|Home | About | Journals | Submit | Contact Us | Français|
The oral spirochete Treponema denticola is associated with human periodontal disease. T. denticola ATCC 35405 and ATCC 33520 are two routinely used laboratory strains. Compared to T. denticola ATCC 33520, ATCC 35405 is more virulent but less accessible to genetic manipulations. For instance, the shuttle vectors of ATCC 33520 cannot be transformed into strain ATCC 35405. The lack of a shuttle vector has been a barrier to study the biology and virulence of T. denticola ATCC 35405. In this report, we hypothesize that T. denticola ATCC 35405 may have a unique DNA restriction-modification (R-M) system that prevents it from accepting the shuttle vectors of ATCC 33520 (e.g., the shuttle plasmid pBFC). To test this hypothesis, DNA restriction digestion, PCR, and Southern blot analyses were conducted to identify the differences between the R-M systems of these two strains. DNA restriction digestion analysis of these strains showed that only the cell extract from ATCC 35405 was able to digest pBFC. Consistently, PCR and Southern blot analyses revealed that the genome of T. denticola ATCC 35405 encodes three type II endonucleases that are absent in ATCC 33520. Among these three endonucleases, TDE0911 was predicted to cleave unmethylated double-stranded DNA and to be most likely responsible for the cleavage of unmethylated pBFC. In agreement with this prediction, the mutant of TDE0911 failed to cleave unmethylated pBFC plasmid, and it could accept the unmethylated shuttle vector. The study described here provides us with a new tool and strategy to genetically manipulate T. denticola, in particular ATCC 35405, and other strains that may carry similar endonucleases.
More than 80% of the adult population at some time in their lives suffers from periodontal disease, which is primarily caused by polymicrobial infections (11, 37). More than 700 different microorganisms have been suggested to colonize the oral flora (13, 27, 36). A body of studies has shown that the presence and burden of oral spirochetes is associated with the severity of periodontal disease (16, 36, 42). In the oral flora, more than 60 different spirochete species have been identified, and they all belong to the genus Treponema (13, 36). Due to their fastidious growth requirements, very few oral treponemes can be reliably cultivated (6, 16). Treponema denticola, an oral spirochete that can be readily cultured, has been used as a model bacterium to study the biology and virulence of oral treponemes because its genome has been sequenced and it can be genetically manipulated (16, 23, 25, 28, 43, 50).
T. denticola ATCC 35405 and ATCC 33520 are two genetically related reference strains that are often used to study the genetics and virulence of spirochetes (17, 21, 25). T. denticola ATCC 33520 shares more than 76% DNA similarity with ATCC 35405 (7). However, these two strains possess many physiological and genetic differences, such as serotype (7, 8), oxygen tolerance (46), and biofilm formation capability (24, 48, 49). In addition, four plasmids have been isolated from several oral treponemes, including ATCC 33520, but none of these plasmids has been isolated from ATCC 35405 (4, 5). Moreover, three shuttle vectors (pKMR4PE, pKMCou, and pBFC) that were derived from the plasmid pTS1 have been successfully transferred into ATCC 33520 but not ATCC 35405 (5, 9, 10, 44). Thus far, there has been no shuttle vector available for the genetic complementation of mutants derived from T. denticola ATCC 35405. ATCC 35405 is more virulent than ATCC 33520, and its genome has been sequenced (3, 12, 14, 43). The lack of a shuttle vector has compromised our efforts to use ATCC 35405 and its genetic information to study the biology and virulence of T. denticola.
