Mesophilic
Aeromonas spp. are ubiquitous waterborne bacteria and pathogens of reptiles, amphibians, and fish (
5). They can be isolated as a part of the fecal flora of a wide variety of other animals, including some used for human consumption, such as pigs, cows, sheep, and poultry. In humans,
Aeromonas hydrophila belonging to hybridization group 1 (HG1) and HG3,
A. veronii biovar sobria (HG8/HG10), and
A. caviae (HG4) have been associated with gastrointestinal and extraintestinal diseases, such as wound infections of healthy humans, and less commonly with septicemia of immunocompromised patients (
30). The swimming motility of all mesophilic aeromonads has been linked to a single polar unsheathed flagellum, expressed constitutively, which is required for adherence to and invasion of human and fish cell lines (
25,
43,
53,
63). Moreover, 50% to 60% of mesophilic aeromonads are able to produce many unsheathed peritrichous lateral flagella when grown in viscous environments or over surfaces (
58), which increase bacterial adherence and are required for swarming motility and biofilm formation (
23). The expression of two distinct flagellar systems is relatively uncommon, although it has been observed with
Vibrio parahaemolyticus (
39),
Azospirillum brasilense (
45),
Rhodospirillum centenum (
32),
Helicobacter mustelae (
49), and
Plesiomonas shigelloides (
29).
Previous reports described two noncontiguous polar flagellum regions for
Aeromonas: (i) a polar flagellum region of
A. caviae, containing five genes that encode two tandem flagellins (FlaA and FlaB), a protein involved in flagellum filament length control (FlaG), a HAP-2 distal capping protein (FlaH), and a putative flagellin chaperone (FlaJ) (
53); and (ii) a polar flagellum region of
A. hydrophila, containing 16 genes which encode chemotaxis (CheV and CheR), hook (FlgE, FlgK, and FlgL), rod (FlgB, FlgC, FlgD, FlgF, FlgG, and FlgJ), L ring (FlgH), and P ring (FlgI) proteins, as well as two specific chaperones (FlgA and FlgN) and the anti-σ
28 factor (FlgM) (
2).
Although some genes have been described, many others are required for the expression and regulation of Aeromonas polar flagella. This work employed transposon mutagenesis and mutant complementation to isolate the A. hydrophila AH-3 chromosomal regions required for polar flagellum expression. Furthermore, we investigated the distribution of these genes among the mesophilic Aeromonas species, characterized several Aeromonas strains with defined mutations in different polar flagellar genes, and studied their motility, presence or absence of both types of flagella, adherence to HEp-2 cells, and ability to form biofilms.
DISCUSSION
Mesophilic
Aeromonas spp. produce a single polar flagellum that is expressed in both liquid and solid media; in addition, 50 to 60% of strains also have an inducible lateral flagellum system that is expressed in high viscosity. Previously, only a few genes required for polar flagellum formation have been published (
2,
53). The isolation of
A. hydrophila AH-3 transposon mutants unable to swim allowed us to genetically characterize three of five
A. hydrophila AH-3 chromosomal regions (regions 2, 3, and 4) containing polar flagellum genes (Fig.
1). Region 2 is composed of
fla and
maf genes;
fli,
flh, chemotaxis, and motor genes are in region 3; and the motor gene
motX is alone in region 4. Polar flagellum loci that have been sequenced in other bacteria, such as
Vibrio or
Pseudomonas, indicated that the gene organization seems highly conserved, although their distributions in chromosomal regions are different between organisms (Fig.
2). Genes coding for flagellins and capping protein (
fla), as well as genes coding for the basal body, hook length regulator, switch, export apparatus, flagellum placement determinant, flagellum number regulator, σ
28 factor, chemotaxis, and motor proteins (
fli and
flh) of
A. hydrophila AH-3, are distributed in two different chromosomal regions (2 and 3). In contrast,
V. parahaemolyticus (Fig.
2) showed all of these genes grouped in a single chromosomal region (
34,
41).
Organization of the
A. hydrophila fla genes in region 2 is identical to that of
A. caviae (
53). A similar gene organization is also observed with
V. parahaemolyticus polar flagellum region 2 (Fig.
2); however, this organism possesses one more flagellin gene (
flaF) and a gene encoding a putative chaperone (
flaI) between
flaH and
flaJ (
30,
36). The
A. hydrophila insertional
flaH mutant and
flaA flaB double mutant showed lateral flagella, absence of polar flagella, and a dramatic reduction in adhesion to HEp-2 cells and ability to form biofilms. All of these effects were rescued by the introduction of pLA-FLA in the mutant strains, suggesting that the
fla genes distributed in this chromosomal region are involved only in polar flagellum biosynthesis or assembly, and both flagellin genes (
flaA and
flaB) are required for optimal polar flagellum function. Recent unpublished work performed with
A. caviae has shown that nonpolar mutants (
flaA flaB and
flaH) do in fact express lateral flagella, in contrast to the previously reported results (
53). The
A. hydrophila AH-3 strains with mutations in
flaH,
flaA,
flaB, and
flaA flaB demonstrated the same phenotypes as their
A. caviae counterparts (
53). In addition we found, downstream of the
A. hydrophila polar flagellum
fla loci and preceded by a σ
28 promoter sequence, an independently transcribed gene not described before for
Aeromonas spp. or
Vibrio spp. This gene encoded a homologue of the Maf proteins reported for
Helicobacter pylori,
Clostridium acetobutylicum, and
Campylobacter jejuni. In all of these bacteria, the genes encoding Maf proteins are linked to either flagellum biosynthesis genes and/or genes involved in sugar biosynthesis and transport (
26,
33). The inactivation of the
A. hydrophila maf-1 gene abolished only polar flagellum formation and did not affect lateral flagella; the wild-type phenotype was restored by introduction of pACYC-MAF (Fig.
