INTRODUCTION

Top Previous Next References

Compared to the extensive literature on the physiology, biochemistry, and ecology of the aerobic red halophilic archaea (family Halobacteriaceae), the aerobic halophilic bacteria have been relatively little studied. Research on the halophilic and halotolerant bacteria often seems to be less glamorous than the study of the archaea, with their unique adaptations, including a highly saline cytoplasm, specialized salt-requiring proteins, and the unique light-driven proton and chloride pumps bacteriorhodopsin and halorhodopsin (171). During early research on the microbiology of hypersaline environments, the halophilic bacteria were often neglected, even though they inhabit a wide range of habitats such as saline lakes, saltern ponds, desert and hypersaline soils, and salted foods, a range much less restricted than the habitats in which the halophilic archaea thrive (276, 284, 285).

However, Kushner (168) clearly states: "Though they are less exciting at first glance than the extreme halophiles the moderately halophilic bacteria, and solute-tolerant microorganisms in general, pose quite sufficiently interesting questions, especially those implied by their ability to grow over wide ranges of solute concentrations. Further work on these relatively little-studied microorganisms may be expected to bring dividends in the form of insight on the relation of internal and external solute concentrations, and on the state of cell-associated ions within the cytoplasm. If the last decade has been that of the extreme halophiles, we can hope that the next one will see their more modest, moderate cousins (in the spiritual sense only) take their proper place in the scientific canon." The moderately halophilic bacteria pose specific questions to the scientist, many of them related to their adaptability to a wide range of salinities. Thus, species such as Salinivibrio costicola and Halomonas halodenitrificans are able to grow over a range of water activities between 0.98 (close to freshwater) to 0.86 (close to saturated NaCl) (168). This in itself is a feat that may be much more difficult to achieve than the rigid, salt-requiring metabolism of the halophilic archaea, which lyse the moment the salt concentration in their environment drops below 10 to 15%.

The occurrence of nonpigmented halotolerant bacteria was probably first mentioned in 1919 by LeFevre and Round in their study of the microbiology of cucumber fermentation brines. One of the bacterial groups isolated grew in 0 to 15% NaCl, whereas other bacteria studied exhibited growth over the range of 5 to 25% (178). An early classic study of halophilic bacteria is the work of Hof (127), who inoculated salt mud from a solar salt facility on Java onto a variety of media of different salinities. In addition to red archaeal types, different types of white colonies were isolated, including endospore-containing Bacillus species able to grow at 24% NaCl. Using media containing between 12 and 18% salt, she isolated a Pseudomonas-type bacterium from salted beans preserved in brines varying in salt concentration from 6 to 29%. This organism, designated Pseudomonas beijerinckii, grew from 3 to 18% salt but not at 0.5%, showing its obligate halophilic character. To quote from Hof's paper: "it may be concluded that most of the important groups of bacteria are able to live in concentrations up to about 15% salt and that many groups are physiologically active even at much higher salt concentrations."

To describe microorganisms according to their behavior toward salt, different classification schemes have been devised. Although several classifications or categories have been proposed (274, 329, 351), the most widely used is that of Kushner, who defined moderate halophiles as organisms growing optimally between 0.5 and 2.5 M salt (168). Bacteria able to grow in the absence of salt as well as in the presence of relatively high salt concentrations (e.g., 8% in the case of Staphylococcus aureus) are designated halotolerant (or extremely halotolerant if growth extends above 2.5 M). A rare case of a bacterium that requires 2 M salt at least (optimal growth at 3.4 M), such as is exemplified by the actinomycete Actinopolyspora halophila (95), is considered a borderline extreme halophile (137, 168).

It should be pointed out that the salt requirement and tolerance of many species vary according to growth conditions such as temperature and medium composition. The growth temperature should be specified, especially for the definition of the lower salt range enabling growth. Thus, Marinococcus halophilus grows at NaCl concentrations as low as 0.01 M at 20°C but at least 0.5 M is required at 25°C (228). Similarly, S. costicola can grow between 0.5 and 4 M NaCl at 30°C but can grow down to 0.2 M at 20°C (168).

Moderately halophilic bacteria constitute a heterogeneous physiological group of microorganisms which belong to different genera. Our current knowledge of the taxonomic status of these bacteria contrasts with early studies, since in 1980 only six moderately halophilic species were included in the Approved Lists of Bacterial Names (312) and most strains used in physiological and biochemical studies were isolated from salted cured foods or unrefined salt or were even laboratory culture contaminants. These moderate halophiles were Vibrio (Salinivibrio) costicola (314), Micrococcus (Nesterenkonia) halobius (237), Paracoccus (Halomonas) halodenitrificans (159), Flavobacterium (Halomonas) halmephilum (60), Planococcus (Marinococcus) halophilus (228), and Spirochaeta halophila (103).

During the last decade, the extensive studies on hypersaline environments that have been carried out in many geographical areas have permitted the isolation and taxonomic characterization of a large number of moderately halophilic species. Thus, moderate halophiles are represented by several methanogenic archaea as well as strictly anaerobic bacteria that have been reviewed recently (182, 232) and are not included in this article. However, most species are gram-negative or gram-positive aerobic or facultatively anaerobic moderately halophilic bacteria (340, 341). Although some gram-negative species were considered members of different genera (Halomonas, Deleya, Volcaniella, Flavobacterium, Paracoccus, Pseudomonas, Halovibrio, or Chromobacterium), phenotypic and phylogenetic data support their close relationship, and they are currently included in the family Halomonadaceae as members of two genera: Halomonas and Chromohalobacter (53, 76). Table 1 shows the features that differentiate the validly published moderately halophilic species included in these two genera.

Besides members of the family Halomonadaceae, several other gram-negative strictly aerobic or facultatively anaerobic species have been described as moderate halophiles belonging to genera that include nonhalophilic species as well, such as Pseudomonas, Flavobacterium, or Spirochaeta, while others are placed in genera represented, at least until now, exclusively by halophilic species: Salinivibrio, Arhodomonas, or Dichotomicrobium. The differential features of these species are shown in Table 2.

Three organisms have been extensively used in physiological and biochemical studies dealing with the mechanisms of haloadaptation and adaptability: Salinivibrio costicola, Halomonas elongata, and Halomonas israelensis.

