Biology of Pseudomonas stutzeri

Pseudomonas stutzeri is a member of the genus Pseudomonas sensu stricto. It is in group I of Palleroni's DNA-rRNA homology group within the phylum Proteobacteria (252, 253). P. stutzeri is now recognized as belonging to the class Gammaproteobacteria. Phylogenetic studies of P. stutzeri strains' 16S rRNA sequences and other phylogenetic markers demonstrate that they belong to the same branch, together with related species within the genus, such as P. mendocina, P. alcaligenes, P. pseudoalcaligenes, and P. balearica. Typically, cells are rod shaped, 1 to 3 μm in length and 0.5 μm in width, and have a single polar flagellum. Under certain conditions, one or two lateral flagella with a short wavelength may be produced. Phenotypic traits of the genus include a negative Gram stain, positive catalase and oxidase tests, and a strictly respiratory metabolism. In addition, P. stutzeri strains are defined as denitrifiers. They can grow on starch and maltose and have a negative reaction in arginine dihydrolase and glycogen hydrolysis tests. The G+C content of their genomic DNA lies between 60 and 66 mol%. DNA-DNA hybridizations enable at least 17 genomic groups, called genomovars, to be distinguished. Members of the same genomovar have more than 70% similarity in DNA-DNA hybridizations. Members of different genomovars usually have similarity values below 50%.

No fluorescent pigments are produced, which differentiates P. stutzeri from other members of the fluorescent group of Pseudomonas spp. Before the use of genomic approaches to identifying bacteria became widespread, P. stutzeri strains were misidentified with other species. This was due to the intrinsic limitations of exclusively phenotypic identification procedures within the former genus Pseudomonas. P. stutzeri was most commonly confused with other Pseudomonas species (P. mendocina, P. pseudoalcaligenes, P. putida); with species actually in other genera (such as Delftia acidovorans and Ralstonia pickettii); or even with the flavobacteria, Alcaligenes or Achromobacter. Mandel proposed the species “Pseudomonas stanieri” for P. stutzeri strains with a low G+C content, around 62% (212); however, G+C content alone is a weak parameter for species differentiation. In some collections, P. stutzeri cultures were labeled P. saccharophila. The strain OX1 (ATCC BAA-172) was classified phenotypically as a P. stutzeri strain (13). It has been intensively studied due to its significant phenotypic characteristics. However, when strain OX1 was characterized taxonomically in detail, it turned out to be a member of the P. corrugata phylogenetic branch (73). Pseudomonas sp. strain OX1 may be confused phenotypically with P. stutzeri because P. stutzeri is phenotypically diverse. However, OX1 is genomically distinct.

The species most closely related to P. stutzeri is P. balearica (formerly genomovar 6 of the species). It shares many basic phenotypic traits with P. stutzeri strains and belongs to the same 16S rRNA phylogenetic branch. However, it can be differentiated chemotaxonomically from P. stutzeri by its ability to grow above 42°C and by a few other biochemical tests (23).

P. chloritidismutans is a member of genomovar 3. However, it has been proposed as the type strain of a new species (404) and is discussed below (see “Physiological properties”). There is always a danger of drawing taxonomic conclusions from the properties of metabolic systems that are involved in the metabolism of unusual substrates or molecules.

The phylogenies of genes of the rrn operon, considered individually or with other housekeeping genes, demonstrate that all P. stutzeri strains are monophyletic. Such phylogenetic studies are currently another good tool for discriminating P. stutzeri from the rest of the bacterial species. P. xanthomarina has recently been described as a new species (289) with only one representative strain. It is located in the same 16S rRNA phylogenetic branch as P. stutzeri and P. balearica, with sequence similarities above 98%. It can be differentiated phenotypically from both species.

Colonies can be distinguished by their unusual shape and consistency (Fig. 1). Freshly isolated colonies are adherent, have a characteristic wrinkled appearance, and are reddish brown, not yellow, in color. They are typically hard, dry, and tenaciously coherent. It is easy to remove a colony in its entirety from a solid surface. Colonies generally resemble craters with elevated ridges that often branch and merge, and they have been described as tenacious, with a coral structure. There may be more mucoid protuberances at the periphery than in other areas. The frequent occurrence of irregular polygon-like structures or concentric zones has also been noted (371). The shapes of colonies are neither uniform nor necessarily constant: they change appearance with time. After repeated transfers in laboratory media, colonies may become smooth, butyraceous, and pale in color. This has been described as colonial dissociation. Strain CMT.9.A hydrolyzes agar. This is a rare property and is mainly restricted to marine bacteria. However, the attack may be limited to what is known as “pitting” of the agar (3). Sorokin et al. give a very detailed description of the colonial morphology, differentiating between R-type and S-type colonies (337). The R-type colonies are stable, but the S type produces both colony types under appropriate conditions. Smooth colonies grown on plates at 30°C and stored at 4°C for 24 h often develop a characteristic wrinkled appearance (A. Cladera, personal communication).

P. stutzeri is grouped with the nonpigmented species of the genus, even though many strains' colonies become dark brown. This is due to the high concentration of cytochrome c in the cells. No diffusible pigments are produced on agar plates.

Cells are typically motile and predominantly monotrichous. In some strains, lateral flagella with a short wavelength are also produced. This particularly occurs in young cultures on complex solid media. These lateral flagella could easily be shed during manipulations incidental to flagellar staining (251). It has been suggested that lateral flagella might be involved in the population's swarming or twitching motility on solid surfaces (319). However, type IV pili may also be responsible for this movement. Statistically, the highest number of flagellated cells is reached at the beginning of the exponential growth phase (192). Seventy percent of cells were flagellated in strain AN11: 38% had only one flagellum, and 31% had one or more additional flagella inserted laterally (80).

Caution should be exercised when only phenotypic traits are used for classification. This can clearly be seen in the case of strain ZoBell. This strain (ATCC 14405) was isolated as a marine bacterium and described by ZoBell and Upham as “Pseudomonas perfectomarinus” in 1944 (412). Subsequently, this organism became the only member of the species P. perfectomarina. Its lack of flagella was emphasized by its assignation to a new species, although the authors who first described this strain stated that it was motile (19, 412). After three passages, enrichment for flagellated bacteria on semisolid tryptone agar enabled a population in which over 80% of cells were flagellated to develop. This revertant strain is motile by means of a single polar flagellum (294).

In a recently published chapter on chemotaxis in Pseudomonas, Parales et al. stated, “All Pseudomonas species are motile by one or more polar flagella and are highly chemotactic” (258). P. stutzeri is no exception. Chemotaxis machinery has not been studied in detail for any Pseudomonas species. Moreover, the ranges of attractants or repellents and environmental conditions to which Pseudomonas spp. respond remain largely unexplored. They seem to be attracted to virtually all of the organic compounds they can use as growth substrates. However, they are also attracted to other compounds that they are unable to metabolize. Ortega-Calvo et al. studied the chemotactic response of several pseudomonads to polycyclic aromatic hydrocarbon-degrading bacteria (243). Strain 9A of P. stutzeri was included in the study. This strain degrades naphthalene, phenanthrene, and anthracene. It was concluded that chemotaxis was positive to naphthalene and to the root exudates of several plants. Chemotaxis may enhance the biodegradation of pollutants in the rhizosphere, at least in laboratory-scale microcosms. Strain KC mineralizes carbon tetrachloride, and motility-enhanced bioremediation in aquifer sediments has been demonstrated (401, 402).