DNA restriction and modification (R-M) systems have been described as “immune systems” to defend against invading foreign DNA (33, 34). In many prokaryotes, R-M systems serve as genetic barriers for gene transformation, conjugation, and transfection (1, 15, 33, 47). The majority of R-M systems consist of a DNA methyltransferase (MTase) and a restriction endonuclease (REase). The MTase enables recognition of self DNA by methylation of specific nucleotides within particular DNA sequences, and the REase cleaves invading unmodified DNA. R-M systems can be divided into four types (I to IV), based on their enzyme compositions, cofactors, and active modes (40). Among these R-M systems, the type II R-M system often protects bacteria and archaea against invading DNA (38). R-M systems were recently identified in the Lyme disease spirochete Borrelia burgdorferi, and disruptions of these systems were able to increase the transformation efficiency of foreign DNA (22, 26, 39). The genome of T. denticola ATCC 35405 encodes three putative type II R-M systems: TDE0227 (MTase)/TDE0228 (REase), TDE0909 (MTase)/TDE0911 (REase), and TDE1268 (REase) (41). In this report, we hypothesize that the existence of these R-M systems may prevent ATCC 35405 from accepting foreign DNA, such as the shuttle vectors of ATCC 33520. To test this hypothesis, DNA restriction digestion, PCR, and Southern blot analyses were conducted to compare the differences between the R-M systems of ATCC 33520 and ATCC 35405. It was found that these R-M systems were absent in ATCC 33520 and that the inactivation of TDE0911, a gene encoding a type II restriction endonuclease, allowed the mutant to accept the unmethylated pBFC shuttle vector.
T. denticola ATCC 35405 and ATCC 33520 strains were grown in oral bacterial growth medium (OBGM) (35) with 10% heated-inactivated rabbit serum at 37°C in an AS-580 anaerobic chamber (Anaerobe Systems, Morgan Hill, CA) with an atmosphere of 80% nitrogen, 10% carbon dioxide, and 10% hydrogen, as previously described (50). The Escherichia coli TOP10 strain (dam+/dcm+; Life Technologies, Carlsbad, CA) was used for routine plasmid constructions and preparations. A E. coli dam/dcm-deficient strain (New England BioLabs, Ipswich, MA) was used to prepare unmethylated plasmids. The oligonucleotide primers used in this study are listed in Table 1, and all primers were synthesized by Integrated DNA Technologies, Inc. (Coralville, IA).
Total genomic DNAs from T. denticola wild-type strains and the isogenic mutant were prepared with the Illustra bacteria genomic prep kit (GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom). Southern blot analysis was carried out following a standard procedure. Briefly, the purified genomic DNAs were first digested with the restriction enzymes ClaI or HindIII, separated on 1.0% agarose gel, and blotted to a Hybond-N+ membrane (GE Healthcare). To prepare DNA probes for Southern blot assays, TDE0228 (925 bp), TDE0911 (766 bp), TDE1268 (659 bp), and ermB, an erythromycin resistance gene (696 bp) (19), were amplified by PCR. Then, the obtained amplicons were labeled with digoxigenin (DIG). To label pBFC, the plasmid was first linearized with NotI and then labeled with DIG. The DNA labeling, hybridization, and detection were carried out using DIG High Prime DNA labeling and detection starter kit I (Roche Diagnosis GmbH, Mannheim, Germany), according to the manufacturer's instructions.
A recently described nonpolar gene inactivation method was applied to inactivate the TDE0911 gene of T. denticola (32). The vector (TDE0911::ermB) for the targeted mutagenesis was constructed by multiple-step PCR as illustrated in Fig. 1. In the first step, the flanking regions of TDE0911 and the erythromycin resistance gene (ermB) were amplified by PCR with three pairs of primers: P1/P2 (flanking region 1), P3/P4 (ermB), and P5/P6 (flanking region 2). The P2, P3, P4, and P5 primers contain several engineered overlapping base pairs (underlined in Table 1). In the second step, flanking region 1 and ermB were fused by PCR using P1 and P4 primers. In the final step, the constructed region 1-ermB fragment and flanking region 2 were further merged by PCR using primers P1 and P6. The final PCR product (flanking region 1-ermB-flanking region 2) was cloned into the pGEM-T-Easy vector (Promega, Madison, WI), generating the construct TDE0911::ermB, in which the entire open reading frame of TDE0911 was deleted and replaced with the promoterless ermB gene. To inactivate TDE0911, 10 μg of TDE0911::ermB plasmid was linearized with NotI and then electroporated into 80 μl of T. denticola ATCC 35405 competent cells. The transformants were selected on OBGM semisolid plates containing erythromycin (60 μg ml−1), and the mutation was confirmed by PCR and Southern blot assays.
