Phylogeny of the staphylococcal major autolysin and its use in genus and species typing

Abstract

The major staphylococcal autolysin Atl is an important player in cell separation and daughter cell formation. In this study, we investigated the amino acid sequences of Atl proteins derived from 15 staphylococcal and 1 macrococcal species representatives. The overall organization of the bifunctional precursor protein consisting of the signal peptide, a propeptide (PP), the amidase (AM), six repeat sequences (R(1) to R(6)), and the glucosaminidase (GL) was highly conserved in all of the species. The most-conserved domains were the enzyme domains AM and GL; the least-conserved regions were the PP and R regions. An Atl-based phylogenetic tree for the various species representatives correlated well with the corresponding 16S rRNA-based tree and also perfectly matched the phylogenetic trees based on core genome analysis. The phylogenetic distance analysis of 18 AtlA proteins of various Staphylococcus aureus strains and 15 AtlE proteins of S. epidermidis revealed that both species representatives formed a relatively homogeneous cluster. Two S. epidermidis strains, M23864:W1 and VCU116, were identified by Atl typing that clustered far more distantly and belonged to either S. caprae and S. capitis or a new subspecies. Here we show that Atl typing is a useful tool for staphylococcal genus and species typing by using either the highly conserved AM domain or the less-conserved PP domain.

Fig 1

Basic organization and similarity of staphylococcal and macrococcal Atl domains. The similarity and identity values shown are based on an alignment of the Atl protein domains of all of the species representatives listed in Fig. 3.

Fig 2

Annotation of staphylococcal and macrococcal Atl domains. Atl is a bifunctional enzyme that is composed of an N-terminal SP, followed by a PP region, an AM domain (red), six GW-containing cell wall binding repeats (R_1a to R_3b), and finally the C-terminal GL domain (blue). The R_1a-to-R_3b subgroup in two repeat species is indicated by white and gray boxes. The checked boxes in AtlCS and AtlSI indicate the absence of GW motifs.

Fig 3

(A) Phylogenetic tree based on staphylococcal and macrococcal Atl proteins. The Atl tree reflects the relationship of 16 staphylococcal and macrococcal species representatives and was reconstructed via SplitsTree4 by neighbor joining (http://www.splitstree.org). The Atl tree revealed seven species groups that were in good agreement with the groups identified by DNA-DNA reassociation studies (20) or by 16S rRNA gene-based analysis (9, 22). Branch length values are indicated. (B) Phylogenetic tree estimated from 16S rRNA gene sequences. The evolutionary relationship of the species analyzed was inferred using the neighbor-joining method (18). The tree shown represents a consensus tree evaluated by 2,000 bootstrap replicates. The tree is drawn to scale, with branch lengths in the same units as the evolutionary distances used to infer the phylogenetic tree. Evolutionary distances were computed by the maximum composite likelihood method (23) as the number of base substitutions per site. All of the steps involved in tree construction were conducted in MEGA5 (24).

Fig 4

Comparative phylogenetic trees based on Atl proteins and on core genome analysis. The tree is based on 12 sequenced staphylococcal species representatives and 1 Macrococcus representative. The core genomes of all of the species representatives comprised 99 genes. Both trees lead to the same phylogenetic relationship. Branch length value is indicated.

Fig 5

Comparative phylogenetic trees of S. aureus strains based on Atl proteins and on core genome analysis. The tree is based on 18 sequenced S. aureus strains with a core genome of 1,071 genes. Both trees lead to the same phylogenetic relationship. Branch length values are indicated.

Fig 6

Comparative phylogenetic trees of S. epidermidis strains and other coagulase-negative species representatives based on Atl proteins and on core genome analysis. The tree is based on 23 sequenced S. epidermidis strains with a core genome of 251 genes. Both trees lead to the same phylogenetic relationship. Branch length values are indicated.