Pseudomonas species have a range of different adhesins that function during initial attachment to a substratum. This leads to biofilm formation. Both flagella and pili seem to be important in the colonization of biotic and abiotic surfaces, particularly in the initial formation of microcolonies. P. aeruginosa's initial biofilm development appears to be conditionally dependent on type IV pili. P. stutzeri possesses both flagella and pili but has not been described as a member of consortia that form natural biofilms. Type IV pili confer twitching motility to P. stutzeri strains (a bacterial movement based on pilus extension/retraction). This is probably at least partly responsible for many colonies' diffuse borders (J. Sikorski, personal communication). These colonies also correspond to strains that have natural transformation ability.

The genomovar concept was originally defined for P. stutzeri as a provisional taxonomic status for genotypically similar strains within a bacterial species. Two strains classified phenotypically as members of the Pseudomonas stutzeri species were included in the same genomovar when their DNA-DNA similarity values were those generally accepted for members of the same species (more than 70% similarity or less than 5°C difference in thermal denaturation temperature [ΔT_m] values). Members of two different P. stutzeri genomovars have 15 to 50% DNA-DNA similarity values or ΔT_m value differences greater than 5°C. Subsequently, this concept has been used taxonomically to group genotypically similar strains in other species, such as Burkholderia cepacia and species in the genera Xanthobacter, Azoarcus, and Shewanella, etc. It provides a useful provisional level of classification.

The methods used to calculate DNA-DNA similarity values have differed from one laboratory to another. Palleroni used ¹²⁵I labeling and/or membrane filters (251). Rosselló et al. used the ΔT_m method, as described previously (291). Sikorski et al. used the method described by Ziemke et al. (411), with digoxigenin and biotin labeling and quantification of the binding ratio in microtiter plates (327). Vermeiren et al. used DNA-DNA thermal reassociation, measured photometrically (380). The results were consistent with the genomovar subdivision of the species, regardless of the method used to estimate the similarity value.

To date, nine different genomovars have been well documented. Eight new genomovars in the species P. stutzeri were put forward recently (327). One reference strain has been proposed for each genomovar and deposited in culture collections. Most strains studied so far are included in genomovar 1 (along with the species' type strain). The genomovars 8 (strain JM300), 9 (strain KC), 10 (strain CLN100), and 18 (strain MT-1) each have only one representative strain. These might be considered genomospecies, sensu Brenner et al. (50). As an example, we can consider strain CLN100, of genomovar 10. It is a representative of a new species from a genomic perspective, sharing many substantial phenotypic and phylogenetic characteristics with members of the P. stutzeri phylogenetic branch. Some phenotypic traits can be used to discriminate CLN100 from the P. stutzeri and P. balearica strains described to date (simultaneous degradation of chloro- and methyl-derivatives of naphthalene and absence of ortho cleavage of catechol, etc.). These characteristics could be the basis for describing CLN100 as the type strain of a new species. However, some of these phenotypic traits could be strain specific; therefore, it was preferred not to define a new species until more strains that are genomically and phenotypically similar to strain CLN100 have been described (114).

Information on genome structure is a very important component of any comprehensive bacterial description. The comparative analysis of bacterial chromosomes on intra- and interspecies levels can provide information about genomic diversity, phylogenetic relationships, and chromosome dynamics. In the genus Pseudomonas, genome structure has been studied only for P. aeruginosa, P. fluorescens, P. putida, and P. stutzeri. Ginard et al. studied 20 strains of P. stutzeri in 1997, representing the seven genomovars known at that time (121). They also studied P. stutzeri's closest relative, P. balearica. The genome of P. stutzeri strains is made up of one circular chromosome. It ranges from 3.75 to 4.64 Mb in size (20% difference in size). In comparison, P. aeruginosa genome sizes, calculated by macrorestriction analysis, range from 6.345 to 6.606 kb, a fluctuation of only about 4%. However, a more recent report on P. aeruginosa genome sizes indicates a 20% fluctuation (from 5.2 to 7.1 Mb) (310). The I-CeuI, PacI, and SwaI low-resolution map of P. stutzeri's type strain enabled 12 genes—including four rrn operons—and the origin of replication to be located (121). The 20 strains' enzyme digests were used to compare rrn backbone organization within the genomovars. The four rrn operons seemed to be at similar locations with respect to the origin of replication, as did the rest of the six genes analyzed. In most genomovar reference strains, rrn operons are not arranged around the origin of replication but are equally distributed along the chromosome. Large chromosomal rearrangements and differences in genome size seem to be responsible for the differences in genome structure. This suggests that they must have played an important role in P. stutzeri diversification and niche colonization. Strains belonging to the same genomovar have similar genome architectures that are well correlated with phylogenetic data (121).

From one to four plasmids were detected in 10 of the 20 strains analyzed in this study (121). The Eckhardt method, using both conventional and pulsed-field gel electrophoresis, turned out to be the most reliable and useful technique for plasmid detection. Seventy-two percent of the plasmids observed were smaller than 50 kb, one plasmid was between 50 and 95 kb, and four plasmids were larger than 95 kb. No two strains shared the same plasmid profile, and no relation was found between genomovars and the distribution of plasmids among the strains. Seven of the 10 plasmid-containing strains were isolated from polluted environments. This is not uncommon in plasmid analyses. A correlation between the degree of contamination and the incidence of plasmid occurrence was found in an environmental study by Baya et al. (20). Naphthalene degradation plasmids are common in Pseudomonas spp. However, in eight of the nine naphthalene-degrading strains of P. stutzeri studied, the catabolic genes were inserted into an I-CeuI chromosomal fragment, as demonstrated by Southern blot hybridizations with nahA and nahH probes. The naphthalene genes seem to be plasmid encoded only in strain 19SMN4 (120, 296).

Genotypic intraspecies relationships in P. stutzeri strains have been determined by various genotyping methods. These are based on restriction fragment length polymorphism (RFLP) analysis of total DNA, PCR amplification of selected genes, or PCR amplification and restriction analysis. These analytical methods differ in discrimination level between strains. They have been applied simultaneously to all P. stutzeri genomovars' reference strains; to P. balearica, the strains most closely related to P. stutzeri; and to related type strains of the genus Pseudomonas. In all methods, computer-assisted analysis generates dendrograms that confirm the consistency of strain clustering with the genomovar subdivisions of the species. Additional typing by multilocus enzyme electrophoresis (MLEE) and multilocus sequence typing (MLST) is discussed below.

Methods based on the electrophoretic patterns of macrorestriction fragments (low-frequency restriction fragment analysis) have been used by two independent groups to examine representative strains (121, 271). The restriction enzymes XbaI and SpeI cut the P. stutzeri genome of the strains studied into 20 to 48 fragments. These fragments were resolved by pulsed-field gel electrophoresis. They are useful for generating fingerprints, which can be used to explore genome structures and to determine the degree of relatedness of strains. No correlation was found between the similarity of macrorestriction patterns and the subdivision of the species into genomovars. This was due to the high discriminatory power of the two enzymes and the heterogeneity of the restriction patterns. However, some patterns allowed clonal variants between strains to be distinguished. In these cases the related strains belonged to the same genomovar. The marked heterogeneity was attributed, at least in part, to large chromosomal rearrangements (121).