The pBFC plasmid is a shuttle vector between E. coli and T. denticola ATCC 33520 (44), and it was kindly provided by R. Limberger (Wadsworth Center). To prepare methylated pBFC, the plasmid was transformed into E. coli TOP10, which contains the dam gene encoding a DNA methyltransferase that methylates the N6 position of the adenine residues in the sequence GATC (18, 20). To prepare unmethylated pBFC, the plasmid was transformed into an E. coli dam/dcm mutant strain. The plasmids were purified using the PureYield plasmid midiprep system (Promega). The plasmid concentrations were measured with a NanoDrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE) and diluted to a final concentration of 1 μg μl−1.
The site-directed mutagenesis of aacC1, a gentamicin resistance gene (45), was carried out using the QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA) with primers P15 and P16, according to the manufacturer's instructions. The resultant mutation was confirmed by DNA sequence analysis. The wild-type and mutated aacC1 genes were amplified by PCR with primers P17 and P18, and the obtained DNA fragments (534 bp) were used for DNA digestion analysis as described below.
To prepare the cell lysates, 5 ml of stationary-phase T. denticola cultures (approximately 1 × 109 cells ml−1) was centrifuged, and the cell pellets were suspended in 1 ml of phosphate-buffered saline (PBS; pH 7.5) buffer. The collected cells were lysed by sonication with the Branson Sonifier 450 (Branson Ultrasonics, Danbury, CT) for 150 s on ice. After centrifugation at 6,000 × g for 5 min, the supernatants were collected and stored at 4°C as crude cell extracts. These extracts were used to detect restriction digestion activity of different T. denticola strains on the methylated and unmethylated plasmids and on the PCR products. For the DNA digestion analysis, the reaction mixtures (40 μl) contained 5 μg DNA, 20 μl crude cell extract, and 4 μl 10× NEB buffer 4 (New England BioLabs). The reaction mixtures were incubated at 37°C for 2 h. The reaction mixtures were analyzed by agarose gel electrophoresis or further detected by Southern blot analysis.
The transformation efficiency of pBFC in T. denticola was measured as previously described (39). Briefly, three T. denticola strains (ATCC 35405, ATCC 33520, and the TDE0911 mutant) were grown to early logarithmic phase (optical density at 600 nm of 0.3 to 0.4), and the cells were enumerated using Petroff-Hausser counting chambers. The cell numbers of these three strains were adjusted to similar amounts before the preparation of competent cells. The preparation of T. denticola competent cells and electrotransformation were conducted as described before (28, 30). To measure the transformation efficiency, 10 μg of methylated or unmethylated pBFC plasmid was electroporated into 80 μl of T. denticola competent cells. The transformants were selected on OBGM semisolid plates containing chloramphenicol (10 μg ml−1). The presence of pBFC in the obtained transformants was confirmed by extracting the plasmid followed by restriction digestion analysis with NotI. The data are expressed as the mean transformation efficiency (transformants/μg pBFC) from three independent experiments.
The shuttle vector pBFC can be transferred into ATCC 33520 but not the ATCC 35405 strain (44). We hypothesized that this plasmid may be disrupted in ATCC 35405 after the transformation. To test this hypothesis, the pBFC plasmid purified from E. coli TOP10 strain was coincubated with the crude cell extracts of ATCC 35405 and ATCC 33520. As hypothesized, the DNA restriction digestion and Southern blot analyses showed that the crude cell extract of ATCC 35405 cleaved the methylated pBFC into at least 10 visible fragments (Fig. 2, lane 2), whereas ATCC 33520 was unable to cleave the plasmid (Fig. 2, lane 4). These results suggest that ATCC 35405 may contain a unique R-M system(s) that is absent in the ATCC 33520 strain.
