INTERNATIONAL JOURNALOF SYSTEMATIC BACTERIOLOGY, Apr. 1978, p. 293-303 Vol. 28, No. 2 0020-7713/78/0028-0293$02.00/0 Copyright 0 1978 International Association of Microbiological Societies Printed in U.S.A. Cytophaga aquatilis sp. nov., a Facultative Anaerobe Isolated from the Gills of Freshwater Fish

WILLIAM R. STROHLT AND LARRY R. TAITtt Department of Biology, Central Michigan University, Mt. Pleasant, Michigan 48859

A facultatively anaerobic, gram-negative, gliding bacterium was isolated from the gills of freshwater fish. Its deoxyribonucleic acid base composition (33.7 mol% guanine plus cytosine), lack of microcysts or fruiting bodies, cell size (0.5 by 8.0 pm), and hydrolysis of carboxymethylcellulose and chitin place it in the genus Cytophaga. This aquatic cytophaga is differentiated from other cytophagas by its fermentation of carbohydrates, proteolytic capabilities, and a number of addi- tional physiological and biochemical tests. The organism was compared to other similar isolates reported from fish, and it appears to belong to a new , for which the name Cytophaga aquatilis is proposed. The type strain of C. aquatilis, N, has been deposited with the American Type Culture Collection under the accession number 29551.

Borg (3) described four strains of facultatively isolates are similar to those strains of faculta- anaerobic cytophagas isolated from at tively anaerobic cytophagas described by Borg the Skagit Fish Hatchery near Seattle, Wash. (3), Pacha and Porter (14), Anderson and Con- He associated those strains with bacterial gill roy (l),and Reichardt (16). disease although their pathogenicity was not MATERIALS AND METHODS proven. Bacterial strains. Thirteen strains of cytophagas While studying “myxobacterial” diseases of were isolated from diseased salmon or trout at the salmonbids, Anderson and Ordal(2) isolated and Platte River Fish Hatchery, Honor, Mich.; the Wolf studied a saprophytic fermentative cytophaga Lake Fish Hatchery, Wolf Lake, Mich.; and from that required COZ for the fermentation of glu- diseased suckers obtained from the holding tanks of cose. They named the organism Cytophaga suc- an Honor, Mich., baitshop. The strains and their cinicans. Pacha and Porter (14) isolated and sources are listed in Table 1. Methods. Three isolation methods were used to characterized a number of facultatively anaero- obtain the cytophagas: (i) salmon gills or fins were bic strains of saprophytic cytophagas from swabbed with sterile cotton swabs and streaked onto salmon, and Anderson and Conroy (I) compared skim milk medium (6), CP medium (5), and Pate and five strains of fermentative cytophagas to Borg’s Ordal medium (15); (ii) aseptically dissected gills or Skagit strains and again associated them with fins were placed onto CP or skim milk medium; and bacterial gill disease of salmon. (iii) aseptically dissected gills were blended with 100 A recent study by Bullock (4) showed that ml of sterile distilled water and plated onto CP and certain cytophaga-like strains, including a group skim milk media. Isolates from the plates were ob- of facultatively anaerobic cytophagas, could be tained by picking the edges of spreading colonies and placing them onto plated Pate and Ordal medium associated with bacterial gill disease as oppor- containing 0.9 or 1.5% agar. Spreading colonies from tunistic invaders, but no single pathogenic spe- these plates were checked for gliding motility. Gliding cies was found. of cells was determined by the use of hanging-drop Trust (22) showed that cytophagas are a part and wet-mount preparations and by observing plate or of the normal gill flora in a number of both wild slide agar mounts under phase and light microscopy. and hatchery-raised salmon, and he indicated A medium of 2% Casitone and Chu no. 10 basal salts that these organisms, along with pseudomonads, (20) was used as a basal medium and a maintenance are the predominant species of on the medium for the cytophagas throughout this investi- gills of healthy salmon. gation. In this report, we describe a new facultatively The temperature for optimal growth was deter- anaerobic species of Cytophaga which was iso- mined by growing the cultures in liquid basal medium at various temperatures from 0 to 37°C and comparing lated from the gills of diseased salmon at the the relative optical densities over a 2-week period. Platte River Fish Hatchery in Michigan. Our Methods for the determination of lauryl sulfate toler- ance, carboxymethylcellulose digestion, alginate liq- t Present address: Dept. of Microbiology, Louisiana State University, Baton Rouge, LA 70803. uefaction, and tyrosine degradation were those of -ftPresent address: Dept. of Microbiology, Wayne State Lewin and Lounsbery (12). The methods of Pacha and University, Detroit, MI 48202. Porter (14) were employed for the determination of 293 294 STROHL AND TAIT INT. J. SYST. BACTERIOL.

