**Population Genetics of the "***Aeromonas hydrophila* **Species Complex"**

Mª Carmen Fusté, Maribel Farfán, David Miñana-Galbis, Vicenta Albarral, Ariadna Sanglas and José Gaspar Lorén *Department of Health Microbiology and Parasitology, Faculty of Pharmacy, University of Barcelona, Barcelona, Spain* 

## **1. Introduction**

38 Studies in Population Genetics

Saumitou-Laprade, P. Vernet, P, Vassiliadis, C. Hoareau, Y. de Magny, G. Dommée, B. and

Schemske, D. W. and R. Lande (1985). "The evolution of self fertilization and inbreeding depression in plants : empirical observations." Evolution 39(1): 41-52. Schoen, D. J., M. O. Johnston, et al. (1997). "Evolutionay history of the mating system in

Schoen, D.J. 2005. Spontaneous deleterious mutation in wild plant species with contrasting

Stebbins, G. L. (1950). Variation and evolution in plants. New York, USA., Columbia

Takebayashi, N. and L. F. Delph (2000). "An association between a floral trait and inbreeding

Uyenoyama, M. K., K. E. Holsinger, et al. (1993). Ecological and genetic factors directing the

Vassiliadis, C. Valero, M. Saumitou-Laprade, P and Godelle, B. 2000 A model for the

Vogler, D. W. and S. Kalisz (2001). "Sex among the flowers: the distribution of plant mating

Waller, D. M. (1986). "Is there a disruptive selection for self-fertilization?" American

Weir, B. S. and C. C. Cockerham (1973). "Mixed self and random mating at two loci."

Winn, A.A., Elle, E., Kalisz, S., Cheptou, P.-O., Eckert,C.G., Goodwillie, C. Johnston,M.O.,

Yampolsky, C. and H. Yampolsky (1922). "Distribution of sex forms in the phanerogamic

Moeller, D.A., Ree, R.H., Sargent, R.D. and Vallejo-Marn, M. 2011 Analysis of inbreeding depression in mixed mating plants provides evidence for selective

evolution of high frequencies of males in an androdioecious plant based on a cross-

Androdioecious Plant Vol. 327 no. 5973 pp. 1648-1650 Science.

*Amsinckia* (Boraginaceae)." Evolution 51(4): 1090-1099.

evolution of self fertilization. O. S. i. e. biology. 9: 327-381.

interference and stable mixed mating *Evolution* (early view).

compatibility advantage of males. Heredity 85, 413

mating systems. *Evolution*, 59: 2370-2377.

depression." Evolution 54(3): 840-846.

systems." Evolution 55(1): 202-204.

Genetical Research 21: 247–262.

Naturalist 128(3): 421-426.

flora." Bibl. Genet. 3: 1-62.

University Press.

Lepart, J. 2009 A Self-Incompatibility System Explains High Male Frequencies in an

Population genetics studies the genetic variability of individuals in a population based on the allele frequencies at several genes or loci and tries to explain this variability in terms of mutation, selection or genetic recombination. The statistical analysis of these frequencies allows models of evolution to be established, which will help us to understand and predict the past and present gene flow in the population (Maynard-Smith, 1991). For the most part population genetics has been designed for diploid organisms with sexual reproduction. In the words of Bruce Levin, "the genetic theory of adaptive evolution was developed by sexually reproducing eukaryotes, for sexually reproducing eukaryotes" (Levin & Bergstrom, 2000). As a consequence, before being applied to prokaryotes, population genetics needs to be adapted.

In theory the haploid nature of bacteria should simplify their analysis, since dominance or over-dominance is not an issue and the genotype can usually be deduced directly from the phenotype. However, central to classical population genetics are infinite population size, random mating, and free recombination. Consequently, as expressed by Maynard-Smith, "the alleles present at one locus are independent of those at other loci. Changes in the frequency of an allele at one locus, therefore, are independent of what is happening elsewhere in the genome: each locus can be treated individually" (Maynard-Smith, 1995). It is true that the size of bacterial populations can be practically infinite but recombination occurs extremely rarely so that changes affecting one locus can lead to the modification of others. In the succinct words of Maynard-Smith, "the genome should be treated as an interrelated whole, and not as a set of independently changing genes". The crux of the problem is knowing the exact level of recombination in bacterial populations, since "it is considerably more challenging to elaborate a theory for a population with little recombination than for one with no recombination, or a lot" (Maynard-Smith, 1995). In bacterial population genetics, sometimes we detect a degree of recombination that is too high for a pure phylogenetic approach, but too low for assessing a random interchange.

Stronger evidence for restricted recombination comes from measurements of linkage disequilibrium: that is, the tendency for particular alleles at different loci to co-occur

Population Genetics of the "*Aeromonas hydrophila* Species Complex" 41

The genus *Aeromonas* Stanier 1943 belongs to the family *Aeromonadaceae* within the class *Gammaproteobacteria* (Martin-Carnahan & Joseph, 2005). Aeromonads are autochthonous inhabitants of aquatic environments including chlorinated and polluted waters, although they can also be isolated from a wide variety of environmental and clinical sources. They cause infections in vertebrates and invertebrates, such as frogs, birds, various fish species and domestic animals. In recent years, some authors have considered *Aeromonas* as an emergent pathogen in humans, producing intestinal and extraintestinal diseases. Aeromonads are facultative anaerobic chemoorganotrophs capable of anaerobic nitrate

The interest in the taxonomy of the genus *Aeromonas* has increased markedly in recent years (Janda & Abbott, 2010), and its classification, with 25 species currently recognized, remains challenging. Novel species are continuously being described, strains and species described so far are being rearranged, and DNA-DNA hybridization studies have observed discrepancies (Janda & Abbott, 2010). Historically, the genus *Aeromonas* has been divided into two groups: nonmotile, psychrophilic species, best represented by *A. salmonicida*, which are generally associated with fish diseases and motile mesophilic species associated with human diseases, including *A. hydrophila*, *A. veronii* and *A. caviae* (Martin-Carnahan & Joseph, 2005*)*. More species have since been described and genealogies have to be adapted

Bacterial species are formally defined as a group of strains that share several phenotypical characteristics and show values of DNA-DNA hybridization ≥ 70% and a 16S rRNA sequence similarity ≥ 97% with their close relatives (Stackebrandt et al., 2002). Indeed, there are hardly any examples in which strains with divergence in the 16S rRNA sequence ≤ 97% are defined as one species (Roselló-Mora & Amann, 2001). In *Aeromonas* 16S rRNA gene sequences are highly similar, being identical in some close related species such as *A. salmonicida*, *A. bestiarum, A. popoffii* and *A. piscicola*, which hampers their utility in defining

Several attempts have been made to generate phylogenies using DNA gene sequences to reconstruct the correct genealogical ties among species in *Aeromonas* (Küpfer et al., 2006; Saavedra et al., 2006; Miñana-Galbis et al., 2009), but the genes chosen for this purpose are not always suitable and do not give congruent phylogenies (Farfán et al., 2010; Silver et al., 2011). Recently, two papers presenting MLST schemes for *Aeromonas* have been published (Martínez-Murcia et al., 2011; Martino et al., 2011) and there is an online MLST database of the genus *Aeromonas,* managed by Keith Jolley and curated by Barbara Cardazzo (http://pubmlst.org/aeromonas). All this accumulated data should help to establish a

Finally, the availability of complete genomes of different species is also useful in this task, but unfortunately in the case of *Aeromonas* only three genomes have been completed, one corresponding to the type strain of *A. hydrophila* subsp. *hydrophila* ATCC 7966 isolated from a tin of milk (Seshadri et al., 2006), the second to a fish pathogen, the strain A449 of *A. salmonicida* subsp. *salmonicida* (Reith et al., 2008) and the third to *A. veronii* isolated from an aquaculture pond sediment (Li et al., 2011). The information given by the genomes of *A. hydrophila* and *A. salmonicida* indicate that while they are of identical size (4,7Mb) and share

reliable clustering of the *Aeromonas* species and elucidate their exact boundaries.

respiration and dissimilatory metal reduction (Martin-Carnahan & Joseph, 2005).

accordingly, although this is not always straightforward.

species in this genus (Figueras et al., 2000; Beaz-Hidalgo et al., 2009).

**2. The genus** *Aeromonas*

(Maynard-Smith et al., 1993; Haubold et al., 1998). Linkage disequilibrium (and the toofrequent occurrence of a particular combination of alleles, which is a manifestation of such disequilibrium) shows that recombination is restricted, but not absent. The determination of the relative importance of mutation in comparison with recombination is central to bacterial population genetics (Feil et al., 1999). Previous studies have demonstrated a wide variety of situations among bacterial species ranging from the clonal diversification of *Salmonella* (Selander et al., 1990) or *Escherichia coli* (Orskov et al., 1990), which are mainly due to mutation, to the frequent recombination found in *Neisseria gonorrhoeae* (O'Rourke & Stevens, 1993) or *Helicobacter pylori* (Salaun et al., 1998). Most of the population studies done with bacterial species suggest that recombination occurs in nature, and indeed may be highly important in generating variation, but that it is infrequent compared to mutation. Consequently, bacterial populations consist largely of independent clonal lineages.

The development of protein electrophoresis was a breakthrough for the study of bacterial population genetics. A pioneer in the field, Milkman (1973) used the methodology to study whether electrophoretic variation is selective or neutral. As described by Selander et al., Multilocus Enzyme Electrophoresis (MLEE) "has long been a standard method in eukaryotic population genetics and systematics" before it was applied "for studying the genetic diversity and structure in natural populations of bacteria. This research established basic population frameworks for the analysis of variation in serotypes and other phenotypic characters and has provided extensive data for systematics and useful marker systems for epidemiology" (Selander et al., 1986).

