New Insights into the Classification and Integration Specificity of Streptococcus Integrative Conjugative Elements through Extensive Genome Exploration

Abstract

Recent genome analyses suggest that integrative and conjugative elements (ICEs) are widespread in bacterial genomes and therefore play an essential role in horizontal transfer. However, only a few of these elements are precisely characterized and correctly delineated within sequenced bacterial genomes. Even though previous analysis showed the presence of ICEs in some species of Streptococci, the global prevalence and diversity of ICEs was not analyzed in this genus. In this study, we searched for ICEs in the completely sequenced genomes of 124 strains belonging to 27 streptococcal species. These exhaustive analyses revealed 105 putative ICEs and 26 slightly decayed elements whose limits were assessed and whose insertion site was identified. These ICEs were grouped in seven distinct unrelated or distantly related families, according to their conjugation modules. Integration of these streptococcal ICEs is catalyzed either by a site-specific tyrosine integrase, a low-specificity tyrosine integrase, a site-specific single serine integrase, a triplet of site-specific serine integrases or a DDE transposase. Analysis of their integration site led to the detection of 18 target-genes for streptococcal ICE insertion including eight that had not been identified previously (ftsK, guaA, lysS, mutT, rpmG, rpsI, traG, and ebfC). It also suggests that all specificities have evolved to minimize the impact of the insertion on the host. This overall analysis of streptococcal ICEs emphasizes their prevalence and diversity and demonstrates that exchanges or acquisitions of conjugation and recombination modules are frequent.

Keywords: Streptococcus; T4SS; integrase; integration site; integrative and conjugative elements.

FIGURE 1

Procedure for identifying candidate ICEs in sequenced genomes. The amino-acid sequences encoded by a chromosome are collected as multifasta files and processed as follows (see Materials and Methods). AAA172 Signature proteins are identified by BlastP search using our reference sequences as query and the set of multifasta files as search database. Resulting hits are filtered and validated. AAA173 The location of genes encoding the validated signature proteins are visualized using Artemis and ICEs are delimited. AAA174 Domain composition of signature proteins are searched using Biomart and the signature proteins are grouped into classes. AAA175 Multiple alignments of signature proteins in each class are performed using Clustal Omega and their phylogeny is analyzed using maximum likelihood (ML) methods and BioNJ.

FIGURE 2

Phylogenetic analysis of ICEs and dICE belonging to the Conj_Tn916 superfamily. (A) Relaxases; (B) coupling proteins; (C) VirB4 proteins. Bootstrap supports are given as followed: ML/BioNJ. X marks the nodes that are not validated with BioNJ. ICE names are colored according to the family of conjugation module they encode: orange = Conj_Tn916, green = Conj_ICESt3 (including elements encoding relaxases related to the one of ICE_SmuA159_tRNAleu in bold dark green). Because of their very close phylogenetic relationships, relaxases (A), CP (B), and VirB4 (C) of only three ICEs representative of the Conj_Tn916 family are shown. Refer to Supplementary Table S1 for ICE/dICE and strain details.

FIGURE 3

Identification of conserved CDSs within the Conj_ICESt3, Conj_Tn1549 and Conj_Tn5252 families of ICEs. (A) Conj_ICESt3 family; (B) Conj_Tn1549 family; (C) Conj_Tn5252 family. Genes encoding conserved proteins in all ICEs of a given family are indicated by arrows with below their putative function (or known homologs).

FIGURE 4

Phylogenetic analysis of ICEs and dICEs belonging to the Conj_Tn5252 superfamily. The trees of CPs are shown in this figure. ML boostrap values for relaxases/CPs/VirB4 (in this order) are given. X marks the nodes that are not validated with other proteins. Families are also supported by BioNJ analysis (data not shown). In mauve are ICEs/dICEs belonging to the Conj_vanG family, in dark blue those belonging to the Conj_Tn1549 family, in light-blue those of the Conj_TnGBS2 family and in purple that of the Conj_Tn5252 family. Elements marked with an asterisk are not integrated in their primary sites but in secondary ones. Refer to Table 1 for ICE/dICE and strain details.

FIGURE 5

Phylogenetic analysis of ICEs and dICEs belonging to the Conj_TnGBS1 superfamily. The trees of CPs are shown in this figure. ML boostrap values for relaxases/CPs/VirB4 is given in this order. Families are also supported by BioNJ analysis (data not shown). Abs means that one sequence was missing. Refer to Supplementary Table S1 for ICE/dICE and strain details.

FIGURE 6

Phylogenetic trees of tyrosine integrases from streptococcal ICEs/dICEs. Bootstrap values are given as followed: ML/BioNJ. X marks the nodes that are not validated with BioNJ. ICE names are colored according to the family of conjugation module they encode: orange = Conj_Tn916, green = Conj_ICESt3, mauve = Conj_vanG, dark-blue = Conj_Tn1549, purple = Conj_Tn5252, and in light-blue those of the Conj_TnGBS2. Brackets gather together closely related integrases sharing the same specificity of integration. Integration specificity of the ICEs/dICEs is indicated. Elements marked with an asterisk are not integrated in their primary sites but in secondary ones. Refer to Supplementary Table S1 for ICE/dICE and strain details.

FIGURE 7

Characterization of ICE/dICE integration loci and their position relative to the integrase CDSs. The genes within (or next to) which an ICE is inserted are in blue. Tyrosine integrases are in red, serine recombinases in green and DDE transposases in yellow. The sizes (in bp) of the DR (or of IRs when specified) are indicated in red when the sequence is inside of the conjugative element (in blue, outside). Numbers represent the number of ICEs integrated in a given target gene.

FIGURE 8

Phylogenetic tree of serine integrases from streptococcal ICE/dICE. Bootstrap values are given as followed: ML/BioNJ. X marks the nodes that are not validated with MP. ICE name are colored according to the family of conjugation module they encode (dark-blue = Conj_Tn1549, purple = Conj_Tn5252, mauve = Conj_vanG), and in light-blue those of the Conj_TnGBS2. Brackets gather together closely related integrases sharing the same specificity of integration. Genes in which the ICEs/dICEs are integrated are indicated. Refer to Supplementary Table S1 for ICE/dICE and strain details. Elements marked with an asterisk are not integrated in their primary sites but in secondary ones.

FIGURE 9

Phylogenetic tree of DDE transposases from streptococcal ICE/dICE. Bootstrap values are given as followed: ML/BioNJ. ICE name are colored according to the family of the conjugation module they encode (black = Conj_TnGBS1, light-blue = Conj_TnGBS2. Refer to Table 1 for ICE/dICE and strain details.

FIGURE 10

Genetic organization of ICE_SagILRI005_rplL and comparison against ICE_SgaUCN34_TnGBS2 and ICE_ Sag018883_rplL. For the comparison, gray shading between the genetic elements represents regions with >50% amino acid sequence identity. The arrows represent the individual ORFs. Putative functions of the conserved genes are those deduced from the functional annotation of the ICE using Agmial.