|Home | About | Journals | Submit | Contact Us | Français|
Transcriptional regulation remains poorly understood in Mycoplasma genitalium, the smallest self-replicating cell and the causative agent of a spectrum of urogenital diseases. Previously, we reported that MG_149, a lipoprotein-encoding gene, was highly induced under physiological hyperosmolarity conditions. In this study we further analyzed MG_149 transcription with a focus on the identification of promoter elements and regulatory mechanisms. We established MG_149 as a genuine osmoinducible gene that exhibited the highest transcript abundance compared to other lipoprotein genes. Using genetic approaches, we demonstrated that the -10 region of the MG_149 promoter was essential for osmoinduction. Moreover, we showed that MG_149 osmoinduction was regulated by DNA supercoiling, as the presence of novobiocin decreased MG_149 expression in a dose-dependent manner. Taken together, these results indicate that DNA supercoiling participates in controlling MG_149 expression during in vivo-like conditions.
Mycoplasma genitalium is the smallest self-replicating prokaryote with a streamlined genome of only 580 kb (Fraser et al., 1995). Because of its limited biosynthetic capabilities, M. genitalium depends on a parasitic lifestyle during which the host supplies essential metabolic precursors (amino acids, nucleotides, and fatty acids) and sterols for mycoplasma growth (Razin et al., 1998). M. genitalium colonizes the surfaces of urogenital epithelial cells and is capable of invading target cells and persisting long term as well (Mernaugh et al., 1993, Baseman & Tully, 1997, Dallo & Baseman, 2000). M. genitalium is considered one of the leading causes of non-gonococcal Chlamydia-negative urethritis in men (Taylor-Robinson & Horner, 2001, Jensen, 2006, Thurman et al., 2010). Also, M. genitalium has been linked to reproductive tract diseases, including cervicitis, endometritis, and pelvic inflammatory disease, as well as other clinical disorders (Baseman & Tully, 1997, Taylor-Robinson, 2002, Baseman et al., 2004, Haggerty, 2008). Consistent with clinical inflammatory manifestations, previous studies show that in vitro infection of human genital epithelial cells or monocyte-derived macrophages with M. genitalium (McGowin et al., 2009b), or stimulation of human monocytes with M. genitalium lipid-associated membrane proteins (LAMP) (You et al., 2008, He et al., 2009), leads to the production of pro-inflammatory cytokines.
Lipoproteins are abundant in mycoplasma cell membranes and have been increasingly recognized as potential virulence determinants (Chambaud et al., 1999, Rottem, 2003). It is known that mycoplasma lipoproteins modulate the host immune system by activating Toll-like receptors (TLR) with subsequent production of pro-inflammatory cytokines (Chambaud et al., 1999). For example, two lipoproteins of M. genitalium, MG_149 and MG_309, have been shown to activate TLR1/2 or TLR2/6 leading to the release of pro-inflammatory cytokines (Shimizu et al., 2008b, McGowin et al., 2009a). Similarly, three lipoproteins of the closely related Mycoplasma pneumoniae, including MPN162 (MG_149 homolog), MPN602 (encoding the subunit b of F0F1-type ATPase), and MPN611, trigger pro-inflammatory cytokine secretion by activating TLR1/6 (Shimizu et al., 2007) or TLR1/2/6 (Shimizu et al., 2005). Further, the synthetic lipopeptides of MPN162 and MPN602 increase chemokine and inflammatory cytokine production and leukocyte infiltration in a mouse model (Shimizu et al., 2008a), indicating that these lipoproteins can induce inflammatory responses in vivo. Apart from the activation of the host immune system, we reported that lipoproteins MG_186 and MPN133 possess nuclease activity and cause pathological effects in cell lines (Li et al., 2010, Somarajan et al., 2010). Both lipoproteins could play essential roles in pathogenesis by degradation of nucleic acids and/or initiation of cell death, given the intracellular residence and complete dependence of M. genitalium and M. pneumoniae on the host for nucleic acid precursors.
