Previous Article | Next Article ![]()
Infection and Immunity, August 2006, p. 4892-4899, Vol. 74, No. 8
0019-9567/06/$08.00+0 doi:10.1128/IAI.02087-05
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Center of Infection and Immunity Amsterdam (CINIMA), Department of Medical Microbiology, Academic Medical Center, Amsterdam, The Netherlands
Received 28 December 2005/ Returned for modification 9 March 2006/ Accepted 11 May 2006
|
|
|---|
|
|
|---|
Although few plasmids of neisseriae have been sequenced, different plasmids have been identified in both pathogenic and commensal neisseriae, albeit with a bias to antibiotic resistance plasmids (7, 22, 24). This limited amount of sequence data restricts insight into neisserial plasmid diversity and evolution. Notable exceptions are the gonococcal plasmids pJD1 (15) and pJD4 (6), the meningococcal plasmids pJS-A (12) and pJS-B (5), and the commensal neisseriae plasmids pNL01 from N. lactamica (32a) and pMIDG2830 from N. flavescens (19).
In order to obtain more insight into the role of plasmids in horizontal gene transfer among neisseriae, we evaluated plasmid content and diversity and nucleotide composition.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Neisserial plasmid classification scheme
|
was transformed by the standard heat shock procedure. Plasmid inserts were sequenced using standard M13 primers or primer walking on cloned fragments or neisserial plasmid DNA according to the manufacturer's instruction (AB). Sequences were analyzed with CodonCode Aligner program (CodonCode Corporation) and analyzed with BLASTN or TBLASTX using the BLOSUM-62 substitution matrix (http://www.ncbi.nlm.nih.gov/BLAST/), the annotation tool Artemis software (www.sanger.ac.uk/), and the web tool NEBcutter version 2 (New England BioLabs, Inc.) (33).
Data analysis.
Clustering analysis of the different open reading frames (ORFs) was carried out using MEGA (Molecular Evolutionary Genetics Analysis) (16) (available from www.megasoftware.net), and genomic dissimilarity analyses of the different plasmids were carried out using 
-web as described previously (32). Tandem repeats and inverted repeats were detected using the Tandem (and Inverted) Repeats Finder at http://tandem.bu.edu/trf/trf.html (3).
Nucleotide sequence accession numbers. The nucleotide sequence data of the various plasmids are available in the EMBL/GenBank/DDBJ nucleotide sequence databases under accession numbers DQ223897 through DQ223899, DQ227323 through DQ227327, and DQ229163 through DQ229166 as depicted in Table 1.
|
|
|---|
Three additional N. lactamica isolates, isolated prior to 2004 and known to carry plasmids, were included (Table 1). From this collection of N. lactamica isolates, a total of 12 plasmids were completely sequenced. Together with sequenced neisserial plasmids available in public databases (pJS-A, pJS-B, pJD1, pNL01, and pMIDG2830 [accession numbers are given in Table 1]), six distinct groups of neisserial plasmids (I to VI) could be distinguished on the basis of nucleotide identity and replication protein similarity. The grouping of the plasmids was supported by intragroup tandem repeat similarity (Table 2). In contrast to groups I and II, groups III and IV comprise plasmids heterogeneous in size, while groups V and VI are represented by only one plasmid each. The single N. meningitidis plasmid identified, pNM12, was approximately 4.2 kb in size on the basis of restriction fragment patterns, which were highly similar (>90% identical fragments) to that of pNL01 (group IV [data not shown]). An AluI restriction fragment of 259 bp was sequenced and was nearly identical (257/259 bp) to the N. lactamica pNL01 plasmid, which is 4,150 bp in size; therefore, pNM12 was not sequenced further. Three N. lactamica plasmids, one plasmid each from groups I, II, and IV, were not sequenced, as their AluI and RsaI restriction enzyme digestion patterns were identical to that of as a sequenced member of the corresponding plasmid group.
