APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1984, p. 352-3600099-2240/84/080352-09$02.00/0Copyright C) 1984, American Society for Microbiology

Growth Potential of Halophilic Bacteria Isolated from Solar SaltEnvironments: Carbon Sources and Salt Requirements

BARBARA J. JAVORScripps Institution of Oceanography, University of California, San Diego, La Jolla, California 92093

Received 12 December 1983/Accepted 15 May 1984

Eighteen strains of extremely halophilic bacteria and three strains of moderately halophilic bacteria wereisolated from four different solar salt environments. Growth tests on carbohydrates, low-molecular-weightcarboxylic acids, and complex medium demonstrated that the moderate halophiles and strains of the extremehalophiles Haloarcula and Halococcus grew on most of the substrates tested. Among the Halobacterium isolateswere several metabolic groups: strains that grew on a broad range of substrates and strains that were

essentially confined to either amino acid (peptone) or carbohydrate oxidation. One strain (WS-4) only grew wellon pyruvate and acetate. Most strains of extreme halophiles grew by anaerobic fermentation and possibly bynitrate reduction. Tests of growth potential in natural saltern brines demonstrated that none of the halobacteriagrew well in brines which harbor the densest populations of these bacteria in solar salterns. All grew best inbrines which were unsaturated with NaCl. The high concentrations of Na+ and Mg2+ found in salterncrystallizer brines limited bacterial growth, but the concentrations of K+ found in these brines had little effect.MgSO4 was relatively more inhibitory to the extreme halophiles than was MgCl2, but the reverse was true forthe moderate halophiles.

Although extremely halophilic bacteria have been inten-sively studied, few investigations have emphasized theirphysiology from an ecological point of view. Questionsremain concerning natural carbon sources, the relative im-portance of aerobic and anaerobic processes of energyconversion, and the role of the ionic environment in limitingthe distribution and survival of extreme and moderate halo-philes in natural brines and salt deposits. Knowledge of theecology of halophilic bacteria is necessary for predictingmicrobial activity in the proposed storage of radioactivewastes in buried salts. An understanding of these microorga-nisms in their natural environments would also provideinsight into the genesis and diagenesis of petroleum associat-ed with evaporite deposits and would provide an empiricalbasis for understanding bacteriostasis by salts.

Hypersaline environments are generally thought to becharacterized by low species diversity, probably becauseeucaryotic organisms are largely or entirely absent. Thenumber and diversity of halophilic bacterial strains andspecies (i) that exist in nature and (ii) that coexist in localhypersaline environments are unknown. It can be hypothe-sized that a relatively large diversity of strains, species, orboth exists within the bacterial floras, a condition thatrequires experimental scrutiny rather than the examinationof gross morphology used for identifying unicellular andmulticellular eucaryotes.Among the extremely halophilic bacteria, three genera

have been described: Halobacterium, Halococcus (5), andHaloarcula (not on the Approved List of Bacterial Names[10]). In the genus Halobacterium, eight or more specieshave been described. The most intensively studied species,Halobacterium halobium, Halobacterium salinarium, andHalobacterium cutirubrum, are thought to be strains of thesame species (23). Because these strains thrive primarily on

amino acids, the ability of Halobacterium marismortui (5,14), Halobacterium volcanii (14), Halobacterium saccharo-vorum (22), Halobacterium vallismortis (6), and Halobacte-rium sodomense (14) to grow on carbohydrates was original-ly viewed as highly unusual. Halococcus morrhuae, the onlyspecies recognized in this genus, is reported to be unable tothrive on carbohydrates (11). The carbon sources that sup-

port the growth of Haloarcula spp. have not been described.A survey of organic substrates metabolized by a repre-sentative sample of extreme halophiles isolated on complexmedium from different solar salt environments would indi-cate how versatile these microorganisms are as hetero-trophs. Such a survey would indicate whether this classicalapproach to bacterial taxonomy can be used to definespecies of halobacteria.Most extreme halophiles require at least 2.5 M NaCl. In

solar salterns, halobacteria color the brines red in NaClcrystallizer ponds with brine densities of ca. 25 to 300 Bd (seeFig. 1) (9). Because of massive NaCl precipitation, the highlyconcentrated brines between 31 and 320 Be density havesodium concentrations of less than 2.5 M. Despite thepresence of other cations, it can be postulated that thedecrease in Na+ alone may account for the failure ofextreme halophiles to survive in these extremely concentrat-ed brines (the bitterns). However, the effects of the increasein concentrations of the other major ions (Mg2+, K+, Cl-,and S042-) should be evaluated.A survey of moderate halophiles isolated from solar salt

ponds showed that at least five genera were represented (24).This study demonstrated that, as a taxonomically diversegroup, moderate halophiles have very diverse metabolicrequirements and capabilities. They may compete well withhalobacteria in some hypersaline environments because theyhave relatively high growth rates at ambient temperatures(17). Their growth potential in medium that reflects thechemical composition of natural brines should be measuredto ascertain the limited success of these microorganisms inthe very concentrated brines of solar salterns.

In this investigation, both extremely halophilic and moder-ately halophilic bacteria were isolated from four differentsolar salt environments of marine origin. The plating effi-ciency of native populations was not determined because itis well established that no one medium composition or set ofgrowth conditions can provide the growth requirements ofthe entire "viable" bacterial flora. Direct microscopiccounts (9) may be in error due to the presence of SrSO4crystals in suspension with the bacteria (B. Javor, SixthInternational Symposiium on Salt, in press). The isolates

352

Vol. 48, No. 2

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

GROWTH POTENTIAL OF HALOPHILIC BACTERIA 353

were screened to establish their physiological diversity andtaxonomic relatedness. This communication reports thegrowth potential of these isolates with different organicsubstrates and in different salt concentrations of both naturalbrines and artificial media.

