reisolated, and Pseudomonas syringae pv syringae, and P. conditions were low stringency (44 to Koch’s postulates were completed for all syringae pv. coronafaciens strains nucle- 55°C and 0.2 × SSC ) (33). three diseases. ated ice (Table 3). The broccolini strains,

Table 2. Disease response of new crucifer pathogens and pathovars of Pseudomonas syringae on crucifers, monocots, and tomato P. syringaev à P. syringaeÃSYÃDOLVDOHQVLVà à P. syringaeÃSYÃPDFXOLFRODà Ãà à à +RVWwà +RVWà à P. syringae SYà P. syringaeÃSYà P. syringae à ([SHULPHQWDOÃKRVWx %URFFROLÃUDDEà %URFFROLà WRPDWRà %URFFROLQLà 9DULRXVà FRURQDIDFLHQVà SYÃV\ULQJDHà 1HJDWLYHÃFRQWUROyà Broccoli raab +z + – – – – – – Broccoli + + – + + – – – Broccolini + + – + + – – – Cauliflower + + – + + – – – Tomato + + + +/– +/– – – – California brome + + – – – + – – Corn – – – – – – – – Oat + + – – – + – – Rye – – – – – + – – Timothy + + – – – – – – v 7 P. syringae strains were applied at 1 × 10 CFU/ml in H2O and Tween 20 (0.05%). Strains tested: P. syringae pv. alisalensis, BS91, BS92, BS130, BS132; P. syringae pv. maculicola from cauliflower, 0649-18, 0788-18, P. syringae pv. maculicola from broccolini BS220, BS221; P. syringae pv. to- mato 0482-1, 0183-4; P. syringae pv. coronafaciens NCPPB600, 345; P. syringae pv. syringae B301Dr. For each pathovar, all strains resulted in the same reaction, except for P. syringae pv. maculicola on tomato. w Host indicates the original host from which the pathogen was isolated. x Experimental host indicates the plants on which pathogenicity was tested. y Negative controls were sprayed with H2O and Tween 20 (0.05%). z Summary of results from all tests. + = plants showed disease symptoms; – = plants showed no disease symptoms; +/- = some plants had symptoms but not all or not in all experiments.

994 Plant Disease / Vol. 86 No. 9 P. syringae pv. tomato, and P. syringae pv. each pathovar of P. syringae (Fig. 1). The broccoli raab and broccoli strains had maculicola did not nucleate ice. Fragments ranging in size from 0.6 to 3.0 a unique profile. The broccoli raab and broccoli strains kbp were amplified with the BOXA1R were sensitive to the bacteriophage PBS1 primer. Distinctive profiles were gener- DISCUSSION (Table 3). P. syringae pv. coronofaciens, P. ated for all of the previously identified Three new foliar bacterial diseases of syringae pv. maculicola, P. syringae pv. pathovars of P. syringae (P. syringae pv. crucifers, on broccoli raab, broccoli, and tomato, and P. syringae pv. syringae were tomato, P. syringae pv. maculicola, P. broccolini, have recently emerged in the not sensitive to the phage. syringae pv. coronafaciens, and P. syrin- Salinas Valley. Phenotypic characteristics Coronatine production genes were found gae pv. syringae) evaluated. All strains as well as fatty acid and genetic tests indi- to be present in all strains of all pathovars tested within a given pathovar had the cate that these strains are LOPAT group I tested except for P. syringae pv. coronafa- identical profiles with only minor poly- fluorescent pseudomonads (24) and are, ciens and P. syringae pv. syringae (Table morphisms occurring in some cases (P. therefore, P. syringae van Hall. Initial re- 3). syringae pv. coronafaciens and P. syrin- search on these diseases did not completely Strains were compared by their carbon gae pv. tomato). The profiles of the broc- characterize the causal agents and their source utilization profiles