TABLE1. Sources and strains of cytophagas isolated from fish in March 1974 Location on fish Strain desig- Source Type of fish nation where obtained ~~~ T I Platte R. Hatchery Salmonb Gills R I Platte R. Hatchery Salmon Gills 0 I Honor, Mich., baitshop' Sucker Gills Q I Platte R. Hatchery Salmon Gills N I Platte R. Hatchery Salmon Gills A I1 Wolf Lake Hatchery Atlantic salmon Fins B I1 Wolf Lake Hatchery Atlantic salmon Fins D I11 Platte R. Hatchery Trout Fins G I11 Platte R. Hatchery Trout Fins P IV Wolf Lake Hatchery Atlantic salmon Fins J IV Platte R. Hatchery Salmon Gills K IV Wolf Lake Hatchery Atlantic salmon Fins S IV Wolf Lake Hatchery Atlantic salmon Fins The group I strains are the strains which are discussed in detail in this paper. The species of the salmon or trout were not recorded. ' From holding tanks outside the baitshop.

gelatin liquefaction, catalase production, casein, mined by the methods of Dworkin and Gibson (9) and starch, and esculin hydrolysis, digestion of chitin, and by placing the culture onto yeast streaks and observing cytochrome oxidase production. Oxidation or fermen- for fruiting structures over a 6-week period. The Cy- tation of carbohydrates was determined by the pro- tophaga pigments were extracted with 100% acetone duction of acid in phenol red basal agar medium or 95% ethanol, and absorption curves were run on a (Difco; 8) with 1%carbohydrate added. For the oxi- Gilford 2400 recording spectrophotometer. Analysis dation of carbohydrates, the acid reaction on plates of and gross characterization of the pigments were ac- aerobically incubated cultures was recorded as posi- complished by the methods of Reichenbach et al. (18). tive. For the fermentation tests, sterile Vaspar was For electron microscope observations of the colo- overlaid on agar deeps, and an acid reaction was nies, frozen-surface replicas were prepared by the considered as positive. Aerobic and anaerobic growth method of Steere (21) with cultures that were fixed for of the cultures was tested on 1%glucose supplemented 1h on plates of basal medium with 3% glutaraldehyde. with Chu no. 10 basal salts alone, basal salts plus 0.5% The agar was sliced approximately 3 mm below the yeast extract, basal salts plus 0.5% KN03, and basal surface and placed onto stubs for insertion into the salts plus 0.1% NH4H2P04.The degradation of native modified Denton DFE-2 freeze-etch unit. Replicas cellulose was tested by placing sterile Whatman no. 1 were shadowed at a 45" angle with platinum-carbon filter paper disks onto plates containing Chu no. 10 and were carbon coated. Electron micrographs were basal salts and 0.1% yeast extract. The cellulose plates taken, using a JEM lOOB electron microscope were incubated aerobically for a 4-week period, and equipped with a 60" tilt goniometer stage. Thin sec- positive results were recorded if dissolution of the tions were prepared by a modified Ryter-Kellenberger paper occurred. Salinity tolerance was tested by plac- technique (19). Plates of cells grown on basal medium ing 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, and 4.0% NaCl into for 48 h were flooded with 0.02% oso4 for 20 to 30 min, the basal medium and observing growth over 2 weeks. and the clumped bacteria were scraped off the agar All other physiological tests were done with standard and transferred to 1%Os04 for 18 h. Acetone dehydra- procedures found in the Difco Manual (€9, except tion (using, in a series, 50, 70, 85, 95, 100, and 100% those done with the Enterotube apparatus (BBL Prod- acetone) and Mollenhauer no. 1 plastic embedding ucts, Inc., Baltimore, Md.). procedure (13) were followed, and thin sections of the The ability of the cytophagas to lyse bacterial cells cytophaga were obtained on a Sorvall MT-2 Porter- was examined by swabbing live cultures of various Blum ultamicrotome with glass knives. Electron mi- bacteria across non-nutrient agar plates and then stab- crographs (magnification approximately x40,OOO) of bing the cytophaga test culture into the center. Posi- thin sections of the cytophaga were taken by using a tive results were indicated by dissolution of the bac- Phillips 300 electron microscope. Measurements of cell terial colony with spreading of the cytophaga colony size were obtained with a Filar micrometer under beyond the swabbed cultures of test bacteria. Lysis of phase microscopy using 24-h-old liquid cultures. cyanobacteria was tested by the methods described by Shilo (20) and by Gromov et al. (10); cultures of RESULTS Plectonema boryanum, Anacystis nidulans, and Nos- The aquatic cytophaga reported here is a fac- toc sp. were used as test cultures. ultatively anaerobic bacterium that uses fer- Antibiotic susceptibilities were determined by plac- ing susceptibility disks of various antibiotics onto basal mentable carbohydrates as suitable substrates medium plates streaked with the Cytophaga cultures; for anaerobic growth. Growth factors in yeast positive results were indicated by zones of inhibition. extract or peptone are required for growth and (See Table 6 for strengths of the antibiotics and results fermentation of the carbohydrates, and anaero- of the susceptibility testing.) bic growth occurs only during fermentation or Microcyst and hiting-body production were deter- during nitrate reduction. VOL. 28, 1978 CYTOPHAGA AQUATILIS, NEW SPECIES 295 A culture stabbed into 2%tryptone agar plates was neither