|
View this table: [in a new window] |
TABLE 2. Tandem repeat counts in the different neisserial plasmids
|
Different groups of neisserial plasmids. (i) Group I. Group I consists of four plasmids almost identical in size (1,986 bp) and sequence, three of which have been isolated from N. lactamica strains isolated in 1975, 1993, and 2004 (pNL750149, pNL932024, and pNL12, respectively). Previously, pJS-A (AJ238491) was identified in Germany in N. meningitidis subgroup IV isolates exclusively (between 1986 and 1988) (12). The largest ORF, encoding a putative replication initiation factor (likely a topoisomerase) of 336 amino acids (pfam02486, E = 2 x 109), shows 31% identity (93/294 amino acid residues) to a putative protein on the 2-kbp cryptic plasmid from the ammonia-oxidizing bacterium Nitrosomonas (34), but no nucleotide identity has been found.
(ii) Group II. Of 18 N. lactamica strains isolated in 2004, 5 contained a plasmid approximately 2.25 kbp in size. These almost identical plasmids, making up plasmid group II, display two noteworthy features. First, all contain a long homopolymeric guanine tract of variable length (10, 12, 12, and 15 guanine residues for plasmids pNL18.1, pNL7.1, pNL15, and pNL11, respectively). This homopolymeric G tract is located upstream of an ORF, p153, encoding a putative protein of 153 amino acids and therefore could be involved in transcriptional phase variation (26, 29). However, it is doubtful whether the p153 sequence encodes a protein, as there is no significant similarity to NCBI database entries, and no putative 35 and 10 promoter domains could be identified.
Second, these four plasmids all contain two inverted neisserial DNA uptake sequences (DUS) (shown underlined) with two adenosine residues in between the DUS (TTCAGACGGCAAGCCGTCTGAA). However, this sequence has the reverse orientation in plasmid pNL15.
(iii) Group III. Group III consists of three different plasmids, of which two have been isolated from N. lactamica isolates (pNL3.2 and pNL9 [this study]) and one from N. meningitidis (pJS-B [5]). The remaining plasmid was previously identified among isolates from a cluster of hypervirulent N. meningitidis ET-37 (plasmid pJS-B, which can also occur as a chromosomal integration [5]). These plasmids vary considerably in size (ranging from 4 kbp [pNL3.2] to 8.3 kbp [pNL9]), and the genetic organization of the plasmids resembles that of the recently described meningococcal disease-associated (MDA) phage (4) (Fig. 1), as well as that of the recently described Eikenella corrodens virulence-associated plasmid pMU1 (1). Their putative replication-associated protein shows similarity to phage replication proteins (COG2946, E = 6 x1066). In addition, pJS-B and pNL9 both contain an ORF (pJS-B_8 and pNL9-8, respectively) with a conserved domain of the zonular occludens toxin Zot (pfam05707, E = 1 x 1048) family of proteins, which is 39% identical to its MDA homolog NMA1799 (Table 3).
![]() View larger version (23K): [in a new window] |
FIG. 1. Comparison of the genetic organizations of N. meningitidis plasmids pNL3.2, pNL9, and pJS-B and the meningococcal disease-associated island with the ORFs NMA1792 to NMA1800. Similar colors indicate similar protein-coding ORFs (except for white). The Eikenella corrodens pMU1 plasmid is depicted at the top of the figure.
|
|
View this table: [in a new window] |
TABLE 3. Properties of the ORFs on plasmids pNL9, pJS-B, and pNL3.2a
|
|
View larger version (11K): [in a new window] |
FIG. 2. Analysis of the upstream region of the phage-associated replication proteins on plasmids from group III (pJS-B_3, pNL9-3, and pNL3.2_2 [see also Table 3]). In the comparison we included the upstream region of the putative phage-associated protein from the meningococcal disease-associated gene cluster (NMA1792), which is thought to be inactive, and the phase-variable promoter from the porA gene (from N. meningitidis MC58 [NMB1429]). The boxes represent regions of similarity between the porA promoter and plasmid sequences. The 35 and 10 domains of the porA promoter are underscored. The putative 35 region of the (inactive) NMA1792 promoter seems to have degenerated.
|
|
View this table: [in a new window] |
TABLE 4. Subclassification of group IV plasmidsa
|
Genome signature comparisons between plasmid DNA and chromosomal DNA display large differences.