In the ribotyping procedure, total DNA is purified and then cleaved by restriction endonucleases. Brosch et al. (51) used the enzymes SmaI and HincII in their study of Pseudomonas strains. Restriction fragments were separated by electrophoresis, transferred to a nylon membrane, and hybridized with a 16S-23S rRNA probe. Nine strains of P. stutzeri clustered together in the dendrogram, which also showed 217 other strains from different Pseudomonas species. Two identical bands were detected by HincII in P. stutzeri. SmaI profiles were more discriminative, distinguishing from four to eight bands. Members of a single genomovar were grouped in the same branch.

Bennasar et al. (25) revealed genetic diversity and the relationships among P. stutzeri strains by rapid molecular typing methods. Repetitive extragenic palindromic PCR and enterobacterial repetitive intergenic consensus PCR analyses, based on DNA consensus sequences, generated fingerprints that were then computer analyzed. Groupings were consistent with the genomic groups that had previously been established by DNA-DNA hybridizations or 16S rRNA sequencing. Members of other Pseudomonas species were clearly different. Sikorski et al. (325) carried out random amplified polymorphic DNA (RAPD) PCR analysis in their study of P. stutzeri isolates from marine sediments and soils in geographically restricted areas (local populations). The results demonstrated the complex composition and high strain diversity of the local populations studied.

Similar genomic relationships have been revealed by PCR amplification of several genes (16S rRNA, internal transcribed spacer region 1 [ITS1], ITS2, and rpoB) and by analyzing the RFLPs generated by several restriction enzymes (25, 133, 325). These methods have confirmed the high genetic diversity of the species, the consistency of genomic groups (genomovars), and the usefulness of the patterns generated for strain identification.

Knowledge of the genomic structure of a population is essential to thoroughly understanding a species' characteristics. Such knowledge is particularly important in studies of population dynamics or habitat colonization, as it is used to elucidate genetic exchange in natural populations. The MLEE technique involves determining allozyme variation in a variety of housekeeping enzymes. Codon changes within enzyme genes, leading to amino acid substitutions, are detected electrophoretically by this technique (314). Thus, the variation in chromosomal genes is recorded, and the degree of gene transfer within a species is estimated. This enables relationships between bacterial isolates to be determined and a phylogenetic framework to be constructed.

Two independent research groups have used the MLEE approach in studies of P. stutzeri (284, 324). In Sikorski's study, 16 P. stutzeri strains belonging to eight different genomovars were analyzed for the allelic profiles of 21 enzymes. A distinctive multilocus genotype was detected in all strains, and up to 11 alleles were detected per locus. In Rius's analysis, 42 P. stutzeri strains from nine genomovars (including 9 strains previously studied by Sikorski et al.) and 20 enzymes were studied. The highest number of different alleles found per locus was 32, and all multilocus genotypes were represented by a single strain. Forty-two electrophoretic types were detected. In both analyses, P. stutzeri was shown to have a highly polymorphic structure. If both groups' results are combined, 49 different P. stutzeri strains have been studied with MLEE. A total of 33 different enzymes were analyzed from these strains. An analysis of this set of 49 strains again demonstrates that all of the multilocus genotypes were represented by a single strain. MLEE studies reveal that P. stutzeri is highly polymorphic. The highest genetic diversity described for a species is revealed (284) by an analysis of the members of genomovar 1 only. An analysis of source and place of isolation showed no clear association in clusters. When two subgroups of P. stutzeri populations (clinical and environmental isolates) were compared, the mean levels of genetic diversity were not significantly different. This indicates that clinical strains come from the same populations as environmental isolates. This may have important epidemiological implications for the microbiology of P. stutzeri infections. However, when two strains were grouped at moderate genetic distances (below 0.55), each pair of strains belonged to the same genomovar.

MLST has been proposed as a good method for population genetic analysis and for distinguishing clones within a species (98). This method employs the same principles as MLEE, as it detects neutral genetic variation from multiple chromosomal locations. This variation is identified by nucleotide sequence determination of selected loci. Cladera et al. (72) attempted to differentiate P. stutzeri populations and to establish the genetic diversity and population structure of the species clearly. They carried out a comparative analysis of gene fragments, using the principles of multilocus sequence analysis. The genes were selected from 26 strains belonging to nine genomovars of the species and from P. balearica strains, the species most closely related to P. stutzeri. Seven representative chromosomal loci were selected, corresponding to three kinds of genes: (i) housekeeping genes that are universally present in bacteria (16S rRNA and ITS1 region, representing the rrn operon, and the gyrB and rpoD genes, which interact with nucleic acid metabolism, coding for gyrase B and DNA-directed RNA polymerase, respectively) and which have been included in previous Pseudomonas taxonomic studies (408); (ii) genes that are characteristic of the species (catA, coding for catechol 1,2-dioxygenase, an enzyme responsible for the ortho cleavage of catechol in species of RNA group I of Pseudomonas, and nosZ, nitrous oxide reductase, a metabolically characteristic gene defining this denitrifying species); and (iii) nahH, coding for catechol 2,3-dioxygenase, responsible for the meta cleavage of catechol, a gene that is considered to be plasmid encoded in the genus Pseudomonas but chromosomally encoded in most naphthalene-degrading P. stutzeri strains studied to date (296).

All loci were highly polymorphic in the 26 strains studied. The number of nucleotide substitutions per nucleotide site varied from 44.2% for catA to 21.8% for nahH. The number of alleles varied in the different loci: 4 in nahH (16 strains), 18 in catA (24 strains), 20 in gyrB (26 strains), 17 in rpoD (26 strains), 18 in nosZ (26 strains), 15 in 16S rRNA (26 strains), and 20 in ITS1 (26 strains). Apart from nahH (a gene that is probably acquired through lateral transfer), the mean number of alleles per locus in the 26 strains was 18.7, an extremely high value. The average number of alleles per locus and strain was 0.72.

In this MLST study (72), the dN/dS ratio—the ratio of nonsynonymous substitutions per nonsynonymous site which resulted in an amino acid replacement (dN) to synonymous substitutions per synonymous site that did not change the amino acid (dS)—was calculated for the genes encoding proteins as a measure of the degree (amount and type) of selection in P. stutzeri populations. Changes are selectively neutral when they are independent of the overlying phenotype and the selection pressure dictated by the phenotype's function. The ratio was less than 0.1 in three genes (gyrB, rpoD, and nosZ). The highest dN/dS ratio corresponded to catA (0.18). All ratios were much less than 1, indicating that these gene fragments are not under selection. In other words, most of the sequence variability identified is selectively neutral. Synonymous substitutions were at least 5.5 times (1/0.18) more frequent than amino acid changes at any locus.

The number of nucleotide substitutions per nucleotide site was higher than in Campylobacter jejuni, Neisseria meningitidis, Streptococcus pneumoniae, Enterobacter faecium, and species of the Bacillus cereus complex. To our knowledge, the number of nucleotide substitutions described for P. stutzeri is the highest recorded to date (145). The average numbers of alleles per locus and strain analyzed in the protein-coding genes were 0.72 for P. stutzeri (an average of 18.7 alleles per locus in only 26 strains), 0.18 for C. jejuni, and 0.43 for the B. cereus complex. These values are in good agreement with previous observations made in MLEE studies of most of the strains analyzed by the MLST technique. In such MLEE studies the genetic diversity was the highest described for a species (284). Therefore, the extremely high genetic diversity of the species manifested by MLEE was corroborated by the MLST study.

