INTERNATIONALJOURNAL OF SYSTEMATICBACTERIOLOGY, Oct. 1988, p. 362-366 Vol. 38, No. 4 0020-7713/88/O40362-05$02.OO/O Copyright 0 1988, International Union of Microbiological Societies

Characterization of campestris pv. pennamericanum pv. nov., Causal Agent of Bacterial Leaf Streak of Pearl MilletT M. QHOBELAS AND L. E. CLAFLIN" Department of Pathology, Kansas State University, Manhattan, Kansas 66506

A survey for bacterial diseases of millet and sorghum was conducted in northern Nigeria during the 1984 growing season. Bacterial diseases were prevalent throughout the area surveyed. Yellow, mucoid bacterial colonies were consistently isolated from pearl millet leaves exhibiting bacterial streak symptoms. The causative organism was characterized as a pathovar of Xanthomonas campestris. The strains isolated formed a homogeneous group of aerobic, motile, gram-negative, rod-shaped organisms which were distinctly different pathologically, serologically, and by membrane protein patterns from other pathovars of X. campestris. The cells measured 0.45 by 2.25 pm, and each cell had one polar flagellum. Optimal growth occurred between 26 and 30°C, and the maximum NaCl tolerance was 3%. Pearl millet (Pennisetum americanum) and Proso millet (Panicurn miliaceum) are the only known hosts. For this pathovar, we propose the name Xanthomonas campestris pv. pennamericanum pv. nov. (the name is derived from Pennisetum americanum, the Latin binomial for pearl millet). The holopathotype strain is strain B6-P, which has been deposited in the American Type Culture Collection, Rockville, Md.

Bacterial diseases of pearl millet [Pennisetum america- B (13), and Xanthomonas phaseoli medium (4). The plates num (L.) Leeke] are generally assumed to be of limited were incubated for 4 to 6 days at 28°C and were examined for economic importance. This assumption is perhaps one of the bacterial colonies. Single colonies of the suspected patho- reasons that bacterial diseases of pearl millet have received gens were transferred to fresh YDCA plates. Verification of limited attention. Bacterial pathogens that have been re- pathogenicity was conducted by inoculating 3- to 4-week-old ported to incite diseases of pearl millet include Pseudomo- pearl millet seedlings (cultivar Serere 3A) grown under nus syringae pv. syringae (17), Pseudomonas avenue (L. E. greenhouse conditions (28°C). The results were recorded 2 Claflin, B. A. Ramundo, J. E. Leach, and I. D. Erinle, weeks after inoculation. The preparation of the inoculum Phytopathology 77:1766, 1987), Xanthomonas annama- and the inoculation method used are described below. The laiensis (20), Xanthomonas rubrisorghi (19), and Xantho- pathogenic isolates were lyophilized for long-term storage, monas penniseti (18); the last three species were subse- and working cultures were maintained on YDCA plates at quently determined to be Erwinia herbicola (7). A high 4°C. percentage of pearl millet observed in a survey Biochemical and physiological characterization. Two iso- conducted in northern Nigeria during the 1984 growing lates of the suspected pathogen were compared with X. season exhibited symptoms similar to those of bacterial leaf campestris pv. graminis (two strains), X. campestris pv. stripe and streak of sorghum (9). Two phytopathogenic holcicola (seven strains) , X. campestris pv. translucens were commonly isolated from leaf samples. One (seven strains), and X. campestris pv. vasculorum (seven isolate was determined to be Pseudomonas avenue, the strains) by using the following tests: Gram reaction (22);acid incitant of bacterial leaf stripe of pearl millet (Claflin et al., production from carbohydrates, using Hugh-Leifson me- Phytopathology 77:1766, 1987). The other was determined to dium (11); occurrence of a yellow water-insoluble pigment be an unknown pathovar of Xunthomonas campestris; this on YDCA (23); casein, gelatin, and Tween 80 hydrolysis (5); organism was determined to be the causal agent of bacterial growth on X. phaseoli medium (4); hydrogen sulfide produc- leaf streak of pearl millet. Chqracterization of the pearl millet tion (6); maximum sodium chloride tolerance (5); minimum bacterial leaf streak agent, X. campestris pv. pennamerica- and maximum growth temperatures (5); malonate and citrate num pv. nov., is reported here. utilization (8); arginine dihydrolase production (24); Kovacs oxidase test (14); and nitrate reduction (8). MATERIALS AND METHODS Dot-immunobinding assay. For the dot-immunobinding as- Isolation of bacterial strains. Symptomatic leaves were say (DIA), the method of Leach et al. (16) was used to rinsed three times in sterile distilled water (20 s per rinse). produce antisera and to assay for serological relationships Sections of leaf tissue (0.5 by 2 cm) were triturated in 4 ml of between the pearl millet pathogen and other X. campestris sterile 12.5 mM PO4 buffer (pH 7.2) by using a sterile mortar pathovars that infect cereal crops. Antiserum was produced and pestle. The resulting suspension was filtered through against the following pathogens: X. campestris pv. penna- two layers of cheesecloth and serially diluted with sterilized mericanum pv. nov. strain B6-PT (T = type strain), X. PO4 buffer. The diluted suspensions were plated onto yeast campestris pv. translucens XT-116, X. campestris pv. hol- dextrose cakium carbonate agar (YDCA) (22),King medium cicoia ATCC 13461, and X. campestris pv. vascuiorum NCPPB 1326. Antisera were cross-absorbed with Pseudo- monas syringae syringae Pseudomonas ave- * Corresponding author. pv. B359 and ? Contribution no. 88-244-5 from the Department of Plant Pathol- nue PA 134 to eliminate cross-reactivity with other bacterial ogy, Kansas Agricultural Experiment Station, Kansas State Univer- pathogens of pearl millet. sity, Manhattan. Pure cultures of the following bacterial pathogens were $ Present address: Agricultural Research Station, Maseru 100, used as antigens in the DIA procedure: X. campestris pv. Kingdom of Lesotho. pennamericanum pv. nov., X. campestris pv. graminis, X.