The genome signature, which is the set of dinucleotide relative abundance values (13, 14), is one of the parameters available to identify putative horizontally transferred DNA. We compare the genome signature dissimilarity scores (
*) between the different neisserial plasmids identified in this study and a representative genome sequence (for N. lactamica we use the preliminary N. lactamica sequence, available as contigs at the Sanger Institute) as described previously (32). Briefly, the
* of the plasmids are compared to the
* of identical-sized fragments of the host genomes. High compositional dissimilarity between plasmid and host genome is reflected by a large percentage of genomic fragments with a lower
* than the
* for the plasmid. We find high compositional dissimilarity between plasmid DNA and chromosomal DNA for most plasmid groups (Table 5), as noted previously for a large collection of prokaryotic plasmids from the Plasmid Database (31). However, group II plasmids do not display a high genomic dissimilarity with the N. lactamica genome sequence.
|
View this table: [in a new window] |
TABLE 5. Compositional analyses based on the genome signatures between the different neisserial plasmids and neisserial speciesa
|
|
|
|---|
The reason for this erratic distribution of plasmids among different neisserial species remains unknown, but the erratic distribution suggests either limited establishment of plasmids in N. meningitidis or more efficient exchange between N. lactamica and N. gonorrhoeae or successful acquisition by N. lactamica and N. gonorrhoeae. Limited establishment might result if plasmids more readily integrate into the chromosome in N. meningitidis. Integration of pJS-B into the meningococcal chromosome was shown by Claus and colleagues (5). Differences in the ability to take up extracellular DNA could explain differences in plasmid acquisition. Although most neisseriae were found to be naturally competent, DUS-mediated transformation has been established only for the pathogenic species (9), and to our knowledge, natural transformation of N. lactamica has not yet been reported. However, plasmid transfer may follow a DUS-independent route for most plasmid groups, as only the compositionally nonanomalous group II plasmids contain copies of the DUS. Finally, different restriction barriers between the different species could influence successful establishment. Only a few restriction and modification systems that are present in N. meningitidis and have no isoschizomer in N. lactamica, are known notably NmeBI, NmeRI, and NmeSI. Only two plasmids (pNL3.1 and pJD4) contain sites for these restriction and modification systems, so it is unlikely that restriction barriers are the main reason for the observed differences in occurrence of all plasmids in N. meningitidis and N. lactamica.
The broad spectrum of plasmids described in this study should harbor suitable vectors for N. lactamica, which may be of interest, as N. lactamica is currently under consideration for its vaccine potential against meningococcal disease (10, 11).
The nucleotide sequences of the different neisserial plasmids display substantial diversity both in length and composition within the relatively small number of tested isolates. Together with the neisserial plasmids with chromosomal sequences available in public databases, we identified six distinct plasmid groups (I to VI) on the basis of nucleotide identity, restriction fragment length polymorphism, and protein similarity. Additionally, we considered intragroup genome signature resemblance and tandem repeat similarity. Group IV plasmids could be subdivided on the basis of the presence or absence of genes encoding Rep and Mob proteins. Plasmids pNL14 and pNL871104, harboring combinations of these genes, appear to be hybrid plasmids on the basis of sequence comparisons. As a consequence, these plasmids may have extended their host range, as they are found in three species, N. meningitidis, N. lactamica, and N. gonorrhoeae.