on Biolog GN colini strains were identical to the P. sy- relationship(s) to each other (5,6,20,21). microplates (Biolog, Inc., Hayward, CA). ringae pv. maculicola strains evaluated. We further characterized the pathogens and All were identified as pathovars of Pseu- domonas syringae. Fatty acid methyl esters (FAME) analysis identified strains as Pseudomonas syringae with a similarity index of 0.80 or higher. Single carbon source tests differenti- ated the broccoli raab strains from all other strains tested (Table 3). The broc- coli raab, broccoli, and P. syringae pv. coronafaciens strains were distinguished from other strains because they did not grow on trigonelline while the broccolini strains, P. syringae pv. tomato, P. syrin- gae pv. syringae, and P. syringae pv. maculicola strains grew on trigonelline. Although P. syringae pv. coronafaciens does not grow on trigonelline, it grows on erythritol, which was not used by the broccoli raab and broccoli strains. Results Fig. 1. Repetitive sequence-based polymerase chain reaction (BOX-PCR) fingerprints from Pseudo- from carbon utilization tests were identi- monas syringae pathovars with the BOX1AR primer. A portion of the PCR products (9 ml) was cal for the broccolini strains, P. syringae loaded in a 1.5% agarose gel, and electrophoresis was performed at 70 V for 12 h in 0.5× TAE buffer. pv. tomato, and P. syringae pv. maculi- K, 1 kb DNA Ladder; sizes are indicated in kilobase pairs. C, water control. Lanes 1 to 3 are P. s y - cola. ringae pv. coronofaciens. Lane 1, 345; 2, ICMP 3114; 3, NCPPB600. Lanes 4 to 11 are P. syringae pv. alisalensis. Lane 4, BS130; 5, BS132; 6, BS136; 7, BS91; 8, BS92; 9, BS101; 10, BS106; 11, Genetic characterization. Primers cor- BS228. Lanes 13 to 17 are P. syringae pv. maculicola.from broccolini. Lane 13, 220; 14, 221; 15, responding to BOX1A subunits of BOX 235; 16, 236; 17, 244. Lanes 18 to 20 are previously identified P. syringae pv. maculicola. Lane 18, elements annealed to genomic DNA and 788-18; 19, 0649-18; 20, 90-32-7A. Lanes 21 to 23 are P. syringae pv. tomato. Lane 21, 0482-1; 22, generated unique genomic fingerprints for 0183-4, 23, 0683-23. Lane 24 is P. syringae pv. syringae, B301Dr.

Table 3. Characterization of Pseudomonas syringae strains P. syringaew P. syringaeÃSYÃDOLVDOHQVLVÃ P. syringaeÃSYÃPDFXOLFRODÃ +RVWw +RVW P. syringae pv. to- P. syringae pv. P. syringae pv. Testy Broccoli raab Broccoli mato Broccolini Various coronafaciens syringae Levan production +z + + + + + + Oxidase – – – – – – – Potato – – – – – – – Arginine – – – – – – – Tobacco HR + + + + + + + KOH reaction + + + + + + + Ice nucleation + + – – – + + Phage sensitivity + + – – – – – Coronatine genes + + + + + – – Carbon utilization Tartaric acid + + + + + – – Erythritol – – – – – + + Trigonelline – – + + + – + Ascorbic Acid – – + + + + + w P. syringae strains tested P. syringae pv. alisalensis, BS91, BS92, BS130, BS132; P. syringae pv. maculicola from cauliflower, 0649-18, 0788-18, P. syringae pv. maculicola from broccolini BS220, BS221; P. syringae pv. tomato 0482-1, 0183-4; P. syringae pv. coronafaciens NCPPB600, 345; P. syrin- gae pv. syringae B301Dr. x Host indicates the original host from which the pathogen was isolated. y Tests performed according to published methods (10,34). z Summary of results from all tests. + = always positive for the test; – = always negative for the test; +/– = sometimes positive and sometimes negative.