362 VOL. 38, 1988 XANTHOMONAS CAMPESTRIS PV. PENNAMERICANUM PV. NOV. 363

TABLE 1. Bacterial strains used for the identification of the pearl millet pathogen

Organism Strain" Host Origin X. albilineans NCPPB 2939= Saccharum oficinarum Fiji X. campestris pv. graminis PDDCC 3473 Lolium tnultiJlorum Switzerland PDDCC 5733T Da c ty lis g 1om e ra ta Switzerland X. campestris pv. holcicola KS 66 Sorghum bicolor Kansas KS 86 Sorghum bicolor Kansas TX- 1 Sorghum bicolor Texas LES 107 Sorghum bicolor Lesotho LES 124 Sorghum bicolor Lesotho NCPPB 1241 Sorghum bicolor Australia PDDCC 3103* Sorghum bicolor New Zealand X. campestris pv. oryzae PDDCC 3128 Oryza sativa India X. campestris pv. oryzicola PDDCC 5743T Oryza sativa Malaysia X. campestris pv. translucens ATCC 10731 Hordeum vulgare Unknown ATCC 9000 Triticum aestivum Canada ATCC 10771 Triticum aestivum Unknown XT-103 Secale cereale Georgia XT-115 Secule cereale Georgia XT-116 Triticum aestivum South Dakota PDDCC 5752T Hordeum vulgare United States X. campestris pv. vasculorum NCPPB 206 Zea mays Republic of South Africa NCPPB 1326 Saccharum oficinarum Zimbabwe PDDCC 253 Saccharum oficinarum India PDDCC 302 Roystonea regia Mauritius PDDCC 304 Saccharum oficinarum Australia PDDCC 327 Thysanolaena maxima Reunion PDDCC 5757T Saccharum oficinarum Mauritius Erwinia herbicola 112Y Unknown United Kingdom