S. costicola was originally isolated from rib bones in a sample of Australian bacon (91, 314). Similar strains have been isolated from saltern ponds near Alicante (96) and other locations in Spain and the Canary Islands (91).

H. elongata was isolated from a solar salt facility on Bonaire, Netherlands Antilles. It is an extremely versatile organism, able to grow at a very wide range of salt concentrations. It can also grow anaerobically with nitrate as the electron acceptor, forming nitrite, and it has also been reported to grow fermentatively on glucose (357). However, glucose fermentation was not confirmed in later studies (83).

H. israelensis (previously designated strain Ba₁) was originally obtained from unrefined solar salt obtained from the Dead Sea (272, 273) and was only recently assigned a species name (128).

The gram-positive moderately halophilic aerobic bacteria, with the exception of two Bacillus species, belong to genera that include only species with halophilic requirements: the genera Halobacillus, Marinococcus, Salinicoccus, Nesterenkonia, and Tetragenococcus. The validly described species currently accepted as moderate halophiles and their characteristics are shown in Table 3. Finally, there are some moderately halophilic actinomycetes that have been recently isolated from different saline soil samples and have been characterized as species of the genera Actinopolyspora or Nocardiopsis. Table 4 shows the differential characteristics of these four moderately halophilic species.

Besides these moderately halophilic species that have been characterized taxonomically and, according to the rules of the code of nomenclature of bacteria, are considered validly described species, there are several other moderately halophilic aerobic bacteria that have been used for other purposes, including physiological, biochemical, and biotechnological studies, but have not been studied taxonomically in detail. Typical examples are "Pseudomonas halosaccharolytica" (120, 121) and "Micrococcus varians subsp. halophilus" (241).

Recent studies based on 16S rRNA sequence analysis have permitted a determination of the phylogenetic position of most moderately halophilic bacteria. During last decade, the old technique of comparison of 16S rRNA oligonucleotide catalogs showed that Spirochaeta halophila belongs to the spirochete phylum and that some Halomonas and Deleya species are members of the gamma subclass of the Proteobacteria (341). They were placed in a new family, Halomonadaceae (76). More recently, complete 16S rRNA sequence analysis confirmed that Spirochaeta halophila is within the Spirochaeta cluster of the spirochete phylum, related to Spirochaeta isovalerica, S. litoralis, S. bajacaliforniensis, and S. aurantia (average similarity, 87.4%) (253). In addition, several studies have identified the phylogenetic position of most gram-negative moderately halophilic aerobic species currently described. Members of the genera Halomonas, Deleya, Halovibrio, and Volcaniella, as well as Paracoccus halodenitrificans, form a monophyletic group within the gamma subclass of the Proteobacteria (53, 56, 200, 204, 290). The levels of 16S rRNA sequence similarity among these species ranged from 91.5 to 100%; although several subgroups, which might represent separate genera, were resolved, they could not be differentiated on the basis of phenotypic or chemotaxonomic features. For these reasons, Dobson and Franzmann (53) proposed placing all members of the above four genera and P. halodenitrificans in a single genus, the genus Halomonas, and emended the description of the family Halomonadaceae. This family now comprises the species of Halomonas and Zymobacter and the moderate halophile, originally isolated from the Dead Sea, Chromohalobacter marismortui (200). All have 15 signature characteristics in their 16S rRNA sequences, including a distinctive cytosine residue at position 486 (53). Arhodomonas aquaeolei represents a deeply branching lineage in the gamma subclass of the Proteobacteria, most closely related to purple sulfur bacteria (particularly species of the genera Ectothiorhodospira and Chromatium). The 16S rRNA sequence analysis supports the placement of this single species in a separate genus (4).

Very recently, the phylogenetic position of six (Salini)vibrio costicola strains revealed that this moderate halophile constitutes a monophyletic branch that is distinct from other Vibrio species and from other species belonging to the gamma subclass of the Proteobacteria (202). Since other phenotypic and genotypic data supported these differences, placement of this species in a separate genus, Salinivibrio, has been proposed (201).

The 16S rRNA sequences of Flavobacterium gondwanense and F. salegens, two moderate halophiles isolated from a hypersaline Antarctic lake, contain the definitive flavobacterial signatures that unequivocally place them in the Flavobacterium-Bacteroides phylum (52). These species cluster with a group of organisms that contains the type species of the genus Flavobacterium, F. aquatile (with 89 and 90% sequence similarity between them and this species) (52).

Recent studies have determined the phylogenetic relationships of moderate halophiles within the gram-positive branch. Farrow et al. (64) showed that Marinococcus halophilus (formerly Planococcus halophilus) forms a distinct line of descent and is only distantly related to the genera Planococcus, Sporosarcina, and Bacillus. The 16S rRNA sequence data confirm the placement of M. halophilus in a separate genus. In addition, they indicated that Sporosarcina halophila, an endospore-forming motile gram-positive moderate halophile, was not closely related to Sporosarcina ureae, the type species of this genus. Later studies permitted the placement of S. halophila in a new genus, Halobacillus, as H. halophilus, closely related to other moderately halophilic species isolated from the Great Salt Lake, H. litoralis and H. trueperi (315). This study also confirmed the placement of Salinicoccus roseus in a separate genus, since it constitutes a deep branch not closely related to other gram-positive bacteria. Bacillus salexigens is closely related to Bacillus pantothenticus, a species that belongs to phylogenetic group I of the genus Bacillus, as well as to Halobacillus halophilus, H. litoralis, and the halotolerant species Bacillus dipsosauri (89). Tetragenococcus muriaticus, a recently described species isolated from a traditionally fermented Japanese fish sauce, is closely related to the halotolerant species Tetragenococcus halophilus (formerly Pediococcus halophilus), which showed a closer phylogenetic relationship to other lactic acid bacteria of the enterococci and lactobacilli than to pediococci (304).

While all these moderately halophilic species belong to the low-G+C group of the gram-positive phylum, only Nesterenkonia halobia (formerly Micrococcus halobius) is within the high-G+C group (218, 316), constituting a cluster that is clearly separate from other species of the genera Micrococcus, Arthrobacter, and Kocuria.