MATERIALS AND METHODSComplex medium (CM), culturing conditions, and protein

measurements have been described previously (10). Techni-cal-grade peptone (Difco Laboratories) was substituted forInolex peptone. A modification of Sehgel and Gibbons (20)medium, designated SG, was also used. It contained all thesalts of CM. The organic source was 7.5 g of Casamino Acids(Difco) plus 5 g of yeast extract (Difco) per liter. Bufferedcomplex medium (CM-B) is described by Tomlinson andHochstein (21). The buffers, MES [2-(N-morpholino)ethane-sulfonic acid], HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid), and MOPS [3-(N-morpholino)propane-sulfonic acid], provided similar growth rates and yields(unpublished data). In any one experiment, medium withonly one of the buffers was used. CM, SG, and CM-B werefiltered with GF/C (Whatman) or Whatman no. 1 filtersbefore autoclaving.

Minimal medium (designated MIN) was employed to testgrowth on individual carbon sources (glycerol, arginine, andacetate). MIN is CM-B without yeast extract and with 200 gof NaCl per liter, 5 mM NH4NO3, and 50 ,ul of Castenholzmicronutrients per liter (2). After autoclaving, 1 ml of 0.3 MNa2HPO4 (pH 7.4 to 7.8) was added per liter. One millilitereach of a filter-sterilized B12 solution (0.1 mg/100 ml ofdistilled water) and a filter-sterilized vitamin mix was addedper liter. The vitamin mix solution contained (per 100 ml ofdistilled water): 1.0 mg of biotin, 10.0 mg of niacin, 5.0 mg ofthiamine, 5.0 mg of p-aminobenzoic acid, 2.5 mg of panto-thenic acid, 25 mg of pyridoxine hydrochloride, 2.0 mg offolic acid, and 2.5 mg of riboflavin. Organic substrates wereadded as described for CM-B.Media prepared from brines from the Exportadora de Sal,

S.A., saltern were first clarified by centrifugation. Peptone(Difco technical grade) or peptone and glycerol (1% each)were autoclaved with the brines. The brines are highlybuffered naturally. The pH ranged from ca. 7.0 to 7.5.For comparative screening tests, organic substrates were

added aseptically to separately autoclaved CM-B. Organicsubstrates were added in 1% [wt/vol] final concentrationsfrom 20% sterile stock solutions (except DNA, which wasadded at a final concentration of 0.5% [wt/vol] from a 10%stock solution). All the carbon sources to be screened wereheat sterilized, except the following, which were filter steril-ized: D-ribose, ot-methyl-D-glucoside, pyruvate, propionate,DNA, and ethanol. Control medium consisted of CM-B towhich distilled water was added in a volume equivalent tothat of the organic substrates.

For the screening tests, a loopful of culture was inoculatedin 3 ml of medium in tubes which were loosely capped.Tubes were agitated at 125 rpm in a New Brunswickincubator at 37°C with a 40-W tungsten bulb and a fluores-cent light placed ca. 50 cm away from the tubes. Growth, asshown by optical density at 750 nm (OD750), was monitoredwith a Beckman DU spectrophotometer equipped with aGilford power supply. Relative specific growth rates fordifferent substrates were determined by comparing theOD750 of cells in a set of experiments in which cells wereinoculated and harvested at the same times (during logarith-mic phase, usually after 2 to 4 days). Comparison with actualgrowth rates (based on protein determinations of cells har-

vested at different times in their growth cycle) showed thatthese two methods gave parallel results (unpublished data).For fermentation tests, the carbon source (when appropri-

ate) and 2 g of NaHCO3 per liter were added to CM-B or SGimmediately after autoclaving. The medium was immediate-ly dispensed aseptically into sterile screwcap tubes, inocu-lated, and incubated at 37°C. Fermentation was also testedon agar plates with CM-B plus glucose, glycerol, or acetatein a Gas-Pak system (BBL Microbiology Systems) at 37°C.

Nitrate and nitrite reduction were determined in culturetubes with inverted Durham tubes. Medium in which thebacteria demonstrated good aerobic growth was supplement-ed with 100 mM KNO3 or NaNO3 or 10 mM NaNO2.Medium was used immediately after autoclaving or wassteamed just before use. The medium was covered with 1 cmof heavy mineral oil, and the tubes were incubated at 37°C.Gram staining, catalase, and oxidase tests have been

described previously (7). Arginine dihydrolase, ornithinedecarboxylase, and lysine decarboxylase were determinedby anaerobic growth in M0ller broth base (Difco) plus thesalts and trace elements of CM. Amino acids were added to a1% [wt/vol] final concentration. The tubes were coveredwith 1 cm of mineral oil.The methods for ionic analyses have been described

(Javor, in press). Lipid analyses were performed by themethods of Ross et al. (19) with the exception that thin-layerchromatography plates were developed in hexane-diethylether-acetic acid (74:25:1) and visualized by charring afterspraying with 3% cupric acetate in 8% phosphoric acid.Spectral analyses of cell lysates in distilled water were donewith a Beckman MVI Acta spectrophotometer. Bacteriorho-dopsin was determined by the method of Javor et al. (10).

All reagents were of analytical grade. DNA (degraded freeacid type IV herring sperm), penicillin G, and bacitracinwere obtained from Sigma Chemical Co., and polymyxin Bwas purchased from Pfizer Inc.Halobacterium halobium Rl, Halobacterium cutirubrum

NRC 34001, Halobacterium salinarium no. 10, and Halobac-terium vallismortis were generous gifts of W. Stoeckenius(University of California, San Francisco). B. Volcani(Scripps Institute of Oceanography) provided Halobacte-rium marismortui and Halobacterium volcanii (this strainwas pale, not red [12]). L. Hochstein (National Aeronauticsand Space Administration) provided Halobacterium sac-charovorum.

RESULTSBrine samples were collected from the Exportadora de

Sal, S.A., saltern in Guerrero Negro, Baja California Sur,Mexico (designated GN); from the muds of natural salt flatsadjacent to this salina (designated GNM); from the La Salinaslough in Baja California (designated LS); and from theWestern Salt Co. saltern in San Diego Bay, Chula Vista,Calif. (designated WS) (see reference 9 and Sixth Interna-tional Symposium on Salt, in press, for descriptions of theseenvironments). Because no one medium or culture conditionis known to support the growth of all halophilic bacteria, thegeneral-purpose, peptone-based medium was employed forboth enrichments and direct plate streaking. Both fast- andslow-forming colonies were selected for eventual study.Strains which subsequently demonstated poor growth onCM were maintained on SG or CM-B with an appropriatecarbon source. All other strains were maintained on CM.