Genome sequencing of M. genitalium and M. pneumoniae revealed the presence of 21 and 46 putative lipoprotein-encoding genes, respectively, and most are of unknown function (Himmelreich et al., 1997). Since the products of these genes appear to be cell surface proteins, they could be critical factors in mediating mycoplasma-host interactions. Available evidence clearly indicates that the expression of mycoplasma lipoprotein genes is tightly controlled at the transcription level. For example, differential expression of lipoprotein genes in M. pneumoniae occurs during oxidative and acidic stress and following mycoplasma contact with human lung epithelial cells (Hallamaa et al., 2008). Also, we reported that lipoprotein genes of M. genitalium are differentially expressed under osmotic shock, with five being significantly induced (Zhang & Baseman, 2011). However, the exact mechanisms by which these genes are regulated have not yet been identified.
Very little is known about transcriptional regulation in mycoplasmas. Transcriptional control in mycoplasmas appears unique, as many regulatory factors commonly found in other bacteria are absent in these genome-streamlined prokaryotes (Herrmann & Reiner, 1998). For example, M. genitalium and M. pneumoniae possess only a single sigma factor (Fraser et al., 1995, Himmelreich et al., 1996), whereas Escherichia coli has seven sigma factors which play key roles in responding to physiological and environmental signals (Helmann, 2002). Also, the Rho termination factor gene is missing in mycoplasmas, and transcription termination occurs at poorly defined sites. It has been shown that genes of unrelated functions can be co-transcribed in mycoplasmas due to the long convergent gene clusters with short or no intergenic regions (Benders et al., 2005). Additionally, the promoter structure of mycoplasma genes is not clearly defined. Although sequence comparisons indicate a strong consensus sequence for the -10 region but not for the -35 region, the role played by these elements for promoter activity has not been clearly examined experimentally (Weiner et al., 2000, Musatovova et al., 2003). Nevertheless, the transcriptome landscape in mycoplasmas appears much more complicated than previously thought. The transcriptome resembles that of eukaryotic cells by exhibiting a high frequency of antisense transcripts, alternative transcripts and dynamic operon structures, as revealed in a recent study on M. pneumoniae (Guell et al., 2009). Hence, transcriptional regulation in mycoplasmas remains largely unknown, although these smallest prokaryotes can differentially express genes in response to various environmental changes (Weiner et al., 2003, Madsen et al., 2006a, Madsen et al., 2006b, Musatovova et al., 2006, Cecchini et al., 2007, Schafer et al., 2007, Madsen et al., 2008).
DNA supercoiling has been recognized as a global mechanism of gene regulation in E. coli and other bacteria (Hatfield & Benham, 2002, Dorman, 2006). DNA supercoiling affects gene expression by modulating promoter activity through alteration of DNA structure and melting energy or by influencing the binding of transcription factors. Environmental factors including temperature, anaerobiosis, osmolarity and growth phase can alter the degree of DNA supercoiling and thus regulate gene expression (Dorman, 1991, Hatfield & Benham, 2002, Travers & Muskhelishvili, 2005). It has been reported that many bacterial pathogens exploit DNA supercoiling as a means of regulating virulence gene expression during infection (Dorman, 2006).
In this report we focused on MG_149, a lipoprotein-encoding gene identified recently as most highly induced when profiling the global transcriptome of M. genitalium during osmotic shock (Zhang & Baseman, 2011). We characterized MG_149 expression under hyperosmolarity conditions and performed detailed genetic analysis of the MG_149 promoter, identifying the -10 region as essential for osmoinduction. Furthermore, we demonstrated that the upregulation of MG_149 can be attributed to DNA supercoiling changes. These data provide important insights into transcriptional regulation of M. genitalium genes during mycoplasma adaptation to host environment.