~ a Strains were obtained from the following sources: American Type Culture Collection, Rockville, Md. (ATCC); Plant Disease Division Culture Collection, Auckland, New Zealand (PDDCC); National Collection of Plant Pathogenic Bacteria, Hertfordshire, United Kingdom (NCPPB); Department of Plant Pathology, Kansas State University, Manhattan (strains KS 66, KS 86, TX-1, LES 107, LES 124, and 112Y); and B. M. Cunfer, Georgia Experiment Station, Experiment (strains XT-103, XT-115, and XT-116). campestris pv. holcicoia, X. campestris pv. ovyzae, X. distilled water (pH 6.75) was added. The samples were campestris pv. translucens, and X. campestris pv. vasculo- boiled for 2 min, allowed to cool, and stored at -20°C. rum. All of the antigen strains listed in Table 1 were checked The buffer system used for gel electrophoresis was the for reactivity with the homologous and heterologous anti- discontinuous sodium dodecyl sulfate system of Laemmli sera. (15). The separating gel contained acrylamide at a final Resolution of bacterial membrane proteins. A method concentration of 12.5% and was prepared from a stock adapted from the method of Ames (2) was used to resolve solution containing 30 g of acrylamide and 0.8 g of N,N'- and differentiate membrane proteins of the phytopathogenic methylene bisacrylamide in 100 ml of distilled water. The bacteria listed in Table 1. Cells were grown with vigorous stacking gel contained 5% acrylamide and was prepared agitation in 0.8% nutrient broth (Difco Laboratories) con- taining 0.5% (wthol) NaCl. A 200-ml portion of a culture in log phase was centrifuged for 20 min at 3,500 X g by using a TABLE 2. Plants used to determine the host range of the pearl Beckman model J-21C centrifuge equipped with a model millet pathogen JA-14 fixed-angle rotor. The bacterial pellet was suspended Scientific name Common name Variety Or Donorb in 10 ml of 10 mM tris(hydroxymethy1)aminomethane hydro- PI" no. chloride-30 mM NaCl (pH 7.2) and sonicated (100 W; Braun Pennisetum americanum Pearl millet Serere 3A 3 Sonic model 1510 sonicator) for 2 min in 20-s intervals with Panicum miliaceum Proso millet 463435 1 40 s of cooling between intervals. The crude sonic extract Zea mays Maize Gold cup 3 was centrifuged for 20 min at 6,000 x g by using a Beckman Sorghum bicolor Sorghum 80B 4 model J-21C centrifuge and a model JA-20 fixed-angle rotor. Saccharum oficinurum Sugarcane 468 5 The supernatant was removed without disturbing the loose Triticum aestivum Wheat Bounty 3 pellet floating above the tightly packed pellet at the bottom Hordeum vulgare Barley Dickson 3 of the tube and then centrifuged at 111,000 x g for 35 min Avenu sutiva Oats Larry 3 with a Beckman model L8-70 ultracentrifuge equipped with Eleusine coracana Finger millet 462979 2 a type 70.1 Ti fixed-angle rotor. The pellet from this centrif- Echinochlou crusgalli-edumlis Japanese millet 274913 1 Brachiaria texanum Texas millet 476984 2 ugation was the membrane fraction (2), and this fraction was Eleusine indica Goosegrass 315700 2 suspended in 0.5 ml of 10 mM tris(hydroxymethy1)aminome- thane hydrochloride-30 mM NaCl (pH 7.2). An equal vol- " PI, Plant introduction. ume of 2~ sample buffer consisting of 1.51 g of tris(hydroxy- ' Plants were obtained from the following sources: Regional Plant Intro- duction Stations at Ames, Iowa (l), and Experiment, Ga. (2); Department of methy1)aminomethane hydrochloride , 20 ml of glycerol, 4.0 Agronomy, Kansas State University, Manhattan (3); D. T. Rosenow, Texas g of sodium dodecyl sulfate, 10 ml of 2-mercaptoethanol, and Agricultural Experiment Station, Lubbock (4); and J. L. Dean, Agricultural 20 mg of bromophenol blue in a final volume of 100 ml of Research and Education Center, Belle Glade, Fla. (5). 364 QHOBELA AND CLAFLIN INT. J. SYST. BACTERIOL.

TABLE 3. Biochemical and physiological characteristics of the pearl millet pathogen and four pathovars of X. campestris Pearl millet X. campestris X. campestris X. campestris X. campestris Characteristic pathogena pv. graminis pv. holcicola pv. translucens pv. vasculorum Gram reaction Growth on X. phaseoli medium Yellow water-insoluble pigment on YDCA Starch hydrolysis Casein hydrolysis Tween 80 hydrolysis Gelatin liquification Arginine dihydrolase Oxidase reaction - Nitrate reduction Citrate utilization V Malonate utilization V Growth at 4°C Growth at 41°C - - V Maximum NaCl tolerance (%) 3 3 3-7 Acid production from: - Adonitol - - L-(+)-Arabinose + V V D-( +)-Cellobiose + V V Dulcitol + V V D-Fructose + V V D-Galactose + V + glucose + + + Glycerol - V V a-Lactose + - V Maltose - - - D-(+)-Mannose + V + D-Mannitol - - D-( +)-Melibiose c my o-Inositol Raffinose a-L-Rhamnose Salicin D-Sorbito1 - - - Sucrose + + + + + D-Xylose + + + + ~ ~ ~~-~- a See text for the number of strains used. ' +, Positive reaction; -, negative reaction; V, reaction variable between strains.