In 1998, Maiden et al. introduced Multilocus Sequence Typing (MLST), an extension of MLEE based on nucleotide sequencing that is able to determine higher levels of discrimination (more alleles per locus). In MLST "alleles are identified directly from the nucleotide sequences of internal fragments of genes rather than by comparing the electrophoretic mobilities of the enzymes they encode" (Maiden et al., 1998). In addition, this method is fully portable between laboratories and data can be stored in a single multilocus sequence database accessible via the internet (http://www.mlst.net). This approach has given "a new dimension to the elucidation of genomic relatedness at the interand intraspecific level by sequence analysis of housekeeping genes subject to stabilising selection. This technique has been mainly used in epidemiology, but it offers the opportunity to incorporate the insights available from population genetics and phylogenetic approaches into bacterial systematics" (Stackebrandt et al., 2002; Pérez-Losada et al., 2005; Robinson et al., 2010).

The results obtained using MLST indicate that despite the high diversity observed in bacteria, it is possible to recognize clusters with a lower degree of variation. The number of sequence types (STs) obtained is less than expected if we consider the product of the individual allelic frequencies. Nevertheless, most of the strains belong to one or a few allelic profiles, whereas most of the STs are represented by one or few strains. In addition, most STs cluster in clonal complexes constituted by closely related genotypes. Typically, each clonal complex consists of a predominant ST and a varied group of less common STs, which have different alleles only in one or two loci (Feil et al., 2004). These differences could correspond to initial stages of clonal divergence from an ancestral genotype that was the origin of the clone (Vogel et al., 2010; Willems, 2010).

## **2. The genus** *Aeromonas*

40 Studies in Population Genetics

(Maynard-Smith et al., 1993; Haubold et al., 1998). Linkage disequilibrium (and the toofrequent occurrence of a particular combination of alleles, which is a manifestation of such disequilibrium) shows that recombination is restricted, but not absent. The determination of the relative importance of mutation in comparison with recombination is central to bacterial population genetics (Feil et al., 1999). Previous studies have demonstrated a wide variety of situations among bacterial species ranging from the clonal diversification of *Salmonella* (Selander et al., 1990) or *Escherichia coli* (Orskov et al., 1990), which are mainly due to mutation, to the frequent recombination found in *Neisseria gonorrhoeae* (O'Rourke & Stevens, 1993) or *Helicobacter pylori* (Salaun et al., 1998). Most of the population studies done with bacterial species suggest that recombination occurs in nature, and indeed may be highly important in generating variation, but that it is infrequent compared to mutation.

Consequently, bacterial populations consist largely of independent clonal lineages.

epidemiology" (Selander et al., 1986).

Robinson et al., 2010).

origin of the clone (Vogel et al., 2010; Willems, 2010).

The development of protein electrophoresis was a breakthrough for the study of bacterial population genetics. A pioneer in the field, Milkman (1973) used the methodology to study whether electrophoretic variation is selective or neutral. As described by Selander et al., Multilocus Enzyme Electrophoresis (MLEE) "has long been a standard method in eukaryotic population genetics and systematics" before it was applied "for studying the genetic diversity and structure in natural populations of bacteria. This research established basic population frameworks for the analysis of variation in serotypes and other phenotypic characters and has provided extensive data for systematics and useful marker systems for

In 1998, Maiden et al. introduced Multilocus Sequence Typing (MLST), an extension of MLEE based on nucleotide sequencing that is able to determine higher levels of discrimination (more alleles per locus). In MLST "alleles are identified directly from the nucleotide sequences of internal fragments of genes rather than by comparing the electrophoretic mobilities of the enzymes they encode" (Maiden et al., 1998). In addition, this method is fully portable between laboratories and data can be stored in a single multilocus sequence database accessible via the internet (http://www.mlst.net). This approach has given "a new dimension to the elucidation of genomic relatedness at the interand intraspecific level by sequence analysis of housekeeping genes subject to stabilising selection. This technique has been mainly used in epidemiology, but it offers the opportunity to incorporate the insights available from population genetics and phylogenetic approaches into bacterial systematics" (Stackebrandt et al., 2002; Pérez-Losada et al., 2005;

The results obtained using MLST indicate that despite the high diversity observed in bacteria, it is possible to recognize clusters with a lower degree of variation. The number of sequence types (STs) obtained is less than expected if we consider the product of the individual allelic frequencies. Nevertheless, most of the strains belong to one or a few allelic profiles, whereas most of the STs are represented by one or few strains. In addition, most STs cluster in clonal complexes constituted by closely related genotypes. Typically, each clonal complex consists of a predominant ST and a varied group of less common STs, which have different alleles only in one or two loci (Feil et al., 2004). These differences could correspond to initial stages of clonal divergence from an ancestral genotype that was the The genus *Aeromonas* Stanier 1943 belongs to the family *Aeromonadaceae* within the class *Gammaproteobacteria* (Martin-Carnahan & Joseph, 2005). Aeromonads are autochthonous inhabitants of aquatic environments including chlorinated and polluted waters, although they can also be isolated from a wide variety of environmental and clinical sources. They cause infections in vertebrates and invertebrates, such as frogs, birds, various fish species and domestic animals. In recent years, some authors have considered *Aeromonas* as an emergent pathogen in humans, producing intestinal and extraintestinal diseases. Aeromonads are facultative anaerobic chemoorganotrophs capable of anaerobic nitrate respiration and dissimilatory metal reduction (Martin-Carnahan & Joseph, 2005).

The interest in the taxonomy of the genus *Aeromonas* has increased markedly in recent years (Janda & Abbott, 2010), and its classification, with 25 species currently recognized, remains challenging. Novel species are continuously being described, strains and species described so far are being rearranged, and DNA-DNA hybridization studies have observed discrepancies (Janda & Abbott, 2010). Historically, the genus *Aeromonas* has been divided into two groups: nonmotile, psychrophilic species, best represented by *A. salmonicida*, which are generally associated with fish diseases and motile mesophilic species associated with human diseases, including *A. hydrophila*, *A. veronii* and *A. caviae* (Martin-Carnahan & Joseph, 2005*)*. More species have since been described and genealogies have to be adapted accordingly, although this is not always straightforward.

Bacterial species are formally defined as a group of strains that share several phenotypical characteristics and show values of DNA-DNA hybridization ≥ 70% and a 16S rRNA sequence similarity ≥ 97% with their close relatives (Stackebrandt et al., 2002). Indeed, there are hardly any examples in which strains with divergence in the 16S rRNA sequence ≤ 97% are defined as one species (Roselló-Mora & Amann, 2001). In *Aeromonas* 16S rRNA gene sequences are highly similar, being identical in some close related species such as *A. salmonicida*, *A. bestiarum, A. popoffii* and *A. piscicola*, which hampers their utility in defining species in this genus (Figueras et al., 2000; Beaz-Hidalgo et al., 2009).

Several attempts have been made to generate phylogenies using DNA gene sequences to reconstruct the correct genealogical ties among species in *Aeromonas* (Küpfer et al., 2006; Saavedra et al., 2006; Miñana-Galbis et al., 2009), but the genes chosen for this purpose are not always suitable and do not give congruent phylogenies (Farfán et al., 2010; Silver et al., 2011). Recently, two papers presenting MLST schemes for *Aeromonas* have been published (Martínez-Murcia et al., 2011; Martino et al., 2011) and there is an online MLST database of the genus *Aeromonas,* managed by Keith Jolley and curated by Barbara Cardazzo (http://pubmlst.org/aeromonas). All this accumulated data should help to establish a reliable clustering of the *Aeromonas* species and elucidate their exact boundaries.

Finally, the availability of complete genomes of different species is also useful in this task, but unfortunately in the case of *Aeromonas* only three genomes have been completed, one corresponding to the type strain of *A. hydrophila* subsp. *hydrophila* ATCC 7966 isolated from a tin of milk (Seshadri et al., 2006), the second to a fish pathogen, the strain A449 of *A. salmonicida* subsp. *salmonicida* (Reith et al., 2008) and the third to *A. veronii* isolated from an aquaculture pond sediment (Li et al., 2011). The information given by the genomes of *A. hydrophila* and *A. salmonicida* indicate that while they are of identical size (4,7Mb) and share

Population Genetics of the "*Aeromonas hydrophila* Species Complex" 43

mean genetic distance of 0.071 (Table 1). The *A. salmonicida* cluster, although comprising five subspecies, constitutes a homogeneous clade with the lowest mean genetic distance of 0.016 (Table 1). The closest relative to *A. salmonicida*, *A. bestiarum,* has a different population structure. Rather than being a single group, it is divided in three clades, one constituted by the great majority of the *A. bestiarum* strains, a second corresponding to *A. popoffii* and the third including some *A. bestiarum* strains together with isolates of *A. piscicola.* The mean genetic distance of the *A. bestiarum* cluster is higher than in *A. salmonicida* 0.029 (Table 1). The *A. hydrophila* group, which exhibits the highest genetic distance 0.037 (Table 1), clearly separates in two clades, one including the *A. hydrophila* subsp. *hydrophila* and *A. hydrophila* subsp. *ranae* and the second constituted by *A. hydrophila* subsp. *dhakensis* and *A. aquariorum.* The presence of different clades in some of these species poses questions about the clade-

All sequences 0.0717 0.0024 *A. bestiarum* 0.0286 0.0014 *A. bestiarum* (clade 1) 0.0157 0.0009 *A. bestiarum* (clade 2) 0.0084 0.0008 *A. bestiarum* (clade 3) 0.0100 0.0008 *A. hydrophila* 0.0370 0.0016 *A. hydrophila* (clade 1) 0.0219 0.0010 *A. hydrophila* (clade 2) 0.0158 0.0012 *A. salmonicida* 0.0163 0.0009

Table 1. Mean genetic distances for all sequences and clusters calculated using the Tamura Nei (1993) distance. Standard error estimates were obtained by the bootstrap method with

*A. bestiarum* vs. *A. salmonicida* 0.0655 0.0033 *A. bestiarum* vs. *A. hydrophila* 0.1042 0.0043 *A. hydrophila* vs. *A. salmonicida* 0.1101 0.0049 *A. bestiarum* (clade 1) vs. *A. bestiarum* (clade 2) 0.0507 0.0027 *A. bestiarum* (clade 1) vs. *A. bestiarum* (clade 3) 0.0279 0.0021 *A. bestiarum* (clade 2) vs. *A. bestiarum* (clade 3) 0.0526 0.0031 *A. hydrophila* (clade 1) vs. *A. hydrophila* (clade 2) 0.0558 0.0030 Table 2. Mean genetic distances among different clusters obtained using the Tamura Nei (1993) distance. Standard error estimates were obtained by the bootstrap method with 500

Estimation of distance frequency within and between clusters from multilocus sequence data provides interesting insights into the population structure of these groups. The frequency distribution of the pairwise genetic distances clearly identifies the AHC species phylogenetic structure (Fig. 2). Within the strains of the *A. bestiarum* group (Fig. 2A), the plot shows three peaks with distance values ranging from 0 to 0.063. The lowest values correspond to pairwise comparisons among isolates within clades (1, 2 and 3) and the highest to those between clades. Distances between groups allowed a clear separation of *A.* 

**TN93 Standard error** 

**TN93 Standard error** 

species relationships.