Table 1 summarizes some of the characteristics of the 21isolates. There are nine strains of extremely halophilic rods(Halobacterium spp.); six strains of extremely halophilicbox-shaped bacteria (Haloarcula spp.; strains GN-8 and

VOL. 48, 1984

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

TABLE 1. Some characteristics of the isolated strains of halophilic bacteria

Strain Morphology Dimensions (p>m) Motility Gas vacuoles" Source brine density bR" Colony color(o BO)GN-1' Box (pleomorphic) 2-3 by 2-3 by 0.5 - - 22.6 - Red-pinkGN-2" Rod 0.5 by 3-9 + + 22.6 + Red-pinkGN-3 Rod 0.5 by 3-12 + + 27.7 + RedGN-4` Straight and curved rods 0.5 by 3 + - 22.6 - WhiteGN-5 Rod 0.5 by 2-6 + - 32.1 - Red-pinkGN-6 Rod 0.5 by 2-5 + - 32.1 - Pale pinkGN-8 Pleomorphic, box? 1-3 by 1-3 by 0.5 - - 32.1 - OrangeGN-9 Pleomorphic, box? 2-3 by 2-6 by 0.5 + - 17.9 ND Red-orangeGNM-3 Rod 0.5 by 6-10 - + ND + Red-pinkWS-4 Rod; coccus on agar 0.5 by 1-3 + + 26.9 - Pale redWS-5 Straight rods 0.5 by 2-3 + - 20.5 - WhiteWS-6 Rod; coccus in "packets" 0.5 by 1-5 - +? 26.9 - Red-pinkLS-It' Pleomorphic coccus 0.5-1 - - 28.7 - CreamLS-2 Pleomorphic coccus 0.5-1 - - 28.7 - Pink

Gas vacuoles were only observed in liquid, anaerobic cultures.Bacteriorhodopsin (bR) was determined in cells grown for several days aerobically and then for several days semi-anaerobically. ND, Not done.GN-7 (from 28.7° Be), GNM-1. and WS-1 (from 26.9° Be) were similar to GN-1. A similar strain from the GN saltern had bacteriorhodopsin (10). WS is the

same saltern referred to as CV in reference 10.d GNM-2 and WS-2 (from 26.9° Be) were similar to GN-2.eWS-3 (from 20.5° Be) was similar to GN-4.LS-3 (from 28.7° BO) was similar to LS-1.

GN-9 are questionable); three strains of extremely halophil-ic, irregular cocci (presumably Halococcus spp. or Halobac-terium volcanii); and three strains of moderate halophiles(possibly Vibrio costicola).

All the strains were gram negative, catalase positive, andoxidase positive. Colonies were circular, convex, entire, andsmooth. All strains produced opaque colonies except GN-2,GN-3, GN-4, GNM-2, GNM-3, WS-2, and WS-3, whichproduced translucent colonies. All the strains except GN-9,WS-4, and LS-2 were positive for arginine dihydrolase,ornithine decarboxylase, and lysine decarboxylase (deter-mined after 11 days of anaerobic growth). All strains grewanaerobically on nitrate, and all but Haloarcula spp. grewanaerobically on nitrite, although no gas formation was

observed. The moderate halophiles grew better anaerobical-ly on nitrite than on nitrate (determined visually). All theextreme halophiles contained bacterioruberin, as demon-strated by the spectra of lysates. The spectra of the lysates ofthe moderate halophiles had a single peak at 406 nm. Severalof the strains produced bacteriorhodopsin.The separation of these strains into extreme halophiles

(archaebacteria) and moderate halophiles (eubacteria) was

determined by lipid analysis. The separation into archaebac-terial and eubacterial strains was confirmed by their growthin the presence of antibiotics: extreme halophiles grew in thepresence of 500 U of penicillin G or 300 U of polymyxin Bper ml, whereas the moderate halophiles could not. Extremehalophiles could not grow in the presence of 1.4 U ofbacitracin per ml, whereas the moderate halophiles did.Strain GN-8 was exceptional in that it could grow in thepresence of 1,500 U of polymyxin B per ml (none of the otherextreme halophiles could) and in the presence of both 1.4and 3.4 U of bacitracin per ml.

All the strains were screened for their ability to growaerobically at the expense of a variety of sugars, sugar

alcohols, and low-molecular-weight acids added to CM-B(Table 2). A comparison was made with seven known strainsof Halobacterium. None of the bacteria could grow on 1%ethanol. Several of the strains were tested for their ability togrow on formate, but none could. Propionate did not supportthe growth of any of the strains in these screening assays.

However, after 3 weeks on MIN plus 1% propionate,Halobacterium marismortui demonstated growth. In all cas-es, growth on sugars resulted in the slight acidification of themedium (several 10ths of a pH unit), and growth on organicacids resulted in a slight alkalinization.

Several trends were noted among the rods. Strains with avery limited ability to attack carbohydrates included GN-2,GNM-2, GNM-3, and WS-2. These strains resemble theHalobacterium halobium-Halobacterium cutirubrum-Halo-bacterium salinarium group. They were also among the fewstrains that synthesized bacteriorhodopsin. However, theGNM strains were less limited in their ability to grow at theexpense of low-molecular-weight acids. GNM-2 was the onlyisolate that grew on 0.5% DNA. Because it could not growon ribose, these results suggest it grew at the expense ofpurines, pyrimidines, or both. Strain WS-4, which could notgrow on sugars or sugar alcohols, grew best in the presenceof pyruvate or acetate. It also grew on MIN supplementedwith 1% acetate. It could barely thrive on peptone (CM) oryeast extract plus Casamino Acids (SG). A more completetaxonomic description of this strain is in preparation.The other rods demonstrated the ability to metabolize

some or all of the carbohydrates and many of the low-molecular-weight acids listed in Table 2. There is not enoughconsistency between the characteristics of any of theseisolates (GN-3, GN-5, GN-6, and WS-6) and those of any ofthe other known strains of Halobacterium in the table tomake this type of screening of taxonomic value. Strain WS-6is unusual because it usually grew in clumps, even when itwas well agitated. In the clumps, the cells tended to bealmost coccoid and encased in multicellular sheathed pack-ets (this made all screening assays difficult). WS-6 is alsounusual in that it was the only rod that may have grown atthe expense of glycolate.Among the box-shaped bacteria, strains GN-1, GN-7,