utilized as a sole carbon source nor produced within 7 days a spreading colony with degraded in the presence of yeast extract. Cata- areas of purple, brown, yellow, and orange pig- lase, indoxyl acetate esterase, and deoxyribonu- mentation. On peptone, Casitone, and nutrient clease were produced, but cytochrome oxidase agar media, colonies spread and were pigmented and urease were not. Ornithine and lysine decar- yellow to orange, whereas on skim milk and CP boxylases and phenylalanine deaminase were media the spreading colony was thinner and not produced; tyrosine was hydrolyzed very nearly colorless. Colonies from anaerobically slowly. Tributyrin and Tweens 40, 60, and 80 grown cultures were a light cream color or non- were all hydrolyzed, although the hydrolysis of pigmented. Nutrient concentrations between 0.1 these lipids was slow in most cases. H2S was not and 2.5% and agar concentrations between 0.75 produced in iron peptone agar and the IMViC and 2.0% did not greatly affect the spreading tests were negative. ability of this organism. Motility of the cells was Conditions for optimum growth of this orga- by gliding and no flagella were present. Copious nism are as follows: temperature, 20°C (range 5 amounts of slime were produced within spread- to 35°C); pH, 6.5 (range 5.5 to 11.0); and NaCl ing colonies (Fig. 1). The cells at the periphery concentration, 0% (tolerance up to 2%). Most of the colony radiated from the central region of carbohydrates and alcohols tested were fer- the colony and were etched into the agar (Fig. mented within 3 to 4 days of inoculation, and 2). As observed in the frozen-surfacereplica (Fig. the reactions were complete within 7 days (Ta- Z), ice crystals formed from free water were in ble 5). Nitrate was reduced to NH3; ammonium direct contact with damaged cells, indicating nitrogen or nitrate could serve as the sole N hydrolysis of nutrients in the medium imniedi- sources. ately adjacent to the cells. On media containing Actinomycin D, aureomycin, bacitracin, chlor- low nutrient concentrations, motile (rotating) amphenicol, erythromycin, kanamycin, nalidixic microcolonies were formed which were con- acid, neomycin, novobiocin, oleandomycin, nected by tracks of slime left behind by gliding streptomycin, sulfathiazole, tetracycline, and tri- clumps of cells. ple sulfas were inhibitory at the strengths listed Five of the 13 strains isolated were able to in Table 6. Vancomycin inhibited gliding, but it hydrolyze chitin only weakly or not at all, and did not inhibit growth. they did not hydrolyze cellulose under aerobic As shown in Table 7, all bacteria tested, except conditions. They did not produce indole or HZS Micrococcus Zysodeikticus strain 4710 and Gaff- or utilize citrate as a sole carbon source, but they kya tetragena strain 4731, were lysed by strain did produce catalase aerobically, hydrolyze N. However, none of the blue-green algae tested starch and gelatin, reduce nitrate to NH3, grow were lysed. Sarcina lutea 4745 and Streptococ- anaerobically on glucose and yeast extract, and cus faecalis 410216 were lysed most rapidly by produce identical pigments and colonial mor- the cytophaga. phologies (Table 2). Cells of all five strains mea- The cell envelope was typical of gram-nega- sured 5 to 15 pm in length (average, 8 pm). They tive organisms (7), and some flexibility of the occurred singly and did not produce sheaths or cell wall was evident (Fig. 3). Division was by filaments. Degenerate coccoid forms and spher- transverse binary fission with septa forming be- oplasts were found in both young and old cul- tween dividing rods. The guanine-plus-cytosine tures although they were more prevalent in the (G+C) content of the deoxyribonucleic acid was older cultures. The coccoid forms were examined found by M. Mandel to be 33.7 mol% for strain by electron microscopy and were found to be N. The major pigments were not carotenoid but shortened rods rather than microcysts. Fruiting instead were similar to the polyene-chromo- bodies were not formed. phoric compounds found in Flexibacter elegans Because the five group I strains (Table 1) were and other cytophagas (17, 18). The absorption so nearly identical morphologically and physio- spectra for the pigments in acetone and ethanol logically, only one strain (N) was studied further solvent systems and the bathochromic shift ob- to elucidate the specific characteristics of the served with the addition of NaOH in ethanol are group. The physiological and biochemical char- given in Fig. 4. The color of the shifted pigments acteristics of strain N are given in Table 3, and changed from yellow to red-violet and could be the morphological characteristics are given in returned to the original spectrum by the addi- Table 4. tion of excess HCl in ethanol (18). The pigment All of the macromolecules examined, except could not be extracted from ether by using native cellulose, were hydrolyzed by the cyto- Na2C03; it resisted photooxidation and gave a phaga. Gelatin, casein, litmus milk, and peptone reddish color with the addition of KOH in meth- were strongly hydrolyzed. Cellulose was not de- anol. The pigments did not give a sky-blue color graded, but 500 ml of carboxymethylcellulose with the presence of saturated SbCL in chloro- was degraded completely within 5 days. Alginate form, but they gave a deep violet reaction when 296 STEZOHL AND TAIT INT. J. SYST.BAC'T~WIOI,.