500 replicates.

replicates.

multiple housekeeping and virulence genes, *A. salmonicida* has acquired several mobile genetic elements, and undergone genome rearrangements and loss of genes in the process of adapting to a specific host. The genome of *A. veronii* is smaller (4,3Mb) and contains fewer virulence genes than the others.

## **3. "***Aeromonas hydrophila* **species complex"**

An example of the taxonomic complexity of the genus *Aeromonas* is the difficulty in discriminating between the phenotypically and genetically closely related species belonging to the "*Aeromonas hydrophila* species complex" (AHC), which includes: *A. hydrophila*, composed by three subspecies: A. *hydrophila* subsp. *hydrophila, A. hydrophila* subsp. *ranae* and *A. hydrophila* subsp. *dhakensis*, *A. bestiarum*, *A. popoffii* and *A. salmonicida,* divided in five subspecies: *A. salmonicida* subsp. *salmonicida*, *A. salmonicida* subsp. *masoucida*, *A. salmonicida* subsp. *achromogenes, A. salmonicida* subsp. *pectinolytica* and *A. salmonicida* subsp. *smithia* (Miñana-Galbis et al., 2002; Martin-Carnahan & Joseph, 2005). Recently, two additional species have been described in this group, *A. aquariorum* and *A. piscicola* (Martínez-Murcia et al., 2008; Beaz-Hidalgo et al., 2009). Members of the AHC were first described as strains producing the enzymes elastase, lecitinase or stapholysin (Abbott et al., 2003). They are genetically closely related and share multiple phenotypic characteristics, which makes discrimination among the species included in this group extremely difficult (Miñana-Galbis et al., 2002).

Several approaches have been used to discriminate among the AHC species: Amplified Fragment Length Polymorphisms (AFLP) (Huys et al., 1997), fluorescent AFLP (FAFLP) (Huys et al., 2002), MLEE (Miñana-Galbis et al., 2004), Random Amplified Polymorphic DNA (RAPD) and MALDI-TOF MS analysis (Martínez-Murcia et al., 2008). Although the results obtained with these methods have been useful for the taxonomy and phylogeny of the AHC, providing a hypothesis for the genealogy of strains and detailing their patterns of descent and degree of genetic variation accumulated over time, only the MLEE study has been used to elucidate their population genetic structure.

Previous studies based on the sequence analysis of several housekeeping genes have demonstrated that the AHC is not monophyletic (Soler et al., 2004; Küpfer et al., 2006; Saavedra et al., 2006; Nhung et al., 2007; Beaz-Hidalgo et al., 2009; Miñana-Galbis et al., 2009). Nevertheless, controversially, other studies have shown the monophylia of this group (Martínez-Murcia et al., 2005; Farfán et al., 2010). This conflict could be due to the incongruence of phylogenies derived from distinct gene sequence analysis.

In order to establish the population structure and divergence of the species included in this group of *Aeromonas* we studied a set of strains representative of the AHC, in which we analyzed the nucleotide sequences (total or partial) of 6 housekeeping genes: *cpn*60 (555 bp), *dna*J (891 bp), *gyr*B (1089 bp), *mdh* (936 bp), *rec*A (1065 bp), *rpo*D (843 bp), giving a total fragment length of 5379 bp.

## **4. Phylogenetic analysis**

The relationships among the analyzed *Aeromonas* isolates are represented as a genealogical tree (Fig. 1). The tree reveals that the AHC splits into three main clusters separated by a

multiple housekeeping and virulence genes, *A. salmonicida* has acquired several mobile genetic elements, and undergone genome rearrangements and loss of genes in the process of adapting to a specific host. The genome of *A. veronii* is smaller (4,3Mb) and contains fewer

An example of the taxonomic complexity of the genus *Aeromonas* is the difficulty in discriminating between the phenotypically and genetically closely related species belonging to the "*Aeromonas hydrophila* species complex" (AHC), which includes: *A. hydrophila*, composed by three subspecies: A. *hydrophila* subsp. *hydrophila, A. hydrophila* subsp. *ranae* and *A. hydrophila* subsp. *dhakensis*, *A. bestiarum*, *A. popoffii* and *A. salmonicida,* divided in five subspecies: *A. salmonicida* subsp. *salmonicida*, *A. salmonicida* subsp. *masoucida*, *A. salmonicida* subsp. *achromogenes, A. salmonicida* subsp. *pectinolytica* and *A. salmonicida* subsp. *smithia* (Miñana-Galbis et al., 2002; Martin-Carnahan & Joseph, 2005). Recently, two additional species have been described in this group, *A. aquariorum* and *A. piscicola* (Martínez-Murcia et al., 2008; Beaz-Hidalgo et al., 2009). Members of the AHC were first described as strains producing the enzymes elastase, lecitinase or stapholysin (Abbott et al., 2003). They are genetically closely related and share multiple phenotypic characteristics, which makes discrimination among the species included in this group extremely difficult (Miñana-Galbis

Several approaches have been used to discriminate among the AHC species: Amplified Fragment Length Polymorphisms (AFLP) (Huys et al., 1997), fluorescent AFLP (FAFLP) (Huys et al., 2002), MLEE (Miñana-Galbis et al., 2004), Random Amplified Polymorphic DNA (RAPD) and MALDI-TOF MS analysis (Martínez-Murcia et al., 2008). Although the results obtained with these methods have been useful for the taxonomy and phylogeny of the AHC, providing a hypothesis for the genealogy of strains and detailing their patterns of descent and degree of genetic variation accumulated over time, only the MLEE study has

Previous studies based on the sequence analysis of several housekeeping genes have demonstrated that the AHC is not monophyletic (Soler et al., 2004; Küpfer et al., 2006; Saavedra et al., 2006; Nhung et al., 2007; Beaz-Hidalgo et al., 2009; Miñana-Galbis et al., 2009). Nevertheless, controversially, other studies have shown the monophylia of this group (Martínez-Murcia et al., 2005; Farfán et al., 2010). This conflict could be due to the

In order to establish the population structure and divergence of the species included in this group of *Aeromonas* we studied a set of strains representative of the AHC, in which we analyzed the nucleotide sequences (total or partial) of 6 housekeeping genes: *cpn*60 (555 bp), *dna*J (891 bp), *gyr*B (1089 bp), *mdh* (936 bp), *rec*A (1065 bp), *rpo*D (843 bp), giving a total

The relationships among the analyzed *Aeromonas* isolates are represented as a genealogical tree (Fig. 1). The tree reveals that the AHC splits into three main clusters separated by a

incongruence of phylogenies derived from distinct gene sequence analysis.

virulence genes than the others.

et al., 2002).

fragment length of 5379 bp.

**4. Phylogenetic analysis** 

**3. "***Aeromonas hydrophila* **species complex"** 

been used to elucidate their population genetic structure.

mean genetic distance of 0.071 (Table 1). The *A. salmonicida* cluster, although comprising five subspecies, constitutes a homogeneous clade with the lowest mean genetic distance of 0.016 (Table 1). The closest relative to *A. salmonicida*, *A. bestiarum,* has a different population structure. Rather than being a single group, it is divided in three clades, one constituted by the great majority of the *A. bestiarum* strains, a second corresponding to *A. popoffii* and the third including some *A. bestiarum* strains together with isolates of *A. piscicola.* The mean genetic distance of the *A. bestiarum* cluster is higher than in *A. salmonicida* 0.029 (Table 1). The *A. hydrophila* group, which exhibits the highest genetic distance 0.037 (Table 1), clearly separates in two clades, one including the *A. hydrophila* subsp. *hydrophila* and *A. hydrophila* subsp. *ranae* and the second constituted by *A. hydrophila* subsp. *dhakensis* and *A. aquariorum.* The presence of different clades in some of these species poses questions about the cladespecies relationships.


Table 1. Mean genetic distances for all sequences and clusters calculated using the Tamura Nei (1993) distance. Standard error estimates were obtained by the bootstrap method with 500 replicates.


Table 2. Mean genetic distances among different clusters obtained using the Tamura Nei (1993) distance. Standard error estimates were obtained by the bootstrap method with 500 replicates.

Estimation of distance frequency within and between clusters from multilocus sequence data provides interesting insights into the population structure of these groups. The frequency distribution of the pairwise genetic distances clearly identifies the AHC species phylogenetic structure (Fig. 2). Within the strains of the *A. bestiarum* group (Fig. 2A), the plot shows three peaks with distance values ranging from 0 to 0.063. The lowest values correspond to pairwise comparisons among isolates within clades (1, 2 and 3) and the highest to those between clades. Distances between groups allowed a clear separation of *A.* 

Population Genetics of the "*Aeromonas hydrophila* Species Complex" 45

0.000 0.025 0.050 0.075 0.100 0.125 0.150

TN93 distance

*A. hydrophila* **group**

**between** *A.hydrophila*

0.000 0.025 0.050 0.075 0.100 0.125 0.150

TN93 distance

*A. salmonicida* **group**

**and** *A.bestiarum*

**within** *A.salmonicida* **between** *A.salmonicida*

0.000 0.025 0.050 0.075 0.100 0.125 0.150

TN93 distance

Fig. 2. Distribution of TN93 distances within and between clusters determined by pairwise

*A. bestiarum* **group**

**between** *A.bestiarum*

**(0.065)**

**and** *A.salmonicida* **between** *A.bestiarum*

**and** *A.bestiarum* **between** *A.hydrophila*

**and** *A.salmonicida*

**(0.110)**

**between** *A.salmonicida* **and** *A.hydrophila*

**and** *A.hydrophila*

**(0.104)**

(best model for the data). Numbers at the branch nodes represent bootstrap values (500 replicates). The scale bar indicates the number of nucleotide substitutions per site. The GenBank/EMBL/DDBJ accession numbers of the nucleotide sequences used in this study are EU306796-EU306797, EU306804-EU306806, EU306814-EU306820, EU306822-EU306824, EU306826-EU306829, EU741625, EU741635-EU741636, EU741642, FJ936120, FJ936135, GU062399, JN711508-JN711610 (*cpn60* gene); FJ936122, FJ936136, JN215529-JN215534, JN711611-JN711731 (*dnaJ* gene); FJ936137, JN215535, JN711732-JN711858 (*gyrB* gene); HM163293-HM163294, HM163305-HM163307, HM163312-HM163318, JN660159-JN660273 (*mdh* gene); JN660274-JN660400 (*recA* gene) and EF465509-EF465510, FJ936132, FJ936138,

JN215536-JN215541, JN712315-JN712433 (*rpoD* gene).