The recent phylogenetic studies of moderate halophiles have been based on the comparison of 16S rRNA sequence data. These studies have been very helpful for improving the classification of moderate halophiles according to a natural (phylogenetic) approach. In particular, results show that they are represented in many of the major bacterial phyla: spirochetes, Proteobacteria, Flavobacterium-Bacteroides, and low-G+C and high-G+C gram-positive organisms. A recent study of African soda lakes showed a wide phylogenetic diversity within the alkaliphilic (and in some cases halophilic) isolates (58). As with the moderate halophiles, these authors concluded that the alkaliphile phenotype is also polyphyletic and might have evolved many times (58).

ECOLOGY

Top Previous Next References

Numerical Taxonomy Approaches to the Analysis of Natural Communities of Moderately Halophilic Bacteria

(i) Salt lakes and brines. The communities of moderately halophilic bacteria in thalassohaline (seawater-derived) hypersaline environments, such as saltern ponds for concentrating seawater, may to a large extent resemble the communities present in seawater. This is not too surprising, since many marine bacteria have a broad salt tolerance. It was even reported that the majority of 30 isolates of marine aerobic heterotrophic bacteria tested could grow at up to 20% NaCl and that some could even multiply in media containing 30% NaCl (72). Moderately halophilic bacteria could also be enriched from seawater by gradual salinity increases: when seawater was periodically amended with salt and nutrients, moderately halophilic bacteria outcompeted the slightly halophilic marine bacteria, which completely disappeared above 15% salt (347). A variety of halophilic bacteria were also isolated from sea sands and seaweeds (235). Thus, the sea contains many moderately halophilic or at least extremely halotolerant bacteria. In a study of Spanish saltern ponds of intermediate salinity (between 15 and 30% sea salts) (Alicante on the Mediterranean coast, Huelva on the Atlantic coast), the dominant types of colonies developing on agar plates were assigned by numerical taxonomy to the genera Salinivibrio, the Pseudomonas-Alteromonas-Alcaligenes group, Acinetobacter, and Flavobacterium (284-286). Most of these isolates should probably be reclassified in the family Halomonadaceae. Most isolates grew optimally at about 10% salts (the concentration that also yielded the largest number of CFU on agar plates) and could be found at salt concentrations up to about 25%. (Salinivibrio species dominated below 15% salt, while bacteria assigned to the Pseudomonas-Alteromonas-Alcaligenes group were especially abundant above 15%. Flavobacterium and Acinetobacter were found in smaller numbers and were evenly distributed up to 30%, while gram-positive cocci were found mostly above 25% salt (286). Most of the 140 isolates from the Huelva salterns clustered in eight phenons, two of which were identified as Halomonas (Deleya) (53), four resembled S. costicola, and the others were tentatively assigned to the genera Flavobacterium and Acinetobacter (187). Similar Acinetobacter-like bacteria (probably to be assigned to the Halomonadaceae) have been found in other hypersaline habitats (265).

(ii) Saline soils. The soil habitat is inherently inhomogeneous, and it can be expected that a wide range of salinities might be present in any one saline soil (101). Saline soils appear to yield mostly halotolerant rather than halophilic microorganisms, presumably reflecting adaptation to periodic episodes of relatively high dilution (267, 268). One early study stated that microorganisms may be unable to multiply in saline soil and that the microbiota of saline soil habitats are passive inhabitants brought by the wind. In this study, soils near the Red Sea, with a salt content varying from 25 to 30% on the surface to 1.5 to 2% at a depth of 50 cm, were examined. The highest bacterial counts were obtained in surface soil, using media without NaCl. The numbers obtained were much smaller than those commonly found in nonsaline soils. The authors stated that "the accepted assumption about the widespread distribution of salt-resistant and halophilic microorganisms in saline soils requires reconsideration" (116).

(iii) Cold saline habitats. Extensive microbiological studies in the Antarctic, especially the cold saline lakes in the Vestfold Hills region and the saline soils of the Dry Valleys, have contributed some interesting insights into the extreme conditions under which halophilic bacteria may occur and thrive.

The Vestfold Hills region is a coastal, ice-free area in east Antarctica, which contains in excess of 300 lakes and ponds. These are relics of seawater catchments isolated some 6,000 years ago by uplift and trapped in valleys and depressions. Some of these lakes are hypersaline; the most saline lakes have a total salt concentration of up to 28%. Salinity may show seasonal variations due to meltstream influx and ice cover melt in the Austral summer. The best studied is Organic Lake, a meromictic lake with a maximum depth of 7.5 m. The lake is stratified, with salt concentrations increasing from 0.8 to 21%, and is anoxic below a depth of 4 to 5 m. The ice cover excludes wind-induced turbulence throughout winter. Thermal profile and the increasing salinity with depth prevent turnover in the ice-free summer period. Temperatures range from 14 to +15°C (

(iv) Alkaline saline habitats. Stable alkaline hypersaline environments are not common and are the result of an unusual combination of geological, geographical, and climatic conditions (102). Most studies on the alkaline saline environments have concentrated on lakes such as Lake Magadi, Kenya, and the Wadi Natrun lakes, Egypt, which are dominated by extremely halophilic archaea.

The preponderance of nonspecialized heterotrophs among the known bacterial halophiles does not necessarily reflect their dominance in salt lakes, saline soils, and other saline habitats but, rather, may be due to the relative ease of culturing these bacteria (136). The halophilic property is probably widespread in the bacterial domain and may occur in a variety of morphological and physiological types. This is nicely illustrated by the studies of Hirsch, who differentiated 104 different morphotypes of bacteria in the hypersaline Solar Lake on the shore of the Sinai peninsula. This small and shallow (maximum depth, about 5 m) thalassohaline lake is stratified in winter, with an epilimnion containing 4.5 to 9% salt, increasing to about 19% at the bottom. In addition to salt stress, temperature may be an important selective factor, since in the upper hypolimnion heliothermal heating can increase the temperature to about 60°C in winter. During the summer season, the lake is mixed and hypersaline (18 to 19%). Morphologically diverse bacteria were observed by direct examination of samples, in enrichments, and in pure cultures; they included cocci, rods, apple-shaped budding bacteria, long flexible filaments, spindle-shaped bacteria, short pointed filaments, and branched hyphae (123). A budding prosthecate bacterium with branching hyphae was obtained in culture and described as Dichotomicrobium thermohalophilum (124). Similar strains have been isolated from a coastal saline lake in Brazil. The nearly tetrahedral mother cells produce up to four hyphae at the tips, on which nonmotile buds are formed. The isolates grow from 0.8-4 to 18-22% salt, depending on the strain, with an optimum at 8 to 14%, and are moderately thermophilic (growing at up to 52 to 65°C). Their metabolism is strictly aerobic, and organic acids are used as carbon and energy sources. However, these bacteria were also found at greater depths in Solar Lake and could be isolated from the anaerobic hypolimnion at a depth of 3.5 m (123, 124). Solar Lake was also the source from which the facultative aerobe Spirochaeta halophila was isolated (103).