Figure 2 shows an analysis of the sequence types (STs) identified among 26 independent strains of P. stutzeri. This analysis led to the assumption that one different ST per strain can be detected. This is the highest possible number of STs. Remarkably, when two strains had an allele in common they belonged to the same genomovar. There was only one exception: strain JM300 (genomovar 8) has an rpoD allele that is identical to strain JD4, one of the two members of genomovar 5. This can be explained by genomovars 5 and 8 having a common ancestor or by a possible lateral gene transfer to JM300, a strain intensively studied due to its natural transformation (206). Another strain, AN10 of genomovar 3, presents a possible recombination event with members of the same genomovar. Strains 19SMN4 and ST27MN3, of genomovar 4, were very closely related in the multilocus sequence analysis. They had identical 16S rRNA, rpoD, and gyrB genes. Both strains were isolated as naphthalene degraders from samples taken in a wastewater treatment lagoon. However, they were from different habitats (water column and sediment). Molecular typing methods (25, 121, 133) and MLEE (284) had previously demonstrated that both strains were genetically related but different. Again, the enormous genetic diversity of the species was demonstrated in this study. Inclusion of nahH in the analysis modifies the topography of the graph, indicating more possible events of lateral gene transfer (Fig. 2).

Several genes have been used as phylogenetic markers in P. stutzeri studies. The most extensively used are the rRNAs, 16S rRNA in particular. However, other genes with different degrees of sequence variation have been studied, because they provide useful information for analyzing different phylogenetic levels. Internal transcribed spacer regions ITS1 and ITS2, between the 16S and 23S rRNAs and between the 23S and 5S rRNAs, respectively, in the rrn operon present more-variable positions and are most useful in determining close relationships. Recently, Yamamoto et al. (408) studied the sequences of other housekeeping genes (gyrB and rpoD). These genes are assumed to be less constant than the 16S rRNA molecule among species of the genus Pseudomonas. In most cases, the study confirmed the phylogenetic branches that were previously defined by the 16S rRNA sequences in the genus.

Phylogenetic tree reconstructions of the same genes used in the MLST method (16S rRNA, ITS1, gyrB, rpoD, nosZ, and catA) were undertaken by Cladera et al. (72). Stability analysis using bootstrap resampling showed that the trees were stable and well defined. Most strains of P. stutzeri clustered in the same phylogenetic branch in the gene trees analyzed. They were usually separated from the other closely related species considered, P. balearica and P. mendocina. Strains belonging to the same genomovar were usually located in the same branch. There were only a few exceptions, which varied depending on the gene analyzed. A consensus phylogenetic tree was constructed for the six genes to deduce a composite molecular phylogeny for P. stutzeri. All P. stutzeri strains are located in the same phylogenetic branch, and members of each genomovar are clustered together, maintaining the genomovar subdivision of the species. This tree is based on a sequence of no less than 4,551 nucleotides, representing at least 9,546 nucleotides from the respective genomes, as there are four copies of the rrn operon in P. stutzeri. Therefore, between 0.2 and 0.25% of the chromosome (depending on the strain's genome size) has been compared pairwise in 24 independent isolates.

There is enormous genetic diversity in P. stutzeri. Despite this, the topologies of the trees and the values of the housekeeping genes' association indices, calculated from MLEE and MLST analyses, indicate that horizontal gene transfer and recombination processes are not enough to disrupt allele associations. This is because there is still a strong linkage disequilibrium among the P. stutzeri isolates. These results suggest that the population structure of P. stutzeri is strongly clonal, indicating that there is no significant level of recombination through independent assortment that might destroy linkage disequilibrium. Some authors have suggested that recombination events explain some of the diversity found in P. stutzeri (324). However, results of studies by Rius et al. (284) and Cladera et al. (72) are clear on this point. They use evidence from linkage disequilibrium analysis to argue strongly against the presence of detectable recombination. In a study on the potential for intraspecific horizontal gene exchange by natural genetic transformation, Lorenz and Sikorski (207) concluded that, with regard to transformation, there is sexual isolation from other Pseudomonas species and other genomovars. Gene transfer between genomovars by transformation is limited by sequence divergence at least; heterogamic transformation was reduced in competent cells. The potential to receive genes can also vary greatly among strains. It appears that some strains have a greater potential than others for gene acquisition. It seems that genomovars are free to diverge in neutral sequence characters as a result of sexual isolation mechanisms. These mechanisms prevent randomization of alleles. Nevertheless, the authors consider this border to not be absolute, and foreign sequences may be acquired and fixed.

A careful analysis of some genes, based on incongruences in the phylogenetic trees and/or what is known as relative codon usage, the codon bias index, or the G+C content of the genes, can help to define some metabolic pathways as genes acquired through horizontal transfer. The following examples are considered below: the aromatic degradative pathway, the nitrogenase system, the ability to use chlorate as a terminal electron acceptor, and the energy-yielding reactions in the oxidation of thiosulfate.

Phenotypic identification based on the characteristics given in the species definition and following dichotomous keys is usually satisfactory (26, 115). At present, studies of nutritional properties are frequently carried out with commercial kits designed to reduce the labor involved in traditional methods. Commercial procedures, such as the API 20NE, Microbact NE, and Biolog GN tests, usually identify P. stutzeri strains correctly. The identification manuals consider important distinguishing characteristics, such as denitrification or maltose utilization, to not be universal (denitrification is 94% positive, maltose is 69% positive, arginine dihydrolase is 2% positive, and gelatin liquefaction is 1% positive in the API strips). It is assumed that some tests may not be in accordance with the species' typical features. The strain sometimes has to be “adapted” to the test, by growing it under similar, but not strictly selective, conditions prior to the test. Denitrification is a good example of this and is considered below. In a study of the presence and identification of P. stutzeri in clinical samples, Holmes (154) stated that routine clinical laboratories have difficulty identifying this species.

A microbial cell expresses some 200 different proteins that can be separated by PAGE. This yields complex banding patterns, which are considered to be highly specific fingerprints (265). Strains with at least 70% DNA similarity tend to have similarities in protein electrophoretograms. Therefore, PAGE is thought to be a sensitive technique for gaining information on the similarities between strains within the same species or subspecies. Individual strains can often be recognized by protein pattern. Under standard growth and PAGE conditions, the patterns are reproducible. Computer-assisted analysis enables the information to be normalized and stored. This method has been used to identify P. stutzeri strains when a wide database is available (366).

The Sherlock microbial identification system is based on analyzing total fatty acid profiles. It gives satisfactory results within the genus Pseudomonas, including P. stutzeri, if the cells are cultured under strictly controlled conditions.

A PCR and an oligonucleotide probe method have been developed specifically for detecting and identifying P. stutzeri. The amplification primers and the probe were designed from the analysis of available 16S rRNA sequences. Positions that were specific for P. stutzeri and differed from the rest of Pseudomonas species were selected from variable regions in the Pseudomonas 16S rRNA. Positions 743 (G) and 746 (A) fulfilled both criteria, and a 21-nucleotide primer was designed (rps743). A second oligonucleotide, fps158 (17-mer), at positions 142 to 158, was selected as a second specific primer. It produced a 625-bp amplicon in PCR. The specificity of the amplicon was further identified with a DNA probe (17-mer) that included 12 bases of the 5′ end of primer rps743 (25).

