Sequence Analysis of Amplified DNA Fragments Containing the Region Encoding the Putative Lipase Substrate-Binding Domain and Genotyping of Aeromonas hydrophila

Abstract

DNA fragments were amplified by PCR from all tested strains of Aeromonas hydrophila, A. caviae, and A. sobria with primers designed based on sequence alignment of all lipase, phospholipase C, and phospholipase A1 genes and the cytotonic enterotoxin gene, all of which have been reported to have the consensus region of the putative lipase substrate-binding domain. All strains showed lipase activity, and all amplified DNA fragments contained a nucleotide sequence corresponding to the substrate-binding domain. Thirty-five distinct nucleotide sequence patterns and 15 distinct deduced amino acid sequence patterns were found in the amplified DNA fragments from 59 A. hydrophila strains. The deduced amino acid sequences of the amplified DNA fragments from A. caviae and A. sobria strains had distinctive amino acids, suggesting a species-specific sequence in each organism. Furthermore, the amino acid sequence patterns appear to differ between clinical and environmental isolates among A. hydrophila strains. Some strains whose nucleotide sequences were identical to one another in the amplified region showed an identical DNA fingerprinting pattern by repetitive extragenic palindromic sequence-PCR genotyping. These results suggest that A. hydrophila, and also A. caviae and A. sobria strains, have a gene encoding a protein with lipase activity. Homologs of the gene appear to be widely distributed in Aeromonas strains, probably associating with the evolutionary genetic difference between clinical and environmental isolates of A. hydrophila. Additionally, the distinctive nucleotide sequences of the genes could be attributed to the genotype of each strain, suggesting that their analysis may be helpful in elucidating the genetic heterogeneity of Aeromonas.

Aeromonas hydrophila is widely distributed in the aquatic environment (1, 14) and has been implicated in human infectious diseases (20), such as gastroenteritis (11) and septicemia (21). A. hydrophila produces a variety of extracellular enzymes, including protease (22), elastase (6), amylase (15), chitinase (24), lipases (2, 9), phospholipases (18, 23), hemolysins (3, 17), and enterotoxins (7, 8, 26, 27).

Two lipase (lip and lipH3) genes, phospholipase C (apl-1) gene, phospholipase A1 (pla) gene, and cytotonic enterotoxin (alt) gene were found from distinct strains of A. hydrophila. The lip and lipH3 genes code for extracellular lipases, which are similar in that they have esterase but not phospholipase activities (2, 9). The apl-1 gene codes for a lipase and exhibits nonhemolytic phospholipase C activity (18). The pla gene codes for a putative lipoprotein; it is thought that this lipoprotein is a lipase with phospholipase A1 activity but is nonhemolytic, noncytotoxic, and nonenterotoxic (23). Although physical properties of these enzymes, such as molecular size and thermal stability as well as the substrate specificities, differ from one another, the lip, lipH3, apl-1, and pla genes show similarities in their amino acid sequences (23). While the DNA size of the alt gene is about half that of the above genes, this gene shows amino acid similarity to the 3′ half of the above genes (8, 23). The alt gene codes for a heat-labile cytotonic enterotoxin; the purified native toxin exhibits the ability to elongate CHO cells and to elicit fluid secretion in rat ligated intestinal loops but had no lipase or phospholipase C activity (8). These lipases, phospholipases, and cytotonic enterotoxin have a consensus amino acid sequence, V-H-F-L-G-H-S-L-G-A, corresponding to the region of putative substrate-binding domains found within bacterial, fungal, porcine, and human lipases (12).

In this study we attempted to define the distribution of the genes which have the region corresponding to the putative lipase substrate-binding domain found in the lip, lipH3, apl-1, pla, and alt genes among clinical and environmental isolates of A. hydrophila belonging to serogroups O11, O16, and O34, which are considered to be the major O serogroups of the genus Aeromonas (19). PCR was applied to amplify the DNA fragments, including the region corresponding to that domain. The PCR products were then sequenced to identify the amplified DNA genes.

Genetic heterogeneity in A. hydrophila has been studied by methods such as random amplified polymorphic-PCR (10, 25) and enterobacterial repetitive intergenic consensus-PCR (10). Also, repetitive extragenic palindromic sequence-PCR (REP-PCR) is one of the useful genotyping methods to examine genetic heterogeneity (4). We compared the REP-PCR DNA fingerprints among the tested A. hydrophila strains and examined the correlation between the genetic heterogeneity and the nucleotide sequences of the above PCR products. In addition, we applied the above experimental analyses on strains of A. caviae and A. sobria to compare with A. hydrophila results.