**within** *A.bestiarum*

**within** *A.hydrophila*

comparisons. Mean distances are indicated within brackets.

0

**B )**

frequency

**C )**

frequency

40 80 120

**A )**

frequency

*bestiarum* from the other species groups. When we plotted the pairwise distance distribution within the *A. hydrophila* group (Fig. 2B) two peaks were shown and again the lowest values are those within the clades and the highest between the clades (0-0.062). *A. hydrophila* is clearly separated from *A. salmonicida* and *A. bestiarum* despite the overlap in the graph (Fig. 2B), which is a consequence of the similar distance values between *A. hydrophila* and *A. salmonicida* (0.110) and *A. hydrophila* and *A. bestiarum* (0.104). Otherwise, species boundaries are objectively defined in the phylogenetic tree (Fig. 1). In the *A. salmonicida* group, the within distance distribution (0-0.030) appears as a single peak, as does the interspecies distance.

Fig. 1. Maximum likelihood tree inferred from the concatenated nucleotide sequences generated with MEGA5 software (Tamura et al., 2011). The tree was constructed using the Tamura Nei distance (TN93) and the rate of variation among sites was modelled with a gamma distribution (shape parameter = 0.5723) assuming heterogeneity of invariant sites

*bestiarum* from the other species groups. When we plotted the pairwise distance distribution within the *A. hydrophila* group (Fig. 2B) two peaks were shown and again the lowest values are those within the clades and the highest between the clades (0-0.062). *A. hydrophila* is clearly separated from *A. salmonicida* and *A. bestiarum* despite the overlap in the graph (Fig. 2B), which is a consequence of the similar distance values between *A. hydrophila* and *A. salmonicida* (0.110) and *A. hydrophila* and *A. bestiarum* (0.104). Otherwise, species boundaries are objectively defined in the phylogenetic tree (Fig. 1). In the *A. salmonicida* group, the within distance distribution (0-0.030) appears as a single peak, as does the interspecies

Fig. 1. Maximum likelihood tree inferred from the concatenated nucleotide sequences generated with MEGA5 software (Tamura et al., 2011). The tree was constructed using the Tamura Nei distance (TN93) and the rate of variation among sites was modelled with a gamma distribution (shape parameter = 0.5723) assuming heterogeneity of invariant sites

distance.

(best model for the data). Numbers at the branch nodes represent bootstrap values (500 replicates). The scale bar indicates the number of nucleotide substitutions per site. The GenBank/EMBL/DDBJ accession numbers of the nucleotide sequences used in this study are EU306796-EU306797, EU306804-EU306806, EU306814-EU306820, EU306822-EU306824, EU306826-EU306829, EU741625, EU741635-EU741636, EU741642, FJ936120, FJ936135, GU062399, JN711508-JN711610 (*cpn60* gene); FJ936122, FJ936136, JN215529-JN215534, JN711611-JN711731 (*dnaJ* gene); FJ936137, JN215535, JN711732-JN711858 (*gyrB* gene); HM163293-HM163294, HM163305-HM163307, HM163312-HM163318, JN660159-JN660273 (*mdh* gene); JN660274-JN660400 (*recA* gene) and EF465509-EF465510, FJ936132, FJ936138, JN215536-JN215541, JN712315-JN712433 (*rpoD* gene).

Fig. 2. Distribution of TN93 distances within and between clusters determined by pairwise comparisons. Mean distances are indicated within brackets.

Population Genetics of the "*Aeromonas hydrophila* Species Complex" 47

*IAS* values different from 0 in all cases, indicating the absence of recombination and again revealing strong linkage disequilibrium when considering both the total population and the different groups of species (Table 5). This is in spite of the high number of alleles per locus and polymorphic sites (Table 3) and huge genetic diversity (Table 4). Values of *VO, VE, IAS* and the 5% critical values as determined by the Monte Carlo process (*L*MC) and from the

**Gene Number of alleles Number of polymorphic sites ± s.e.\*\***  *cpn60* 95 144 0.075 ± 0.009 *dnaJ* 102 237 0.082 ± 0.007 *gyrB* 108 266 0.063 ± 0.005 *mdh\** 113 225 0.059 ± 0.009 *recA\** 106 282 0.071 ± 0.009 *rpoD* 110 212 0.090 ± 0.010

Table 3. Sequence variation at six loci. Standard error estimates were obtained by the

\* genetic diversity at individual loci (*hj)*; \*\* mean genetic diversity (H) ± standard error (s.e.) *h*j = (*n*/*n*-1)∑*p*ij2, where *p*ij is the frequency of the *i*th allele at the *j*th locus and *n* the number of loci.

Data were calculated by using R statistical software (R Development Core Team, 2010). Table 4. Genetic diversity (*h*) at six loci for all STs and major species sets.

All STs 127 0.9931 0.9959 0.9966 0.9978 0.9945 0.9973 0.9959 ± 0.0007 *A. bestiarum* 44 0.9704 0.9863 0.9863 0.9905 0.9852 0.9873 0.9843 ± 0.0029 *A. hydrophila* 42 0.9930 0.9884 0.9954 1.0000 0.9930 0.9930 0.9938 ± 0.0015 *A. salmonicida* 41 0.9744 0.9902 0.9878 0.9890 0.9707 0.9951 0.9846 ± 0.0039

All STs 0.0635 0.0248 0.3131 0.0253 0.0255 < 1.00 x 10-04 *A. bestiarum* 0.2314 0.0924 0.3010 0.0986 0.1001 < 1.00 x 10-04 *A. hydrophila* 0.1382 0.0369 0.5495 0.0395 0.0405 < 1.00 x 10-04 *A. salmonicida* 0.1916 0.0908 0.2222 0.0978 0.0988 < 1.00 x 10-04 For the meaning of acronyms, see text. The *LMC* and *PMC* results were obtained from 10000 resamplings. Data were calculated by using R statistical software except for *Lpara* , which was determined with the

During the last years, with the availability of the first DNA sequence data of individual genes, evidence of recombination at the molecular level has accumulated for *Aeromonas* in genes such as *dnaJ*, *gyrB* and *recA* (Silver et al., 2011). Incongruence between trees reconstructed from individual genes appeared as further proof of recombination at the gene level. Nevertheless, multilocus analysis with gene sequences also revealed a clear clonal

**STs** *hj\** **H ± s.e.\*\*** *cpn60 dnaJ gyrB mdh recA rpoD*

*VO VE IAS Lpara LMC PMC*

parametric approach (*Lpara*) of Haubold et al. (1998) are shown in Table 5.

\* full-length sequence gene; \*\* nucleotide diversity () ± standard error (s.e.)

Calculated by using MEGA5 software (Tamura et al., 2011).

bootstrap method with 500 replicates.

LIAN 3.5 program (Haubold & Hudson, 2000).

Table 5. Multilocus linkage disequilibrium analysis of the AHC.

Although to our knowledge there is no function that could be used to define a level of divergence for distinguishing species, it seems clear that, in bacteria, there is always an exponential relationship between interspecies recombination and sequence divergence (Roberts & Cohan, 1993; Zawadzki et al., 1995; Vulic et al., 1997). Values of divergence among clades (Table 2) seem to correspond to those of a genetically isolated biological species. The lowest value corresponds to *A. bestiarum* clade 3, indicating that its isolates are undergoing speciation, but species boundaries are not as well-defined as in the other clades.

#### **5. Linkage disequilibrium and population structure**

Linkage equilibrium is characterized by the statistical independence of alleles at all loci. A common method used in bacterial population genetics to quantify the degree of linkage between a set of loci in MLSA data is to use the index of association (*I*A) (Brown et al., 1980; Maynard-Smith et al., 1993). This index compares the ratio of variance calculated from the pairwise distribution of allele mismatches in the data (*V*O) and the ratio expected under a null hypothesis of linkage equilibrium (*V*E). In this case

$$E(V\_{\bullet}) = \sum h\_{\circ} \text{(1-h\_{\circ})}\_{\prime\prime}$$

where *h*j is an unbiased estimator of the population genetic diversity equivalent to heterozygosis in diploid organisms (Table 4). The index of association (*I*A) originally used by Brown et al. (1980) is computed as

$$I\_{\rm A} = V\_{\rm O} / \, V\_{\rm E} \, 1\_{\rm A}$$

which would give a value of zero if there is no association between loci (Maynard-Smith et al., 1993). Values of this index significantly different from zero reflect strong linkage disequilibrium (lower rates of recombination). The value of *I*A depends on the number of loci analyzed. Haubold & Hudson (2000) subsequently proposed a standardized index of association (*IAS*) defined as

$$I\_{\Lambda} \mathbf{s} = \begin{pmatrix} 1/l \ \mathbf{-1} \end{pmatrix} I\_{\Lambda \mathbf{s}}.$$

where *l* is the number of loci studied. This index has the advantage of being comparable between studies as long as it can be assumed that the neutral mutation parameter = 2*N*eµ is constant (Hudson, 1994). Two methods are commonly used for determining whether the computed *IA* represents a significant deviation from linkage equilibrium (*IA* ~ 0): resampling from randomized data sets using Monte Carlo simulations or using the parametric method described by Haubold et al. (1998). Both methods are perfectly good alternatives.