A second set of primers, fps158 and rps1271, was developed by Bennasar et al. (24). These primers produced a 1,159-bp amplicon containing a BamHI restriction site. The specificity of the amplicon for P. stutzeri was then corroborated by restriction, giving two fragments, of 695 and 465 bp, respectively. A slightly modified set of primers in the same region was used successfully by Sikorski et al. (325).

The three methods permit good molecular differentiation of P. stutzeri from other species. They have been used to identify P. stutzeri and to detect it in environmental samples, as indicated below (see “Occurrence and Isolation Procedures”).

The species is well-defined phenotypically and chemotaxonomically. However, some of its distinguishing traits are lacking in well-documented strains (starch hydrolysis, arginine dihydrolase activity, and motility, etc.). In addition, many biochemical properties are extremely variable within the species and are not correlated with the genomovar groupings. DNA-DNA similarity values of more than 70% (or less than 5°C difference in thermal denaturation temperatures) are required to definitively assign a strain to a given species. In P. stutzeri, a polyphasic taxonomic approach is needed for assigning a new strain to the species: the strain has to agree with the basic phenotypic traits of the species, has to be placed in the same branch as P. stutzeri reference strains in the phylogenetic trees of one or more housekeeping genes, and has to show DNA-DNA similarity values of more than 70% with a reference strain of a recognized genomovar. If the last condition is not fulfilled, the strain can be proposed as a new genomovar within the species. If it can be phenotypically distinguished from P. stutzeri strains, it can be proposed as a new species.

The species has a wide range of growth temperatures. Temperatures from 4°C (strain NF13 grew at 4, 22, and 35 but not 55°C [297]) to 45°C (CMT.9.A grows at 45°C) have been cited for individual strains. However, growth at these extreme temperatures seems to be limited to selected strains. Strains that grow at low temperatures are mainly those isolated from cold habitats. Most strains grow at 40°C and 41°C, some at 43°C. The optimum temperature for growth is approximately 35°C. Palleroni et al. (251) subdivided P. stutzeri into two biotypes: one clustered around 62% G+C that does not tolerate a temperature of 43°C, and a second of around 65 to 66% G+C that grows at 43°C or higher.

Some strains (NF13, MT-1) have been isolated from the deep-sea bottom. Organisms adapted to the deep-sea environment have to grow under conditions of 2°C and 100-MPa pressure. On 28 February 1996, a sediment sample was obtained from the Mariana Trench by the unmanned submersible Kaiko. It seems likely that this was the first time sediment samples were collected from the world's deepest point without any microbiological contamination from other depths (351). The analysis of amplified 16S rRNA sequences from DNA directly extracted from these sediment samples demonstrated the presence of bacteria belonging to the P. aeruginosa branch (Mariana bacteria no. 2 [D87347] and no. 11 [D87346]). Pressure-regulated gene clusters were also amplified. Therefore, in addition to being barotolerant, the bacteria from the Mariana sediment may be barophilic microorganisms. Barophilic microorganisms were isolated by maintaining the conditions of 100 MPa and 4°C. Twenty-eight strains were selected. Strain MT-1, isolate HTA208, was grown on marine agar at 28°C and pH 7.6. Its 16S RNA sequence affiliates the strain with the P. stutzeri phylogenetic branch. It was able to grow at a hydrostatic pressure of 30 to 60 MPa, and slight growth occurred at 100 MPa. The growth rate of the P. stutzeri type strain was strongly affected by hydrostatic pressure. It must be clarified whether the isolated bacteria are active or inactive under high hydrostatic pressure and low temperature or whether their presence is simply a result of settling of flocculated organic matter.

As mentioned above, no strain tolerates acidic conditions: all fail to grow at pH 4.5. This is probably the reason why there is a negative reaction to the oxidation/fermentation test for the use of carbohydrates. Many P. stutzeri strains give a neutral result, as the medium is not buffered and acidification inhibits further growth, even when the strain might be able to use the added sugar.

P. stutzeri strains grow well under atmospheric oxygen. However, microaerophilic conditions have to be established when nitrogen-fixing strains are cultured as diazotrophs. All strains described to date are facultatively anaerobic with nitrate. Some strains are also anaerobic, with chlorate or perchlorate as terminal electron acceptors. Both anaerobic properties are discussed in the following section.

The denitrification process carried out by bacteria makes use of N oxides as terminal electron acceptors for cellular bioenergetics under anaerobic, microaerophilic, and occasionally even aerobic conditions (for reviews, see references 45, 77, 184, 263, and 420). During the denitrification process, which involves a pathway of four successive steps, several metalloproteins catalyze the reduction of nitrate to nitrite, nitric oxide (NO), and finally nitrous oxide (N₂O) to dinitrogen (N₂). The metalloenzymes include nitrate reductase (EC 1.7.99.4), nitrite reductase (EC 1.7.2.1 and EC 1.9.3.2), nitric oxide reductase (EC 1.7.99.7), and nitrous oxide (N₂O) reductase (EC 1.7.99.6) (152).

In contrast to the assimilatory reduction of nitrate or nitrite to ammonia for biosynthetic purposes, denitrification in bacteria is a dissimilatory transformation, associated with energy conservation (420). In other words, the enzymatic electron transfer is coupled to ATP synthesis via proton translocation and the formation of a membrane potential (347). The bacterial process of denitrification is normally a facultative trait. It provides bacteria with a respiratory pathway for anaerobic life (184, 420). The distribution of denitrification capabilities among the prokaryotes does not follow a clear pattern (263). The former genus Pseudomonas is one of the largest taxonomic clusters of known denitrifying bacteria. This fact has largely favored the use of species of the genus Pseudomonas as model organisms for studying the denitrification process. Within the genus Pseudomonas, and probably also within the prokaryotes, much of the relevant work, advances in the biochemical characterization of denitrification, and essential genetics using highly interdisciplinary approaches have been achieved with P. stutzeri. Denitrification is a stable trait for P. stutzeri; it is one of the most active denitrifying, heterotrophic bacteria, and it has been considered a model system for the denitrification process (420).