MATERIALS AND METHODS

Bacterial strains and genes.

Aeromonas strains used in this study are described in Fig. 1. The profiles of the reference genes, lip, lipH3, apl-1, pla, and alt, are listed in Table 1.

FIG. 1.

Aeromonas strains in this study and comparison of the REP-PCR DNA fingerprints between the nucleotide and the amino acid sequence groups. For the source of strains, all feces were obtained from patients with diarrhea except strain 014, which was isolated from feces of a healthy adult. Environmental sources included food, river water and soil, and seawater. Lipase activity was detected on a tributyrin agar medium. The low-molecular-weight fragments of the REP-PCR DNA fingerprints are shown in the left side of photographs. The details of the nucleotide sequence groups and the deduced amino acid sequence groups are shown in Fig. 3 and 4, respectively.

TABLE 1.

Lipase, phospholipase, and cytotonic enterotoxin genes cloned from A. hydrophila strains

Gene	Coding protein	DNA size of ORFa (bp)	Original strain	GenBank accession no.	Reference
lip	Extracellular lipase	2,253	MCC-2	U63543	9
lipH3	Extracellular lipase	2,052	H3	S65123	2
apl-1	Phospholipase C	2,052	JMP636	U14011	18
pla	Phospholipase A1	2,415	AH-3	AF092033	23
alt	Cytotonic enterotoxin	1,104	SSU	L77573	8

^aORF, open reading frame.

Detection of lipase activity.

Lipase activity of the tested strains was detected on tributyrin agar medium containing 0.5% peptone (Difco), 0.3% yeast extract (Difco), and 1.5% agar (Oxoid) supplemented with 1% tributyrin (18). After inoculation the tributyrin agar medium was incubated at 37^°C for 48 h. Escherichia coli ATCC 11775 was used as a negative control strain of lipase activity.

Preparation of template DNA solution for PCR and REP-PCR.

The bacterial cells were picked from colonies grown on tryptic soy agar (Eikenkagaku Co., Tokyo, Japan), suspended in sterile distilled water, and heated in boiling water for 10 min. After centrifugation of the suspension at 9,600 × g for 10 min, the supernatant was used as the template DNA solution for PCR and REP-PCR.

PCR amplification and nucleotide sequencing.

The primers were designed to detect the genes having the region corresponding to the lipase putative substrate-binding domain and to amplify this region on the basis of a multiple alignment of nucleotide sequences of the lip, lipH3, apl-1, pla, and alt genes (Fig. 2). The first PCR was performed with the primer pair AHL-L1 (5′-TCGGTSAARGARAACGCYTAYG-3′) and AHL-R1 (5′-VGTRKYGAACTTGAACAGRGCRTC-3′); the nucleotide symbols K, R, S, V, and Y correspond to G or T, A or G, C or G, A, C, or G, and C or T, respectively. Nested PCR was performed by using the first PCR product as the template DNA with the primer pair AHL-L2 (5′-TGAACCTGAGCTACCTGACC-′) and AHL-R2 (5′-CGAACTTGAACAGAGCATCC-3′), and this nested PCR product was used as the DNA template for the direct sequencing described below. The first PCR mixtures contained 2 μl of 10× EX Taq buffer (Takara Bio Co., Shiga, Japan), 1.6 μl of 25 mM MgCl₂ (Takara Bio Co.), 1.6 μl of 2.5 mM deoxynucleoside triphosphates (dNTPs; Takara Bio Co.), 0.24 μl of 50 μM primers (AHL-L1 and AHL-R1), 0.05 μl of EX Taq (5 U/μl; Takara Bio Co.), 2 μl of template DNA solution, and sterile distilled water to 20 μl. The first PCR amplification was performed by following the touchdown PCR method program (13): preincubation at 94°C for 3 min followed by 2 cycles at 94°C for 45 s (denaturation), 62°C for 30 s (annealing), and 72°C for 90 s (extension), and then 2 identical cycles but with an annealing temperature of 60°C followed by 6 identical cycles with the annealing temperature reduced 1^°C per cycle from 58°C to 53°C, 25 identical cycles with the annealing temperature at 52°C, and a final extension at 72°C for 10 min. The nested PCR mixtures contained 5 μl of 10× EX Taq buffer (Takara Bio Co.), 4 μl of 25 mM MgCl₂ (Takara Bio Co.), 4 μl of 2.5 mM dNTPs (Takara Bio Co.), 1.5 μl of 10 μM primers (AHL-L2 and AHL-R2), 0.25 μl of EX Taq (5 U/μl; Takara Bio Co.), 5 μl of a 10,000-fold dilution of the first PCR product as template DNA, and sterile distilled water to 50 μl. The nested PCR amplification was performed by using the following temperature program: preincubation at 94°C for 3 min and then 35 cycles at 94°C for 30 s (denaturation), 65°C for 30 s (annealing), and 72°C for 60 s (extension); the final extension was at 72°C for 10 min. The first and nested PCR-amplified DNA fragments were confirmed by electrophoresis in 2.5% agarose gel and visualized with ethidium bromide staining.