FIG. 1. Frozen-surface replica of strain N. Note the fibrillar nature of the slime (S) and the undulated surface (U) of the cells. Magnification, X43,500. VOL. 28, 1978 CYTOPHAGA AQUATILIS, NEW SPECIES 297

FIG. 2. Frozen-surface replica from colony edge of strain N showing cells “etched” into agar surface and forming tonguelike extensions radiating from the colony center. Note ice-crystal damage in close proximity to cells (0.The bottom of the micrograph is towards the center of the colony. Magnification, X25,OOO. 298 STROHL AND TAIT INT.J. SYST.BACTERIOL.

TABLE2. General characteristics of the five strains studied Reaction“ Characteristic Strain T Strain R Strain 0 Strain Q Strain N Gliding motility ...... + + + + + - Microcysts ...... - - - - Fruiting bodies ...... - - - - - Spreading colony ...... + + + + + Colony pigments: 2% Tryptone ...... Y,O,P,b y,o,p,b Yl0,Plb Y,O,P,b Y,O,P,b Nutrient agar ...... o,b o,b o,b o,b o,b CP medium ...... Y Y Y Y Y Autolysis in old cultures ... + + + + + Anaerobic growth ...... + + + + + Nitrate reduced to NH3 ..... + + + + + Growth at 4°C ...... ------Cellulose degradation ...... - - - - Chitin degradation ...... + Starch hydrolysis ...... + Casein hydrolysis ...... + Gelatin hydrolysis ...... + Peptone utilization . . , . . , . . + Catalase production ...... + H2S production ...... - Indole production ...... - Citrate utilization ...... - Glucose oxidationb ...... + Galactose oxidation ...... - Sucrose oxidation ...... - Cellobiose oxidation ...... - Lactose oxidation ...... - Maltose oxidation ...... + Mannitol oxidation ...... -

‘‘ Abbreviations and symbols: b, brown; 0,orange; p, purple; y, yellow; +, positive result; -, negative result. Acid produced in 6 days. treated with concentrated sulfuric acid. With the base ratio, proteolytic capabilities, and in the exception of negative reactions with 50% etha- requirement by F. succinicans of COZ for fer- nolamine in ethanol and in the SbCb reaction, mentation of carbohydrates. the pigments appeared similar to the flexirubin Our fermentative cytophaga is similar, how- pigment found in F. elegans (18). ever, to many isolates reported by various re- searchers (Table 8). Borg (3) reported four DISCUSSION strains of fermentative cytophagas which differ The unicellular gliding bacterium was placed from our cytophaga only in their inability to in the genus Cytophaga because of its gliding hydrolyze gelatin. Pacha and Porter (14) isolated motility, its lack of microcyst or fruiting-body several strains of saprophytic cytophagas, in- production, its 33.7 mol% G+C base ratio, its cluding six fermentative strains, from the sur- weak degradation of chitin, and its strong depo- faces of freshwater fish. These differ very little lymerization of carboxymethylcellulose. It is dif- from our organism and may well be considered ferent from any presently known species of Cy- as strains of C. aquatilis. Bullock (4) studied tophaga due to a combination of its fermentative many strains of cytophaga-like bacteria from and nitrate-reducing abilities, its freshwater or- fish. Among his 55 strains were three faculta- igin, its starch, lipid, and strong protein hydro- tively anaerobic strains capable of carrying out lytic capabilities, its production of catalase, and fermentation of glucose. These strains appear to its lack of cytochrome oxidase. The only other be identical to our organism, and Reichardt (16) fermentative freshwater member of the family reported a strain “S-K,” isolated from a fresh- is Flexibacter succinicans (2), water habitat, which appears to be similar to our formerly known as Cytophaga succinicans, fermentative cytophaga. However, with the lim- which was also isolated from fish. Our cytophaga ited knowledge available about this organism, it differs greatly from F. succinicans in chitin and is impossible to determine its relationship to C. carboxymethylcellulose depolymerization, G+C aquatilis. All of these organisms are similar in VOL. 28, 1978 CYTOPHAGA AQUATILIS, NEW SPECIES 299