Plant Disease / September 2002 995 clarified the relationships among the causal P. syringae pv. maculicola is the most it was important to control as many pa- agents of these three crucifer diseases. We important pseudomonad causing disease on rameters as possible within and among demonstrated that a novel pathogen caused crucifers. P. syringae pv. maculicola tests. We have conducted this experiment bacterial blights on both broccoli raab and (originally designated Bacterium maculi- numerous times, controlling for as many broccoli and that this pathogen was distinct colum) was first described as a leaf spot factors as possible and are confident that from P. syringae pv. maculicola causing pathogen from cauliflower in 1911 by the P. syringae pv. alisalensis is also a bacterial leaf spot on broccolini. McCulloch (29) and since that time it has pathogen of tomato. Data suggested that a single unique been described on at least 25 members of Host range data from strains of P. syrin- pathogen caused bacterial blight of both the Brassicaceae including broccoli (3,42). gae pv. coronafaciens resulted in disease broccoli raab and broccoli. This new It is a common leaf spot pathogen of cruci- on some but not all of the known hosts for pathogen is referred to as P. syringae pv. fers and is the causal agent of bacterial leaf this pathovar. It caused disease when alisalensis in the rest of this manuscript. P. spot on broccoli and cauliflower in the sprayed on oat, rye, and brome, but not on syringae pv. alisalensis infected all cruci- Salinas Valley (40). We correctly hypothe- common timothy or corn. Corn and com- fers tested to date (21) including broc- sized that bacterial leaf spot on broccolini mon timothy are natural hosts for some but colini. Its ability to cause disease on cruci- was caused by P. syringae pv. maculicola. not all strains of P. syringae pv. coronafa- fers suggested that it was similar to other Results from phenotypic, genetic, and host ciens (3,32). However, the P. syringae pv. coronatine producers such as P. syringae range tests were identical for the strains coronafaciens strains tested did not cause pv. maculicola and P. syringae pv. tomato. causing bacterial leaf spot on broccolini disease on crucifers, which distinguishes However, neither of these pathogens and reference P. syringae pv. maculicola them from the P. syringae pv. alisalensis caused disease on broccoli raab. Interest- strains. The recent emergence of this strains. In addition, the presence of coro- ingly, P. syringae pv. alisalensis infected pathogen on broccolini may be because natine production genes in the P. syringae grasses including California brome, oat, this is a relatively new crucifer commer- pv. alisalensis strains also distinguishes and common timothy, as well as cruci- cialized in 1998 (H. Bouzar, personal them from the P. syringae pv. coronafa- fers. P. cichorri is the only other phytopa- communication). ciens strains, in which the coronatine pro- thogenic pseudomonad known to cause P. syringae pv. maculicola and P. syrin- duction genes were not present. disease on both crucifers and grasses. It gae pv. tomato are closely related (12) and Analysis of distances between repetitive was isolated from crucifers but caused some authors have suggested that these extragenic sequences using rep-PCR proto- disease on oat when artificially inocu- names are synonyms of one pathovar cols is a powerful tool for distinguishing lated (3). Here we described the first (13,36). There may be overlap in the host related bacterial taxa. It has been used to pathovar of P. syringae with a host range range of these pathogens. For example, successfully distinguish and identify bacte- that included both crucifers and mono- strains of P. syringae pv. tomato caused rial plant pathogens both in situ and in vivo cots. This unique host range was suffi- disease on cauliflower in some (13,27,36) (31). Specifically, BOX-PCR (rep-PCR cient to designate this organism as a new but not all studies (40). Our host range data using the BOX repetitive extragenic se- pathovar (9); however, additional genetic supported the hypothesis that these are two quences) is able to distinguish pathovars of (BOX-PCR), phenotypic (ice nucleation separate but closely related pathovars P. syringae (14,19,28,39,42). BOX-PCR and phage sensitivity), and biochemical (27,40) because a clear difference in host banding patterns can be unique fingerprints (carbon source utilization) data further range was obtained for reference strains of that correspond to a single pathovar. Alter- supported this designation. P. syringae pv. maculicola including the natively, several banding patterns may be Taxonomy and subsequent nomenclature new strains from broccolini and P. syringae generated by different strains of a single of P. syringae and related pathogens has pv. tomato. P. syringae pv. tomato did not pathovar (26,42). For example, Zhao et al. been in flux since the 1980 publication of cause disease on broccoli, broccolini, or (42) used BOX-PCR to help assign the the Approved Lists of Bacterial Names cauliflower while all P. syringae pv. macu- causal agent of bacterial spot on spinach (35) when all plant pathogens that resem- licola strains caused disease. However, we (Spinacia oleracea), mustard (Brassica bled P. syringae were lumped into this used only a few well characterized strains juncea), turnip (Brassica napus), and kale species regardless of host range. Plant in our studies. More variability might be (Brassica oleracea var. acephala) to P. pathologists used the infrasubspecific epi- seen in analysis of a greater number of syringae pv. maculicola. This pathovar had thet pathovar to differentiate among plant strains. In contrast, reactions varied among two possible distinct profiles, A or B (42). pathogens with distinct host ranges. Re- reference strains of P. syringae pv. maculi- This indicates that P. syringae pv. maculi- search is currently being conducted to cola. Some caused severe necrotic flecking cola may have converged from at least two differentiate and designate independent and chlorosis (BS 119, 0788-18) on tomato evolutionary lines. In this research, three species from pathovars of P. syringae. while others (BS121, 0649-18) resulted in strains used as positive controls for P. Distinct genomospecies can be distin- both asymptomatic and light flecking reac- syringae pv. maculicola and five formerly guished using DNA-DNA hybridization tions. uncharacterized P. syringae pv. maculicola techniques but these groups do not always We have also seen differences in the re- strains from broccolini had identical band- coincide with species based on phenotypic action of P. syringae pv. alisalensis on ing patterns. The profiles from group B tests (12). Because genetic data and pheno- tomato. In the experiments described here, (42) appeared similar to the profile of the typic data do not yet concur, we followed which were conducted at two locations strains from broccolini. The broccolini the current convention and proposed that (USDA and UCCE), P. syringae pv. strains were isolated in the Salinas Valley. the new pathogen of broccoli raab is a alisalensis was pathogenic on tomato. This The limited number of P. syringae pv. pathovar of P. syringae. However, as soon result was inconsistent with data previ- maculicola strains and the limited geo- as phenotypic characteristics can be corre- ously published from work done at the graphic distribution of pathogen origin lated to the genetic data, this pathogen UCCE location (21). In the studies re- probably resulted in our observation of should be characterized and placed in its ported in this paper, the inoculum concen- only one group. However, it would be appropriate genomospecies. P. syringae pv. tration was carefully adjusted for each interesting to know if group A occurs in the alisalensis may be a member of geno- experiment, and bacteria were applied as Salinas Valley. mospecies 3 because it phenotypically water suspensions. For the previous ex- In addition to observing a unique finger- resembled the genomotype (P. syringae pv. periments, overnight nutrient broth cultures print for the P. syringae pv. maculicola tomato) and an additional member of this were used as inoculum without adjusting strains tested, unique fingerprints were genomospecies (P. syringae pv. maculi- the population density (21). Because of the obtained for P. syringae pv. tomato and P. cola; 12). variability in pathogenicity tests on tomato, syringae pv. coronafaciens. Although the