FIG. 2.

Primer annealing sites and the region corresponding to the putative lipase substrate-binding domain in the nucleotide sequence alignment of the lip, lipH3, apl-1, pla, and alt genes. Identical nucleotides are shaded black. Primer annealing sites are marked by > or < symbols.

The nested PCR product was sequenced by using a Thermo Sequenase Cy5.5 Dye terminator cycle sequencing kit (Amersham Biosciences) with primer AHL-R2. After the cycle sequencing reaction the products were analyzed on a Gene Rapid sequencer (Amersham Biosciences), and the results were analyzed by using SEQ 4 × 4 Basecaller software (Amersham Biosciences).

REP-PCR DNA fingerprints.

REP-PCR was performed with the following primer pair designed by Wong and Lin (28): REP-1D (5′-NNNRCGYCGNCATCMGGC-3′) and REP-2D (5′-RCGYCTTATCMGGCCTAC-3′); the nucleotide symbols M and N correspond to A and C and to A, C, G, and T, respectively. The REP-PCR mixtures contained 2 μl of 10× EX Taq buffer (Takara Bio Co.), 1.6 μl of 25 mM MgCl₂ (Takara Bio Co.), 1.6 μl of 2.5 mM dNTPs (Takara Bio Co.), 2 μl of 10 μM primers (REP-1D and REP-2D), 0.1 μl of EX Taq (5 U/μl; Takara Bio Co.), 2 μl of template DNA solution, and sterile distilled water to 20 μl. The REP-PCR amplification was performed by using the following temperature program: preincubation at 94°C for 3 min and then 30 cycles at 94°C for 30 s (denaturation), 60°C for 60 s (annealing), and 70°C for 3 min (extension); the final extension was at 72°C for 16 min. REP-PCR amplification fragments were electrophoresed in 2.5% agarose gels and were photographed on a UV light transilluminator after staining with ethidium bromide.

RESULTS

Lipase activity of strains.

A clear halo indicating lipase activity appeared around colonies of all A. hydrophila, A. caviae, and A. sobria strains tested on tributyrin agar medium (Fig. 1). A halo did not appear around colonies of E. coli.

PCR amplification and nucleotide sequence analysis.

The first PCR with primer pair AHL-L1 and AHL-R1 and also the nested PCR with primer pair AHL-L2 and AHL-R2 allowed amplification of DNA fragments from all A. hydrophila, A. caviae, and A. sobria strains tested (Fig. 2). Sequence analysis with primer AHL-R2 was used to determine the nucleotide sequence of the target region in all of the nested PCR amplified fragments. The nucleotide sequences of the amplified fragments from the various Aeromonas strains differed and were sorted into 38 distinct nucleotide sequence groups (Fig. 3). In the A. hydrophila strains, 35 distinct nucleotide sequence groups were found, and 23 of them were present in only one strain. One identical sequence group was found in three A. caviae strains, and two distinct sequence groups were found in three A. sobria strains. Although the DNA size of most of the nucleotide sequence groups in the A. hydrophila strains and group b-36 in the A. caviae strains was 111 bp, four groups (b-31, b-32, b-33, and b-34) from A. hydrophila strains had 108-bp fragments, and group b-35 in A. hydrophila and both groups b-37 and b-38 in A. sobria had 117-bp fragments. One of the nucleotide sequence groups (b-08) of A. hydrophila had a sequence identical to the corresponding region of the pla gene. A 30-bp nucleotide sequence encoding the amino acids V-H-F-L-G-H-S-L-G-A, which corresponds to the putative lipase substrate-binding domain in the lip, lipH3, apl-1, pla, and alt gene products, was found in all sequence groups. When the amino acid sequence was deduced and aligned on the basis of this consensus sequence, 15 distinct amino acid sequence groups were sorted (Fig. 4). There were 12 distinct deduced amino acid sequence groups in A. hydrophila strains, and 7 of the groups (aa-C, aa-E, aa-F, aa-G, aa-H, aa-K, and aa-L) were unique and were found in only one strain each. The amino acid sequence groups aa-J and aa-K, which were deduced from 108-bp nucleotide sequences, were able to align with other groups but had a 1-amino-acid deletion. Groups aa-L, aa-N, and aa-O, which were deduced from 117-bp nucleotide sequences, were able to align but had a 3-amino-acid insertion and a 1-amino-acid deletion relative to the other sequence groups. Although the amino acid sequence group aa-I was deduced from a 111-bp nucleotide sequence group, its sequence had a 1-amino-acid insertion and one deletion. In the deduced amino acid sequence alignment, the N-terminal three amino acids (positions −18 to −16) were identical in all strains tested. The C-terminal 23 deduced amino acids (positions −1 to 22), which contained the putative lipase substrate-binding domain, were also identical in all A. hydrophila strains, but two amino acids (positions 12 and 14) in all A. caviae and all A. sobria strains differed from those of A. hydrophila. The amino acid sequence groups aa-B and aa-D in A. hydrophila were identical to the pla and alt genes, respectively, in the deduced amino acid sequences of the region. Although the nucleotide sequences of the apl-1, lipH3, and lip genes were almost identical to those of the amplified fragments of the tested strains within the aligned region, the deduced amino acid sequences of these genes differed from those of the tested strains because of some deletions in the nucleotide sequences.