As already mentioned, MG_149 was the most highly induced gene during osmotic shock (Zhang & Baseman, 2011) and was reported to be dispensable during in vitro growth (Glass et al., 2006), thus making further genetic analysis possible. We focused on MG_149 because it is also the major LAMP component that can activate TLRs and may contribute to pathogenesis of M. genitalium in vivo (Shimizu et al., 2008b). To further characterize MG_149 expression under hyperosmolarity conditions, we performed Northern blot analysis with total RNA isolated from M. genitalium cultures grown in the presence or absence of NaCl for 1 h (Fig. 1A). MG_149 appeared as a monocistronic mRNA with a single band close to 900 bases. The expression of MG_149 occurred at very low levels without NaCl treatment. However, in the presence of NaCl, MG_149 expression increased dramatically in an osmolarity-dependent manner (1.8-, 3.1- and 4.8-fold for 0.1, 0.2, and 0.3 M NaCl, respectively, as determined by densitometry; Fig. 1A), with maximal induction following exposure to 0.2 and 0.3 M NaCl. Hereafter, we used 0.2 M NaCl for osmotic shock experiments, as this concentration gave optimal induction of MG_149 with less inhibition on the growth of M. genitalium cells (Zhang & Baseman, 2011). The dependence of MG_149 transcription on hyperosmolarity was further validated when osmotically shocked M. genitalium cells were returned to SP-4 medium without NaCl addition for 1 h. Fig. 1B shows that this resulted in the restoration of MG_149 expression to control levels, indicating that induction of MG_149 is reversible. Additionally, the induction of MG_149 by NaCl is not a salt-specific effect, as the use of sucrose to achieve equivalent osmotic pressures of SP-4 medium produced a similar expression pattern (Fig. S1).
Previously, we reported that MG_149 and four other lipoprotein-encoding genes (MG_067, MG_068, MG_439 and MG_440) were upregulated after osmotic shock (Zhang & Baseman, 2011). Since lipoproteins may act as virulence determinants for mycoplasma pathogenesis, we compared MG_149 transcript levels with the other four lipoprotein genes. We treated M. genitalium with or without 0.2 M NaCl for 1 h and measured transcripts using quantitative Real-Time PCR (qRT-PCR). MG_149 exhibited the highest basal expression, even under control conditions (Fig. 2). With NaCl treatment, MG_149 demonstrated the greatest increase in transcript expression and abundance, although the other four genes also responded with varying degrees of upregulation. These quantitative data reinforce the selective induction of these lipoprotein genes under osmotic shock and identified MG_149 as the most responsive.
The MG_149 locus is positioned in a long cluster of genes orientated in the same direction. MG_149 (nucleotides 188607–189452) is flanked upstream by MG_148 (nucleotides 187302–188531) and downstream by MG_478 (nucleotides 189492–189956, previously annotated as MG_149.1) and separated from MG_148 by an intergenic region of 76 bp. Since Northern blot analysis revealed MG_149 is a monocistronic transcript, we identified the MG_149 promoter by mapping the TSS. Both primer extension (PE) and 5′ RACE (Rapid Amplification of cDNA Ends) produced a single TSS for MG_149 regardless of NaCl treatment (Fig. 3 and Fig. S2). The TSS is located 20 nucleotides upstream of the translational start of MG_149. Based on this information, we identified a hexamer of TATAAT located 5 nucleotides upstream of the TSS as the putative -10 region (Pribnow box, Fig. 3), which agrees well with the consensus -10 sequence of σ70 promoters that we previously described for M. genitalium genes (Dhandayuthapani et al., 1998, Musatovova et al., 2003). Also, we noted a weak -35 consensus region of TTAGAA upstream of this putative -10 region (Fig. 3).
To experimentally characterize the MG_149 promoter, we generated a series of deletion mutants with progressively truncated intergenic regions starting at the 5′ end (Figs. 4 and and5).5). These mutants were produced with potential promoter regions precisely deleted on the chromosome using homologous recombination. Moreover, we placed the selectable marker tetM438 immediately upstream of MG_149 but in opposite orientation, so that it would not affect MG_149 expression (Fig. 4A). Individual mutants were picked from single colonies and confirmed by PCR amplification. All mutants produced a single amplification product about 2 kb larger than that of wild type due to the tetM438 insertion (Fig. 4B). We tested the effects of promoter deletions on the expression of MG_149 by exposing deletion mutants to 0.2 M NaCl for 1 h (Fig. 5). Using Northern blot analysis, we showed that the deletion of sequences covering the putative -35 region and beyond (PD2 and PD3), and the spacer region (PD4), did not significantly alter NaCl-induced MG_149 transcription (3.0-, 2.9- and 2.8-fold for PD2, PD3 and PD4, respectively, compared to 3.0-fold for wild type and PD1, as determined by densitometry). In contrast, deletion of the sequence covering the putative -10 region (PD5 and PD6) totally abolished MG_149 expression. Importantly, retention of the putative -10 region (PD4) enabled MG_149 induction in the presence of 0.2 M NaCl, indicating that the -10 region alone is essential for promoter activity.