GNM-1, and WS-1 thrived at the expense of nearly all thesugars, sugar alcohols, and low-molecular-weight acids list-ed in Table 2. WS-1 differed from the other strains in itsability to attack lactate and its inability to metabolize a-methyl-D-glucoside (an analog of glucose used in phosphatetransport system studies). GN-8 was characterized by rela-

354 JAVOR

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

VOL.48,1984~~~~~GROWTH POTENTIAL OF HALOPHILIC BACTERIA 355

TABLE 2. Carbon sources utilized by bacterial isolates

Growth of culture:"

Substrate -or medium a

Z Z z Z Z Z Z Z Z ZA c

D-Glucose e 0 e m e e e 0 e m 0 e 0 0 e e e e 0 0 0 m m m m e e eD-Fructose e 0 e m e e e 0 e 0 0 e 0 0 0 e m m 0 0 0 m m m m 0 0 0D-Galactose e 0 e m e e 0 0 e 0 0 m 0 0 m e m 0 0 0 0 m m m m 0 0 0oa-MGb m 0 0 m 0 m m 0 m 0 0 0 0 0 0 m m 0 0 0 0 m m 0 m 0 0 0Sucrose e 0 m m e e e 0 e 0 0 e 0 0 m e e e 0 0 0 m m m e m m mD-Xylose e 0 0 0 e e e 0 e 0 0 m 0 0 0 e e 0 0 0 0 m m 0 0 0 0 0D-Ribose e 0 0 0 e e 0 0 e 0 0 m 0 0 0 m 0 0 0 0 0 e m 0 0 0 0 0Glycerol e e m e e e e m e e m e e 0 0 e e e e m 0 e m e m e e eMannitol m 0 0 0 0 m 0 0 m 0 0 m 0 0 0 0 0 0 0 0 0 m 0 0 0 0 0 0Sorbitol mO0 0 00 mO 0 mO0 0 mO0 0 mO 00 0 00 mO m m0 0 0Pyruvate m m m e e m e e m m m m 0 e m e e e m 0 0 e m m m e e eAcetate m 0 m 0 e m e 0 m m 0 m 0 e 0 e 0 e 0 0 0 e m 0 0 e e eCitrate m 0 0 0 0 m 0 0 m 0 0 m 0 0 0 0 0 0 0 0 0 m 0 0 0 0 0 0Succinate e 0 m 0 e m e 0 m m m m m 0 0 e e e 0 0 0 e m 0 0 0 0 0Lactate 0 0 0 e 0 0 m 0 0 0 m m 0 0 0 e e e 0 m 0 0 0 0 0 e 0 mcm e e m m 0 e 0 m e m m e e 0 0 e 0 0 e 0 e m m m 0 e e eSG 0 0 e 0 0 mO0 m eMIN + ± + --+ - - + - - + - - +.. .. + + ± +--

glycerolMIN + + -+ --+ - - + - - + - - +.. . . + + +----

arginineMIN + +

acetatea Cells were grown aerobically. 0. No growth above that of the control (CM-B without an additional carbon source): m, moderate growth (0D750 20 to 200%

greater than that of the control): e, excellent growth (OD750) greater than 200% that of the control): ±. turbidity seen within 1 month: and -. no turbidity seen with-in 1 month. For the moderate halophiles, media contained 15% NaCI.

bocs-Methyl-D-glucoside.

tively low yields and extremely poor growth on CM and SG.GN-9 was characterized by very slow growth rates (up to1-month for colonies to appear on plates) and low yields.SG produced better growth than did CM. Because strainGN-9 grew so poorly under all the conditions tried, some ofthe other growth tests were not done with it.Three strains of extremely halophilic cocci (LS-1, LS-2,

and LS-3) were isolated from the La Salina environment.These organisms tended to be irregular in shape, often likeflakes. No coccus was ever detected in enrichments or fromdirect streaking of brines from the other hypersaline environ-ments. No rods or box-shaped bacteria were seen in themicroscopic examination of the La Salina brines, on platesfrom direct streaking, or after enrichments of these brines.There is considerable diversity among the three strains. LS-1grew on nearly all the carbohydrates and low-molecular-weight acids listed in Table 2. It grew well on peptone. LS-2grew on most of the substrates listed in Table 2. It grewpoorly on peptone alone, but well on SG. LS-3 metabolizedmany of the carbohydrates and most of the low-molecular-weight acids, but it grew poorly on both CM and SG. LS-3was unique among the cocci in its ability to metabolizeglycolate. The LS isolates showed little lysis in distilledwater.With two exceptions, the three strains of moderate halo-

philes (GN-4, WS-3, and WS-5) had similar characteristics.

Of the substrates listed in Table 2, they could only grow onglucose, sucrose, glycerol, pyruvate, acetate, lactate (exceptWS-3), and peptone. GN-4 is the only moderately halophilicisolate that grew in minimal medium with 1% glycerol. Themoderate halophiles all grew in CM-B with 1% choline,whereas none of the extreme halophiles could. These areprobably strains of V. costicola (24).The results of one set of experiments to demonstrate

fermentative growth of extreme halophiles under anaerobicconditions are presented in Table 3. Many strains grewfermentatively on CM, carbohydrates, pyruvate, or combi-nations of these. Most strains that had somewhat bettergrowth on glycerol than on glucose under aerobic conditionshad somewhat better growth on glucose than on glycerolunder anaerobic conditions. The Halobacterium halobium-like strains (GN-2, GNM-2, GNM-3, and WS-2) grew wellanaerobically on SG. Strain WS-4, which grew well aerobi-cally on acetate and pyruvate, failed to show significantanaerobic growth on these substrates. In another experi-ment, strains GN-3, LS-1, LS-3, and Halobacterium volcaniidemonstrated significant anaerobic growth on acetate. Otherstrains that grew on acetate aerobically (GN-1 and GN-6) didnot grow on it anaerobically. Fermentative growth in thelight was similar to that in the dark. Yields after 6 days ofgrowth were much lower than typical yields during aerobicgrowth after 2 to 4 days at the same temperature (see the

