Short tandem repeat sequences in the Mycoplasma genitalium genome and their use in a multilocus genotyping system

Abstract

Background: Several methods have been reported for strain typing of Mycoplasma genitalium. The value of these methods has never been comparatively assessed. The aims of this study were: 1) to identify new potential genetic markers based on an analysis of short tandem repeat (STR) sequences in the published M. genitalium genome sequence; 2) to apply previously and newly identified markers to a panel of clinical strains in order to determine the optimal combination for an efficient multi-locus genotyping system; 3) to further confirm sexual transmission of M. genitalium using the newly developed system.

Results: We performed a comprehensive analysis of STRs in the genome of the M. genitalium type strain G37 and identified 18 loci containing STRs. In addition to one previously studied locus, MG309, we chose two others, MG307 and MG338, for further study. Based on an analysis of 74 unrelated patient specimens from New Orleans and Scandinavia, the discriminatory indices (DIs) for these three markers were 0.9153, 0.7381 and 0.8730, respectively. Two other previously described markers, including single nucleotide polymorphisms (SNPs) in the rRNA genes (rRNA-SNPs) and SNPs in the MG191 gene (MG191-SNPs) were found to have DIs of 0.5820 and 0.9392, respectively. A combination of MG309-STRs and MG191-SNPs yielded almost perfect discrimination (DI = 0.9894). An additional finding was that the rRNA-SNPs distribution pattern differed significantly between Scandinavia and New Orleans. Finally we applied multi-locus typing to further confirm sexual transmission using specimens from 74 unrelated patients and 31 concurrently infected couples. Analysis of multi-locus genotype profiles using the five variable loci described above revealed 27 of the couples had concordant genotype profiles compared to only four examples of concordance among the 74 unrelated randomly selected patients.

Conclusion: We propose that a combination of the MG309-STRs and MG191-SNPs is efficient for general epidemiological studies and addition of MG307-STRs and MG338-STRs is potentially useful for sexual network studies of M. genitalium infection. The multi-locus typing analysis of 74 unrelated M. genitalium-infected individuals and 31 infected couples adds to the evidence that M. genitalium is sexually transmitted.

Figure 1

Clustering dendrogram of unrelated M. genitalium clinical strains based on genotype profiles at five genomic loci. The dendrogram was constructed using categorical coefficient and UPMGA clustering and included 30 strains from Scandinavian patients (SC) and 44 from New Orleans patients (NO). The genotypes at three PLP STR loci were determined by direct sequencing alone. For specimens containing mixed STR sequences, the genotypes are listed in the order from predominant to minor type. Only 4 sets of two patients had matching genotype profiles at all 5 loci (highlighted in grey shading). Note that three sets of two patients (nos. 58 and 14NO, 104F and 7NO, and 66NO and 199) and one set of three patients (nos. 27, 170 and 83F), which showed identical MG309-STR and MG191-SNP genotypes, could be distinguished by MG307-STRs, MG338-STRs or rRNA-SNPs.

Figure 2

Representative sequence chromatograms illustrating how mixed STR sequences can be detected by direct sequencing. (A) Partial sequence chromatograms of MG309-STRs in specimen no. 57 which contained a mixture of 12 and 11 repeats. (B) Partial sequence chromatograms of MG309-STRs in specimen no. 198.2 which contained a mixture of 15, 16 and 17 repeats. The PCR product from each specimen was directly sequenced in two separate sequencing reactions from forward and reverse directions, respectively. The AGT/AAT repeat units are underlined. In direct sequencing of the PCR products, the minority population (11 repeats in panel A and 16 and 17 repeats in panel B) is three bases out of alignment with the majority population (12 repeats in panel A and 15 repeats in panel B), resulting in a mixture of two or three chromatogram peaks as indicated by asterisks. The sequences flanking the repeat region can be determined accurately by sequencing from two directions. The estimated individual sequences shown above the original sequence readouts were confirmed by sequencing of plasmid clones for both specimens (no. 57 in additional file 2 and no. 198.2 in additional file 2 and Figure 3B).

Figure 3

Variation in repeat numbers and distribution patterns of the MG309 AGT/AAT repeat sequence. (A) MG309 AGT/AAT repeat sequences in selected patient specimens obtained from unrelated patients, sequential specimens from the same patients and specimens from concurrently infected couples. (B) MG309 AGT/AAT repeat sequences among selected plasmid clones from two sequential specimens from one patient (no. 198). The published sequence from the M. genitalium G37 genome (GenBank accession number NC_000908) was chosen as the reference sequence (G37). The boxed regions in each panel contain the MG309 AGT/AAT repeat units. Hyphens indicate gaps introduced to optimize alignment. Numbers on the left of each sequence refer to the specimen codes (panel A) or plasmid clones (panel B). Repeat units that differ from those of G37 are highlighted in yellow.

Figure 4

Molecular typing of M. genitalium based on rRNA-SNPs. (A) SSCP analysis of a 246-bp DNA fragment covering the three polymorphic sites for types 2, 3 and 4 of the rRNA operon [13]. The gel was sliver stained. (B) Distribution of M. genitalium rRNA-SNP genotypes between New Orleans (n = 44) and Scandinavia (n = 30) patients. The DNA sequences for types 2 to 5 are available from GenBank with accession numbers AY374418 to AY374421. By the generalized Fisher's exact test, p = 0.0004 for the overall distribution. By traditional Fisher's exact test, p = 0.0089 for type 2 and p = 0.0002 for type 3.

Figure 5

Clustering dendrogram of M. genitalium genotype profiles in concurrently infected couples. Eighteen couples were Scandinavia (SC) and 13 were from New Orleans (NO). The dendrogram was constructed using categorical coefficient and UPMGA clustering. Each couple was assigned the same number. F – female; M – male. One female (12-F) had two male partners 12-M1 and 12-M2. For specimens containing mixed sequences, the genotypes are listed in the order from predominant to minor type. The 4 couples showing mismatching profiles were highlighted in four different colours.