TABLE3. Physiologic characteristics of strain N TABLE4. Morphological characteristics of the five ~- strains studied Test Result Deoxyribonucleic acid base composition Characteristic Result (mol% G + C; buoyant density) ..... 33*7 Fruiting bodies present ...... - (0/5) Salinity tolerance (%) ...... 0-2.0 Microcysts present ...... - (0/5) Temp range for optimal growth ("C) . . 20-25 Cells in trichomes ...... - (0/5) Upper temp limit for growth ("C) ..... 35 Cells in sheath ...... - (0/5) Lower temp for growth ("C) ...... 5 Cells in filaments ...... - (0/5) pH range for growth ...... 5.5-11.0 Helical forms present ...... - (0/5) pH for optimal growth ...... 6.5 Branching present ...... - (0/5) Nitrate reduction to NH3 ...... + Spheroplasts present ...... + (5/5) Nitrogen fixation ...... - Gram reaction ...... - (0/5) Sodium dodecyl sulfate tolerance (0.01%) - Carotenoids as major pigments ...... - (strain N) Litmus milk reaction ...... Hard curd Pigment peak (acetone solvent) Utilization of NH4-nitrogen ...... + (nm) ...... 446 Citrate utilization ...... - Shoulders (nm) ...... 429,461,473 Methyl red test ...... - Fluorescence at 260 nm ...... - (strain N) Indole production ...... - Length of cells (pm) ...... 5-15 (5/5) Acetylmethylcarbinol production ..... - Width of cells (pm) ...... 0.5-1.0 (5/5) HS production ...... - Binary fission by septation ...... + (5/5) Catalase ...... + Flagella present ...... - (0/5) Cytochrome oxidase ...... - Gliding motility ...... + (5/5) Urease ...... - Spreading on nutrient agar ...... f (5/5) Deoxyribonuclease ...... + Colonies etched into agar ...... + (5/5) Indoxyl acetate esterase ...... + Ruthenium-red-positive slime present + (strain N) Lysine decarboxylase ...... - Flexuous cells ...... + (5/5) Phenylalanine deaminase ...... - Sblfur deposits ...... - (0/5) Ornithine decarboxylase ...... - Polysaccharide deposits ....., ..... + (5/5) Blood hemolysis ...... - Polv-P-hydroxvbutvrate deposits ...: - (0/5) Tributyrin hydrolysis ...... + Tween 40 hydrolysis ...... + TABLE5. Oxidative and fermentative production of Tween 60 hydrolysis ...... + acid from carbohydrates and alcohols by strain N" Tween 80 hydrolysis ...... + Pectin hydrolysis ...... + Substrate Oxidation Fermen-tation Tyrosine hydrolysis ...... + Gelatin hydrolysis ...... + Carbohydrate Casein hydrolysis ...... + Glucose ...... + + Esculin hydrolysis ...... + Galactose ...... W + Alginate hydrolysis ...... - Arabinose ...... W + Starch hydrolysis ...... + Xylose ...... W + Glycogen hydrolysis ...... + Lactose ...... W + Agar hydrolysis ...... - Trehalose ...... W Cellulose degradation ...... - Fructose ...... W + Carboxymethylcellulose digestion ..... + Maltose ...... W + Chitin degradation ...... +, weak Mannose ...... W + Sorbose ...... - - Sucrose ...... W + that they were isolated from fish or aquatic Raffinose ...... V W habitats. Those that were tested hydrolyzed chi- Cellobiose ...... W + tin but not cellulose, fermented glucose, lacked Alcohol cytochrome oxidase, were indole negative, hy- Glycerol ...... V + drolyzed gelatin, starch, casein, and tributyrin, Dulcitol ...... V + - - were catalase positive, and had an optimum Sorbitol ...... Mannitol ...... W + temperature of about 20 to 25°C (Table 8). - - The production of H2S was-detected in only Inositol ...... the strains of Pacha and Porter, nitrate reduc- a Symbols: +, positive, acid formed within 6 days; tion was variable among the strains, and tyrosine W, weakly positive, acid formed after 7 days; -, neg- degradation was variable with the strains of ative; V, variable. Pacha and Porter and of Bullock. Isolation and more complete characterization of additional istics of our five strains and the additional tests strains may show that these factors are variable on strain N should constitute the description of in the species. At this point, however, character- the species. 300 STROHL AND TAIT INT. J. SYST. BACTERIOL.