VOL. 48, 1984

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

TABLE 3. Growth yields of bacterial isolates by anaerobic fermentation"Growth yield from culture:

Carbon .Z z

source XXA

000000 0 0 0 0 0 ~~~~~~~~~~ ~ ~ -~~ -~~~ - Q -

Control 0.068 0.030 0.065 0.040 0.019 0.052 0.009 0.006 0.066 0.057 0.038 0.065 0.030 0.060 0.036 0.061 0.008 0.027 0.096 0.106 0.069 0.094Glucose 0.105 ND 0.120 0.112 0.031 0.067 0.039 ND 0.092 0.051 ND 0.107 ND ND 0.065 0.093 0.040 0.049 ND 0.143 0.071 0.194Fructose 0.071 ND 0.051 0.067 0.019 0.016 0.030 ND 0.021 0.021 ND 0.069 ND ND ND 0.115 0.022 0.024 ND 0.146 0.041 0.101Glycerol 0.087 0.053 0.101 0.105 0.042 0.047 0.050 0.006 0.075 0.068 0.024 0.077 0.021 ND ND 0.126 0.038 0.078 0.110 0.157 0.077 0.146Pyruvate 0.081 0.042 0.096 0.098 0.046 0.064 0.047 0.031 0.079 0.076 0.046 0.084 0.041 0.071 0.138 0.102 0.061 0.090 0.109 0.155 0.077 0.154SG i0.068 1.ll9 0.057 0.074 0.o32 0.062 0.025 0.014.0054 0.171 0.101 0.057 0.123 0.038 0.08010.75 0.040 0.043 0.2501.0.73 0.0610.216

a OD750 of cells grown for 6 days under conditions described in the text.b OD750 of cells grown in CM-B plus acetate was 0.054.

ordinates of Fig. 3-6 for comparison). When the experimentwas repeated with 12 days of incubation, cell densities wereapproximately twice those achieved after 6 days. The resultsof the experiments to show fermentative growth on thesesubstrates under anaerobic conditions in liquid medium wereconfirmed by growth tests on solid medium in a Gas-Paksystem. In addition to these tests, most strains grew anaero-bically on arginine, ornithine, and lysine in M0ller broth.For a study of the distribution of halophilic bacteria in

nature, the isolates were grown in Exportadora de Sal brines

E

supplemented with peptone or glycerol and peptone. Theionic analyses of the brines are presented in Fig. 1. None ofthe bacteria had optimal growth rates in brines saturatedwith NaCl (.25.50 Be). Most of the rods and box-shapedbacteria showed the greatest growth potential in ca. 21 to 220Bd brines (3.0 to 3.5 M Na+, 0.42 to 0.45 M Mg2+, 4.2 to 4.5M Cl-, 0.16 M S042-; Fig. 2a). Strain GN-5 had a widerrange of growth potential between ca. 17 and 220 Be (Fig.2b). Strains GN-6 and GN-8 had a growth potential similar tothat of the coccoid strains, with maximum yields in lo arith-mic phase in ca. 180 Be brines (2.5 M Na+, 0.33 M Mg +, 3.3M Cl-, 0.15 M S042-; Fig. 2c). Strain LS-1 had a goodgrowth rate in 130 Be brine (1.9 M Na+, 0.25 M Mg2+, 2.2 MCl-, 0.12 M S04-), but strains LS-2 and LS-3 did not. Forcomparison, several strains of known species of Halobacte-rium were tested for this medium. Halobacterium marismor-tui, Halobacterium cutirubrum, and Halobacterium salinar-ium had a 21 to 220 Be optimum, whereas Halobacterium

0

0

0IC)

00

BRINE DENSITY (°BeI)FIG. 1. Concentrations of major ions in the Exportadora de Sal,

S.A., saltern, 13 September 1982. Each curve is defined by theanalysis of 26 different brines in the range of 11.1 to 34.5O B6. Thecurves were extrapolated to normal seawater concentration. (A)Na+, Mg2+, K+, Ca2+, Cl-, and S042-. (B) Sr2+.

300°0Be

FIG. 2. Growth potential of halophilic bacterial isolates in sal-tern brines plus 1% peptone. Open circles represent repeatedexperiments.

356 JAVOR

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

GROWTH POTENTIAL OF HALOPHILIC BACTERIA 357

0I)

0

0.3r

0.21

0.1

0.2

0.1

M Na(

C.LS-

0.3

0.2

0.1

3 4 5M Na(

FIG. 3. Growth potential of halopwith different concentrations of NaCl.

volcanii had an 180 Be optimum.had optimal growth potentials in 13The strains were tested for groN

to NaCl concentration in SG. Mosigreatest growth rate in -3.5 M DGN-3 in Fig. 3a. Strains GN-6,optimally in 3.0 to 3.5 M NaCl (Fbroad tolerance of NaCl (Fig. 3c),LS-3 showed the .3.5 M NaClhalophiles grew best in SG with 2.To determine whether any o

detrimental to the survival of thetrated natural brines, several straitenriched Exportadora de Sal brine

- b. Na+,Mg2' , or K+ were added in concentrations found inbrines up to ca. 280 Be (the addition of sulfate salts caused aprecipitate). A comparison of four strains grown in 16.20 BeGN-6 brines (2.2 M Na+, 0.29 M Mg2+, 2.8 M Cl-, 0.14 M S042-)

//=WS-4 is shown in Fig. 4. Strain GN-2 thrived well in high-NaCl0/o \ brines, but it did rather poorly in Mg2+-enriched brines. KCI

slightly enhanced growth potential. Strain GNM-3 appeared3 4 5 to require high Mg2+ concentrations for greatest growth