The genetic structure of the AHC, which has been previously determined using enzyme electrophoresis (MLEE), has revealed a clear clonal structure with strong linkage disequilibrium among 15 different protein loci (Miñana-Galbis et al., 2004). Additionally, this study demonstrated the usefulness of MLEE for separating the strains belonging to *A. bestiarum* and *A. salmonicida*, which are almost indistinguishable by phenotypic characteristics and 16S rRNA sequence analysis and present borderline DNA-DNA homology values.

In a multilocus sequence analysis (using six housekeeping genes) of a new set of AHC strains in which we included representatives of *A. piscicola* and *A. aquariorum*, we obtained

Although to our knowledge there is no function that could be used to define a level of divergence for distinguishing species, it seems clear that, in bacteria, there is always an exponential relationship between interspecies recombination and sequence divergence (Roberts & Cohan, 1993; Zawadzki et al., 1995; Vulic et al., 1997). Values of divergence among clades (Table 2) seem to correspond to those of a genetically isolated biological species. The lowest value corresponds to *A. bestiarum* clade 3, indicating that its isolates are undergoing speciation, but species boundaries are not as well-defined as in the other clades.

Linkage equilibrium is characterized by the statistical independence of alleles at all loci. A common method used in bacterial population genetics to quantify the degree of linkage between a set of loci in MLSA data is to use the index of association (*I*A) (Brown et al., 1980; Maynard-Smith et al., 1993). This index compares the ratio of variance calculated from the pairwise distribution of allele mismatches in the data (*V*O) and the ratio expected under a

*E*(*V*O) = ∑ *h*j (1-*h*j), where *h*j is an unbiased estimator of the population genetic diversity equivalent to heterozygosis in diploid organisms (Table 4). The index of association (*I*A) originally used by

*I*A =*V*O/*V*E-1, which would give a value of zero if there is no association between loci (Maynard-Smith et al., 1993). Values of this index significantly different from zero reflect strong linkage disequilibrium (lower rates of recombination). The value of *I*A depends on the number of loci analyzed. Haubold & Hudson (2000) subsequently proposed a standardized index of

*I*AS = (1/*l*-1) *I*A, where *l* is the number of loci studied. This index has the advantage of being comparable between studies as long as it can be assumed that the neutral mutation parameter = 2*N*eµ is constant (Hudson, 1994). Two methods are commonly used for determining whether the computed *IA* represents a significant deviation from linkage equilibrium (*IA* ~ 0): resampling from randomized data sets using Monte Carlo simulations or using the parametric method

The genetic structure of the AHC, which has been previously determined using enzyme electrophoresis (MLEE), has revealed a clear clonal structure with strong linkage disequilibrium among 15 different protein loci (Miñana-Galbis et al., 2004). Additionally, this study demonstrated the usefulness of MLEE for separating the strains belonging to *A. bestiarum* and *A. salmonicida*, which are almost indistinguishable by phenotypic characteristics and 16S rRNA sequence analysis and present borderline DNA-DNA

In a multilocus sequence analysis (using six housekeeping genes) of a new set of AHC strains in which we included representatives of *A. piscicola* and *A. aquariorum*, we obtained

described by Haubold et al. (1998). Both methods are perfectly good alternatives.

**5. Linkage disequilibrium and population structure** 

null hypothesis of linkage equilibrium (*V*E). In this case

Brown et al. (1980) is computed as

association (*IAS*) defined as

homology values.

*IAS* values different from 0 in all cases, indicating the absence of recombination and again revealing strong linkage disequilibrium when considering both the total population and the different groups of species (Table 5). This is in spite of the high number of alleles per locus and polymorphic sites (Table 3) and huge genetic diversity (Table 4). Values of *VO, VE, IAS* and the 5% critical values as determined by the Monte Carlo process (*L*MC) and from the parametric approach (*Lpara*) of Haubold et al. (1998) are shown in Table 5.


\* full-length sequence gene; \*\* nucleotide diversity () ± standard error (s.e.) Calculated by using MEGA5 software (Tamura et al., 2011).

Table 3. Sequence variation at six loci. Standard error estimates were obtained by the bootstrap method with 500 replicates.


\* genetic diversity at individual loci (*hj)*; \*\* mean genetic diversity (H) ± standard error (s.e.) *h*j = (*n*/*n*-1)∑*p*ij2, where *p*ij is the frequency of the *i*th allele at the *j*th locus and *n* the number of loci. Data were calculated by using R statistical software (R Development Core Team, 2010).

Table 4. Genetic diversity (*h*) at six loci for all STs and major species sets.


For the meaning of acronyms, see text. The *LMC* and *PMC* results were obtained from 10000 resamplings. Data were calculated by using R statistical software except for *Lpara* , which was determined with the LIAN 3.5 program (Haubold & Hudson, 2000).

Table 5. Multilocus linkage disequilibrium analysis of the AHC.

During the last years, with the availability of the first DNA sequence data of individual genes, evidence of recombination at the molecular level has accumulated for *Aeromonas* in genes such as *dnaJ*, *gyrB* and *recA* (Silver et al., 2011). Incongruence between trees reconstructed from individual genes appeared as further proof of recombination at the gene level. Nevertheless, multilocus analysis with gene sequences also revealed a clear clonal

Population Genetics of the "*Aeromonas hydrophila* Species Complex" 49

polymorphisms. This result agrees with the presence of mutations under positive selection spreading quickly through a population. These changes do not contribute to polymorphism but have a cumulative effect on divergence and the fixed NS/S is consequently greater than polymorphic NS/S (Egea et al., 2008). Two of the McDonald-Kreitman tests were not significant, *A. bestiarum* versus *A. salmonicida* and *A. piscicola* (*A. bestiarum* clade 3) versus the

> *A. best* **clade 3**

*A. hyd* **clade 1** 

*Fixed Polymorphic NI P* **S NS S NS** 

88 15 596 43 0.423 0.015757 \*

21 3 659 48 0.510 0.402179 ns

*A. hyd*  **clade 2**  *A. salm*

*A. best* **clade 2** 

*A. bestiarum* clade 1 --- 0.1703 0.4532 0.1366 0.1068 0.1907 *A. bestiarum* clade 2 0.7460 --- 0.1174 0.0863 0.0683 0.1051 *A. bestiarum* clade 3 0.5245 0.8099 --- 0.1077 0.0833 0.1480 *A. hydrophila* clade 1 0.7854 0.8529 0.8228 --- 0.1264 0.1264 *A. hydrophila* clade 2 0.8240 0.8798 0.8571 0.6428 --- 0.0993 *A. salmonicida* 0.7239 0.8264 0.7716 0.7982 0.8343 ---

Table 6. Pairwise estimates of population differentiation, *NST* (lower-left) and gene flow, *Nm*

*A. bestiarum* vs. *A. hydrophila* 83 26 1235 88 0.227 0.000000 \*\*\* *A. bestiarum* vs. *A. salmonicida* 57 7 964 77 0.650 0.325391 ns *A. hydrophila* vs. *A. salmonicida* 118 36 1141 86 0.247 0.000000 \*\*\*

*A. hydrophila* (clade 1*)* vs. *A. hydrophila* (clade 2) 56 16 803 48 0.209 0.000009 \*\*\* Acronyms are for synonymous substitutions (S); nonsynonymous substitutions (NS); neutrality index (*NI*); *P*-value from Fisher's exact test (*P*). \* 0.01<*P*<0.05; \*\* 0.001<*P*<0.01; \*\*\* *P*<0.001; not significant (ns).

Developments in gene sequence analysis have greatly enhanced the study of bacterial population genetics. Gene-wide approaches to mapping bacterial diversity, which have already proved effective for gaining fresh insight into bacterial evolution, have the potential to reveal the phenotypic basis of genetic diversity in the AHC and to investigate the dynamics of this complex bacterial community. The general objective of the work described in this chapter has been to evaluate the suitability of combining population genetics and phylogenetic approaches for the delineation of bacterial species in the AHC, considered by

The results obtained from the linkage disequilibrium analysis and sequence divergence show that the AHC is composed of four robust groups that basically correspond with the

other strains of the *A. bestiarum* group.

(upper-right).

(clade 2)

(clade 3)

**7. Conclusions** 

*A. best* **clade 1** 

Calculated using DnaSP v5.10 software (Librado & Rozas, 2009).

Calculated using DnaSP v5.10 software (Librado & Rozas, 2009).

many specialists a taxonomically tangled group.

Table 7. McDonald-Kreitman Test for molecular evidence of selection.

*A. bestiarum* (clades 1 and 3) vs. *A. bestiarum* 

*A. bestiarum* (clade 1 and 2) vs. *A. bestiarum*

structure in this bacterial group (Table 5). The question is that although bacteria are capable of accumulating gene fragments from other bacterial species or mutations, the recombinant segments are not long enough to break the clonal structure of the population. While the absence of linkage (*I*AS ~ 0) is difficult to explain without assuming high levels of recombination, linkage disequilibrium does not exclude the presence of significant levels of recombination (Touchon et al., 2009).

In our study we have also determined the presence of recombinant fragments in the *recA* (in four *A. bestiarum* strains) and *dnaJ* genes (five strains, 2 *A. bestiarum*, 2 *A. hydrophila* and 2 *A. salmonicida*). However, although these strains cluster separately when the corresponding tree is constructed, revealing the different origin of the gene fragments, they group together with the other strains when a concatenated tree is generated. This confirms that recombination is not sufficient to break the genetic cohesion of this group.