Similarly, we also generated a set of M. genitalium mutants with point mutations introduced in the putative -10 or -35 regions, and the MG_149 promoter was further characterized with these mutants in the absence or presence of 0.2 M NaCl (Figs. 4 and and6).6). We found that a single A to G mutation at the fifth base of the hexameric TATAAT of the -10 region totally abolished the expression of MG_149 (Fig. 6, -10B). However, the same mutation at the second base of TATAAT showed no effect when compared to wild type (3.0-fold for -10A and WT, Fig. 6). This could be due to a more important role played by downstream nucleotides in promoting the formation of “open-complexes” when transcription is initiated. By contrast, three mutants with mutations at either single or double positions in the putative -35 region exhibited no significant difference in MG_149 expression (3.0-, 3.1- and 3.0-fold for -35A, -35B and -35C, respectively, as determined by densitometry, Fig. 6). Together, these results further indicate that the -10 region is critical for the activity of the MG_149 promoter, whereas the -35 region is not.
To test whether the -10 region is sufficient for MG_149 promoter activity, we generated a transcriptional fusion mutant of M. genitalium placing the expression of MG_186, a lipoprotein gene encoding a nuclease (Li et al., 2010), under the control of the -10 region of MG_149 promoter (-10). This was achieved by replacing 100 bp of MG_186 upstream sequence (containing promoter region) with the -10 region of MG_149 using homologous recombination (Fig. S3A). Similarly, a promoter-less MG_186 mutant was also generated by removing 100 bp of MG_186 upstream sequence (ΔP) and was used as a control. These strains were confirmed by PCR amplification and DNA sequencing (Figs. S3B and S3C). qRT-PCR data revealed that fusion of the -10 region of MG_149 promoter significantly increased the basal level of MG_186 compared to wild type and control strains (ΔP) (Fig. 7). This is consistent with the observed high basal expression of MG_149 mentioned early (Fig. 2). Under osmotic shock, while the expression of MG_186 was repressed in the wild-type strain, the transcript level of MG_186 was increased ~2.0 fold in the promoter fusion mutant (-10) (Fig. 7).
To investigate mechanism(s) by which MG_149 osmoinduction is controlled, we considered alternative non-protein mechanisms because of the paucity of regulatory factors in M. genitalium. Since high osmolarity is known to cause a rapid increase in negative DNA supercoiling in other bacteria (Higgins et al., 1988, Alice & Sanchez-Rivas, 1997), we tested whether DNA supercoiling participates in MG_149 osmoinduction by using novobiocin, an inhibitor of DNA gyrase (Gellert et al., 1976). If DNA supercoiling is involved in transcriptional regulation of MG_149, we reasoned that the presence of novobiocin should reduce MG_149 expression upon osmotic shock. We simultaneously treated cultures of M. genitalium with novobiocin at sub-inhibitory concentrations (minimal inhibitory concentration, MIC: 16 μg ml−1) in the presence or absence of 0.2 M NaCl for 1 h. The presence of novobiocin showed no significant influence on the basal expression of MG_149. However, we observed that MG_149 osmoinduction was reduced by novobiocin in a dose-dependent manner (2.4-, 2.2-, and 1.8-fold for 2, 4 and 8 μg ml−1 novobiocin, respectively, compared with 2.7-fold for NaCl only control, as determined by densitometry) (Fig. 8A). Similar results were also obtained using sub-inhibitory concentrations of norfloxacin (MIC: 25 μg ml−1), another inhibitor of DNA gyrase (Fig. S4). As a control, chloramphenicol (MIC: 6 μg ml−1), an antibiotic that inhibits protein synthesis and does not affect DNA supercoiling, showed minimal effect on MG_149 osmoinduction (Fig. S5).