TABLE6. Antibiotic susceptibility of strain N to provide further understanding of the ecology Antibiotic Dosage Inhibition of the disease. Surface-replica preparations (Fig. 1 and 2) Actinomycin D ...... + showed that the cytophaga has an undulated Ampicillin ...... surface and produces great amounts of slime, Aureomycin ...... which probably facilitates the gliding motility of Bacitracin ...... Chloramphenicol ...... these organisms. The cells at the colony edge Col ymycin ...... are “etched” into the agar with zones of ice- Erythromycin ...... crystal damage around the cells (Fig. 1). It is Kanamycin ...... possible that these ice-crystal zones are due to Lincomycin ...... the formation of water released into the medium Methicillin ...... during hydrolysis of nutrients. That idea pro- Nalidixic acid ...... vides an explanation as to why ice-crystal dam- Neomycin ...... age is seen only in close proximity to the cells. Novobiocin ...... Further ultrastructural analysis of the cyto- Oleandomycin ...... Penicillin G ...... phaga will be published elsewhere (W. R. Strohl Polymyxin B ...... and L. R. Tait, manuscript in preparation). Streptomycin ...... + The major pigments have been found to be Sulfadiazine ...... similar to the “flexirubin” pigment found in F. Sulfathiazole ...... + elegans (18). Reichenbach (personal communi- Tetracycline ...... + cation) has indicated that these pigments are Triple sulfa ...... + common to freshwater cytophagas. They were -a Vancomycin ...... formerly thought to be carotenoids (2, 11, 12). “ Inhibits gliding. In the past, determination of cytophaga pig- ments as carotenoids was usually done by using the absorption spectra and the cyanolipid H2S04 It may be noted that only three out of Bul- test. Since neither of these tests distinguishes lock’s 55 strains were fermentative, yet all of his flexirubin-like pigments from carotenoids, it freshwater isolates showed a striking consistency should become standard procedure to utilize fur- among the basic physiological tests. It is very ther screening tests such as the KOH in ethanol possible that a glucose-oxidative,strictly aerobic spectral shift and the 50°C-NaBH4-induced species of Cytophaga or Flexibacter exists which bathochromic shift as proposed by Reichenbach is similar to the fermentative cytophaga. et al. (18). Because all five of our strains were isolated We regard our five isolates (strains T, R, O., from gills of fish displaying symptoms of bacte- Q, and N) as belonging to a new species, for rial gill disease, it is possible that these orga- which we propose the name Cytophaga aqua- nisms were pathogenic to those fish under the ti&. A formal description of the species is given prevailing environmental conditions. Since no below with a description of the type strain. Many tests were devised to test this hypothesis, it of the species characteristics are based on one cannot be determined if the cytophagas were the disease-causing organisms. Anderson and Con- TABLE7. Lysis of microogranisms by strain N roy (1) and Bullock (4) have suggested that aquatic “myxobacteria” should be considered as Test organism Lysis opportunistic pathogens, only producing gill dis- Rhodospirillum rubrum 1221” ...... + ease when the fish are stressed by environmental Escherichia coli 4411 ...... + conditions. Since Trust (22) found that cytopha- Serratia marcescens 4471 ...... + gas are part of the normal gill flora of healthy Proteus mirabilis 4482 ...... + salmon, he concluded also that environmental Micrococcus lysodeikticus 4710 ...... - Staphylococcus epidermidis 4722 ...... + conditions must “predispose the gill to patholog- Sarcina lutea 4745 ...... + ical changes by these organisms” common to the Streptococcus faecalis 410216 ...... + fish and that it acts as an opportunistic invader Clostridium perfringens 42CP3 ...... + when the fish are placed under stressed condi- Caulobacter sp. CP4 ...... + tions. Since hatchery-raised fish are often Gaffiya tetragena 4731 ...... - stressed due to factors such as overcrowding, Plectonema boryanum BPBl ...... - temperature, and low dissolved-oxygen tension, Anacystis nidulans BAN1 ...... - it is possible that the salmon are killed off by Nostoc SP. BNSl ...... - pathogenic effects produced by the cytophagas “Strain numbers are those of the Department of under these conditions. Much more research Biology, Central Michigan University, Mount Pleas- must be done in this area with similar organisms ant, Mich. I?IG. 3. Thin-section micrograph of strain N showing typical gram-negative cell envelope (CE) with the cytc gplasmic membrane (CM), peptidoglycan layer (PL), and lipopolysaccharide outer membrane (OM). Note the well developed mesosomes (M) and nuclear region (NR). Magnification, ~40,000. 301 302 STROHL AND TAIT INT. J. SYST. BACTERIOL. i 12-