Many of the aerobic moderately halophilic bacteria can use nitrate as an alternative electron acceptor. Thus, Halomonas elongata can grow anaerobically by reducing nitrate to nitrite (357). Other well-known nitrate reducers are H. halodenitrificans and Bacillus halodenitrificans. The latter was isolated from a solar saltern in the south of France by enrichment in medium containing 1.06 M NaNO₃. It grows at NaCl concentrations between 0.35 to 4.25 M (optimum, 0.5 to 1.35 M) and is unusually tolerant to nitrite: growth is possible in 0.58 M NaNO₂. Since nitrous oxide reductase is absent, N₂O is the sole product of nitrate and nitrite reduction (49).

Considerable diversity also exists with respect to the carbon and energy sources used. Hydrocarbons can be used up to quite high salinities. A series of enrichment cultures successfully produced mineral oil degraders by using water from the Great Salt Lake, Utah, as an inoculum at salinities up to 17.2%. Experiments with radiolabeled hexadecane showed decreasing degradation rates with increasing salinity (360).

Aromatic compounds such as benzoate may be degraded by versatile halophilic bacteria such as H. halodurans, which cleaves aromatic rings by ortho cleavage (290). Even more exotic compounds may be utilized at high salinities, as illustrated by strain JD6.5, an Alteromonas type of organism growing at 2 to 24% salt and degrading several highly toxic organophosphorus compounds (38). Another Halomonas isolate degraded formaldehyde and proved highly tolerant to high formaldehyde concentrations (8, 250, 251).

Many additional metabolic functions may exist within the highly diverse group delineated by the common denominator "aerobic, halophilic bacteria". Thus, the moderately halophilic Thiobacillus halophilus, isolated from a Western Australian hypersaline lake (Lake O'Grady North), grows at salt concentrations of up to 4 M. It is a chemoautotroph that oxidizes reduced sulfur compounds (365). Aerobic halophilic methylotrophic bacteria were described as well (57). Many species of cyanobacteria are also moderate halophiles. However, they will not be discussed within the framework of this review, which deals primarily with heterotrophs. Whether moderately halophilic counterparts exist for all types of bacterial metabolism that occur in freshwater and marine environments is still unknown. One function that seems to be missing in the heterotrophic bacterial communities at high salt concentrations is the ability to fix molecular nitrogen. This fact was already recognized by Hof (127), who pointed out that N₂ fixers could be isolated on 0 and 3% salt but not on 6% and higher. Large numbers of bacteria able to grow in a nitrogen-free medium could be isolated from saline soil (Granada, Spain), but nitrogenase activity was not detected (275).

It is evident that very few attempts have been made to isolate specialized and unusual types of halophilic aerobic bacteria. Therefore, one may assume that many more physiological types remain to be discovered.

A few attempts have been made to assess microbial activities in the Great Salt Lake, Utah, by incubating water samples from the less saline southern part (8.5% salt) and the more saline northern half (22% salt) with [¹⁴C]glucose, [¹⁴C]glycerol, or [¹⁴C]acetate and monitoring the appearance of the radioactive label as ¹⁴CO₂. Lowered rates of breakdown of the three substrates were found at the higher salinity (66), a finding that parallels the lowered dissimilation of [¹⁴C]hexadecane by Great Salt Lake water samples with increasing salinity (360).

Measurements of [methyl-³H]thymidine incorporation by heterotrophic communities in increasingly saline saltern ponds in Spain showed that the growth rate was highest between 5 and 10% salt. This was much higher than in the community of archaea present in ponds with salt concentrations exceeding 20% (104). A similar finding was reported from the Eilat saltern ponds at the coast of the Red Sea, where the estimated doubling times of the bacteria in ponds of low to intermediate salinity was 1.1 to 12 days, based on the thymidine incorporation rate (245).

In recent years, the characterization of 16S rRNA genes isolated directly from the environment has been used to obtain information on prokaryotic community structure. The technique has not yet been extensively applied to hypersaline environments. However, 16S rRNA genes were amplified by PCR from a saltern crystallizer pond in Spain. As expected, in view of the high salinity, archaeal sequences were recovered most frequently while bacteria represented only a minor component. One cluster of bacterial sequences showed about 82% identity to Rhodopseudomonas marina (alpha subclass of the Proteobacteria) (13, 14). To enable comparison of the prokaryotic communities in the salt concentration gradient presented by the salterns (preparation ponds of 6.4 and 9.2% salt, concentrator ponds of 13.3 and 21.6% salt, and an NaCl precipitation pond of 30.8% salt), fingerprinting of the community was performed with different restriction endonucleases. The highest similarities were found between the two concentrator ponds and between the two preparation ponds. The bacterial community decreased in complexity with increasing salinity. The bacterial genes isolated from the crystallizer pond showed little similarity to the genes isolated from the less saline ponds. Thus, it is improbable that the bacteria in the crystallizer pond represent only a carryover of inactive cells from the previous evaporation stage (13, 191).