FIG. 3.

Comparison of the nucleotide sequence groups in the amplified fragments and the reported genes: the lip, lipH3, apl-1, pla, and alt genes. The groups are sorted by their nucleotide sequences. The identical nucleotides are shaded black. The details of deduced amino acid sequence groups are shown in Fig. 4.

FIG. 4.

Comparison of the deduced amino acid sequence groups in the amplified fragments and the reported genes: the lip, lipH3, apl-1, pla, and alt genes. The amino acid sequences were deduced and aligned based on the sequence of putative lipase substrate-binding domain. Identical amino acids are shaded black.

REP-PCR DNA fingerprints.

REP-PCR was used to compare the DNA fingerprints of the tested strains. The REP-PCR DNA fingerprints in Fig. 1 are arranged by the nucleotide sequence groups described above. The DNA fingerprinting patterns closely resembled one another between the strains sorted in the following nucleotide sequence groups: group b-14 (A. hydrophila 104 and 105), group b-18 (A. hydrophila 096, 097, 098, and 099), group b-20 (A. hydrophila 082 and 084), group b-30 (A. hydrophila 048 and 102), group b-36 (A. caviae 055 and 056), and group b-37 (A. sobria 080 and 012). The strains of other groups showed different patterns.

Source of strains, sequence groups, and DNA fingerprints.

All the strains sorted into the deduced amino acid sequence groups aa-A and aa-D were clinical isolates. Twenty-two of the 31 strains sorted into the deduced amino acid sequence group aa-B were environmental isolates, and there were only five environmental isolates sorted into the other groups. Seven of the deduced amino acid sequence groups were found in only one strain each. All of these strains were clinical isolates, but one was an environmental isolate. One clinical isolate with one environmental isolate was sorted into each of the nucleotide sequence groups b-30 and b-31 (Fig. 1).