To provide further evidence that DNA supercoiling is involved in MG_149 osmoinduction, we examined the expression kinetics of MG_149 along with MG_122 (topA), a gene known to be transcriptionally sensitive to DNA supercoiling change (Dorman & Corcoran, 2009). M. genitalium cultures were treated with 0.2 M NaCl or 0.2 M NaCl plus novobiocin (8 μg ml−1) for 20, 40 and 60 min, and qRT-PCR was performed to compare transcript levels of these genes with control cultures (Fig. 8B). The expression of MG_122, which increased with salt and decreased with novobiocin, is consistent with the known effects of DNA supercoiling change. The rapid induction of MG_149 (3.5-fold) at 20 min post NaCl treatment and the significant downregulation in the presence of novobiocin further indicates that MG_149 osmoinduction is regulated by DNA supercoiling.
Finally, we tested whether the -10 region of MG_149 promoter is sensitive to DNA supercoiling change. We treated PD4 mutant with 0.2 M NaCl or 0.2 M NaCl plus novobiocin (8 μg ml−1) for 1 h and measured MG_149 transcripts with Northern blot. Our result shows that the presence of novobiocin dramatically reduced the osmoinduction of MG_149 compared to the culture treated with NaCl only (Fig. 8C, 1.2-fold vs. 3.9-fold as determined by densitometry). Together with observations described above, these results have convincingly demonstrated that MG_149 osmoinduction is mediated by DNA supercoiling.
To identify the structural element of MG_149 that confers DNA supercoiling responsiveness, we examined G+C content in the DR located between the -10 region and the TSS (Fig. 9). Previously, a high G+C content in DR was responsible for promoter responsiveness to DNA supercoiling (Figueroa-Bossi et al., 1998, Schneider et al., 2000). We generated a mutant of M. genitalium with two Gs of the DR being substituted by Ts (Fig. 9, designated DRm) in order to test if lowering the DR G+C content affects MG_149 osmoinduction (Fig. 9). Interestingly, lowering the G+C content showed little effect on MG_149 osmoinduction compared to wild type (3.9-fold vs. 4.2-fold as determined by densitometry). Furthermore, MG_149 expression was decreased similarly in novobiocin-treated DR mutant and wild-type strain (2.1-fold for both, as determined by densitometry), indicating that sequence features other than DR were responsible for MG_149 osmoinduction.
The transcription of many lipoprotein-encoding genes, as with other genes in M. genitalium and M. pneumoniae, is known to be tightly regulated (Hallamaa et al., 2008). Although selective induction of these genes appears to be dependent on the presence of specific environmental cues, no clear mechanism has been identified so far. In this regard, conserved heat shock genes are the only examples known to be regulated by CIRCE (Controlling Inverted Repeat of Chaperone Expression) and HrcA (Weiner et al., 2003, Musatovova et al., 2006). In our efforts to investigate the mechanisms controlling transcription of lipoprotein genes, we focused on MG_149 because it is a gene uniquely sensitive to osmolarity change (Zhang & Baseman, 2011) and could be a virulence factor (Shimizu et al., 2008b). We first confirmed that the induction of MG_149 was dependent on hyperosmolarity conditions and reversible once osmotic stress was relieved (Fig. 1). We then demonstrated that MG_149 displayed the highest transcript abundance of all five lipoprotein genes under both control and NaCl conditions (Fig. 2). Collectively, MG_149 represents a physiologically relevant and responsive gene, thus providing a good candidate for transcriptional analysis.
We investigated the promoter elements essential for MG_149 osmoinduction, as Northern blot analysis revealed that MG_149 was expressed as a monocistronic transcript. MG_149 produced only a single TSS using either RNA from NaCl-treated or control samples (Fig. 3 and Fig. S2), which differs from the presence of heterogeneous TSS frequently identified in lipoprotein genes of M. pneumoniae (Weiner et al., 2000). By sequence comparison, we identified the putative -10 region (TATAAT) and -35 region (TTAGAA) of the MG_149 promoter. It is not surprising that the putative -35 region deviates from the consensus sequence (TTGACA) of σ70 promoters. Previous studies show that a strong consensus sequence exists only for the -10 region but not the -35 region for mycoplasmal promoters (Dhandayuthapani et al., 1998, Weiner et al., 2000, Musatovova et al., 2003). It appears that the -35 region may not be a prerequisite for RNA polymerase recognition, but this notion has not been clearly proven experimentally. To characterize the MG_149 promoter, we generated a series of M. genitalium mutants with progressive deletions or point mutations in the promoter region using homologous recombination (Figs. 4 and and5).5). The placement of tetM438 in the opposite direction of MG_149 retained the intactness of genetic context, so that MG_149 expression would not be disturbed. Our results unambiguously indicated that the -10 region was essential for MG_149 promoter activity, whereas the -35 region was not (Figs. 5 and and6).6). This conclusion was further supported by the transcriptional fusion of the -10 region to MG_186 (-10) leading to the induction of MG_186 in the presence of NaCl (Fig. 7) as opposed to downregulation driven by its endogenous promoter (Fig. 7). Our results partially agree with a genetic analysis of ldh and ackA promoters of M. pneumoniae (Halbedel et al., 2007). Both studies conclude that the -10 region is critical for promoter activity, while the -35 is not. However, the promoters of ldh and ackA produced only constitutive expression of the reporter gene and failed to respond to the presence of glycerol or glucose (Halbedel et al., 2007). The failed response of these latter constructs to environmental signals could be explained by the loss of the original genome context, resulting from random transposon integration.