10- Lu V Z 8- Ly

9m 6- Lua

t-1 4 4- Lu OL A 2- L

4 560

WAVELENGTH (nm) FIG. 4. Absorption spectra of strain-Npigrnents in ethanol (A) and acetone (B). Bathochromic shift induced by the addition of NaOH in ethanol is shown as (C).

TABLE8. Comparison of strain N with some other facultatively anaerobic cytophagas isolated from fish or wate?

Borg’s iso- and Bullock’s Reichardt’s Test Strain N lates (2) p~~~~~~~-isolates* (14) isolate (16)

Cellulose degradation ...... - - Chitin decomposition ...... + + Starch hydrolysis ...... + + + Gelatin liquefaction ...... - + + Indole production ...... - - - Casein hydrolysis ...... + + Nitrate reduction ...... + V V H2S production ...... + Tributyrin hydrolysis ...... + + Catalase production ...... + + + Urease production ...... - - Cytochrome oxidase ...... - - Esculin hydrolysis ...... + + Tyrosine hydrolysis ...... + V V Anaerobic growth ...... + + + + + Glucose fermentation ...... + + + + + Pectin hydrolysis ...... + Bacteriolysis ...... + + + Optimal temp 20 to 25OC ...... + + + + + Symbols: +, positive; -, negative; V, variable. Three fermentative strains (4.53,4.26, and 4.7). strain (strain N) although several of the distin- characteristics are based on five strains. Flexible guishing factors are based on five strains. rods, 0.5 to 0.75 pm in diameter, 5 to 15 pm in Cytophaga aquatilis sp. nov. (a. qua’ti. lis. L. length. Gram negative. Motile by gliding; afla- adj. aquatilis living in water). The following gellate. Both elongated and coccoid forms are VOL. 28, 1978 CYTOPHAGA AQUATILIS, NEW SPECIES 303