4). This fact together with the knowledge that
A. hydrophila AH-3 polar flagellins are glycosylated (unpublished observation), similarly to the
A. caviae polar flagellins (
25,
53), may suggest that the encoded protein (Maf-1) is involved in posttranslational polar flagellum glycosylation, but their exact role in flagellar biosynthesis remains unknown.
The
A. hydrophila polar flagellum region 3 (Table
3) showed an organization similar to that of the genes downstream of
flaM in
V. parahaemolyticus polar flagellum region 2, with the absence of the motor genes (
34,
41) (Fig.
2). No master regulatory genes encoding homologues of
V. parahaemolyticus FlaK, FlaL, and FlaM,
V. cholerae FlrA, FlrB, and FlrC (
52), or
P. aeruginosa FleQ, FleS, and FleR (
4,
54) were found upstream of
A. hydrophila fliE. In contrast to the
Vibrio polar flagellum systems, we found two genes (
pomA and
pomB) which encode orthologues of the MotA and MotB motor proteins of
Pseudomonas (
15). The
A. hydrophila pomA and
pomB genes are transcribed independently, according to the RT-PCR results, while in
P. aeruginosa, these genes are cotranscribed with some chemotaxis genes (
15).
Strains with mutations in the
fliM,
flhA, and
fliA genes were able to produce lateral flagella but were unable to produce polar flagella, and they also showed a large reduction in adhesion to HEp-2 cells and ability to form biofilms. The data obtained from these mutants suggest that they are required only for the production of polar flagella but not for production of lateral flagella. The MotA-MotB complex constitutes the stator of the flagellum motor and is involved in the formation of a proton or sodium-conducting channel to generate the force necessary to drive the flagella (
31,
61). The
pomB insertion mutant had both flagellum types and was fully able to swim, adhere, and form biofilms as well as the wild-type strain. This situation was similar for
P. aeruginosa motB mutants (
18,
64) but not for
V. cholerae or
V. parahaemolyticus motB mutants, which were able to produce polar flagella but were nonmotile (
8,
24). In
Pseudomonas, there are two sets of
motAB-like genes,
motAB and
motCD, distributed in different chromosomal regions, as well as another gene,
motY, which contributes to proton-driven flagellar motility (Fig.
2). Loss of either
motAB-like gene still resulted in motile bacteria in aqueous environments, and only mutations of both sets of genes encoding the MotA or MotB homologue were sufficient to abolish motility (
18,
64). The data obtained from the
pomB mutant suggest that PomB is not essential for swimming motility, leading to two different possible explanations: the stator of lateral flagella can supply its function, or another
pomAB-like locus is present in
A. hydrophila. Further studies are required in order to completely understand the motility process in
A. hydrophila AH-3.
Region 4 of
A. hydrophila AH-3 polar flagella includes a gene that encodes a homologue of the sodium-driven motor MotX of
V. alginolyticus and
V. parahaemolyticus, which is involved with MotA, MotB, and MotY in torque generation of polar flagella (
8,
40), although its exact function is unknown. This
Aeromonas gene has both a putative σ
28 promoter sequence and a putative terminator sequence, suggesting that it is transcribed independently, as in
Vibrio spp. Inactivation of the
A. hydrophila motX gene did not affect either lateral or polar flagellum formation, but mutants were unable to swim in liquid media, as described for
V. parahaemolyticus, and showed a reduction (50%) in adhesion to HEp-2 cells and ability to form biofilms (37% reduction) in comparison with the wild-type strain.
In contrast to master regulatory genes for other polar flagellated bacteria, such as
Vibrio and
Pseudomonas (Fig.
2), the
Aeromonas flrA to
flrC genes are located in an independent chromosomal region (region 5). The central domain of
Aeromonas FlrB and FlrC homologues contains a σ
54 interaction domain, which is present in σ
54 activators, and carboxyl- and amino-terminal domains of FlrB and FlrC homologues, respectively, which define these two proteins as members of the two-component family of bacterial signal transducers. A characteristic histidine kinase domain was found in the FlrB homologue protein, and their N termini contain a PAS domain, usually associated with proteins that play a role in detection and adaptation to environmental change (
62). A PAS domain was also found in
V. cholerae FlrB and
P. aeruginosa FleS N termini, in contrast to FlaL of
V. parahaemolyticus. The
Aeromonas flrC mutant was able to produce lateral flagella but unable to produce polar flagella and had a large reduction in adhesion to HEp-2 cells and ability to form biofilms. The data obtained from this mutant suggest that it is required only for the production of polar flagella but not for production of lateral flagella. Plasmids pLA-FLR and pACYC-FLR1 were able to fully complement the
flrC defects when introduced independently into this mutant strain. With the possible exception of the motor genes, which require further investigation, the rest of the structural genes for polar and lateral flagellum formation in
A. hydrophila AH-3 are clearly independent, and further studies are required in order to completely understand the motility process in
A. hydrophila AH-3.
By comparing previous results with results described in this work for mutants unable to produce polar and lateral flagella (
25), mutants able to produce polar but not lateral flagella (
23), and mutants able to produce lateral but not polar flagella, we can conclude that both types of flagella contribute to HEp-2 cell adhesion and biofilm formation in
A.hydrophila AH-3. In
V. parahaemolyticus, only the polar flagella seem to be involved in these pathogenic features (
20).