#### **6. Gene flow and divergence**

The existence of barriers to gene flow such as geographical separation, ecological adaptation or the accumulation of genetic differences ultimately leads to distinct lineages. These processes are usually more complicated in prokaryotes, since the boundaries of their species are sometimes distorted by gene transfer between divergent organisms. In addition, the mechanisms that can contribute to the cohesion of groups are very different in bacteria. We have determined the divergence among the different AHC clades using the nucleotide fixation index, *NST* (Nei & Kumar, 2000). The use of the *NST* as a measure of population differentiation under certain circumstances has been criticized (Jost, 2008), but it is still used for describing the average amount of such differentiation observed from multiple locus data (Ryman & Leimar, 2009). In our study, the determination of the *NST* values indicated a high level of interclade genetic differentiation (*N*ST = 0.8025). On average, most of the 80% of the total variance of nucleotide diversity was attributable to genetic differentiation among clades, whereas about 20% was found within populations. High levels of diversification were also found among most of the AHC groups of species (Table 6), with values always higher than 0.7 except in the case of *A. bestiarum* clade 3 / *A. bestiarum* clade 1 (*NST* = 0.5) and *A. hydrophila* clade 2 / *A. hydrophila* clade 1 (*NST* = 0.6). The clear divergence between the different clades described suggests they form coherent groups in which the phenomenon of recombination, if present, fails to break this consistency.

In bacterial populations it seems reasonable to equate the effect of lateral gene transfer (LTG) from other species with the product *Nm* (effective population size x migration rate) determined from *NST*. Indeed, assuming the limitations of the Wright island model (Wright, 1940), the value of *NST* at the equilibrium (1/(2*Nm*+1)) allows *Nm* to be calculated. The values obtained for all AHC clades (*NST* = 0.8, *Nm* ~ 0.13) again suggest that gene flow (in this case, the lateral gene transfer) is insufficient to counteract their genetic differentiation.

The McDonald-Kreitman Test (MKT, McDonald & Kreitman, 1991) was applied to our data sets to detect signs of selection (Table 7). Under neutrality, the ratio of nonsynonymous-tosynonymous fixed substitutions (between species) should be the same as the ratio of nonsynonymous-to-synonymous polymorphism (within species). We observed a high level of fixed replacements between most species studied. In all comparisons the ratio of nonsynonymous-to-synonymous substitutions was higher for fixed differences than for polymorphisms. This result agrees with the presence of mutations under positive selection spreading quickly through a population. These changes do not contribute to polymorphism but have a cumulative effect on divergence and the fixed NS/S is consequently greater than polymorphic NS/S (Egea et al., 2008). Two of the McDonald-Kreitman tests were not significant, *A. bestiarum* versus *A. salmonicida* and *A. piscicola* (*A. bestiarum* clade 3) versus the other strains of the *A. bestiarum* group.


Calculated using DnaSP v5.10 software (Librado & Rozas, 2009).

Table 6. Pairwise estimates of population differentiation, *NST* (lower-left) and gene flow, *Nm* (upper-right).


Acronyms are for synonymous substitutions (S); nonsynonymous substitutions (NS); neutrality index (*NI*); *P*-value from Fisher's exact test (*P*). \* 0.01<*P*<0.05; \*\* 0.001<*P*<0.01; \*\*\* *P*<0.001; not significant (ns). Calculated using DnaSP v5.10 software (Librado & Rozas, 2009).

Table 7. McDonald-Kreitman Test for molecular evidence of selection.

## **7. Conclusions**

48 Studies in Population Genetics

structure in this bacterial group (Table 5). The question is that although bacteria are capable of accumulating gene fragments from other bacterial species or mutations, the recombinant segments are not long enough to break the clonal structure of the population. While the absence of linkage (*I*AS ~ 0) is difficult to explain without assuming high levels of recombination, linkage disequilibrium does not exclude the presence of significant levels of

In our study we have also determined the presence of recombinant fragments in the *recA* (in four *A. bestiarum* strains) and *dnaJ* genes (five strains, 2 *A. bestiarum*, 2 *A. hydrophila* and 2 *A. salmonicida*). However, although these strains cluster separately when the corresponding tree is constructed, revealing the different origin of the gene fragments, they group together with the other strains when a concatenated tree is generated. This confirms that

The existence of barriers to gene flow such as geographical separation, ecological adaptation or the accumulation of genetic differences ultimately leads to distinct lineages. These processes are usually more complicated in prokaryotes, since the boundaries of their species are sometimes distorted by gene transfer between divergent organisms. In addition, the mechanisms that can contribute to the cohesion of groups are very different in bacteria. We have determined the divergence among the different AHC clades using the nucleotide fixation index, *NST* (Nei & Kumar, 2000). The use of the *NST* as a measure of population differentiation under certain circumstances has been criticized (Jost, 2008), but it is still used for describing the average amount of such differentiation observed from multiple locus data (Ryman & Leimar, 2009). In our study, the determination of the *NST* values indicated a high level of interclade genetic differentiation (*N*ST = 0.8025). On average, most of the 80% of the total variance of nucleotide diversity was attributable to genetic differentiation among clades, whereas about 20% was found within populations. High levels of diversification were also found among most of the AHC groups of species (Table 6), with values always higher than 0.7 except in the case of *A. bestiarum* clade 3 / *A. bestiarum* clade 1 (*NST* = 0.5) and *A. hydrophila* clade 2 / *A. hydrophila* clade 1 (*NST* = 0.6). The clear divergence between the different clades described suggests they form coherent groups in which the phenomenon of

In bacterial populations it seems reasonable to equate the effect of lateral gene transfer (LTG) from other species with the product *Nm* (effective population size x migration rate) determined from *NST*. Indeed, assuming the limitations of the Wright island model (Wright, 1940), the value of *NST* at the equilibrium (1/(2*Nm*+1)) allows *Nm* to be calculated. The values obtained for all AHC clades (*NST* = 0.8, *Nm* ~ 0.13) again suggest that gene flow (in this case, the lateral gene transfer) is insufficient to counteract their genetic differentiation. The McDonald-Kreitman Test (MKT, McDonald & Kreitman, 1991) was applied to our data sets to detect signs of selection (Table 7). Under neutrality, the ratio of nonsynonymous-tosynonymous fixed substitutions (between species) should be the same as the ratio of nonsynonymous-to-synonymous polymorphism (within species). We observed a high level of fixed replacements between most species studied. In all comparisons the ratio of nonsynonymous-to-synonymous substitutions was higher for fixed differences than for

recombination is not sufficient to break the genetic cohesion of this group.

recombination, if present, fails to break this consistency.

recombination (Touchon et al., 2009).

**6. Gene flow and divergence** 

Developments in gene sequence analysis have greatly enhanced the study of bacterial population genetics. Gene-wide approaches to mapping bacterial diversity, which have already proved effective for gaining fresh insight into bacterial evolution, have the potential to reveal the phenotypic basis of genetic diversity in the AHC and to investigate the dynamics of this complex bacterial community. The general objective of the work described in this chapter has been to evaluate the suitability of combining population genetics and phylogenetic approaches for the delineation of bacterial species in the AHC, considered by many specialists a taxonomically tangled group.

The results obtained from the linkage disequilibrium analysis and sequence divergence show that the AHC is composed of four robust groups that basically correspond with the

Population Genetics of the "*Aeromonas hydrophila* Species Complex" 51

Beaz-Hidalgo, R.; Alperi, A.; Figueras, M.J. & Romalde, J.L. (2009). *Aeromonas piscicola* sp.

Brown, A.D.; Feldman, M.W. & Nevo, E. (1980). Multilocus structure of natural populations

Egea, R.; Casillas, S. & Barbadilla, A. (2008). Standard and generalized McDonald-Kreitman

Farfán, M.; Miñana-Galbis, D.; Garreta, A.; Lorén, J.G. & Fusté, M.C. (2010). Malate

Feil, E.J.; Maiden, M.C.J.; Achtman, M. & Spratt, B.G. (1999). The relative contributions of

Feil, E.J.; LI, B.C.; Aanensen, D.M.; Hanage, W.P. & Spratt, B.G. (2004). eBURST: Inferring

multilocus typing data. *Journal of Bacteriology,* Vol. 186, No. 5, pp. 1518-1530. Figueras, M.J.; Soler, L.; Chacón, M.R.; Guarro, J. & Martínez-Murcia, A.J. (2000). Extended

*Journal of Systematic and Evolutionary Microbiology,* Vol. 50, No. 6, pp. 2069-2073. Janda, J.M. & Abbott, S. (2010). The genus *Aeromonas:* taxonomy, pathogenicity, and

Jost, L. (2008). GST and its relatives do not measure differentiation. *Molecular Evolution*, Vol.

Haubold, B.; Travisano, M.; Rainey, P.B. & Hudson, R.R. (1998). Detecting linkage disequilibrium in bacterial populations. *Genetics,* Vol. 150, No. 4, pp. 1341-1348. Haubold, B. & Hudson, R.R. (2000). LIAN 3.0: detecting linkage disequilibrium in

Hudson, R.R. (1994). Analytical results concerning linkage disequilibrium in models with

Huys, G.; Kämpfer, P.; Altwegg, M.; Kersters, I.; Lamb, A.; Coopman, R.; Lüthy-Hottenstein, J.;

*International Journal of Systematic Bacteriology,* Vol. 47, No. 4, pp. 1165-1171. Huys, G.; Kämpfer, P.; Albert, M.J.; Kühn, I.; Denys, R. & Swings, J. (2002). *Aeromonas* 

Küpfer, M.; Kuhnert, P.; Korczak, B.M.; Peduzzi, R. & Demarta, A. (2006). Genetic

Lan, R. & Reeves, P.R. (2001). When does a clone deserve a name? A perspective on bacterial species based on population genetics. *Trends in Microbiology,* Vol. 9, No. 9, pp. 419-424. Levin, B.R. & Bergstrom, C.T. (2000). Bacteria are different: observations, interpretations,

genetic transformation and conjugation. *Journal of Evolutionary Biology,* Vol. 7, No. 5,

Vancanneyt, M.; Janssen, P. & Kersters, K. (1997). *Aeromonas popoffii* sp. nov., a mesophilic bacterium isolated from drinking water production plants and reservoirs.

*hydrophila* subsp. *dhakensis* subsp. nov., isolated from children with diarrhoea in Bangladesh, and extended description of *Aeromonas hydrophila* subsp. *hydrophila* (Chester 1901) Stanier 1943 (Approved Lists 1980). *International Journal of Systematic* 

relationships of *Aeromonas* strains inferred from 16S rRNA, *gyrB* and *rpoB* gene sequences. *International Journal of Systematic and Evolutionary Microbiology,* Vol. 56,

speculations, and opinions about the mechanisms of adaptive evolution in

of *Hordeum spontaneum*. *Genetics,* Vol. 96, No. 2, pp. 523-536.