The halophilic bacteria form a versatile group, adapted to life at the lower range of salinities and with the possibility of rapid adjustment to changes in the external salt concentration. In contrast, the halophilic archaea (family Halobacteriaceae) are generally found at higher salinities and their requirement for high salt concentrations for the maintenance of structural cell components makes them strictly dependent on the constant presence of high salt concentrations (3 to 4 M for most species). Accordingly, the two groups occupy different niches, which seem to overlap very little (247). When heterotrophic bacteria from Spanish saltern ponds of increasing salinity were enumerated on agar plates, only a narrow salinity range (25 to 32% total salt) was found to contain the two groups (288). A similar conclusion was reached in studies in which amino acid incorporation by the heterotrophic communities in saltern ponds in Eilat, Israel, was measured. When inhibitors specifically directed against the halophilic archaea (bile salts such as taurocholate or deoxycholate and the protein synthesis inhibitor anisomycin) or bacteria (chloramphenicol or erythromycin) were tested, the contribution of each group could be distinguished. Up to a salinity of about 25%, all amino acid incorporation activity was inhibited by the bacterial inhibitors, while above 25%, inhibitors known to act on the archaea completely inhibited all activity (243, 244, 246). Similarly, aphidicolin, an inhibitor of DNA replication in halophilic archaea, completely inhibited thymidine incorporation in saltern brines with salinities exceeding 25% (245, 246).

Competition between halophilic bacteria and archaea has also been studied in laboratory model systems. When samples from the subterranean saline well that supplies the saltern ponds of La Malá near Granada, Spain, were subjected to gradual salinity changes with additional nutrient enrichment, marine-type slightly halophilic bacteria, halotolerant types, moderate halophiles, and extremely halophilic archaea were enriched, depending on the final salt concentration achieved. Marine and moderately halophilic types were most abundant between 3 and 30% salt. At 25°C, hardly any development of extremely halophilic archaea was observed at the higher salinities, but at 35°C, dense communities of archaea were obtained at 35 and 40% salt, suggesting that temperature is an important factor in determining the outcome of the competition (40). The importance of temperature was confirmed in chemostat studies in which competition was examined in continuous culture, with a mixed natural community from a Spanish saltern as the inoculum. Salt concentration, temperature, and nutrient concentration (dilution rate) were used as variables (287). At low salt concentrations, the moderately halophilic bacteria won the competition, while at the highest salinities, the pigmented archaea outcompeted the bacteria. Within the intermediate salt concentration range (20 to 30%), temperature was the decisive factor determining the outcome, with the bacteria being favored by low temperatures. Also differences in the affinity for low concentrations of organic nutrients determined the result of the competition experiments: bacterial strains that grew slowly in batch cultures often predominated in chemostat cultures at low dilution rates thanks to their higher affinity for nutrients. In nutrient-rich batch cultures, the halophilic bacteria generally grew faster than the archaea, even at the high NaCl concentrations preferred by the archaeal types (287). The average growth rate of pigmented archaea (39 strains tested) in complex medium was 0.02 h¹ at 25% salt, while nonpigmented bacterial strains (93 strains tested) grew much more rapidly (average optimal growth rate, 0.06 h¹ at 10% salt, with a broad optimum between 2 and 25%) (288).

Several strains of moderately halophilic bacteria may be involved in the precipitation of CaCO₃ (calcite, aragonite) and other minerals (41). No direct evidence exists for the active involvement of these bacteria in the formation of mineral deposits in nature, either recent or ancient, but the phenomenon observed in cultures is of sufficient interest to be discussed here.

Most studies on the subject used Halomonas halophila as the model organism, and the minerals formed were identified by X-ray diffraction. In a test of 27 isolates, all caused the formation of CaCO₃ crystals as long as conditions allowed bacterial growth. Most crystals were spherical and consisted of calcite and magnesium calcite, in which MgCO₃ forms a significant part of the crystal (75 to 85% Ca and 25 to 15% Mg); the ratio between these minerals depended on the salinity and medium composition. High magnesium concentrations inhibited CaCO₃ precipitation by H. halophila. Growth at low salinity favored crystal formation, while at low temperature and/or at high salinity, crystal formation was repressed. The bacteria influenced the type of CaCO₃ crystal formed in vitro, and this effect may be species specific (42, 68, 280, 281). Calcification commences with a nucleus formed by aggregation of a few calcified bacterial cells and subsequent accumulation of more calcified cells and carbonate, which holds the bacteria together. This results in the formation of spherical bioliths of about 50 µm in diameter (282). To what extent the precipitation of CaCO₃ was triggered by a local increase in pH or whether the bacteria served only as crystallization nuclei was never ascertained.

Moderately halophilic isolates assigned to the genera Flavobacterium and Acinetobacter make calcite, and Acinetobacter also produces aragonite. High temperatures and low ionic strengths favor crystal formation. Flavobacterium makes magnesium calcite with 0.04 to 0.32 mol% magnesium; Acinetobacter produces magnesium calcite with up to 14% aragonite at the highest salinities (69). Similarly, 63 strains of Salinivibrio isolated from an inland saltern in Spain were found to be involved in crystal formation (279).

Induction of lysogenic phages by mitomycin C led to the isolation of phage F9-11 from an H. halophila strain obtained from soil. This phage replicates over a wide range of salinities (2.5 to 15%) (20, 92).

A water sample from Lake Chaplin, Canada, produced a bacteriophage that lysed a bacterium designated Pseudomonas strain G3, which is able to grow from 0.25 to over 3 M NaCl. The phage also infects S. costicola and two unidentified halophilic bacteria and is stable in the absence of salt (150). Another phage (designated UTAK) active against an S. costicola strain was isolated from the salterns of Alicante, Spain. The burst size of this phage was maximal at 1 to 2 M NaCl (80 to 105 phages per cell) and decreased to an average of only 12.5 phages per cell at 0.5 M. It was thus suggested that intracellular phage replication may be controlled by the salinity of the medium (96). Additional bacteriophages were isolated from the saltern in the course of this study, attacking morphologically different hosts, but most of these phages were not studied further.

Two phages lysing Tetragenococcus halophilus involved in soy sauce fermentation were characterized: phages 7116, having an isometric head and a contractile tail, and D-86, having an isometric head and a noncontractile tail. Both could propagate at all salinities in which their host could grow. Phage D-86 was stable at all salinities between 0.03 and 2.6 M, while phage 7116 was specifically unstable between 0.04 and 0.1 M salt while being stable in both the lower (0.01 to 0.03 M) and the higher (0.2 to 2.6 M) salt range. This effect, which is known for other (nonhalophilic) phages as well, is probably due to the salting-in effect of Na⁺, causing destruction of phage protein and DNA-protein complexes (331).