from the same stock solution. The final concentration of sodium dodecyl sulfate was 0.1% in both gels and in the electrode buffer. Samples (10 to 20 1.11) were boiled for 2 min prior to loading on the gel. The gels were run at a constant current of 30 mA for 6 h and were stained overnight with Coomassie blue (0.125% Coomassie blue R-250, 50% meth- anol, 10% acetic acid). The gels were destained in methanol- acetic acid-distilled water (5:1:4, vol/vol) on a tabletop shaker for 6 to 10 h with several changes of destaining solution. Plant inoculation procedure. Seeds of the plants listed in Table 2 were planted in a mixture of soil, sand, and peat (1: 1:l).Seedlings (2 to 3 weeks old) were inoculated by using a Hagborg device (10) with strains of X. campestris pv. pennamericanum pv. nov., X. campestris pv. graminis, X. campestris pv. holcicola, X. campestris pv. translucens, and X. campestris pv. vasculorum (Table 1) to determine the host range of the pearl millet pathogen. All inoculations were performed under greenhouse conditions (28°C). Pathogenic- ity results were determined 2 weeks after inoculation. The 1. Membrane protein patterns Of x. campesfris Pv. culorum (lanes A and B), X. campestris pv. holcicola (lanes C and inoculum was prepared by growing to log phase in lom1 D), pearl millet pathogen (lane E), X. campestris pv. translucens of nutrient broth with vigoro~sagitation at 28°C- A cell (lanes F and G), X. campestris pv. graminis (lane H), X. albilineans suspension was centrifuged at 1,200 g for 20 min, and the (lane I), X. campestris pv. ovyzae (lane J), Erwinia herhicola (lane Pellet was suspended in Po, buffer. Cell Concentrations were K), and SDS-7 Dalton mark VII-L molecular weight marker (lane L) adjusted to lo8 colony-forming units per ml with PO, buffer in a 12.5% polyacrylamide gel slab. K Da, Kilodaltons. VOL. 38, 1988 XANTHOMONAS CAMPESTRIS PV. PENNAMERICANUM PV. NOV. 365

TABLE 4. Reactions of bacterial strains with antisera prepared against four X. campestris pathovars in the DIA

Reactions with the following antisera: Antigen Pearl millet X. cumpestris X. campestris X. compestris pathogen pv. translucens pv. holcicola pv. vasculorum Pearl millet pathogen + (106))" + (108) - - X. campestris pv. holcicola - - + (106) + (109 X. campestris pv. translucens + (108) + (lo6) - - X. campestris pv. vasculorum - - + (106) + (lo6) X. campestris pv. graminis + (108) - - - Control (12.5 mM phosphate buffer) - - - - +, Positive reaction in the DIA; -, no reaction in the DIA. The numbers in parentheses are the minimal numbers of colony-forming units of the pathogen per ml detectable by the DIA.