*Molecular Biology and Evolution,* Vol. 16, No. 11, pp. 1496-1502.

infection. *Clinical Microbiology Reviews,* Vol. 23, No. 1, pp. 35-73.

multilocus data. *Bioinformatics,* Vol. 16, No. 9, pp. 847-848.

*Nucleic Acids Research,* Vol. 36, No. 2, pp. W157-W162.

*and Applied Microbiology,* Vol. 33, No. 8, pp. 427-435.

pp. 471-479.

17, No. 18, pp. 4015-4026.

*Bacteriology,* Vol. 52, No. 3, pp. 705-712.

No. 12, pp. 2743-2751.

pp. 535-548.

nov., isolated from diseased fish. *Systematic and Applied Microbiology,* Vol. 32, No. 7,

test: a website to detect selection by comparing different classes of DNA sites.

dehydrogenase: a useful phylogenetic marker for the genus *Aeromonas*. *Systematic* 

recombination and mutation to the divergence of clones of *Neisseria meningitidis*.

patterns of evolutionary descent among clusters of related bacterial genotypes from

method for discrimination of *Aeromonas* spp. by 16S rDNA RFLP analysis. *International* 

phenotypically described species *A. hydrophila*, *A. bestiarum*, *A. popofii* and *A. salmonicida*. The average divergence between these clusters seems to exclude a significant influence of recombination in the genetic structure of this bacterial group and therefore they are valid taxonomic units, despite the extensive variability within some of them. Phenotypic characteristics lead to the differentiation of five *A. salmonicida* subspecies, but the lack of a consistent signal in our multilocus sequence analysis only allowed the possible differentiation of *A. salmonicida* subsp. *pectinolytica*. Similarly, it is impossible to differentiate between *A. hydrophila* subsp. *hydrophila* and *A. hydrophila* subsp. *ranae*. These results are in agreement with those obtained in a previous study with isolates belonging to the *A. veronii* Group, in which the authors failed to achieve the differentiation of biovars within these *Aeromonas* species (Silver et al., 2011). Nevertheless, *A. hydrophila* subsp. *dhakensis* strains, which cluster together with *A. aquariorum* isolates, exhibited the divergence levels of a biological species and hence deserve full species status. Consequently, the *A. aquariorum* isolates should be reclassified. Finally, in the *A. bestiarum* group we distinguished a clade (clade 3) that includes *A. piscicola* isolates as well as several strains probably misclassified as *A. bestiarum.* This clade seems to constitute an incipient new species with low values of differentiation and species boundaries less well defined than in *A. bestiarum* (clade 1) or *A. popofii* (clade 2).

It has been frequently postulated that in bacterial populations, lateral gene transfer is so common that it precludes the existence of true biological species. One of the aims of this study has been to verify if this hypothesis is applicable to our AHC data. Three lines of evidence suggest the contrary. First of all, using the Tamura Nei model, which best fits our data, we found considerable interspecific nucleotide diversity, suggesting a high degree of divergence that hampers recombination among AHC species. Secondly, the linkage disequilibrium analysis of six loci reveals a strong disequilibrium with *IAS* values, suggesting little or null influence of recombination in the genetic structure of AHC species. Thirdly, the *NST* values obtained reflect a high degree of differentiation between clades. In short, the genetic structure of the AHC appears to confirm that the entities phenotypically described as species form cohesive groups in which genetic recombination plays a limited role in reducing genetic variation and can be defined as biological species.

Like other authors (Lan & Reeves, 2001; Vinuesa et al., 2005), we agree that a combination of phylogenetic and population genetic studies is currently the best theoretical and practical approach to delineate species as natural and discrete lineages in the bacterial world.

## **8. Acknowledgments**

We are very grateful to Katri Berg (University of Helsinki, Finland), Margarita Gomila (Universitat de les Illes Balears, Mallorca, Spain) and Carmen Gallegos (Hospital Sant Llàtzer, Mallorca, Spain) for kindly providing some of the strains used in this study. This research was supported by the project CGL-2008-03281/BOS from the Ministerio de Educación y Ciencia, Spain.

## **9. References**

Abbott, L.S.; Cheung, W.K.W. & Janda, M.J. (2003). The genus *Aeromonas*: Biochemical characteristics, atypical reactions, and phenotypic identification schemes. *Journal of Clinical Microbiology,* Vol. 41, No. 6, pp. 2348-2357.

phenotypically described species *A. hydrophila*, *A. bestiarum*, *A. popofii* and *A. salmonicida*. The average divergence between these clusters seems to exclude a significant influence of recombination in the genetic structure of this bacterial group and therefore they are valid taxonomic units, despite the extensive variability within some of them. Phenotypic characteristics lead to the differentiation of five *A. salmonicida* subspecies, but the lack of a consistent signal in our multilocus sequence analysis only allowed the possible differentiation of *A. salmonicida* subsp. *pectinolytica*. Similarly, it is impossible to differentiate between *A. hydrophila* subsp. *hydrophila* and *A. hydrophila* subsp. *ranae*. These results are in agreement with those obtained in a previous study with isolates belonging to the *A. veronii* Group, in which the authors failed to achieve the differentiation of biovars within these *Aeromonas* species (Silver et al., 2011). Nevertheless, *A. hydrophila* subsp. *dhakensis* strains, which cluster together with *A. aquariorum* isolates, exhibited the divergence levels of a biological species and hence deserve full species status. Consequently, the *A. aquariorum* isolates should be reclassified. Finally, in the *A. bestiarum* group we distinguished a clade (clade 3) that includes *A. piscicola* isolates as well as several strains probably misclassified as *A. bestiarum.* This clade seems to constitute an incipient new species with low values of differentiation and species boundaries less well

It has been frequently postulated that in bacterial populations, lateral gene transfer is so common that it precludes the existence of true biological species. One of the aims of this study has been to verify if this hypothesis is applicable to our AHC data. Three lines of evidence suggest the contrary. First of all, using the Tamura Nei model, which best fits our data, we found considerable interspecific nucleotide diversity, suggesting a high degree of divergence that hampers recombination among AHC species. Secondly, the linkage disequilibrium analysis of six loci reveals a strong disequilibrium with *IAS* values, suggesting little or null influence of recombination in the genetic structure of AHC species. Thirdly, the *NST* values obtained reflect a high degree of differentiation between clades. In short, the genetic structure of the AHC appears to confirm that the entities phenotypically described as species form cohesive groups in which genetic recombination plays a limited

Like other authors (Lan & Reeves, 2001; Vinuesa et al., 2005), we agree that a combination of phylogenetic and population genetic studies is currently the best theoretical and practical

We are very grateful to Katri Berg (University of Helsinki, Finland), Margarita Gomila (Universitat de les Illes Balears, Mallorca, Spain) and Carmen Gallegos (Hospital Sant Llàtzer, Mallorca, Spain) for kindly providing some of the strains used in this study. This research was supported by the project CGL-2008-03281/BOS from the Ministerio de

Abbott, L.S.; Cheung, W.K.W. & Janda, M.J. (2003). The genus *Aeromonas*: Biochemical

characteristics, atypical reactions, and phenotypic identification schemes. *Journal of* 

approach to delineate species as natural and discrete lineages in the bacterial world.

role in reducing genetic variation and can be defined as biological species.

*Clinical Microbiology,* Vol. 41, No. 6, pp. 2348-2357.

defined than in *A. bestiarum* (clade 1) or *A. popofii* (clade 2).

**8. Acknowledgments** 

Educación y Ciencia, Spain.

**9. References** 


Population Genetics of the "*Aeromonas hydrophila* Species Complex" 53

Miñana-Galbis, D.; Urbizu-Serrano, A.; Farfán, M.; Fusté, M.C. & Lorén, J.G. (2009).

Nei, M., & Kumar, S. (2000). Genetic polymorphism and evolution, In: *Molecular Evolution* 

Nhung, P.H.; Hata, H.; Ohkusu, K.; Noda, M.; Shah, M.M.; Goto, K. & Ezaki, T. (2007). Use

O'Rourke, M. & Stevens, E. (1993). Genetic structure of *Neisseria gonorrhoeae* populations: a

Orskov, F.; Whittam, T.S.; Cravioto, A. & Orskov, I. (1990). Clonal relationships among

Pérez-Losada, M.; Brownw, E.B.; Madsen, A.; Wirth, T.; Viscidi, R.P. & Crandall, K.A. (2006).

Reith, M.E.; Singh, R.K.; Curtis, B.; Boyd, J.M.; Bouevitch, A.; Kimball, J.; Munholland, J.;

Roberts, M.S. & Cohan, F.M. (1993). The effect of DNA sequence divergence on sexual

Robinson, A.; Falush, D. & Feil, E.J. (eds.) (2010). *Bacterial Population Genetics in Infectious* 

Roselló-Mora, R. & Amann, R. (2001). The species concept for prokaryotes. *FEMS* 

Ryman, N. & Leimar, O. (2009). GST is still a useful measure of genetic differentiation – a comment on Jost's D*. Molecular Ecology*. Vol. 18, No. 10, pp. 2084-2087. Saavedra, M.J.; Figueras, M.J. & Martínez-Murcia, A.J. (2006). Updated phylogeny of the

Salaun, L.; Audibert, C.; Le Lay, G.; Burocoa, C.; Fauchère, J.L. & Picard, B. (1998). Panmintic

Seshadri, R.; Joseph, S.W.; Chopra, A.K.; Sha, J.; Shaw, J.; Graf, J.; Haft, D.; Wu, M.; Ren, Q.;

Selander, R.K.; Caugant, D.A.; Ochman, H.; Musser, J.M.; Gilmour, M.N. & Whittam, T.S.

markers. *FEMS Microbiology Letters,* Vol. 161, No. 2, pp. 231-239.

genus *Aeromonas*. *International Journal of Systematic and Evolutionary Microbiology,* 

structure of *Helicobacter pylori* demonstrated by the comparative study of six genetic

Rosovitz, M.J.; Madupu, R.; Tallon, L.; Kim, M.; Jin, S.; Vuong, H.; Stine, O.C.; Ali, A.; Horneman, A.J. & Heidelberg, J.F. (2006). Genome sequence of *Aeromonas hydrophila* ATCC 7966: jack of all trades. *Journal of Bacteriology,* Vol. 188, No. 23, pp.