DISCUSSION

We examined the distribution of the genes which have the region corresponding to the putative lipase substrate-binding domain by PCR in clinical and environmental isolates of A. hydrophila and also A. caviae and A. sobria strains belonging to serogroups O11, O16, and O34. The PCR primers were designed based on a multiple alignment of nucleotide sequences of the lip, lipH3, apl-1, pla, and alt genes. Therefore, the primers enabled the first PCR and subsequent nested PCR to amplify DNA fragments of the predicted size from the multiple alignment of nucleotide sequences of those genes from all Aeromonas strains tested. To identify the genes amplified from each strain, we sequenced all nested PCR-amplified DNA fragments. The sequenced region includes the target site that apparently corresponds to the putative lipase substrate-binding domain found in the lip, lipH3, apl-1, pla, and alt genes. Sequencing of the target region in the amplified DNA fragments identified nucleotide sequences corresponding to the putative lipase substrate-binding domain, and their deduced amino acid sequences were identical to the consensus sequence V-H-F-L-G-H-S-L-G-A, found in the lip, lipH3, apl-1, pla, and alt genes. Because lipase activity was observed on tributyrin agar medium and a region corresponding to the putative lipase substrate-binding domain was found in all strains, A. hydrophila and also A. caviae and A. sobria strains probably have genes coding for proteins with lipase activity. The alignment of the nucleotide sequences of the target region showed that the nucleotide sequences were sorted into 35 different groups by their sequence patterns. While the sequences of many groups were similar to each other with some substitutional nucleotides, few other sequence groups had unique sequences with either additional or fewer nucleotides in the target region. The nucleotide sequences of some of the amplified DNA fragments were very similar to the corresponding regions of previously reported genes, indicating the presence of polymorphic sequences of the reported genes. Consequently, nucleotide sequence analysis of only the amplified fragment is not sufficient for conclusive identification of the gene. In addition, the amplified DNA fragments that have unique nucleotide sequences may be from unknown genes. Because some nucleotide substitutions, additions, and deletions were found at single or multiple loci in the sequence alignment of the amplified DNA fragments, the identification of the genes from which the DNA fragments were amplified may not be possible with PCR or restriction fragment length polymorphism analysis. Cascón et al. have reported that the lipH3 gene was detected by PCR from all A. hydrophila strains included in DNA hybridization group 1 (5). However, the primers used in the PCR were designed from lipH3 gene sequences that were highly homologous to the lip, apl-1, and pla genes. Therefore, the PCR with these primers probably detected the lip, apl-1, and pla genes as well as unreported genes instead of specifically detecting the lipH3 gene. Granum et al. have reported that the alt gene was detected by PCR from all A. hydrophila strains tested (16), but they also used primers that may detect the lip, lipH3, apl-1, and pla genes as well as other homologous unreported genes.

When deduced amino acid sequences of the N- and C-terminal regions of the putative substrate-binding domain sequence in the target region were aligned, the deduced amino acid sequence patterns were categorized into 12 groups in A. hydrophila strains. The deduced amino acid sequences of the target region were identical to those of the pla gene and the alt gene in more than half and five of the strains, respectively. However, the deduced amino acid sequences in other strains were not identical to those of any reported genes. In addition, the amino acid sequences of the lip, lipH3, and apl-1 genes differed markedly in the target region from those deduced from the amplified DNA fragments of Aeromonas strains. The differences in the nucleotide sequences did not result in variations in the amino acid sequences of the target region that are identical to the corresponding region of the pla or alt gene product. However, variations in the sequence, such as nucleotide substitutions, insertions, or deletions in other regions besides the target region in this study, may result in alterations in the amino acid sequence. Therefore, sequence analysis of only the amplified target region does not allow conclusive identification of the genes as pla or alt. These findings indicate that identification of the amplified DNA fragments from the tested strains will require complete nucleotide sequence analysis of the corresponding structural genes.

In the alignment analysis of the deduced amino acid sequences, identical sequences in all A. hydrophila strains were found in the N-terminal 3 and the C-terminal 23 amino acids, with the latter region containing the putative lipase substrate-binding domain sequence. These consensus sequences were also found in the groups with the unique amino acid sequences that contained two additional or one fewer amino acid than other sequences, suggesting the presence of highly conserved regions in A. hydrophila strains. In contrast, the deduced amino acid sequences of the amplified fragments differed markedly from those regions thought to be highly conserved in the lip, lipH3, and apl-1 genes. Therefore, the lip, lipH3, and apl-1 genes may be strain-specific genes of each strain in which these genes were found. The results of these alignment analyses indicate that lipase-like proteins in A. hydrophila have both conserved and variable regions of amino acid sequence. Since all strains showed lipase activity, these variable regions appear not to be associated with the lipase activity of the lipase-like protein in each strain. Thus, when the other regions of the gene that were not analyzed in this study are sequenced, more diversity of deduced amino acid sequence should be found among the genes, and when the number of analyzed strains is increased, more diversity should be found in the aligned sequences among the strains. Therefore, the amplified DNA fragments should sort into more nucleotide sequence groups. Accordingly, the genes from which the amplified DNA fragments came may actually be unknown genes containing nucleotide sequences that are highly homologous to the reported genes: the lip, lipH3, apl-1, pla, and alt genes. It is possible that the genes encoding proteins with lipase activity are classified as homologs, and distinct homologs (those with both distinctive amino acid sequence and the consensus amino acid sequence, including the putative lipase substrate-binding domain) seem to be widely distributed in A. hydrophila. It is interesting to consider how many distinct homologs are distributed in A. hydrophila and how the characteristics of the homologs may differ.