Cultures of Actinopolyspora halophila grown on low-salt media (10 to 12%, near the lower limit required for growth) showed holes resembling viral plaques (95). This phenomenon was not investigated further.

All bacteriophages isolated thus far from the moderate halophiles are double-stranded DNA phages with distinct heads and tails. Most are almost equally stable in the presence and the absence of salt and can retain infectivity for weeks in dilute solutions (151), in contrast to the halophilic archaeal phages, which, similar to their hosts, are inactivated in the absence of salt. The bacterial phages are "halophilic" only to the extent that their hosts are: the phages multiply only when the halophilic host is growing. In view of the relatively low salt concentration within the cytoplasm of halophilic bacteria compared to the outside medium (see below), the process of attachment and infection of the host cells may occur at high salt concentrations while intracellular multiplication occurs in a low-salt medium (151).

Another approach to obtain information on the role that bacteriophages may play in regulating the community sizes of halophilic bacteria in nature is based on direct electron microscopic observation and enumeration of phages in environmental samples. This approach was used in a study of Spanish salterns of different salinities (104). While many infected cells (probably of halophilic archaea) were seen at salinities above 25%, infected cells were not observed in the lower salinity range. At the lower salinities, bacterivory by protozoa was estimated to be much more important than phage lysis. In the lower salinity range (up to about 15%) around 1 × 10⁷ to 2 × 10⁷ bacterial cells and 5 × 10⁷ to 7 × 10⁷ virus-like particles, most of them with icosahedral heads, were counted per ml of brine. Viral abundance, as well as prokaryote abundance, increased with salinity. Also in the Dead Sea, virus-like particles (head and tail or spindle-shaped) were abundantly found. In view of the dominance of halophilic archaea in the lake, these virus-like particles were most probably derived from lysis of archaea, but the possibility that bacterial viruses were involved cannot be ruled out (249).

Another factor which may cause a decrease in halophilic bacterial numbers, in addition to lysis by bacteriophages and predation by protozoa, is the existence of halophilic predatory bacteria of the genus Bdellovibrio. With Vibrio parahaemolyticus and V. alginolyticus as hosts, halophilic Bdellovibrio strains that grew from 1 to 15.5% salt were isolated (303). Thus, predatory bacteria should also be considered as potential regulators of the halophilic bacterial community size in nature.

PHYSIOLOGY

Top Previous Next References

Requirement for Salt

The common denominator for all moderately halophilic bacteria is their requirement for salt and their ability to tolerate high salt concentrations. Salt requirement and tolerance are highly variable among the different species. Moreover, these parameters are by no means constant, since they may vary according to the growth temperature and the nature of the nutrients available (174). The salinity range of many isolates has been investigated in complex media. This fact led to the classification of some organisms as halotolerant rather than halophilic. Thus, H. elongata was originally described as extremely halotolerant (357). However, in minimal medium it requires at least 0.5 M NaCl, thus behaving like a true halophile (23). An examination of the specific requirement for Na⁺ and Cl ions, as well as the tolerance toward other salts, is also necessary. In most cases, a minimum concentration of Na⁺ is essential for growth. This may be due in part to the requirement for Na⁺ gradients to drive transport processes in the cell membrane. Certain species may also possess a primary respiration-driven outward sodium pump (see below). S. costicola (optimal growth at 0.8 to 1.5 M NaCl, and growing up to 3.3 M in a peptone-based complex medium) had a minimum requirement for 0.5 M NaCl in media based on NaCl as the sole salt. Addition of high concentrations of compounds such as glucose or glycerol lowered the NaCl requirement to 0.3 M, but no further lowering of the sodium concentration required was achieved (1). Other cations may be tolerated in high concentrations. Thus, "Micrococcus varians subsp. halophilus" can grow in 1.5 to 2 M LiCl, RbCl, or CsCl in the presence of 60 mM Na⁺ (146). There does not seem to be an absolute requirement for Cl ions: H. elongata grew as well on NaBr and NaNO₃ (but not on NaI or Na₂SO₄) as on NaCl (358). H. halophila grew well on NaCl, NaBr, Na₂SO₄ and Na₂S₂O₃ but not on other sodium salts (264). The moderately halophilic Pseudomonas sp. strain 40 can grow in 1 to 4 M NaCl, 1 to 2 M NaNO₃, or 1 M Na₂SO₄ but not in 1 to 4 M KCl (238).

Salt requirement and tolerance may be temperature dependent. In certain halophilic archaea such as Haloferax volcanii, the minimum and optimum salt concentrations shifted to higher values with increasing temperature (219), and a similar phenomenon was observed in halophilic bacteria as well. Thus, the optimum salt concentration for growth of H. halophila at 32 and 42°C was 7.5%, whereas the optimal concentration for growth at 22°C was 5% (264). H. elongata grew in complex medium at 20 and 30°C at salt concentrations between 0.05 and 3.4 M. At 40°C, no growth was obtained at 0.05 M, but growth was possible between 0.375 and 4.5 M. In defined medium with glucose and alanine as organic nutrients, salt tolerance was decreased, growth occurred within a narrower salt range than in complex medium, and a higher salt concentration was needed for optimum growth (358).

Marinococcus halophilus (NaCl range, 0 to 5.5 M; optimum, 1 M at 35°C) grew in the virtual absence of NaCl at 20°C. At 25°C, at least 0.5 M was required and could not be replaced by KCl or by nonionic solutes (168, 171, 174, 227, 228). Somewhat different behavior was observed in S. costicola (optimum, 1 M NaCl at 30°C): at higher or lower growth temperatures, both the optimum and the lower limit of NaCl concentrations were higher (3). While at 30°C cells grew from 0.5 to 5 M salt, at 20°C 0.2 M salt was sufficient for growth. This lower limit could not be reduced further (168, 171). Many halophiles may thus prove to grow at a wider range of NaCl concentrations when tested at a greater range of temperatures.