The main function of R-M systems is to defend their hosts against foreign DNA, which is often achieved by recognition of self DNA via methylation at defined sites and disrupting unmethylated invading DNA (33). To test the influence of methylation on the cleavage of pBFC, an experiment similar to the one described above was conducted by using the unmethylated plasmid. The unmethylated pBFC was cleaved by the crude cell extract of ATCC 33520 (Fig. 2, lane 5). Interestingly, the lack of methylation protected the plasmid from cleavage by ATCC 35405 to some extent, since fewer fragments were detected (Fig. 2, lane 3) than when ATCC 35405 extract was incubated with the methylated pBFC (Fig. 2, lane 2). Collectively, these results imply that both strains ATCC 33520 and ATCC 35405 have R-M systems that are able to disrupt the unmethylated DNA and that ATCC 35405 may encode a unique restriction endonuclease(s) that cleaves methylated pBFC at defined sites.
The genome of ATCC 35405 was sequenced, and it encodes three putative type II R-M systems: TDE0227 (MTase)/TDE0228 (REase), TDE0909 (MTase)/TDE0911 (REase), and TDE1268 (REase) (41, 43). Since the type II R-M systems typically protect bacteria against invading DNA (38), we reasoned that some of these systems may be absent in ATCC 33520. To test if ATCC 33520 contains the above systems, PCR and Southern blot analyses were conducted to detect the genes encoding three endonucleases (TDE0228, TDE0911, and TDE1268). Both PCR (data not shown) and Southern blot analyses revealed that these three genes were present in the genome of ATCC 35405 (Fig. 3, lanes 3 and 4) but were absent in the ATCC 33520 strain (Fig. 3, lanes 1 and 2). These results further demonstrate that ATCC 35405 and ATCC 33520 contain different type II R-M systems.
The functions of the above three type II endonucleases were recently described in REBASE (41). It was predicted that TDE1268 (TdeI) belongs to the methylation-dependent restriction enzymes and recognizes GmATC at the N6 position of methylated adenine. TDE0228 (TdeII) was predicted to belong to a family of nicking endonucleases that recognize the asymmetric sequence 5′-CTCTTC-3′ and cleave one strand of DNA. REBASE also predicted that TDE0911 (TdeIII) belongs to a family of REases that recognize unmethylated GGNCC. The predicted digestion profile of the pBFC plasmid sequence (44) by TDE0911 corresponds to the above DNA restriction digestion pattern (Fig 2, lane 3), suggesting that TDE0911 is responsible for the cleavage of unmethylated pBFC. If this is the case, we would expect the inactivation of TDE0911 may block the cleavage of unmethylated pBFC and the mutant may be able to accept the unmethylated plasmid.
To test if TDE0911 is involved in pBFC cleavage, the corresponding gene was inactivated by targeted mutagenesis with the vector of TDE0911::ermB. A total of 63 Ermr colonies appeared 10 days after plating, and PCR analysis with a pair of primers specific to ermB (P9 and P10) showed that only 8 colonies contained ermB. One clone (TdΔ911) was further confirmed by Southern blotting and PCR analysis. As expected, for the TDE0911-specific probe, two bands (3.3 kb and 5.8 kb) were detected in the chromosomal DNA of ATCC 35405 treated with ClaI (Fig. 4 a, lane 1) and one band (4.8 kb) when treated with HindIII (Fig. 4a, lane 2). In contrast, no positive band was detected in the TdΔ911 mutant (Fig. 4a, lanes 3 and 4), indicating that the TDE0911 gene was deleted in the mutant. For the ermB-specific probe, one positive band was detected in the chromosomal DNA of TdΔ911 treated with ClaI (9.1 kb) and one with HindIII (6.1 kb) (Fig. 4b, lanes 3 and 4), but none of them appeared in ATCC 35405 (Fig. 4b, lanes 1 and 2). The sizes of detected fragments corresponded to the one with which ermB is integrated to the locus of TDE0911. These results showed that the whole TDE0911 gene was deleted and replaced with ermB as expected. This conclusion was further confirmed by PCR analysis with a pair of ermB-specific primers (P3/P4) in combination with one primer (P19) flanking the mutant allele (Fig. 4c).