For over 50 years, it has been known that bacteria can reduce chlorate (ClO₃⁻) and perchlorate (ClO₄⁻) under anaerobic conditions. Many nitrate-reducing bacteria in pure cultures reduce chlorate and perchlorate [which is usually referred to as (per)chlorate] by means of membrane-bound respiratory nitrate reductases and assimilatory nitrate reductases. In all cases, chlorite (ClO₂⁻) is produced as a toxic end product. For many years, there was no evidence that these bacteria could grow using this metabolism. It is now known that specialized bacteria that can grow by anaerobic reductive dissimilation of (per)chlorate into innocuous chloride have evolved. For a recent review, see the work of Coates and Achenbach (75). All known dissimilatory (per)chlorate-reducing bacteria (DCRB) are facultatively anaerobic or microaerophilic bacteria. Some, but not all, are able to respire nitrate. The DCRB are phylogenetically diverse. Some isolates belong to the Gammaproteobacteria. The strains PK, CFPBD, PDA, and PDB have been affiliated with the P. stutzeri phylogenetic branch by 16S rRNA sequence analysis. They also seem to be affiliated with genomovars 3 and 1 or 5 in the phylogenetic trees. Strain AW1 is a member of P. stutzeri genomovar 3. Wolterink et al. have proposed this strain as the type and only member of a new species, P. chloritidismutans (404). Chlorate and perchlorate reduction specifically involves a c-type cytochrome in the transfer of electrons to (per)chlorate. A periplasmic oxygen-sensitive perchlorate reductase has been characterized recently (179). In addition to perchlorate, this reductase also reduces chlorate, nitrate, iodate, and bromate. The next biochemical step is the quantitative dismutation of chlorite into chloride and O₂ by chlorite dismutase. Chlorite dismutase is a highly specific enzyme that does not act on other analogous anions. Studies with a chlorite dismutase-specific immunoprobe indicated that this enzyme is present on the outer membrane of all DPRB and that it is highly conserved among these organisms, regardless of their phylogenetic affiliation. Expression of the chlorite dismutase gene (cld) is constitutive in strains PDA and PK. Genomic organization has been studied in strain PK: a gene encoding a c-type cytochrome lies between cld and a transposase gene. The transposase gene is followed by the chlorate reductase operon, with the gene order clrABDC (α subunit, β subunit, chaperone protein, and γ subunit). A recent comparison of a phylogenetic tree based on 16S rRNA with a tree developed from the cld gene sequences of 11 diverse DPRB demonstrated significant discrepancies. The results of this comparison supported evolution through horizontal gene transfer (22). As mentioned above, strain AW1 is a member of genomovar 3. The definition of this strain as a new species (P. chloritidismutans) was based only on its ability to use chlorate, rather than nitrate, as the terminal electron acceptor. However, careful adaptation to nitrate use enabled a denitrifying variant to be isolated. The only difference between this new species and P. stutzeri is its ability to dechlorinate (74). This characteristic is most probably acquired through horizontal gene transfer. This case demonstrates the danger of drawing taxonomic conclusions from homologies that are found in metabolic systems involved in the use of unusual substrates.

Extensive nutritional studies of carbon substrates (more than 150) used by P. stutzeri strains have been carried out by Stanier et al (340). Conventional methodologies and commercial kits were later used by other investigators (131, 295). These showed that intraspecies heterogeneity was high but P. stutzeri strains clustered separately from other phenons in the P. aeruginosa group, as pointed out by Palleroni and Doudoroff in their review on the genus Pseudomonas (253a).

In an exhaustive phenotypic study, Rosselló-Mora et al. (295) analyzed a total of 327 biochemical characteristics in 48 strains. The expanded physiological analysis did not improve the situation that was previously observed in the 102-characteristic study (291). This high degree of intraspecies physiological heterogeneity had already been observed by Palleroni and coworkers and by others (113, 251, 340, 371) and in other later numerical studies (115, 291). Numerical analysis of the data did not lead to genomovar-specific clusters but reflected considerable heterogeneity within genomovars. Characterization of individual genomic groups on the basis of biochemical tests was not possible.

All strains tested were positive for the following activities with the indicated substrates: growth on gluconate, d-glucose, d-maltose, starch, glycerol, acetate, butyrate, isobutyrate, isovalerate, propionate, fumarate, glutarate, glycolate, glyoxylate, dl-3-hydroxybutyrate, itaconate, dl-lactate, dl-malate, malonate, oxaloacetate, 2-oxoglutarate, pyruvate, succinate, d-alanine, d-asparagine, l-glutamate, l-glutamine, l-isoleucine, and l-proline and hydrolysis of l-alanine-para-nitroanilide. One hundred seven characteristics were variable in the 48 strains tested. The 48 phenotypic tests enabling the best discrimination between genomic groups of Pseudomonas stutzeri were selected. They are listed in the original publication by Rosselló-Mora and coworkers (295).

The mineralization, or cometabolism, of xenobiotics or anthropogenic substrates by specialized strains merits special attention. It is described in “Biodegradation and useful properties for biotechnological applications,” below.

The phylogenetic diversity of sulfur- and thiosulfate-oxidizing bacteria has been of interest to several authors. The evolution of such bacteria can be clarified by molecular investigation of the genes encoding these pathways' enzymes, together with an analysis of the 16S rRNA phylogenies of bacteria that have these properties. Strain NF13 was isolated by Ruby and coworkers from the Galapagos rift hydrothermal vents (190, 300). It was assumed that H₂S could be the predominant energy source for chemosynthesis in this habitat. Enrichment cultures under selective conditions for sulfur-oxidizing bacteria yielded several physiological groups of strains. One group, represented by strain NF13, was able to grow heterotrophically in media containing peptone or yeast extract either with or without thiosulfate. Acid was produced when thiosulfate was present. This group was unable to utilize CO₂ as the sole source of carbon. Of all the sulfur-oxidizing bacteria isolated, only strain NF13 was found to fix nitrogen. Strain NF13 was studied phylogenetically by Lane and coworkers using the 16S rRNA sequence (190). It was classified as an unnamed “thiobacillus” within the Gammaproteobacteria. Recent analysis (A. Cladera, personal communication) clearly placed isolate NF13 in the P. stutzeri phylogenetic branch. The phenotypic properties analyzed by Ruby and coworkers (300) were in accordance with ascription to the species: gram-negative, denitrifier, motile, nonfermentative.

More recently, Sorokin and coworkers analyzed the anaerobic oxidation of thiosulfate to tetrathionate by obligately heterotrophic bacteria (337). Strains were isolated from seawater and freshwater and from a sulfide-oxidizing bioreactor. All isolates were obligately heterotrophic. In phenotypic and genomic identification analyses, they were classified as members of P. stutzeri (seven strains) and P. balearica (one strain). Of the seven P. stutzeri isolates, four were ascribed by DNA-DNA hybridizations to genomovars 3, 4, and 5. The remaining isolates seemed to cluster in another genomic group. Strain ATCC 27951, of “Flavobacterium lutescens,” was included in this study (it had previously been reclassified as P. stutzeri by Bennasar et al. [25]). All seven strains oxidized thiosulfate to tetrathionate using nitrite, nitrate, or N₂O as an electron acceptor. Thiosulfate oxidation under anaerobic conditions was much slower than in the presence of oxygen. In addition, it was controlled by the availability of an organic electron donor. The oxidation of thiosulfate to tetrathionate (yielding one electron) instead of sulfate (yielding eight electrons) does not generate enough energy to support autotrophic growth (a high-energy-requiring process). It is therefore not surprising that tetrathionate-forming isolates are obligate heterotrophs (337). They grow under these circumstances as chemolithoheterotrophs, as the provision of thiosulfate increased their growth yield in acetate-limited continuous cultures.

Sijderius, in 1946, studied some P. stutzeri strains with this ability in his dissertation, “Heterotrophe Bacterien, die Thiosulfaat oxydereeren” (322). Whether this chemolithotrophy is restricted to specialized strains or is a general property of the species has not yet been studied. All strains studied were isolated from sulfur-rich environments, with the exception of strain ATCC 27951, which was isolated (surprisingly) from a yogurt in Algeria.