Many of the moderately halophilic bacteria have simple growth requirements, and minimum growth requirements have been determined for several species. Thus, H. halophila grows well on a medium containing inorganic salts, including nitrate as the nitrogen source, and glucose as the only carbon and energy source (264). H. halodenitrificans could grow aerobically on any of a number of organic carbon sources in the presence of thiamine. Under anaerobic conditions (with nitrate as the electron acceptor), methionine had to be supplied as well. Methionine could be replaced by glycine betaine or by vitamin B₁₂ but not by dimethylglycine. It was suggested that the bacterium may be deficient in the cobalamine-dependent path for methionine synthesis and is therefore unable to produce glycine betaine anaerobically (126). However, H. halodenitrificans also produces ectoine as compatible solute anaerobically and cannot synthesize glycine betaine de novo (83); hence, this explanation for the effect of methionine may not be valid. S. costicola has more complex growth requirements. The earliest designed synthetic medium contained glucose, L-cysteine or cystine, glutamate, arginine, valine, isoleucine, and salts (70). Glucose, cyst(e)ine, and NaCl were essential, and omission of any of the other components led to decreased growth. A simpler formulation was based on glucose, glutamate, two vitamins (biotin and thiamine), choline (as a precursor of glycine betaine), and salts (148).

Most moderate halophiles have more demanding nutritional requirements at high salt concentrations. Complex media stimulate growth at high salt concentrations. The effect may be due to the presence of compatible solutes or their precursors that can be accumulated or to the fact that other growth factors may be synthesized more slowly under the high-salt conditions (136). Thus, the salt tolerance of S. costicola in defined medium could be extended by including 2% sodium glutamate (148), and its growth in 4 M (but not in 3 M) salt required the presence of nutrients such as glycine betaine (296). The widest salt range for growth was found in proteose peptone and tryptone medium, Casamino Acids alone gave a narrower range (0.4-2.5 M), while in a defined medium no growth was obtained above 2.2 to 2.3 M (71).

Surveys of heavy metal sensitivity and tolerance to 10 heavy metal ions in moderate halophiles (224), both from culture collection strains and from fresh isolates, showed a very heterogeneous response among the taxonomic groups (the Halomonas group, Acinetobacter, Flavobacterium, moderately halophilic cocci), as well as among the strains included in each group. All were sensitive to mercury, silver, and zinc and tolerant to lead. The response to arsenic, cadmium, chromium, and copper was very heterogeneous. Acinetobacter strains proved the most metal tolerant, and Flavobacterium strains were the most sensitive. The influence of salinity and yeast extract concentrations in the test medium on the toxicity of the heavy metals tested was also examined. In general, lowering the salinity led to enhanced sensitivity to cadmium and, in some cases, to cobalt and copper. However, increasing the salinity resulted in a decrease only in the cadmium, copper, and nickel toxicities. Reduction in the yeast extract concentration resulted in an increased sensitivity to all metals, but only a slight decrease in the toxicities of nickel and zinc was found when the yeast extract concentration was increased (224).

Different S. costicola strains were compared for heavy metal tolerance. All proved sensitive to cadmium, copper, silver, zinc, and mercury. All tolerated lead, and most were also tolerant to nickel and chromium, so that multiple tolerance to the three metal ions chromium, nickel, and lead emerged as the major pattern (90). On the basis of these studies, several metal concentrations were proposed to discriminate between heavy metal-tolerant strains and those that were sensitive (222), thereby facilitating the isolation of metal-tolerant strains from polluted hypersaline habitats. In a recent study, the isolation and taxonomic characterization of a large number of heavy metal-tolerant halophilic strains from different geographical sites in Spain has been attempted. A total of 222 metal-tolerant (to mercury, cadmium, copper, chromium, or zinc) moderately halophilic strains were selected for a detailed taxonomic analysis. Most isolates were assigned to the genus Halomonas, and approximately 30% of the strains displayed multiple resistances (278).

Salinity-dependent cadmium tolerance was documented in Pseudomonas sp. strain 40. In 1 M NaCl, poor growth was obtained in the presence of 2 mg of CdCl₂ per ml and no growth was possible at 2.5 mg/ml. However, in 2 to 4 M NaCl and 2.5 mg of CdCl₂ per ml, moderate growth was observed. NaNO₃ and Na₂SO₄ enhanced cadmium toxicity. Cadmium ions react with chloride ions to form complexes whose nature depends on the chloride concentration: at 1 M NaCl, most of the cadmium appears as a mixture of CdCl₂ (35%) and CdCl₃ (45%); at 2 M NaCl, the anionic complexes CdCl₃ (47%) and CdCl₄² (33%) predominate and are probably less toxic (238).

Osmotic balance can be achieved by the accumulation of salts, organic molecules, or a combination thereof. A fourth possibility, that the cell is able to control water movement in and out and maintain a hypoosmotic state of their intracellular space, has been proposed for S. costicola and H. elongata (308, 351, 359).

Much of the controversy in the literature about the nature of the real intracellular environment of the halophilic bacteria originates from the difficulties in the estimation of the cell volume. Intracellular solute concentrations are generally determined by analysis of cell pellets and thus depend on the precise assessment of what fraction of the pellet volume is occupied by the intracellular space. Volume determinations are based mostly on the distribution of radioactive marker molecules labeling the total water space, the water space excluding the cytoplasm, and the water space excluding the whole cell volume, including the outer layers and periplasmic space (133). Cell-impermeable markers that have been successfully used so far include inulin (105, 193, 308, 359), dextran (46, 203, 298), and blue dextran (146). Permeable solutes such as tritiated water or ethylene glycol are also used to calculate the total water space in the cell pellets (170). An exact determination of the intracellular and extracellular water spaces is essential. Errors are especially large for sodium and chloride, which are generally abundant in the medium, and thus a small error in the determination of the spaces results in a large error in their apparent concentrations. Accurate estimation of the cytoplasmic volume also suffers from the lack of reliable methods to differentiate between the periplasmic space and the osmotically active cytoplasmic space (133). Small molecules such as raffinose (105), sorbitol, and sucrose (308), which have been used in this type of experiment, can be expected to penetrate the outer membrane and become distributed in the periplasmic space as well. This periplasmic space can have a considerable size: in S. costicola, it was estimated to occupy 38% of the total cellular space (308). All calculations are based on the (not necessarily true) assumption that the molecules used as markers are not taken up and metabolized by the cells or bound to the envelopes and other cellular structures. Indeed, it was suggested that dextran and inulin may bind to cell envelopes (130) and that smaller fragments of unpurified dextran may be taken up by the cells, causing large errors in the calculation of the cytoplasmic solute concentrations. That a proper knowledge of the intracellular water space under different growth conditions is essential and that a comparative approach based, e.g., on the ion content per unit of cell protein is insufficient is clearly shown by the finding that in H. elongata the cell volume per unit of protein is inversely related to salinity, decreasing from 2.62 to 2.06 µl/mg of protein in cells grown from 0.175 to 1.37 M NaCl (203). A similar phenomenon was observed in H. canadensis: cells grown at 0.6 M NaCl had a volume of 4.94 µl/mg of protein, decreasing to 2.69 µl/mg of protein at 4.35 M (193).