We explored the possible mechanisms controlling MG_149 expression. Since only a single sigma factor exists in M. genitalium, we focused on alternative mechanisms and proposed that DNA supercoiling may participate in MG_149 osmoinduction. It is known that high osmolarity can cause a rapid increase in negative DNA supercoiling in bacteria such as E. coli and Bacillus subtilis, as revealed by the topological status of reporter plasmids (Higgins et al., 1988, Alice & Sanchez-Rivas, 1997, Conter et al., 1997). Additionally, a regulatory role of DNA supercoiling for osmolarity-sensitive genes has been demonstrated in E. coli and other bacteria (Higgins et al., 1988, Cheung et al., 2003, Fournier & Klier, 2004). Using novobiocin, we clearly showed that DNA supercoiling is involved in MG_149 osmoinduction, as increasing drug concentrations gradually decreased MG_149 expression in the presence of NaCl (Fig. 8A). Monitoring the kinetics of MG_122 (topA), a gene known to be sensitive to DNA supercoiling (Cheung et al., 2003, Dorman & Corcoran, 2009), along with MG_149, further supported this conclusion (Fig. 8B). Also, the rapid response of MG_149 transcription to NaCl alone or NaCl and novobiocin, shown by the maximal changes observed at 20 min post NaCl treatment, is consistent with direct regulation by DNA supercoiling. Finally, MG_149 osmoinduction was dramatically decreased by novobiocin in PD4 mutant, the promoter deletion mutant having only the -10 region (Fig. 8C). Taken together, these data convincingly demonstrate that MG_149 is regulated by DNA supercoiling.
We analyzed DNA features that confer MG_149 transcriptional responses to DNA supercoiling changes. Previously, high G+C content within the DR was shown to be responsible for promoter sensitivity to DNA supercoiling (Pemberton et al., 2000, Schneider et al., 2000). High G+C content could impose a kinetic restraint to either DNA untwisting or DNA melting, or both, which is important for transcriptional initiation. Increase in negative DNA supercoiling is known to drive DNA unwinding and overcome this barrier (Dorman, 2006). As a result, mutations that lower G+C of DR could decrease promoter responsiveness to negative DNA supercoiling. Our results indicated that mutations lowering G+C in DR did not appreciably alter MG149 expression patterns (Fig. 9). The sensitivity of MG_149 promoter to DNA supercoiling could be intrinsic (Lim et al., 2003), which is supported by the transformed osmolarity responsiveness of MG_186 (Fig. 7) and by the significant difference of MG_149 from other osmoinducible lipoprotein genes in the possible promoter region (Fig. S6). Interestingly, a lower fold change was observed for the transcriptional fusion mutant (-10) compared to that of MG_149 in wild type strain, which suggests sequence downstream of the MG_149 translational start may be involved in full osmoinduction as previously reported for proU in Salmonella typhimurium (Overdier & Csonka, 1992).
In summary, we have further characterized MG_149 osmoinduction by determining its TSS and promoter properties. Moreover, we provide evidence that the selective induction of MG_149 in responsive to osmotic stress is regulated by DNA supercoiling. DNA supercoiling can be affected by many environmental factors and has been shown to regulate virulence genes when pathogenic bacteria adapt to host environments. Therefore, it represents an appealing mechanism for transcription regulation in M. genitalium and other mycoplasmas. Closer examination of the regulatory role played by DNA supercoiling could yield interesting insights into how these minimal cells adapt to the in vivo environment.