Pyoverdines are important pigments from a taxonomic and physiological perspective, as they function as efficient siderophores. Their production is enhanced under conditions of iron starvation. Although they are not pigmented, some strains of P. stutzeri synthesize siderophores. P. stutzeri ATCC 17588 produces desferriferrioxamines E and D2 (224, 225). A different strain (RC7) produces a catechol-like siderophore (63). No siderophores have been detected in P. stutzeri YPL-1 (202). Siderotyping is also a powerful technique for discriminating species within the genus Pseudomonas (225). It is interesting to note that the internal heterogeneity of the nonfluorescent species P. stutzeri is also reflected in its siderophore production capacity.

The best studied siderophore is produced by strain KC, the reference strain and only member of genomovar 9 (316). It is highly likely that this is a secondary siderophore in this strain (200). P. stutzeri KC can degrade carbon tetrachloride (CT) (trade names, Freon 10 and Halon 104) to carbon dioxide, chloride ions, and other nonvolatile compounds, such as formate. Chloroform is not formed in this process. CT is used in the manufacture of fluorocarbon propellants, as a cleaner and degreasing solvent, and as a fumigant. The EPA estimates that 10% of U.S. groundwater may be contaminated with CT. CT is a toxic, carcinogenic, and ozone-depleting xenobiotic compound detected in groundwater. For CT transformation to occur, P. stutzeri KC must be grown in an anaerobic, slightly alkaline medium at around pH 8. The high pH lowers iron solubility, limiting the iron concentration of the growth medium. The medium must also contain nitrate, an electron donor (such as acetate), and trace levels of copper. Strain KC secretes the molecule PDTC (pyridine-2,6-bisthiocarboxylate), which has a siderophore function. PDTC has been implicated in the uptake of other transition metals in addition to iron (200). It is believed to be an iron chelator that is fortuitously involved in CT transformation. Because of its ability to effectively scavenge iron, strain KC has a competitive advantage over bacteria that lack this ability. The iron chelator alone may transform CT. This is indicated by the fact that the supernatant alone (containing the chelator) from washed cells transforms CT. In its active form and in the presence of copper, PDTC transforms CT to carbon dioxide and other nonvolatile products, including formate and chloride ions.

A pdt locus corresponding to the PDTC biosynthetic pathway in the strain KC genome has been mapped and cloned as a 25,746-bp insert in the pT31 cosmid (199, 200, 315). The pdt operon was sequenced, and a low-density PDTC microarray consisting of 17 PCR-amplified ORFs was printed on chemically modified glass substrates. This was the first published attempt to apply microarray technology to the parallel detection of multiple target genes in P. stutzeri for environmental monitoring (235).

After years of controversy, it now seems that several strains that are unambiguously classified as true Pseudomonas species can be added to the list of nitrogen fixers, on the basis of physiological properties, nitrogenase assays, phylogenetic studies, and the detection of nifH by hybridization or PCR amplification and sequencing (66). All well-classified nitrogen-fixing Pseudomonas strains described in the literature are members of P. stutzeri (92). Former Pseudomonas strains from other species able to fix N₂ were transferred to other genera in the α- and β-Proteobacteria (“Pseudomonas paucimobilis,” “P. diazotrophicus,” and “P. saccharophila,” etc.). P. stutzeri's simultaneous capacity for nitrogen fixation and denitrification may be of relevance to overall nitrogen cycling in several ecosystems. P. azotofigens has only one strain (6H33b^T) and has been described recently (140) as a novel nitrogen-fixing species isolated from a compost pile.

Pseudomonas stutzeri A15 is a member of genomovar 1. It is a particularly predominant diazotrophic strain. It was isolated from the rice paddy rhizosphere and is widely used as a rice inoculant in China (136, 270, 380, 409). Previously identified as an Alcaligenes faecalis strain, it has been widely studied physiologically, biochemically, and genomically. Strain A15 is able to colonize and infect rice roots and to grow endophytically (92). It may provide rice plants with fixed nitrogen and hence promote plant growth. Another strain, P. stutzeri CMT.9.A, was isolated from the roots of a Sorghum nutans cultivar in Germany (189). It has not been assigned to any known genomovar. P. stutzeri JM300, the only member of genomovar 8, is a denitrifying soil isolate (18, 60). Strain ZP6b, classified as P. stutzeri “var. mendocina,” was isolated from the rhizosphere of capers (Capparis spinosa) in Spain (6). Functional nitrogenase activity was tested by the acetylene reduction assay and by the incorporation of ¹⁵N. Rates of acetylene reduction in the nitrogenase assay are in the range of those reported for other strains that are presently considered to be Pseudomonas species (66). Nitrogen fixation occurs at low oxygen tension under microaerobic conditions, as has been observed in many aerobic diazotrophs. The optimal pO₂ for nitrogen fixation by strain CMT.9.A is 0.01 atm. In addition to nitrogen-fixing ability, some P. stutzeri strains, such as JM300 and CMT.9.A, show hydrogenase activity (18), supplying additional energy for metabolism.

The nifH gene, encoding the nitrogenase protein, is considered to be the diagnostic gene for nitrogen fixation. It has been detected in all diazotrophic P. stutzeri strains studied to date. A nifHDK probe from Klebsiella pneumoniae gave a positive signal against the EcoRI-restricted total DNA of strain ZP6b (6). Both nifH and nifDK probes from Azospirillum brasilense also gave a positive signal against strain A1501 (a derivative of strain A15). Partial nifH sequences from strains A15 and CMT.9.A are identical. The nifH phylogeny in the domain Bacteria is largely congruent with phylogenetic trees based on 16S rRNA. They are within the cluster of other strains of the Gammaproteobacteria (380). Some strains have been analyzed for the presence of alternative nitrogenase systems. Genomic DNA from strains CMT.9.A and JM300 did not hybridize with a nifDK probe encoding the large subunit of Mo-nitrogenase in Azotobacter chroococcum (105). In addition, no signal was detected with the genes of the alternative nitrogenases anfDGK and vnfDGK. The other known pseudomonad that fixes nitrogen, P. azotofigens 6H33bT, has nifH and nifD genes that are closely related to the corresponding genes of strain A15.

Deduced amino acid sequences for the P. stutzeri A15 nifHDK operon had the highest identity (87 to 91%) with the respective A. vinelandii homologs (92). Moreover, the gene organization in the nifH region of P. stutzeri A15 was identical to the gene organization of A. vinelandii. The close relationship between the species has been revealed by a detailed phylogenetic analysis of A. vinelandii. This led to the suggestion that Azotobacter could be considered a Pseudomonas in disguise (277).

The regulation of nif genes has been studied only for strain A1501. This strain is similar to A15 (if not identical). Strain A1501 was reisolated from rice roots inoculated with strain A15. Results suggest that P. stutzeri has a large number of NtrC family response regulators, as does P. aeruginosa. Differences were detected in a P. stutzeri NtrB mutant that had impaired nitrogen fixation. No such differences were found in the case of Azotobacter spp. Moreover, NifA controls NtrBC expression. This may reflect the fact that the regulatory circuit is different from other organisms. The on/off switch of nitrogenase activity and the role of the nif genes also reflect interesting features of this particular strain. A study to identify which genes are switched on during rice colonization and switched off during free-living growth on a synthetic medium has been conducted by Rediers et al. (276). Some of the corresponding genes are involved in stress response, chemotaxis metabolism, and global regulation. Others encode putative proteins that have either unknown functions or no significant homology to known proteins.