present in old cultures, the latter superficially myxobacterium. J. Bacteriol. 81: 130-138. resembling microcysts. Chitin may be weakly 3. Borg, A. F. 1960. Studies on myxobacteria associated with diseases in salmonoid fishes. Wildl. Dis. 8:l-85. hydrolyzed (two of five strains are positive), but 4. Bullock, G. L. 1972. Studies on selected myxobacteria cellulose and agar are not. Starch, casein, gelatin, pathogenic for fishes and on bacterial gill disease in and peptone are hydrolyzed. Catalase is pro- hatchery-reared salmonoids. Technical paper 60. U.S. duced, but HZS and indole are not. Citrate is not Department of the Interior, Bureau of Sport Fisheries and Wildlife, Washington, D.C. utilized. Nitrate is reduced to ammonia and can 5. Carlson, R. V., and R. E. Pacha. 1968. Procedure for be used as a sole N source. Acid is produced the isolation and enumeration of myxobacteria from from glucose and maltose oxidatively but not aquatic habitats. Appl. Microbiol. 16:795-796. from sucrose, cellobiose, lactose, galactose, or 6. Christensen, P. J., and F. D. Cook. 1972. The isolation and enumeration of cytophagas. Can. J. Microbiol. mannitol. Anaerobic growth occurs in the pres- 18: 1933-1940. ence of glucose and yeast extract or during ni- 7. Costerton, J. W., J. M. Ingram, and K. J. Cheng. trate reduction. Colonies are pigmented yellow 1974. Structure and function of the cell envelope of or orange to light brown when grown aerobically gram-negative bacteria. Bacteriol. Rev. 38:87-110. 8. Difco Laboratories. 1953. Difco manual. Difco Labora- and are nonpigmented under anaerobic condi- tories, Inc., Detroit, Mich. tions. Additional characteristics of strain N are 9. Dworkin, M., and S. M. Gibson. 1964. A system for listed below. Alginate is not used as a carbon studying microbial morphogenesis: rapid formation of and energy source. Carboxymethylcellulose is microcysts in Myxococcus xanthus. Science 146:243-244. strongly hydrolyzed. Deoxyribonuclease and in- 10. Gromov, B. V., 0. G. Ivanov, K. A. Mamkaeva, and doxy1 acetate esterase are produced, but cyto- I. A. Avilov. 1972. A flexibacter that lyses blue-green chrome oxidase and urease are not. Acetylmeth- algae. Mikrobiologiya 41: 1074-1079. ylcarbinol is not produced; the methyl red test 11. Leadbetter, E. R. 1974. , p. 99-127. In R. E. Buchanan and N. E. Gibbons (ed.), Bergey’s manual is negative. Tweens 40,60, and 80, and tributyrin of determinative bacteriology, 8th ed. The Williams and are hydrolyzed and are used as sole carbon Wilkins Co., Baltimore, Md. sources. Strongly bacteriolytic but not cyano- 12. Lewin, R. A., and D. M. Lounsbery. 1969. Isolation, bacteriolytic. Ferments and weakly oxidizes a cultivation, and characterization of flexibacteria. J. Gen. Microbiol. 58: 145-170. wide range of carbohydrates and alcohols. The 13. Mollenhauer, H. H. 1963. Plastic embedding mixtures G+C content of the deoxyribonucleic acid is 33.7 for use in electron microscopy. J. Stain Technol. mol% (based on a buoyant-density analysis). 39: 111-114. The major pigments produced resemble flexiru- 14. Pacha, R. E., and S. Porter. 1968. Characteristics of myxobacteria isolated from the surface of freshwater bin and are not carotenoid. The temperature for fish. Appl. Microbiol. 16: 1901-1906. optimal growth is 20°C; no growth occurs at 0 or 15. Pate, J. L., and E. J. Ordal. 1967. The fine structure of 37°C. Chondrococcus columnaris. I. Structure and formation Type strain. The only strain of the species of mesosomes. J. Cell Biol. 35:l-13. now available, strain N, is designated as the type 16. Reichardt, W. 1974. Zur Okophysiologie einiger Gewas- serbakterien aus der Flauobacterium-Cytophaga- strain; a culture of this strain has been deposited Gruppe. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. with the American Type Culture Collection un- Hyg. Abt. I Orig. 227:85-93. der the accession number 29551. 17. Reichenbach, H., and H. Kleinig. 1972. Die Carotinoide der Myxobakterien. Zentralbl. Bakteriol. Parasitenkd. ACKNOWLEDGMENTS Infektionskr. Hyg. Abt. I Orig. 220:458-463. 18. Reichenbach, H., H. Kleinig, and H. Achenbach. 1974. We thank Manley Mandel for analyzing the deoxyribonu- The pigments of Flexibacter elegans: novel and che- cleic acid of strain N and Russell Steere for the use of his mosystematically useful compounds. Arch. Microbiol. laboratory and his technical assistance in preparing the frozen- 101:131-144. surface replicas. We also thank Ellis Brockman and James 19. Ryter, A., and E. Kellenberger. 1958. Etude au micro- Lampky for their cooperation, patience, and helpful assistance scope electronique de plasmas contenant de l’acide des- throughout this investigation and John Larkin for his assist- oxyribonucleique. I. Les nucleoides des bacteries en ance in the preparation of this manuscript. croissance active. Z. Naturforsch. 13b:597-605. REPRINT REQUESTS 20. Shilo, M. 1970. Lysis of blue-green algae by myxobacter. J. Bacteriol. 104:453-461. Address reprint requests to: William R. Strohl, Dept. of 21. Steere, R. L. 1973. Preparation of high-resolution freeze- Microbiology, 508 Life Sciences Building, Louisiana State etch, freeze-fracture, frozen-surface, and freeze-dried University, Baton Rouge, LA 70803. replicas in a single freeze-etch module, and the use of stereo electron microscopy to obtain maximum infor- LITERATURE CITED mation from them. In E. L. Benedetti and P. Favard 1. Anderson, J. I. W., and D. A. Conroy. 1969. The (ed.), Freeze-etching techniques and applications. So- pathogenic myxobacteria with special reference to fish ciete Francaise de Microscopie Electronique, Paris. diseases. J. Appl. Bacteriol. 32:30-39. 22. Trust, T. J. 1975. Bacteria associated with the gills of 2. Anderson, R. L., and E. J. Ordal. 1961. Cytophaga salmonoid fishes in freshwater. J. Appl. Bacteriol. succinicans sp. n., a facultatively anaerobic, aquatic 38: 225-233.