Wild type M. genitalium strain G37 (ATCC 33530) and mutants were grown in SP-4 broth at 37°C in tissue culture flasks (Corning). For osmotic shock experiments, NaCl (5 M stock solution) was added to SP-4 medium of exponential phase M. genitalium cultures to achieve the required final concentrations. Mycoplasma cultures were grown at 37°C for specific time periods prior to RNA extraction. For general cloning purposes, TOP10 E. coli cells were grown at 37°C in LB broth or LB agar plates containing 100 μg ml−1 ampicillin.
RNA isolation was performed as previously described (Zhang & Baseman, 2011).
For Northern blot analysis, 3 μg of RNA from each sample were resolved by a formaldehyde denaturing gel and transferred by capillary action to a Zeta-Probe GT nylon membrane (Bio-Rad). Probes were generated by random priming of the full length MG_149 PCR product with the large (Klenow) fragment DNA Polymerase I (New England Biolab) in the presence of [α-32P] dATP (PerkinElmer). Membranes were hybridized at 42°C overnight, washed twice for 10 min at 59°C as described previously (Church & Gilbert, 1984), and scanned using a Typhoon 9400 PhosphorImager (Molecular Dynamics), or exposed to X-ray film (Kodak). To quantify changes in levels of RNA, the intensity of bands on digital images of Northern blots was analyzed by using ImageJ available at http://rsb.info.nih.gov/ij/index.html. The values were expressed in arbitrary units, and the fold change in the signal intensity was determined after the background value was subtracted.
qRT-PCR was performed as previously described (Zhang & Baseman, 2011). To measure transcript abundance of lipoprotein genes, 1.5 μg of DNase I-treated RNA were mixed with 1 μl of reverse primers (2 uM each) and transcribed. The obtained cDNAs were 50 X diluted with distilled water before adding to SYBR green PCR master mix (Applied Biosystems). Absolute quantification mode was used to obtain Ct values for each gene and genomic DNA standards (five serial 10-fold dilutions of M. genitalium genomic DNA, 108 to 104 copies/reaction). Copy numbers of each transcript were calculated by comparing Ct values with corresponding standard curves. For monitoring the kinetic expression of MG_122 and MG_149, we applied the comparative threshold cycle method to compare amounts of transcripts under different experimental conditions. 16S ribosomal RNA was used as the endogenous control to normalize data. Relative levels of transcripts were expressed as fold changes (n-fold) compared to control values, and calculations were completed by RQ Manager 1.2 (Applied Biosystems).
For primer extension, oligonucleotide MG_149PE (antisense primer) was end-labeled with [γ-32P] ATP using T4 polynucleotide kinase (Promega). About 15–20 ng of radiolabeled oligonucleotides were added to 25 μg of total mycoplasma RNA, and the mixture was heated to 58°C for 20 min and cooled slowly to allow annealing of primers with RNA. For each reaction, 9 μl of reverse transcriptase extension mix containing 5 μl of AMV primer extension 2 X buffer, 1.4 μl of 40 mM sodium pyrophosphate and 1 μl of AMV Reverse Transcriptase were added and incubated at 42°C for 30 min. The cDNA yield was ethanol precipitated and dissolved in 10 μl of double distilled H2O and 5 μl of loading buffer. Sequencing reactions were performed with Sequenase 7-deaza-dGTP Sequencing Kit according to manufacturer’s instruction (USB) using MG149_PE. A plasmid harboring a fragment amplified with MG_149S and MG_149PE was used as the template. Five μl of each primer extension sample were analyzed on 6% sequencing gels alongside sequencing reactions.
5′ RACE was performed with a 5′/3′ RACE kit (Roche) designed for amplification of the 5′/3′ end of messenger RNAs by reverse transcription PCR. Briefly, 2 μg of total mycoplasma RNA were used to synthesize the first strand cDNA by Transcriptor Reverse Transcriptase using MG_149 specific primer SP1. The cDNA was enriched by QIAquick PCR purification column (Qiagen) and further treated with terminal transferase to add a homopolymeric A-tail at the 3′ end in the presence of dATP. The obtained cDNA was used as template for the amplification of targeted cDNA with oligo dT-anchor primer and MG_149 specific primer SP2. Finally, the PCR product was TA cloned into pCRII plasmid (Invitrogen) and five individual plasmids were sequenced to map the 5′ end.