A 50-kb plasmid that may carry some genetic information for nitrogen fixation in strain CMT.9.A has been identified. When this plasmid is cured, there is an associated loss of N₂ fixation capability (189). No plasmid was detected in strain A1501. This is consistent with the assumption that nif genes have a chromosomal localization in strain A1501. The origin of the nitrogen-fixing genes in some strains of P. stutzeri can be explained by a plausible lateral gene transfer acquisition. This hypothesis seems to be supported by nifH phylogenies (380).

A number of bacteria have been shown to be capable of oxidizing reduced phosphorous compounds when these are provided as the sole source of phosphorous. Inorganic phosphite and hypophosphite can be used as such a source. They are oxidized to phosphate by several species, including members of the genus Pseudomonas. In a screening of bacteria that oxidize reduced phosphorous compounds to phosphate, Metcalf and Wolfe, in 1998, were able to isolate 10 bacterial strains after enrichment under selective conditions (223). Strain WM88 was studied in detail. It is a P. stutzeri strain that is closely related to the genomovar 3 reference strain DSM50227. Related collection strains, including four known P. stutzeri strains, were not able to oxidize hypophosphite. However, strain DSM50227 (a clinical isolate) was able to oxidize phosphite. The genes required by strain WM88 have been cloned and studied, together with the respective enzymes. Two operons in the Pho regulon (htx and ptx) are needed to use phosphite and hypophosphite as alternative P sources (395). An Alcaligenes faecalis strain was isolated recently. This strain has htx genes that are virtually identical to their homologs in P. stutzeri, indicating that horizontal gene transfer may have occurred. However, pdtx in A. faecalis is different from the homolog in P. stutzeri (400). Whether the htx genes are widely distributed in P. stutzeri or restricted to strain WM88 remains to be studied.

Pseudomonas stutzeri is a ubiquitous bacterium with a high degree of physiological and genetic adaptability. It is present in a large number of different natural environments (see “Habitats and ecological relevance,” below). Like other Pseudomonas species (e.g., P. putida), P. stutzeri is involved in environmentally important metabolic activities. Some of its major tasks are metal cycling and degradation of biogenic and xenobiotic compounds (oil derivatives—aromatic and nonaromatic hydrocarbons—and biocides).

The composition of the bacterial rhizosphere population, and in particular that of the diazotrophic bacteria, is of major interest. New isolation media and enrichment conditions have been developed with low oxygen tensions simulating rhizosphere conditions. This has led to the conclusion that the genus Pseudomonas is dominant or predominant in association with wheat, barley, and wetland rice (66, 93, 208). The role of diazotrophic P. stutzeri strains in soils might be more relevant than previously considered. A recent study involving PCR and denaturing gradient gel electrophoresis analysis of the N₂-fixing bacterial diversity in soil revealed a high percentage of nifH genes identical to those of P. stutzeri (93). Molecular analysis of diazotroph diversity in the rhizosphere of smooth cordgrass (Spartina alterniflora) suggests that P. stutzeri-related strains are present in the Spartina rhizosphere. Recently, analysis of bacterial populations in the rhizosphere of cordgrass, based on PCR amplification of nifH sequences and separation of the amplicons by denaturing gel electrophoresis, revealed nifH sequences highly similar to those of strains A1501 (a derivative of strain A15) and CMT.9.A (208, 209). The activity of an aromatic amino acid aminotransferase and the production of indole-3-acetic acid in P. stutzeri A15 have also been reported. This may be involved in the production of growth-regulating substances in plants in addition to their nitrogen-fixing ability (261).

As mentioned above, many P. stutzeri strains have been isolated from contaminated soil sites, where degradative and contaminant-resistant strains have to develop relevant ecological activities. Some strains, such as KC, and several methyl-naphthalene-degradative strains have been isolated in our laboratory from groundwaters contaminated with aircraft fuel (JetA1). The efficacy of strain KC in detoxifying groundwaters has been shown through bioaugmentation.

Most strains isolated from marine environments and initially classified in the genus Pseudomonas have been transferred to other genera after an analysis of their phylogenies. These transfers include P. doudoroffii to Oceanimonas doudoroffii, P. nautica to Marinobacter hydrocarbonoclasticus, P. stanieri to Marinobacterium stanieri, P. elongata to Microbulbifer hydrolyticus, and P. marina to Cobetia marina. Not many species within the genus Pseudomonas sensu stricto have been detected in marine waters. For a strain to be considered of marine origin, it must have the physiological characteristic of requiring, or at least tolerating, NaCl. P. stutzeri (including strain ZoBell, formerly P. perfectomarina), P. balearica, and P. xanthomarina (isolated from ascidian specimens in the Sea of Japan [289]) seem to be true marine Pseudomonas species. In addition, P. alcaliphila and P. aeruginosa (181) have been isolated from marine waters. Further research is required to define whether the latter pseudomonads might be considered marine bacteria or allochthonous to the ecosystem.

Marine strains of P. stutzeri are located in the water column and in sediment. The most relevant strains studied in detail are ZoBell (isolated from the water column in the Pacific ocean and studied as a model denitrifier in marine environments), AN10 (isolated from polluted Mediterranean marine sediment and studied as a naphthalene degrader), NF13 (isolated from a sample taken at 2,500- to 2,600-m depth in the Galapagos rift from near a hydrothermal vent and studied as a strain that oxidizes sulfur chemolithotrophically), and strains MT-1 and HTA208 (isolated from deep-sea samples taken at the Mariana Trench at 10,897-m depth). The main ecological role of these strains seems to be denitrification, besides their specific physiological properties.

The study by Sikorski et al. (325) is the only one in which a large number of P. stutzeri strains have been isolated from the same sample, in this case marine sediment from the shore of the North Sea. This enabled a genetic study of the populations present in a single habitat to be undertaken.

The ability of P. stutzeri to oxidize thiosulfate to tetrathionate both aerobically and anaerobically was not known before the work of Sorokin et al. (337). Several strains were isolated from the Black Sea at more than 100 m in depth. It was suggested that this widespread bacterium could be important in the turnover of thiosulfate in marine environments and that it may compete with thiosulfate disproportionation and reduction by thiosulfate-reducing bacteria.

Spartina marshes support high rates of macrophyte primary production and microbially mediated nutrient cycling. The possible ecological role of P. stutzeri in such marshes seems to be its contribution to global carbon and nitrogen budgets. Primary production and decomposition in Spartina marshes are nitrogen limited (208). In these systems, diazotrophy is a key source of new nitrogen, and denitrification completes the nitrogen cycle. P. stutzeri participates in both processes. Direct molecular analysis of diazotrophic diversity in the rhizosphere of Spartina alterniflora demonstrates that gene sequences of nifH are highly similar to those of P. stutzeri. In addition, they are located in the same phylogenetic branch as many other sequences of nif genes obtained from marine microorganisms.

To screen bacteria with unusual metabolic properties, such as the degradation of anthropogenic compounds for bioremediation purposes, it is common to examine samples taken from wastewater treatment plants or to design bioreactors imitating the conditions of a treatment plant. Naphthalene degraders, thiosulfate oxidizers, chlorobenzoate degraders, and cyanide oxidizers have been isolated in this way. It has been demonstrated that P. stutzeri is also distributed in wastewater. However, no attempt has been made to quantify P. stutzeri in such habitats or to determine its relevance.