by using a Klett-Summerson photoelectric colorimeter (red Increasing antiserum dilutions above 1:20,000 resulted in a filter) prior to inoculation. loss of sensitivity without any gain in specificity. Antiserum developed for the pearl millet pathogen reacted only with its RESULTS AND DISCUSSION homologous antigen at all of the optimal dilutions, indicating the serological identity of this pathogen. However, cross- Morphological, biochemical, and physiological characteris- reactivity was evident at 10' colony-forming units per ml tics of the bacterial leaf streak pathogen. Cells of the bacterial with X. campestris pv. graminis and X. campestris pv. leaf streak pathogen were gram-negative, motile rods with translucens antigens. A summary of the results of all anti- single polar flagella. The cells measured approximately 0.45 gen-antiserum reactions is shown in Table 4. by 2.25 km. Growth on YDCA was moderate, with full Sodium dodecyl sulfate-polyacrylamide gel electrophoresis colony development requiring 3 to 4 days. Colonies were of membrane proteins. The use of polyacrylamide gel elec- yellow, circular, convex, and mucoid with entire margins. trophoresis has played a significant role in the of Strains grew at 36°C but not at 4 or 41°C. The biochemical phytopathogenic bacteria. Carlson and Vidaver (3) revised and physiological characteristics of the bacterial leaf streak Corynebacterium pathogen were very similar to those of other X. campestris the taxonomy of plant pathogens by using pathovars (Table 3), confirming our assumption that the sodium dodecyl sulfate-polyacrylamide gel electrophoresis pearl millet pathogen is a pathovar of X. campestris. Our of cellular proteins, while Kersters and De Ley (12) used a results further emphasize the inadequacy of these techniques similar method for differentiating among several gram-nega- to differentiate pathovars of X. campestris, as reported tive phytopathogenic bacteria. The membrane protein pat- previously (5). terns of the phytopathogenic bacteria examined differed DIA. Serological techniques, such as immunofluorescent distinctly from each other. The use of a 12.5% polyacryl- staining (21) and the enzyme-linked immunosorbent assay amide slab gel enabled good resolution of protein bands at (l),are routinely used in the detection of phytopathogenic molecular weights ranging from 14,000 to 66,000. The pro- xanthomonads. The DIA technique was utilized to differen- tein patterns were consistent and reproducible for all patho- tiate between pathovars of X. campestris. The success of the vars except X. campestris pv. vasculorum, which showed a technique was attributed primarily to the high antiserum wide range of heterogeneity. All of the strains listed in Table specificity and titer, as all of our antisera had agglutination 1 were analyzed by this method. Different strains of X. titers exceeding 1: 1,280. The optimal homologous antigen- campestris pv. holcicola and some strains of X. campestris antiserum reactions in the DIA occurred when the antisera pv. vasculorum revealed nearly identical protein patterns. were diluted 1:20,000, coupled with antigen dilutions of lo8, These patterns differed distinctly from the patterns for other lo', and lo6 colony-forming units per ml. At antiserum pathovars by having a major protein band at a molecular dilutions of 1: 10,000 or less, cross-reactivity was observed weight of about 29,000 (Fig. 1). The pearl millet pathogen between all pathovars of X. campestris used in the DIA. (Fig. 1, lane E) also had a clearly different membrane protein