We also applied the above experimental analysis to strains of A. caviae and A. sobria. DNA fragments that contained the region corresponding to the putative lipase substrate-binding domain were PCR amplified from all three A. caviae and three A. sobria strains. The amino acid sequences deduced from the target region in the amplified DNA fragments from the A. sobria and A. caviae strains differed between the species and were not identical to those of any of the A. hydrophila strains. The deduced amino acid sequences of both the A. caviae and A. sobria strains were identical to that of the A. hydrophila strains within the N-terminal 3 and C-terminal 23 amino acids, which are the previously described highly conserved sequences, except for two amino acid substitutions. These substituted amino acids were found in the same two distinct positions in both the A. sobria and A. caviae strains, suggesting species-specific variations in the sequence. Therefore, the fragments amplified from A. caviae and A. sobria are probably gene homologs that differ from those in the A. hydrophila strains.

REP-PCR DNA fingerprinting was applied to examine the relationship between the genetic diversity of the PCR-amplified DNA fragments and the genetic heterogeneity in Aeromonas strains. The results demonstrated that there was marked heterogeneity in the tested strains. A close resemblance of the DNA fingerprinting patterns was found in some strains that sorted into the same nucleotide sequence groups, and therefore the genotype may be closely related to the nucleotide sequence of the amplified fragment in these strains. However, cases in which the nucleotide sequences of the amplified DNA fragments sorted into the same nucleotide sequence groups but showed different DNA fingerprinting patterns suggest that the DNA sequence in other regions of the gene that were not analyzed in this study may differ. Therefore, additional nucleotide sequence differences are likely to be found in other regions of the gene.

Further, the relationship between the deduced amino acid sequence groups and source of strains tested was examined. More than half of the A. hydrophila strains were sorted into the amino acid sequence group aa-B. Twenty-two of the 31 A. hydrophila strains sorted into this group were environmental isolates, and there were only five environmental isolates sorted into the other groups. This finding leads to the hypothesis that the amino acid sequence of the group aa-B is identical to a region of an original amino acid sequence of the genes that encode a protein with lipase activity; therefore, these gene homologs encoding a protein with lipase activity were derived from a gene distributed among environmental strains as the origin. In addition, the clinical isolates were sorted into many distinct amino acid sequence groups, unlike the environmental isolates; moreover, six of the seven strain-specific amino acid sequence groups were found in the clinical isolates. These findings suggest that there are many gene homologs encoding a protein with lipase activity in the clinical isolates, unlike the case with the environmental isolates. This suggestion leads to the additional hypothesis that the clinical isolates of A. hydrophila strains are variants that were derived from the environmental A. hydrophila strains as a consequence of evolutionary genetic changes. One of the evolutionary genetic changes in A. hydrophila probably appeared as the genetic diversity among the amplified DNA fragments in this study. Besides, the clinical isolates of A. hydrophila possibly acquired some virulence while the genetic variations and additions involved in the evolutionary genetic changes have occurred. Some clinical isolates were sorted into the amino acid sequence group aa-B. Additionally, some clinical and environmental isolates were sorted into the same nucleotide sequence groups. These incompatibilities could be elucidated by analyzing the diversity of the nucleotide sequences of the other regions of the gene. In addition to this analyzing, the investigation of diversity of the other genes, especially virulence genes, may clarify the evolutionary relationship between the clinical and environmental isolates of A. hydrophila. Meanwhile, the relationship between the sources of strains and the REP-PCR DNA fingerprinting patterns was not observed. Since the genetic heterogeneity of A. hydrophila appears to be complicated, the genetic differences between the clinical and environmental isolates were probably hidden in the distinctive overall genetic heterogeneity of each strain.

In this study we amplified DNA fragments of genes that appear to encode proteins with lipase activity from all tested A. hydrophila, A. caviae, and A. sobria strains. These amplified DNA fragments are probably portions of reported or unknown gene homologs, and the gene homologs appear to be widely distributed in the A. hydrophila, A. caviae, and A. sobria strains. The distinctive gene homologs from the various species are likely to differ. Additionally, the distinctive gene homologs probably differ between clinical and environmental isolates of A. hydrophila. In order to identify these homologs, sequencing of the complete structural genes would be necessary. Furthermore, this analysis may also clarify the relationship of the diversity in the genes to the genotyping of Aeromonas strains. Further investigations of the diversity in the gene homologs may be helpful to elucidate the genetic heterogeneity of these strains. In addition, analysis of homologs of other genes, especially virulence genes, may clarify the evolutionary genetic variation of Aeromonas.