(1986). Methods of multilocus enzyme electrophoresis for bacterial population

*and Phylogenetics,* pp. 231-264, Oxford University Press, New York.

*Evolutionary Microbiology,* Vol. 57, No. 6, pp. 1232-1237.

*The Journal of Infectious Diseases,* Vol. 162, No. 1, pp. 76-81.

evolution of a fish pathogen. *BMC Genomics,* Vol. 9, No. 427.

*Disease*, Wiley & Sons, ISBN 978-0-470-42474-2, New Yersey.

isolation in *Bacillus. Genetics,* Vol. 134, No. 2, pp. 401-408.

*Microbiology Reviews*, Vol. 25, No. 1, pp. 39-67.

Vol. 56, No. 10, pp. 2481-2487.

*Microbiology,* Vol. 6, No. 3, pp. 198-208.

*Microbiology,* Vol. 59, No. 8, pp. 1976-1983.

project.org/.

8272-8282.

*Aer. popoffii* by multilocus enzyme electrophoresis (MLEE). *Environmental* 

Phylogenetic analysis and identification of *Aeromonas* species based on sequencing of the *cpn60* universal target. *International Journal of Systematic and Evolutionary* 

of the novel phylogenetic marker *dnaJ* and DNA-DNA hybridization to clarify interrelationships within the genus *Aeromonas*. *International Journal of Systematic and* 

non-clonal pathogen. *Journal of General Microbiology,* Vol. 139, No. 11, pp. 2603-2611.

classic enteropathogenic *Escherichia coli* (EPEC) belonging to different O groups.

Population genetics of microbial pathogens estimated from multilocus sequence typing (MLST) data. *Infection, Genetics and Evolution,* Vol. 6, No. 2, pp. 97-112. R Development Core Team (2010). *R: a language and environment for statistical computing,* R

Foundation for Statistical Computing, Vienna, Austria, URL http://www.r-

Murphy, C.; Sarty, D.; Williams, J.; Nash, J.H.E.; Johnson, S.C. & Brown, L.L. (2008). The genome of *Aeromonas salmonicida* subsp. *salmonicida* A449: insights into the

prokaryotes. *Proceedings of the National Academy of Sciences of the United States of America,* Vol. 97, No. 13, pp. 6981-6985.


Li, Y.; Liu, Y.; Zhou, Z.; Huang, H.; Ren, Y.; Zhang, Y.; Li, G.; Zhou, Z. & Wang, L. (2011).

Librado, P. & Rozas, J. (2009). DnaSP v5: A software for comprehensive analysis of DNA

Maiden, M.C.; Bygraves, J.A.; Feil, E.; Morelli, G.; Russell, J.E.; Urwin, R.; Zhang, Q.; Zhou,

Martínez-Murcia, A.J.; Soler, L.; Saavedra, M.J.; Chacón, M.R.; Guarro, J.; Stackebrandt, E. &

Martínez-Murcia, A.J.; Saavedra, M.J.; Mota, V.R.; Maier, T.; Stackebrandt, E. & Cousin, S.

Martínez-Murcia, A.J.; Monera, A.; Saavedra, M.J.; Oncina, R.; López-Alvarez, M.; Lara, E. &

Martino, M.E.; Fasolato, L.; Montemurro, F.; Rosteghin, M.; Manfrin, A.; Patarnello, T.;

Maynard-Smith, J. (1991). The population genetics of bacteria. *Proceedings of the Royal Society* 

Maynard-Smith, J.; Smith, N.H.; O'Rourke, M. & Spratt, B.G. (1993). How clonal are

Maynard-Smith, J. (1995). Do bacteria have population genetics?, In: *Population Genetics of* 

McDonald, J.H. & Kreitman, M. (1991). Adaptive protein evolution at the *Adh* locus in

Milkman, R. (1973). Electrophoretic variation in *Escherichia coli* from natural sources. *Science,* 

Miñana-Galbis, D.; Farfán, M.; Lorén, J.G. & Fusté, M.C. (2002). Biochemical identification

Miñana-Galbis, D.; Farfán, M.; Fusté, M.C. & Lorén, J.G. (2004). Genetic diversity and

polymorphism data. *Bioinformatics,* Vol. 25, No. 11, pp. 1451-1452.

Staley, J.T. (eds.), Vol. 2, part B, pp. 557-578, Springer, New York.

*Systematic and Applied Microbiology,* Vol. 34, No. 3, pp. 189-199.

*B Biological Sciences,* Vol. 245, No. 1312, pp. 37-41.

pp. 1-12, Cambridge University Press, Cambridge.

*Drosophila*. *Nature,* Vol. 351, pp. 652-654.

*America,* Vol. 97, No. 13, pp. 6981-6985.

*Microbiology,* Vol. 8, No. 4, pp. 259-269.

1169-1175.

77, No. 14, pp. 4986-5000.

Vol. 90, No. 10, pp. 4384-4388.

Vol. 182, No. 116, pp. 1024-1026.

Vol. 193, No. 13, pp. 3389-3390.

prokaryotes. *Proceedings of the National Academy of Sciences of the United States of* 

Complete genome sequence of *Aeromonas veronii* strain B565. *Journal of Bacteriology,* 

J.; Zurth, K.; Caugant, D.A.; Feavers, I.M.; Achtman, M. & Spratt, B.G. (1998). Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. *Proceedings of the National Academy of Sciences of the United States of America,* Vol. 95, No. 6, pp. 3140-3145. Martin-Carnahan, A. & Joseph, S.W. (2005). Genus I. *Aeromonas* Stanier 1943, 213AL, In:

*Bergey's Manual of Systematic Bacteriology,* Garrity, G.M.; Brenner, D.J.; Krieg, N.R. &

Figueras, M.J. (2005). Phenotypic, genotypic, and phylogenetic discrepancies to differentiate *Aeromonas salmonicida* from *Aeromonas bestiarum*. *International* 

(2008). *Aeromonas aquariorum* sp. nov., isolated from aquaria of ornamental fish. *International Journal of Systematic and Evolutionary Microbiology,* Vol. 58, No. 5, pp.

Figueras, M.J. (2011). Multilocus phylogenetic analysis of the genus *Aeromonas*.

Novelli, E. & Cardazzo, B. (2011). Determination of microbial diversity of *Aeromonas* strains on the basis of multilocus sequence typing, phenotype, and presence of putative virulence genes. *Applied and Environmental Microbiology,* Vol.

bacteria? *Proceedings of the National Academy of Sciences of the United States of America,* 

*Bacteria,* Baumberg, S.; Young, J.P.W; Wellington, E.M.H. & Saunders, J.R. (eds.),

and numerical taxonomy of *Aeromonas* spp. isolated from environmental and clinical samples in Spain. *Journal of Applied Microbiology,* Vol. 93, No. 3, pp. 420-430.

population structure of *Aeromonas hydrophila*, *Aer. bestiarum*, *Aer. salmonicida* and

*Aer. popoffii* by multilocus enzyme electrophoresis (MLEE). *Environmental Microbiology,* Vol. 6, No. 3, pp. 198-208.


**4** 

*Russia* 

**Minisatellite DNA Markers in Population Studies** 

The discovery of an anonymous multiallelic locus in 1980 demonstrated for the first time that the human DNA contains hypervariable regions (Wyman & White, 1980). Eight alleles of a polymorphic locus that was not associated with any known gene were identified during the blot hybridization of total human genomic DNA treated with the restrictase EcoRI with a human DNA fragment (16 tbp in length) isolated from a phage genomic library. The multiallelic nature of the polymorphism at this locus did not stem from a variation in restriction sites; rather, it originated from a variable number of tandem repeats in the short core DNA sequence. Later, other similar polymorphic sites were detected near the 5 end of the insulin gene (Bell et al., 1982), in the Harvey *ras* oncogene (Capon et al., 1983), in the globin pseudogene (Proudfoot et al., 1982), and within the -globin cluster (Weller et al., 1984). In 1985, Jeffreys et al. published the results of their research, in which they described a fourfold repeat of 33 nucleotides in one of the introns of the human myoglobin gene (Jeffreys et al., 1985). These polymorphic regions consisted of tandem repeats of a short sequence (11–60 bp) and were termed variable number tandem repeats (VNTRs) (Kendrew

The other term for VNTR loci, minisatellites, was attributed based on the similarity of some of their properties with those of highly repetitive satellite DNA sequences. Tandem satellite DNA repeats are combined into a huge and structurally diverse group arranged into continuous clusters with monomeric units positioned in a head-to-tail configuration. The differences between minisatellite loci and satellite DNA are the greater variability in the length of the repeating unit of the latter (varying from 10 to 1000 bp) and chromosomal localization: satellite DNA is located in the regions of near-centromere heterochromatin in metaphase chromosomes and in the chromocenters of interphase nuclei, whereas minisatellite sequences are distributed evenly over most regions of all chromosomes (Miklos

The classification of Jeffreys et al. discriminates between microsatellites (with an elementary link of 2–6 bp), minisatellites (up to 100 bp), midisatellites (100–400 bp), and macrosatellites (up to several thousand bp) (Jeffreys et al., 1994). Later, a great emphasis was placed on loci that were generally classified as microsatellites with elementary links of less than 10 bp (short tandem repeats, STRs) and minisatellites with links of more than 10 bp in size (VNTRs) (Schlotterer C., 1998; Gemayel et al., 2010). These two types of variable DNA were

used as genetic markers in genomic and population studies.

**1. Introduction** 

& Lawrence, 1994).

& John, 1979).

Svetlana Limborska, Andrey Khrunin and Dmitry Verbenko *Institute of Molecular Genetics, Russian Academy of Sciences, Moscow* 

genetics and systematics. *Applied and Environmental Microbiology,* Vol. 51, No. 5, pp. 873-884.