Cl rates. Strain LS-1 was inhibited nearly equally by the highNa+ and Mg2+ concentrations found in 280 Be brines. It was

d. slightly inhibited by the higher KCI concentrations. TheWS-3 moderate halophiles had a response similar to that of LS-1 to

increased Na+, Mg2+, and K+ concentrations.The effects of Mg2+ as chloride and sulfate salts were

tested on several strains grown in CM with 3.0 M NaCl.3 4 5 Among the extreme halophiles, chloride was better tolerated

Cl than sulfate, although the opposite was true for the threemoderate halophiles. Among the box-shaped bacteria tested

hilic bacterial isolates in SG (strains GN-1, GN-7, and GNM-1), the best growth potentialwas recorded in medium with 0.3 to 0.5 M MgCl2 or 0.2 to0.3 M MgSO4 (Fig. 5a). Strain GN-2, a rod, showed optimal

The moderate halophiles growth potential in medium with -0.9 M MgCI2 (Fig. Sb),to 150 Be brines (Fig. 2d). whereas strains GNM-2 and GNM-3 thrived best in 0.7 Mwth potential with respect MgCl2 (data not shown). There was little change in growthtof the strains showed the potential in these three strains, as MgSO4 concentration4aCl, as shown for strain varied from that of the control medium (81 mM) to nearly 0.9GN-8, and WS-4 grew M. Another rod, strain GN-3, grew optimally in medium

ig. 3b). Strain LS-1 had a with .81 mM Mg (Fig. Sc). The LS strains responded in aalthough strains LS-2 and manner similar to that of GN-3 with respect to MgCl2 andoptimum. The moderate MgSO4 concentrations (Fig. 5d). The moderate halophiles5 to 3.5 M NaCl (Fig. 3d). also showed maximal growth potential in medium with .81,ne cation is particularly mM Mg (Fig. Se).bacteria in highly concen- For all the rod- and box-shaped extremely halophilicns were grown in peptone- bacteria tested, an increase in KCl concentration from thats to which chloride salts of of the control medium (27 mM) to 277 mM was accompanied

0.4 aF G -l 0.4 bF GNM-3

0.2 F

3.0 4.0M NaCI

0.5 !.0 1.5M MgCI2

100 200mM KCI

0In0r0

_

0.2 F

3.0M NaCl

4.0

I

0.5 1.0 1.5M MgC12

00 200

0.4rmM KCI

0.2k

C.

LS-II

3.0 4.0M NaCI

0.5 1.0 1.5M MgC12

100 200mM KCI

3.0 4.0M NaCI

II

0.5 1.0 1.5M MgC12

100 200mM KCI

FIG. 4. Growth potential of halophilic bacterial isolates in peptone-enriched saltern brines to which different concentrations of NaCl (@),MgCl2 (O), or KCI (A) were added.

0

o0

0.6

0.4

0.2

VOL. 48, 1984

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

0.2_

0.5 2+M Mg2

a.GN-1

0.4

0.2~

1.0

- bGN -2

0.4F

02F

0.5 2 1.0

M Mg

C.GN-3

0.5 2+M Mg2

1.0

0.1 0.2 0.3 0 0.1 0.2 0.3 0 0.1 0.2 0.3M K+ M K+ M K+

0.8 r

d.LS-1

l-1 0.5 IDM Mg2+

0.3 _a ffit

0.2 -e.

o.- GN-4

0.5 1.0M Mg2+

.

0 0.1 0.2 0.3 0 0.1 0.2 0.3MK+ MK+

FIG. 5. Growth potential of halophilic bacterial isolates in CM with 3.0 M NaCI to which different concentrations of MgSO4 (0), MgCI2(O), or KCI (A) were added.

by a slight increase in potential growth rate (Fig. 5). In thehalococcus LS-1 and in all three moderate halophiles,growth potential remained the same for this range of KCIconcentrations. Strain LS-2 was inhibited by KCI concentra-tions greater than 0.1 M (data not shown). Strain LS-3 wasnot tested.

DISCUSSION

Heterotrophy: general conclusions. Among the extremelyhalophilic bacterial isolates, several major groups can berecognized according to the substrates they metabolize:strains that can oxidize a broad range of substrates, andstrains that are largely confined to the oxidation of aminoacids, carbohydrates, or pyruvate and acetate. Both theExportadora de Sal and Western Salt solar salt ponds harbormixed populations of the versatile and "specialist" strains ofbacteria. Among the three genera of halobacteria, onlymembers of the genus Halobacterium have thus far demon-strated narrow substrate specificity. It remains to be investi-gated whether there are strains of extreme halophiles thatspecialize in attacking other classes of organic compoundssuch as lipids, aromatics, and Cl compounds.Low-molecular-weight carboxylic acids, particularly the

products of glycolysis and the tricarboxylic acid cycle, havegenerally been excluded in tests of substrate utilization byextremely halophilic bacteria. Pyruvate supported thegrowth of nearly all the extreme halophiles, and in strainWS-4, it produced the most luxuriant growth of all thesubstrates tested. The box-shaped bacteria appeared to bethe most versatile in metabolizing the four other carboxylicacids listed in Table 2. It is noteworthy that propionateactually inhibited growth of most of the halobacteria, al-though it has been shown to stimulate CO2 fixation (3, 13).The inability of some strains to metabolize the substrates

tested may be due to an inability to take up the substratefrom the medium, or there may be an enzymatic block in the

degradative pathway. However, all the enzymes of thetricarboxylic acid cycle have been demonstrated in halobac-teria (1). Strains such as GN-8 and LS-2, which grew well onxylose but not at all on ribose, may lack a single isomerase orepimerase.

The taxonomic description of the genus Halococclus (11)states that Halococcus morrhiuae cannot grow at the ex-

pense of carbohydrates. The authors used a medium inwhich glucose was autoclaved with the salts and yeast

extract. Heat sterilization of glucose with the medium makesit unsuitable for bacterial growth (unpublished data). Theresults with the coccoid strains of this study suggest thatthey can grow on a wide range of substrates, although strainsLS-2 and LS-3 grew poorly on amino acid-based medium(CM). A more rigorous appraisal of strains of Halococcusmorrhiuae would clarify the metabolic diversity found in thisspecies. One or more of the coccoid LS strains might requirea new species assignment in this genus.