TABLE 5. Pathogenicity test results for the pearl millet pathogen and four pathovars of X. campestris

~~~~ ~ - Pearl millet X. campestris X. campestris X. campestris X. campestris Host pat hogen pv. graminis pv. holicolo pv. translucens pv. vnsculorum Pearl millet +" - - - - Proso millet + - - - - Maize - - + - + Sorghum - - + - - Sugarcane - - - - + Wheat - - ND + ND Barley - - ND + ND Oats - - ND + ND Finger millet - - - - - Japanese millet - - - - - Texas millet - - - - - Goosegrass - - - - -

a +, Pathogenic response; -, nonpathogenic response; ND, not determined. 366 QHOBELA AND CLAFLIN INT. J. SYST.BACTERIOL. pattern than Xanthomonas albilineans (lane I), X. campe- Gibbons (ed.), Bergey’s manual of determinative bacteriology, stris pv. oryzae (lane J), Erwinia herbicola (lane K), and X. 8th ed. The Williams & Wilkins Co., Baltimore. campestris pv. oryzicola (data not shown). The patterns 8. Fahy, P. C., and G. J. Persley. 1983. Plant bacterial diseases, a diagnostic guide. Academic Press, Inc., New York. exhibited by X. campestris pv. graminis (lane H) and X. 9. Fredericksen, R. A. (ed.). 1986. Compendium of sorghum dis- campestris pv. translucens (lanes F and G) differed slightly eases. American Phytopathological Society, St. Paul, Minn. from those of the pearl millet pathogen by not having a 10. Hagborg, W. A. F. 1970. A device for injecting solutions and protein band at an approximate molecular weight of 22,000 suspensions into thin leaves of plants. Can. J. Bot. 48:1135- (Fig. 1). 1136. Host range tests. The pearl millet pathogen exhibited a 11. Hugh, R., and E. Leifson. 1953. The taxonomic significance of very narrow host range; pearl millet and proso millet were fermentative versus oxidative metabolism of carbohydrates by the only hosts (Table 5). Primarily due to its distinctive various gram-negative bacteria. J. Bacteriol. 66:24-26. pathogenicity and other nonpathogenic characteristics, our 12. Kersters, K., and J. De Ley. 1975. Identification and grouping of bacteria by numerical analysis of their electrophoretic protein results indicate that the pearl millet pathogen constitutes a pat terns. J. Gen. Microbiol. 87:333-342. new pathovar of X. campestris. Thus, we propose the name 13. King, E. O., M. K. Ward, and D. E. Raney. 1954. Two simple Xanthomonas campestris pv. pennamericanum pv. nov.; media for the demonstration of pyocyanin and fluorescin. J. the name is derived from Pennisetum americanum, the Latin Lab. Clin. Med. 44:301-307. binomial for pearl millet. 14. Kovacs, N. 1956. Identification of Pseudornonas pyocyanea by the oxidase reaction. Nature (London) 178:703. 15. Laemmli, U. K. 1970. Cleavage of structural proteins during the ACKNOWLEDGMENTS assembly of the head of bacteriophage T4. Nature (London) Special thanks are due to B. A. Ramundo for technical assistance. 227:680-685. This research was supported by the International Crops Research 16. Leach, J. E., B. A. Ramundo, D. L. Pearson, and L. E. Claflin. Institute for the Semiarid Tropics, by grant 603-0024-G-00-3029from 1987. Dot-immunobinding assay for detecting Xanthomonas the Agency for International Development, Washington, D.C., and campestris pv. hofcicolu in sorghum. Plant Dis. 71:30-33. by the Kansas Agricultural Experiment Station. 17. New South Wales Department of Agriculture Biology Branch 38. 1964. Annual report 1964, 34th plant disease survey of the twelve months ending 1964. New South Wales Department of LITERATURE CITED Agriculture, Sydney, Australia. 1. Alvarez, A. M., and K. Lou. 1985. Rapid identification of 18. Rajagopalan, C. K. S., and G. Rangaswami. 1958. Bacterial Xanthomonas campestris pv. cumpestris by ELISA. Plant Dis. leafspot of Pennisetum typhoides. Curr. Sci. 27:30-31. 69: 1082-1086. 19. Rangaswami, G., N. N. Prasad, and K. S. S. Easwaran. 1961. 2. Ames, G. F.-L. 1974. Resolution of bacterial proteins by poly- Two new bacterial diseases of sorghum. Andhra Agric. J. 8:269- acrylamide gel electrophoresis on slabs. J. Biol. Chem. 249:634- 272. 644. 20. Rangaswami, G., N. N. Prasad, and K. S. S. Easwaran. 1961. A 3. Carlson, R. R., and A. K. Vidaver. 1982. Taxonomy of Coryne- bacterial leaf blotch disease of cumbu (Pennisetum typhoides bucterium plant pathogens, including a new pathogen of wheat, Stapf and Hubbard). Madras Agric. J. 48:18&181. based on polyacrylamide gel electrophoresis of cellular pro- 21. Schaad, N. W. 1978. Use of direct and indirect immunofluores- teins. Int. J. Syst. Bacteriol. 32:315-326. cence tests for identification of Xanthomonas campestris. Phy- 4. Claflin, L. E., A. K. Vidaver, and M. Sasser. 1987. MXP, a topathology 68:249-252. semiselective medium for Xunthomonus campestris pv. pha- 22. Schaad, N. W. 1980. Identification schemes. I. Initial identifica- seoli. Phy topathology 77:730-734. tion of common genera, p. 1-11. In N. W. Schaad (ed.), 5. Dye, D. W. 1962. The inadequacy of the usual determinative Laboratory guide for identification of plant pathogenic bacteria. tests for the identification of Xunthomonas spp. N. Z. J. Sci. 5: American Phytopathological Society, St. Paul, Minn. 393416. 23. Starr, M. P., and W. L. Stephens. 1964. Pigmentation and 6. Dye, D. W. 1968. A taxonomic study of the genus Erwiniu. 1. taxonomy of the genus Xanthomonas. J. Bacteriol. 87:293-303. The “amylovora” group. N. Z. J. Sci. 11:591-607. 24. Thornley, M. J. 1960. The differentiation of Pseudomonas from 7. Dye, D. W., and R. A. Lelliott. 1974. Genus 11. Xanthomonus other gram-negative bacteria on the basis of arginine metabo- Dowson 1939, 187, p. 243-249. In R. E. Buchanan and N. E. lism. J. Appl. Bacteriol. 23:37-52.