996 Plant Disease / Vol. 86 No. 9 P. syringae pv. tomato fingerprint was had one to five polar flagella (21). Strains 7. Cintas, N. A., Koike, S. T., and Bull, C. T. distinct, one strain had a slight polymor- glistened on MGY plates and fluoresced 2000. Rappini bacterial blight declines with delayed replanting in the Salinas Valley of phism. In the research by Zhao et al. (42), with a bluish green pigment on KMB. All California. Phytopathology 90:S15. two P. syringae pv. tomato strains, includ- strains were gram negative. In LOPAT 8. Cupples, D. A., Moore, R. A., Morris, V. L. ing the well characterized strain DC3000, tests, strains were positive for levan pro- 1990. Construction and use of a nonradioac- had banding patterns corresponding to duction and tobacco hypersensitivity, but tive DNA hybridization probe for detection of group A of P. syringae pv. maculicola. negative for oxidase, arginine dihydrolase, Pseudomonas syringae pv. tomato on tomato However, the banding patterns of three and ability to rot potato, identifying the plants. Appl. Environ. Microbiol. 56:1743- 1749. other strains of P. syringae pv. tomato organism as P. syringae. Strains did not 9. Dye, D. W., Bradbury, J. F., Goto, M., Hay- were unique and appear to be similar to reduce nitrate nor liquefy gelatine and ward, A. C., Lelliot, R. A., Schroth, M. N. those presented in this study. were variable for urease activity (21). 1980. International standards for naming BOX-PCR successfully distinguished P. Mannitol, inositol, sorbitol, glycerol, glu- pathovars of phytopathogenic bacteria and a list of pathovar names and pathotype strains. syringae pv. alisalensis from P. syringae cose, sucrose, arabinose, mannose, D (–), Rev. Plant Pathol. 59:153-168. pv. maculicola, P. syringae pv. tomato, and tartrate, citrate, succinate, malonate, L (–), 10. Fahy, P. C., and Hayward, A. C. 1983. Media P. syringae pv. coronafaciens. These data L (+), and DL-tartrate, and DL-lactate were and methods for isolation and diagnostic supported the hypothesis that the P. syrin- utilized as carbon sources. Strains were not tests. Pages 337-378 in: Plant Bacterial Dis- gae pv. alisalensis is a new pathovar of P. able to utilize erythritol, trigonelline, D (+) eases. P. C. Fahy, and G. J. Persley, eds. Aca- syringae. The ability of rep-PCR to distin- tartrate, meso-tartrate, molybdate, acetate, demic Press, NY. 11. Fanizza, G., Damato, G., and Bianco, V. V. guish this pathovar from other P. syringae propionate, maleate, rhamnose, and L (+) 1992. A multivariate morphometric study on pathovars may be useful for rapid identifi- lactate. Utilization of L (–) lactate and landrace differentiation in broccoli raab cation from field infections. growth on MacConkey agar were weakly (Brassica rapa L.). Acta Hortic. 318:195-200. To date, P. syringae pv. maculicola was positive (21). The strains produced ice 12. Gardan, L., Shafik, H., Belouin, S., Broch, R. found on broccolini from only one com- nuclei and were sensitive to phage PBS1. Grimont, F., and Grimont, P. A. D. 1999. DNA relatedness among the pathovars of Pseudo- mercial transplant greenhouse. This disease All of the strains had DNA to which the monas syringae and description of Pseudo- has not been found on broccolini in the coronatine production genes from DC3000 monas tremae sp. nov. and Pseudomonas field. However, P. syringae pv. maculicola hybridized. None of the strains induced cannabina sp. nov. (ex Sutic and Dowson is a pathogen of many crops grown in rota- pitting on polypectate gel (21). Hosts were 1959). Int. J. Syst. Bacteriol. 49:469-478. tion with broccolini in the Salinas Valley the following: Natural hosts–broccoli raab 13. Hendson, M., Hildebrand, D. C., and Schroth, including cauliflower and broccoli (40). P. and broccoli; Artificial inoculation– N. 1992. Relatedness of Pseudomonas syrin- gae pv. tomato, Pseudomonas syringae pv. syringae pv. alisalensis has a broader host broccolini, cauliflower, arugula (Eruca maculicola, and Pseudomonas syringae pv. range infecting many crucifers as well as sativa), bok choy (Brassica campestris antirrhini. J. Appl. Bacteriol. 