It was predicted that TDE0911 recognizes and cleaves targeted DNAs containing the sequence GGNCC (41). To confirm this prediction, aacC1, which contains one putative recognition site (G97GCCC) of TDE0911 (45), was treated with the crude cell extracts of ATCC 35405 and the TdΔ911 mutant. As shown in Fig. 5, the crude cell extract of ATCC 35405 cleaved aacC1 into two fragments (lane 2), as expected, but the extract of the TdΔ911 mutant did not (lane 4). In addition, a mutation in the recognition site (G97GCCC mutated to A97GCCC) blocked the cleavage (lanes 1 and 3). The cleaved fragments were further cloned and sequenced, and it was determined that the cleavage occurred at the site 5′- GGCCC3′/3′-C▲CGGG5′ (where the triangles represent the cleavage sites).
As mentioned above, TDE0911 recognizes and digests unmethylated DNA. If this is a sole endonuclease that cleaved the unmethylated pBFC, we would expect that the TdΔ911 mutant would fail to cleave the unmethylated DNA. To test this hypothesis, the DNA restriction digestion analysis described above was carried out. As expected, the crude cell extract of T. denticola ATCC 35405 cleaved the unmethylated pBFC into 4 fragments (Fig. 6, lane 3); however, there was no cleaved fragment observed in the sample treated with the crude cell extract of TdΔ911 (Fig. 6, lane 5). These results suggest that TDE0911 is a sole endonuclease that cleaves unmethylated pBFC.
The failure to cleave the unmethylated DNA suggests that the TdΔ911 mutant may be able to accept unmethylated pBFC. To test this speculation, the methylated and unmethylated pBFC plasmids were electroporated into the TdΔ911 mutant. In addition, the strains ATCC 33520 and ATCC 35405 were included as controls. Consistent with the previous reports (44), the methylated pBFC plasmid could be transferred into ATCC 33520 (Fig. 7 c) but not the ATCC 35405 strain (Fig. 7a). As speculated, it was found that the unmethylated plasmid was successfully transferred into the TdΔ911 mutant (Fig. 7f), but not the ATCC 33520 strain (Fig. 7d). The observed phenotype is consistent with the above in vitro DNA digestion analyses, in which the unmethylated pBFC was digested by ATCC 33520 but not the TdΔ911 mutant. The transformation efficiency of the unmethylated pBFC in the TdΔ911 mutant (6.0 ± 1.7 transformants/μg DNA [mean ± standard error of the mean]) was similar to that of the methylated plasmid in the ATCC 33520 strain (7.3 ± 2.1 transformants/μg DNA) (Table 2).
The lack of sufficient genetic tools is a bottleneck to study the biology and virulence of the oral spirochete T. denticola. Although targeted and transposon mutagenesis are currently available, their efficiencies are very low (2, 29, 31, 50). Thus far, only a few mutants have been constructed and genetically complemented. It has been speculated the R-M system, which has been referred to as a microbial “immune system,” may prevent T. denticola from accepting foreign DNA. Consistently, the genome of ATCC 35405 encodes several R-M systems (41). The functions of these systems remain elusive. The study reported here is the first step to investigate the roles of these R-M systems in T. denticola, and it provides us with a new tool and strategy to genetically manipulate T. denticola.
We thank R. Limberger for providing T. denticola strains and shuttle vectors.
This research was supported by Public Health Service grants DE018829 and DE019667 to C. Li.
Published ahead of print on 20 May 2011.