During the analysis of cell pellets obtained by centrifugation, anaerobic conditions may develop in the densely packed cells during handling and washing, potentially leading to loss of substantial amounts of potassium and gain in the amount of sodium (308). The effect is reversible: when dense cell suspensions of S. costicola were aerated in the presence of an energy source, potassium ions were taken up while sodium was released (168). Perhaps variations in harvesting and handling may explain the differences in the estimated intracellular ion concentrations in H. halodenitrificans as published by different authors (31, 298) (Table 5). Centrifugation of cells through a layer of silicone oil may avoid some of the problems involved in conventional centrifugation techniques. Oils of different densities must be used to adjust for the density of the cells, which depends on the salt concentration of the medium (171). The method has not been widely applied in the study of moderate halophiles. A different approach toward the estimation of intracellular ion concentrations, yet to be applied to the aerobic halophilic bacteria, is the use of X-ray microanalysis in the electron microscope. This method, in which individual cells are analyzed, was recently used to measure ion contents in the halophilic anaerobe Haloanaerobium praevalens (252).

The phase of growth can also have a major influence on the results of the internal ion concentration measurements; stationary-phase cells may have a much higher intracellular sodium concentration than exponentially growing cells. This may also explain some of the unusually high apparent intracellular sodium concentrations reported in the literature.

Table 5 summarizes some of the reported estimates of intracellular ion concentrations of halophilic bacteria. Most analyses are limited to Na⁺ and K⁺, intracellular Cl concentrations have seldom been determined, and data on divalent cations are scarce. Great variations in the intracellular ion concentrations are obvious, both among different species of moderate halophiles and within the same species, depending on the growth conditions and on the method used. A few general trends are clear, however (176). (i) The intracellular K⁺ concentration is generally higher than that in the medium. (ii) The Na⁺ concentration inside the cells is generally lower (to different extents) than that outside. (iii) The apparent intracellular Na⁺ and K⁺ concentrations increase with increasing external NaCl concentration in a nonlinear fashion. (iv) Generally the sum of the concentrations measured is insufficient to balance the osmotic pressure of the medium. However, taking into account the presence of organic osmotic solutes as well (see below), such a balance may be achieved.

In certain gram-positive bacteria, the apparent intracellular cation concentrations are similar to those of the growth medium. This was reported for the haloalkaliphilic Bacillus haloalkaliphilus (361). However, the reported value of 0.37 g of intracellular water per g (dry weight) is rather high and may overestimate the cytoplasmic ion concentrations. Moreover, this strain produces ectoine and other yet unknown organic osmotic solutes (83). In "Micrococcus varians subsp. halophilus," the apparent intracellular Na⁺ concentration was approximately equal to that in the medium over the range from 1 to 2 M, while in cells grown at 4 M NaCl, 2.1 M Na⁺ was measured intracellularly (28). The presence of high intracellular Na⁺ concentrations in this organism was confirmed by Kamekura and coworkers, who also showed that other monovalent cations added to the growth medium (K⁺, Li⁺, Rb⁺, and Cs⁺) were not excluded from the cytoplasm (146, 176).

In many Halomonas species (H. elongata, H. canadensis, and H. halodenitrificans), the sum of the apparent intracellular Na⁺ and K⁺ concentrations is much lower than the medium concentration. In H. halodenitrificans, the sum of intracellular Na⁺ and K⁺ concentrations remained low and constant (about 0.1 M Na⁺ and 0.3 M K⁺ in exponentially growing cells) over a wide range of medium NaCl concentrations. In stationary-phase cells, a drastic increase in the intracellular Na⁺ concentration was observed (up to 0.5 and 1.1 M in cells grown in 1 and 3 M NaCl, respectively) and the intracellular K⁺ concentration decreased to about 0.1 M (298). In H. canadensis, the intracellular Na⁺ concentration increased when the cells reached the stationary phase (193). In "Pseudomonas halosaccharolytica," the apparent internal salt concentration was relatively independent of the salt concentration in which the cells were grown, but in this organism the measured sum of the intracellular concentrations was quite high (1.4 to 1.9 M Na⁺ + K⁺ in cells grown between 1 and 3 M NaCl) (192). In S. costicola, the intracellular Na⁺ + K⁺ concentration was significantly lower than outside only at the highest medium salinities. However, upon treatment with cetyltrimethylammonium bromide, the cellular Na⁺ and K⁺ concentrations did not equilibrate with the external medium, possibly indicating that the ions may partially occur in a bound state (308).

Below we summarize a few data on the intracellular abundance of individual ions.

(i) Sodium. The apparent intracellular Na⁺ concentrations are often far too high to enable the generally salt-sensitive cytoplasmic enzymes to be active (see below). However, the assessment of the true intracellular Na⁺ concentration is problematic, as discussed above. In addition, Na⁺ and other ions may be bound to the outer cell layers, in amounts increasing with external salinity (133). It was thus suggested that much of the cell-associated Na⁺ in "P. halosaccharolytica" is not cytoplasmic (192).

(ii) Potassium. In most halophilic bacteria, K⁺ is accumulated to a few tenths of 1 M (133). H. elongata seems to be an exception, with K⁺ concentrations as low as