73:455-464. grasses. Since 1995, we have found this subsp. chinensis), cabbage (Brassica ol- 14. Hollaway, G. J., Bretag, T. W. 1997. Survival pathogen each year in broccoli raab plant- eracea subsp. capitata), Chinese cabbage of Pseudomonas syringae pv. pisi in soil and ings. This pathogen was also found in the (Brassica campestris subsp. pekinensis), on pea trash and their importance as a source of inoculum for a following field pea crop. same commercial broccoli field in two Japanese mustard (Brassica campestris Aust. J. Exp. Agric. 37:369-375. consecutive years. In the Salinas Valley, subsp. nipposinica), red mustard (Brassica 15. Keane, P. J., Kerr, A., and New, P. B. 1970. crucifers can be planted back-to-back with juncea subsp. rugosa), and tah tsai (Bras- Crown gall of stone fruit. II. Identification as little as a week between incorporation of sica campestris subsp. nari-nosa) tomato, and nomenclature of Agrobacterium isolates. previous crop and planting of the subse- California brome, oat, and common timo- Aust. J. Biol. Sci. 23:585-595. 16. King, E. O., Ward, M. K., and Raney, D. E. quent crop. Residue from infected crops is thy. Corn and rye were not hosts. 1954. Two simple media for the demonstra- a significant inoculum source for up to 4 ACKNOWLEDGMENTS tion of pyocyanin and fluorescin. J. Lab. Clin. Med. 44:301-307. months after incorporation (7). Rotation Special thanks to Diana Henderson, Thi Duong, with crops, for which soils are fumigated E. J. Rowland, and Rose Vega for technical sup- 17. Klement, Z., Rudolf, K., Sands, D. C. 1990. annually, is currently used to control soil- port on this project; R. Bunch, H. Bouzar, and H. Methods in Phytobacteriology. Akadémiai borne diseases in vegetable production Azad for strains, seed, and discussion; and E. Kiadó, Budapest. 18. Koike, S. T. 1997. Broccoli raab as a host of fields. The fumigant methyl bromide will Little and J. Barak who reviewed the manuscript prior to submission. Alternaria brassicae in California. Plant Dis. no longer be available after 2005, and dis- 81:552. eases like the ones described here may LITERATURE CITEDÃ 19. Koike, S. T., Barak, J. D., Henderson, D. M., become more important if fumigation is no 1. Anonymous 2002. Vegetables and Melons and Gilbertson, R. L. 1999. Bacterial blight of longer used. Considering the tight cropping Outlook. USDA, Economic Research Service, leek: a new disease in California caused by Pseudomonas syringae. Plant Dis. 83:165- sequence and overlap in the host range of Bull. VGS-289. 2. Ausubel, F. M., Brent, R., Kingston, R. E., 170. the two pathogens characterized here, tools Moore, D. D., Seidman, J. G., Smith, J. A., 20. Koike, S. T., Cintas, N. A., and Bull, C. T. for rapid identification would be useful in Struhl, K. 1987. Current Protocols in Molecu- 2000. Bacterial blight, a new disease of broc- making recommendations for rotation lar Biology. Wiley-Interscience, New York. coli, caused by Pseudomonas syringae, in crops. For example, if a bacterial disease 3. Bradbury, J. F. 1986. Guide to Plant Patho- California. Plant Dis. 84:370. 21. Koike, S. T., Henderson, D. M. Azad, H. R., occurs on broccoli, rapid identification of genic Bacteria. CAB International, Slough, UK. Cooksey, D. A., and Little, E. L. 1998. Bacte- P. syringae pv. maculicola would suggest 4. Canteros de Echenique, B. I., Zagory, D., rial blight of broccoli raab: a new disease that a grower could grow broccoli raab or a Stall, R. E. 1985. A medium for cultivation of caused by a pathovar of Pseudomonas syrin- cover crop with grasses without risk of the B-strain of Xanthomonas campestris pv. gae. Plant Dis. 82:727-731. disease developing from the same patho- citri, cause of cancrosis B in Argentina and 22. Koike, S. T., and Saenz, G. S. 1997. First gen. In contrast, this would not be true if Uruguay. Plant Dis. 69:122-123. report of powdery mildew, caused by Erysi- 5. Cintas, N. A., Bull, C. T., Koike, S. T., and phe cruciferarum, on broccoli raab in Cali- the pathogen was identified as P. syringae Bouzar, H. 2001. A new bacterial leaf spot fornia. Plant Dis. 81:1093. pv. alisalensis. This research provided the disease of broccolini, caused by Pseudomonas 23. Lauritzen, E. 2000. Monterey County Agri- initial information needed to develop rapid syringae pathovar maculicola, in California. cultural Commissioner 1998 Crop Report, identification methods using BOX-PCR Plant Dis. 85:1207. CA. and significant host range data for two 6. Cintas, N. A., Koike, S. T., Bull, C. T. 2001. 24. Lelliott, R. A., Billing, E., and Hayward, A. C. Emerging bacterial pathogens of crucifers in 1996. A determinative scheme for the fluores- pathogens found in this region. the Salinas Valley of California, Pages 272- cent plant pathogenic Pseudomonads. J. Appl. Description of P. syringae pv. alisalen- 274 in: Plant Pathogenic Bacteria, S. H. De- Bacteriol. 29:470-489. sis pv. nov. The bacterium formed colonies Boer, ed. Kluwer Academic Publishers, Bos- 25. LeStrange, M., Mayberry, K. S., Koike, S. T., that were circular, entire, and convex and ton, MA. and Valencia, J. 1996. Broccoli production in