In the ionic environments in which the moderate halo-philes coexist with the extreme halophiles, the two types ofmicroorganisms apparently can compete for many of thesame organic substrates. It is noteworthy that the Vibriostrains could grow at the expense of choline and none of thehalobacteria could. Rafaeli-Eshkol (15) reported that anunidentified moderate halophile oxidized choline to betaineand that choline acted as a protective substance in saltresistance. The ability of moderate halophiles to attackcholine may be a useful tool for enrichments of thesemicroorganisms in extremely hypersaline environments.

Anaerobic growth. The description of the genera Halobac-terium and Halococcus in Bergey's Maniual of Determina-tive Bacteriology (5) states that fermentation is never foundin organisms of these genera. Tomlinson and Hochstein (21)demonstrated glucose fermentation in Halobacterium sac-

charovorum. Since then, several other species have provedto be capable of carbohydrate fermentation in the presenceof air. Additionally, Hartmann et al. (8) showed that Halo-

oO

I.-

a0

0.6 _

0.4

0.2

0

358 JAVOR

0.4A- .

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

GROWTH POTENTIAL OF HALOPHILIC BACTERIA 359

bacterium halobium grew fermentatively at the expense ofarginine under anaerobic conditions.

Several general conclusions may be drawn about fermen-tative growth under anaerobic conditions by the isolates inthis investigation: basal medium (CM-B with no addedsubstrates) supported at least some growth in nearly all theisolates; some strains could grow relatively well on one ormore of the substrates that supported good growth underaerobic conditions; and all strains had much slower orpoorer growth than under aerobic conditions.Many strains that grew fermentatively on carbohydrates in

the presence of air (determined by the acidification of themedium) grew especially well on glucose, totally by sub-strate-level phosphorylation in the absence of air. Tomlinsonand Hochstein (21) demonstrated that 02 consumption byHalobacterium saccharovorum in the presence of glucosewas 18% of the theoretical amount required for its completeoxidation. In the presence of galactose, it was 83%. It ispossible that fructose, like galactose, requires more oxygenfor its oxidation and therefore was generally a poorersubstrate for fermentation. The measurement of 0, con-sumption, growth rates, and fermentation products in thepresence of glycerol, pyruvate, and acetate by strains thatdemonstrated relatively good anaerobic growth on thesesubstrates would provide further evidence of the importanceof fermentative metabolism in extreme halophiles.Although all the isolates grew anaerobically in the pres-

ence of nitrate, growth may have occurred by fermentationrather than by nitate reduction. Assays for nitrate reductasewould confirm that the strains can grow by dissimilatorynitrate reduction.

Effects of salts. Although halobacteria are found in thegreatest numbers in saltern brines of the density range of ca.27 to 300 Be (9), they can barely grow in these elevatedsalinities under culture conditions. These results suggest oneor more of the following: (i) the bacteria may be passivelyconcentrated by brine evaporation, (ii) their distribution inlower salinities may be limited by brine shrimp grazing or bycompetition from other bacteria, or (iii) growth studies underlaboratory conditions may not be a good reflection ofbacterial activity in nature. This question will be at leastpartially answered by uptake studies in natural populations.The Halococclus strains demonstrated salt requirements

intermediate between those of Halobacteriium and Haloar-cula and those of the moderate halophiles. Of the threecoccoid strains, LS-1 most closely resembled Halobacte-riium volcanii in morphology, paucity of carotenoids, sub-strates metabolized, salt tolerance and optima, and sub-strates that stimulated and inhibited CO2 fixation (B. Javor,submitted for publication). Because the epithet bacterium isusually reserved for rods, Halobacterium volcanii might bebetter classified as Halococcius volcanii. Clarification of thetaxonomic status of this species requires further biochemicaland metabolic comparisons with known strains of Halococ-cus.The abundance of extremely halophilic cocci in the La

Salina slough, a body of water whose salinity is that ofseawater at times, is explained by the ability of thesebacteria to remain viable in dilute solutions (16, 18). BothHalobacterium and Haloarcula lyse in dilute solutions. Thereason for the lack of Halococcus in the three other environ-ments sampled is not clear. The inability to synthesizebacteriorhodopsin may be a partial explanation. An ecologi-cal study of a solar salt environment where Halococciuscompetes successfully with Halobacteriuim and Haloarcuilamay provide an answer.

The natural brines in which the extreme halophiles demon-strated the greatest growth potential (18 to 220 Be) supportedfair growth in the moderate halophiles. Only in the NaCl-saturated brines (.25.50 Bd) did the moderate halophiles failto demonstrate appreciable growth. Rodriguez-Valera et al.(17) found similar salinity overlaps and limits by plating outsaltern bacteria on artificial medium. Thus, although theNaCl-crystallizing ponds of solar salterns are suboptimalenvironments for halobacteria, they are entirely hostile forthe eubacteria.

Studies on the Na+, Mg2+, and K+ requirements andsensitivities of the isolates showed that both Na+ and Mg2+concentrations found in natural brines of up to ca. 280 Becould affect the distribution of individual strains but that K+did not appear to be especially stimulatory or inhibitory.However, it was difficult to separate the effects of Cl- ongrowth potential. Strains such as GNM-3 demonstrated anextremely high Mg2+ requirement for optimal growth, acharacteristic previously described only in the Dead Seastrain Halobacterium sodomense (14). Most other strains ofHalobacterium require 0.1 to 0.5 M Mg, and some aretolerant of higher concentrations (12).

It is noteworthy that chloride salts of magnesium wererelatively more inhibitory to the moderate halophiles thanwere sulfate salts, in contrast to the response of the extremehalophiles. Sulfate has a very low activity coefficient inconcentrated brines (4). It is possible that MgSO4 is lessinhibitory than MgCIl only because it is relatively lessdissociated. Alternatively, the true cell wall of the moderatehalophiles may be less permeable to the divalent sulfate ionthan to the monovalent chloride ion.

It is difficult to assess the relative importance that each ionplays in preventing halobacteria from thriving in the bitternsbrines. Javor (9) suggested that the low water potential ofsuch concentrated brines may be the actual limiting factor.The study of these bacteria with reference to the chemistryof their natural habitats should stimulate future investiga-tions concerning such phenomena as divalent cation trans-port and regulation and physiological activity at differentwater potentials.