In group III, the different plasmids vary in length considerably. The genetic organizations of the ORFs encoding putative phage-associated proteins (including a putative toxin and a phage replication-associated protein) on the larger plasmids resemble that of the recently identified, integrated, MDA phage from N. meningitidis Z2491 (4). It is tempting to hypothesize that these N. lactamica plasmids may in fact be phage replicative forms isolated during the plasmid DNA isolation procedure. Further experiments are to be performed to confirm or refute this hypothesis. The identification of a putative phage toxin-encoding gene on N. lactamica plasmids highly similar (39% identical) to that of the MDA phage raises questions concerning the association between the MDA phages and pathogenic potential, as found recently in a study by Bille and coworkers (4). Obviously, N. lactamica lacks other virulence factors important in the pathogenicity of invasive disease (the capsule is one clear example).
The presence of disease-associated sequences in commensal bacteria indicates that commensal strains may form a reservoir of latent virulence factors, which may contribute to the variability of these factors in pathogenic neisseriae. As a clear example, Linz and coworkers demonstrated the frequent exchange of tbpB sequences between commensal neisseriae and N. meningitidis, which contributed significantly to the antigenic variation of the transferrin binding protein TbpB in meningococci (17). In a more recent study, we have shown that N. lactamica also contains genomic islands encoding putative virulence-associated genes with high similarity to genes on genomic islands in meningococci (32a).
The phase variation-associated sequence C6-10N10G7 upstream of the ORF encoding phage replication-associated protein is another feature in group III plasmids. The variation among the different plasmids in its homopolymeric cytidine tract and the striking resemblance with the phase-variable porA promoter in N. meningitidis (30) are indicative of a phase-variable expression of the Rep protein on these plasmids. Of note, a similar phase variation-associated sequence was previously identified upstream of NMA1792 by Snyder and coworkers (27). In vitro verification of phase variation for the putative phage replication proteins is required, but these sequences indicate that plasmids (or replicative forms of phages), neisserial phages, and even Eikenella corrodens plasmids potentially contain phase-variable genes.
A homopolymeric nucleotide tract of variable length was also identified in group II plasmids. Association of the expression of the ORF downstream of this homopolymeric guanidine tract with phase variation is uncertain, since the ORF shows no similarity to any database entry and we were unable to identify a putative promoter in the sequences flanking this guanidine tract.
Most of the plasmids have very high genomic dissimilarity scores compared to their respective genomic contexts. How this relates to the plasmid host histories is unclear. A high genomic dissimilarity score may reflect either a recent acquisition of the mobile element, a bias supporting extracellular stability, or an atypical genetic organization required to maintain its selfish and/or mobile behavior (31). In contrast, plasmids of group II have a genome signature similar to that of the N. lactamica genome. Also, group II plasmids contain a dyad neisserial DUS, as well as tandem repeat sequences that are relatively overrepresented in the meningococcal, gonococcal, and preliminary N. lactamica genome sequences, in contrast to plasmids of other groups. Together, these results suggest that plasmids of group II may have a longer evolutionary relationship with their host than plasmids of the other plasmid groups.
Carriage of N. lactamica may assist in the development of natural immunity to N. meningitidis by induction of cross-reactive antibodies (8). Moreover, the incidence of meningococcal disease is lower in communities with high N. lactamica carrier rates (20). It has been previously established that commensal and pathogenic neisseriae share a common gene pool (17, 18). Our data indicate that commensal neisseriae may function as a reservoir for putative virulence-associated sequences located on plasmids (e.g., those of group III) and on chromosomal genomic islands (32a). Possibly, during carriage of N. lactamica early in life, these pathogenicity-associated factors may contribute to the natural immunity to meningococcal infection later in life (25).
This research was part of the Network of Excellence "European Virtual Institute for Functional Genomics of Bacterial Pathogens" funded by the Sixth Framework Programme of the European Commission (proposal/contract no. 512061).
Supplemental material for this article may be found at http://iai.asm.org/. ![]()
|
|
|---|

-Web, an online tool to assess composition similarity of individual nucleic acid sequences. Bioinformatics 21:3053-3055.This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»