Plant Disease / September 2002 997 California. University of California. Division Bruijn, F. J. 1998. Characterization of the di- 37. Verhoeven D. T., Verhagen, H., Goldbohm, R. of Agriculture and Natural Resources. Publi- versity of ecologically important microbes by A., van den Brandt, P. A., van Poppel, G. cation 7211. rep-PCR genomic fingerprinting. Mol. Mi- 1997. A review of mechanisms underlying an- 26. Little, E. L., Bostock, R. M., Kirkpatrick, B. crobiol. Ecol. Man. 3.4.3:1-27. ticarcinogenicity by brassica vegetables. C. 1998. Genetic characterization of Pseudo- 32. Ribeiro, R. de L. D., Durbin, R. D., Arny, D. Chem. Biol. Interact. 103:79-129. monas syringae pv. syringae strains from C., and Uchytil, T. F. 1977. Characterization 38. Versalovic, J., Schneider, M., De Bruijn, F. J., stone fruits in California. Appl. Environ. Mi- of the bacterium inciting chocolate spot of and Lupski, J. R. 1994. Genomic fingerprint- crobiol. 64: 3818-3823. corn. Phytopathology 67:1427-1431. ing of bacteria using repetitive sequence- 27. Little, E. L., Gilbertson, R. L. 1997. Pheno- 33. Sambrook, J., Fritsch, E. F., and Maniatis, T. based polymerase chain reaction. Methods typic and genotypic characters support place- 1989. Molecular cloning: a laboratory man- Mol. Cell. Biol. 5:25-40. ment of Pseudomonas syringae strains from ual, 2nd ed. Cold Spring Harbor Laboratory 39. Weingart, H., and Völksch, B. 1997. Genetic tomato, celery, and cauliflower into distinct Press, Cold Spring Harbor, N.Y. fingerprinting of Pseudomonas syringae pathovars. Pages 542-547 in: Pseudomonas 34. Schaad, N. W., 1988. Laboratory Guide for pathovars using ERIC-, REP-, and IS50-PCR. syringae Pathovars and Related Pathogens. K. Identification of Plant Pathogenic Bacteria. J. Phytopathol. 145:339-345. Rudolf, T. J.Burr, J. W.Mansfield, D. Stead, A. 2nd ed. American Phytopathological Society, 40. Wiebe, W. L., and Campbell, R. N. 1993. Vivian, J. von Kietzell, eds. Kluwer Academic St. Paul, MN. Characterization of Pseudomonas syringae Publishers, Boston, MA. 35. Skerman, V. B. D., McGowan, V., and Sneath, pv. maculicola and comparison with P. s . t o - 28. Louws, F. J., Fulbright, D. W., Stephens, C. T., P. H. A., eds. 1989. Approved lists of bacterial mato. Plant Dis. 77:414-419. and De Bruijn, F. J. 1994. Specific genomic names. American Society for Microbiology, 41. Xu, G.-W., and Gross, D. C. 1988. Evaluation fingerprints of phytopathogenic Xanthomonas Washington, DC. of the role of syringomycin in plant patho- and Pseudomonas pathovars and strains gener- 36. Takikawa, Y., Nishiyama, N., Ohba, K., Tsu- genesis by using Tn5 mutants of Pseudomo- ated with repetitive sequences and PCR. Appl. yumu, S., and Goto, M. 1994. Synonymy of nas syringae pv. syringae defective in syrin- Environ. Microbiol. 60:2286-2295. Pseudomonas syringae pv. maculicola and gomycin production. Appl. Environ. 29. McCulloch, L. 1911. A spot disease of cauli- Pseudomonas syringae pv. tomato. Pages Microbiol. 54:1345-1353. flower, USDA, Bur. Plant Ind. Bull. 225:1-15. 199-204 in: Plant Pathogenic Bacteria, 8th In- 42. Zhao, Y. F., Damicone, J. P., Demezas, D. H., 30. Paisley, R. 1995. Page M-5 in: MIS Whole ternational Conference on Plant Pathogenic Rangaswamy, V., Bender, C. L. 2000. Bacte- Cell Fatty Acid Analysis by Gas Chromatog- Bacteria M. Lemattre, S. Freigoun, K. Ru- rial leaf spot of leafy crucifers in Oklahoma raphy. MIDI, Inc., Newark, DE. dolph, J. G. Swings eds. Les Colloques Edi- caused by Pseudomonas syringae pv. maculi- 31. Rademaker, J. L. W., Louws, F. J., and De tions, INRA, Paris, France. cola. Plant Dis. 84:1015-1020.

998 Plant Disease / Vol. 86 No. 9