In an evaluation of the ecological importance of the resultsof this study, the following conclusions should be consid-ered: (i) in the saltern environments, a fairly wide diversityof species and strains was present (although their respectiveroles in natural populations were not determined); (ii) all thehalobacteria isolated were facultative anaerobes; (iii) classi-cal screening methods (carbon source utilization) can onlybroadly define species of halobacteria; (iv) halobacteria grewpoorly in the concentrated brines which harbored the great-est density of these microorganisms, suggesting that theywere passively concentrated by brine evaporation; and (v)Halococcus was only isolated from a euryhaline environ-ment, and conversely, Halobacterium and Haloarcula wereonly isolated from stenohaline environments. The genusHalococcus needs taxonomic reassessment.To more fully develop the picture of microbial processes

in hypersaline environments, future studies should addressthe nature of the organic chemistry of natural brines, micro-bial metabolic and turnover rates in native environments,and the isolation and culture of halophilic bacteria underselective conditions.

ACKNOWLEDGMENTSI thank Monica Armstrong for technical assistance and the

management of Exportadora de Sal, S.A., and Western Salt Co. fortheir cooperation.

VOL. 48, 1984

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

This work was supported by National Science Foundation grantPCM-8116330 and Petroleum Research Fund grant PRF 13704-AC2.

LITERATURE CITED

1. Aitken, D. M., and A. D. Brown. 1969. Citrate and glyoxylatecycles in the halophil, Halobacterium cutiriubruim. Biochim.Biophys. Acta. 177:351-354.

2. Castenholz, R. W. 1970. Laboratory culture of thermophiliccyanophytes. Schweiz. Z. Hydrol. 32:538-551.

3. Danon, A., and S. R. Caplan. 1977. CO, fixation by Halobacte-rium halobium. FEBS Lett. 74:255-258.

4. Edgerton, M. E., and P. Brimblecombe. 1981. Thermodynamicsof halobacterial environments. Can. J. Microbiol. 27:899-909.

5. Gibbons, N. E. 1974. Halobacteriaceae, p. 269-273. In R. E.Buchanan and N. E. Gibbons (ed.), Bergey's manual of deter-minative bacteriology, 8th ed. The Williams & Wilkins Co.,Baltimore.

6. Gonzalez, C., C. Gutierrez, and C. Ramirez. 1978. Halobacte-rium vallismortis sp. nov. An amylolytic and carbohydratemetabolising extremely halophilic bacterium. Can. J. Microbiol.24:710-715.

7. Harrigan, W. F., and M. E. McCance. 1966. Laboratory meth-ods in microbiology. Academic Press, Inc., New York.

8. Hartmann, R., H.-D. Sickinger, and D. Oesterhelt. 1980. Anaer-obic growth of halobacteria. Proc. NatI. Acad. Sci. U.S.A.77:3821-3825.

9. Javor, B. J. 1983. Planktonic standing crop and nutrients in asaltern ecosystem. Limnol. Oceanogr. 28:153-159.

10. Javor, B., C. Requadt, and W. Stoeckenius. 1982. Box-shapedhalophilic bacteria. J. Bacteriol. 151:1532-1542.

11. Kocur, M., and W. Hodgkiss. 1973. Taxonomic status of thegenus Halococcus Schoop. Int. J. Syst. Bacteriol. 23:151-156.

12. Mullakhanbhai, M. F., and H. Larsen. 1975. Halobacteriumvolcanii spec. nov., a Dead Sea halobacterium with a moderatesalt requirement. Arch. Microbiol. 104:207-214.

13. Oren, A. 1983. Bacteriorhodopsin-mediated CO2 photoassimila-tion in the Dead Sea. Limnol. Oceanogr. 28:33-41.

14. Oren, A. 1983. Halobacterium sodomense, sp. nov., a Dead Seahalobacterium with an extremely high magnesium requirement.Int. J. Syst. Bacteriol. 33:381-386.

15. Rafaeli-Eshkol, D. 1968. Studies on halotolerance in a moderate-ly halophilic bacterium. Biochem. J. 109:679-685.

16. Rodriguez-Valera, F., F. Ruiz-Berraquero, and A. Ramos-Cor-menzana. 1979. Isolation of extreme halophiles from seawater.Appl. Environ. Microbiol. 38:164-165.

17. Rodriguez-Valera, F., F. Ruiz-Berraquero, and A. Ramos-Cor-menzana. 1981. Characteristics of the heterotrophic bacterialpopulations in hypersaline environments of different salt con-centrations. Microb. Ecol. 7:235-243.

18. Rodriguez-Valera, F., A. Ventosa, E. Quesada, and F. Ruiz-Berraquero. 1982. Some physiological features of a Halococcussp. at low salt concentrations. FEMS Microbiol. Lett. 15:249-252.

19. Ross, H. N. M., M. D. Collins, B. J. Tindall, and W. D. Grant.1981. A rapid procedure for the detection of archaebacteriallipids in halophilic bacteria. J. Gen. Microbiol. 123:75-80.

20. Sehgel, S., and N. E. Gibbons. 1960. Effect of some metal ionson the growth of Halobacterium cutirubrum. Can. J. Microbiol.6:165-169.

21. Tomlinson, G. A., and L. I. Hochstein. 1972. Studies on acidproduction during carbohydrate metabolism by extremely halo-philic bacteria. Can. J. Microbiol. 18:1973-1976.

22. Tomlinson, G. A., and L. I. Hochstein. 1976. Halobacteriumsaccharovorum sp. nov., a carbohydrate-metabolizing, ex-tremely halophilic bacterium. Can. J. Microbiol. 22:587-591.

23. Torsvik, R., and I. Dundas. 1982. The classification of halobac-teria. Methods Enzymol. 88:360-368.

24. Ventosa, A., E. Quesada, F. Rodriguez-Valera, F. Ruiz-Berra-quero, and A. Ramos-Cormenzana. 1982. Numerical taxonomyof moderately halophilic gram-negative rods. J. Gen. Microbiol.128:1959-1968.

360 JAVOR

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from