To generate M. genitalium mutants with deletions in the MG_149 promoter region, a series of knock-out plasmids with progressive deletions of the intergenic region upstream of MG_149 were constructed. First, a 1 kb BamHI-NotI fragment harboring MG_148 to its stop codon was amplified with primer PDL5′ and PDL3′ (Table S1), and this PCR product was cloned into pMTnTet438 to create pM_ut. Next, the NotI-EcoRI fragment including tetM438 and the MG_148 region was cleaved and ligated with pCRII to create pCR_ut. Then, a series of DNA fragments (2.3 kb) which differed only in the MG_149 promoter region were amplified using primer PDR3′ in combination with primers PD1 to PD6, respectively. These fragments were separately cloned into pCR_ut by using EcoRI and SpeI to create corresponding knock-out plasmids pCRII-utd of PD1 to PD6.
To generate M. genitalium mutants with point mutations in the putative -10 and -35 regions of MG_149 promoter and the DR, site-directed mutagenesis was performed on PD1 knock-out plasmid with various primer pairs (Table S1) according to the instructions provided by the QuickChange Site-Directed Mutagenesis Kit (Stratagene). The obtained plasmids were verified to carry the desired mutations by DNA sequencing (Nucleic Acids Core Facility of UTHSCSA).
To generate M. genitalium strains with the promoter region of MG_186 being removed or fused to the -10 region of MG_149 promoter, two knock-out plasmids were constructed. First, tetM438 was moved from pMTnTet438 to pCRII by using NotI and XhoI. Next, a BamHI-NotI fragment harboring 1 kb of the upstream sequence of MG_186 was amplified with primer 5′ UMG_186 and 3′ UMG_186. This fragment was digested and ligated upstream of tetM438 on pCRII. The obtained plasmid was then used for the insertion of MG_186 fragments by using XhoI and XbaI downstream of tetM438. A fragment harboring the full length MG_186 was amplified with primer pair 5′ MG_186 (ΔP) and 3′DMG_186. The insertion of this fragment created the knockout plasmid for generating promoter-less MG_186. Similarly, a fragment fusing the -10 region of MG_149 to MG_186 was amplified with primer pair 5′ DMG_186 (-10) and 3′ DMG_186 using the MG_186 full length PCR product as DNA template. The insertion of this fragment downstream of tetM438 created the knock-out plasmid for generating the -10 region fusion of MG_186 (-10).
Transformation of M. genitalium was performed by electroporation as previously described (Pich et al., 2006). Briefly, strain G37 was grown in a 75 cm2 tissue culture flask to mid-exponential phase. Surface-attached mycoplasma cells were harvested and forced to pass through 0.45 μm filters (Millipore) to reduce clumps. Filtered cells were regrown in 40 ml SP-4 medium in a 150 cm2 tissue culture flask for 24 h. Electroporation buffer (8 mM HEPES pH7.2, 272 mM sucrose) was used to wash attached mycoplasmas three times, and then cells were scraped and resuspended in electroporation buffer. To perform electroporation, 5 μg plasmid DNA previously dissolved in 20 μl electroporation buffer were mixed with 90 μl mycoplasma cell suspension in a cuvette (Bio-Rad) and placed on ice for 15 min. Cells were electroporated (2.5kV, 129Ω) for 5 milliseconds. After cooling on ice for 15 min, 900 μl SP-4 medium were added to the cuvette, which was further incubated at 37°C for 4 h. Two hundred μl of cell suspension were plated on SP-4 agar with tetracycline (2 μg ml−1) and grown at 37°C in a humidified incubator for two weeks. Single colonies were picked and transferred into SP-4 broth for propagation.
A two-tailed-Student t test was used to analyze all qRT-PCR data in the current study.
This work was supported by the National Institutes of Health Grant No. AI045429 and the Frost Bank Trusts. The authors are grateful to Dr. Oxana Musatovova for guidance with qRT-PCR and primer extension and to Rose Garza for assistance with finalizing the manuscript.