Bacterial strains and media

Bacterial strains utilised in this study are listed in Table 1. E. coli strains were grown in Luria-Bertani medium at 37°C. Clostridium spp were cultured under anaerobic condition in an anaerobic cabinet (MG1000 Anaerobic Work Station, Don Whitley Scientific Ltd) containing an atmosphere of 80% nitrogen, 10% hydrogen and 10% carbon dioxide. Antibiotics were used at the following concentrations: erythromycin (Em), 20 μg/ml, thiamphenicol (Tm), 15 μg/ml for Clostridium spp; Em, 500 μg/ml, chloramphenicol (Cm), 25 μg/ml, tetracycline (Tc), 10 μg/ml for E. coli.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Bacterial strains and plasmids used in this study.

https://doi.org/10.1371/journal.pone.0122411.t001

C. acetobutylicum ATCC 824 and its mutant derivatives were grown in Clostridial Growth Medium (CGM) [23] for routine manipulations, or Supplemented Clostridium Basal Medium (CBMS) for fermentation assays. CBMS was based on Clostridium Basal Medium (CBM) as previously described[24] with two modifications. Glucose was added at a final concentration of 50 g/l, and 5 g/l calcium carbonate was added as a buffering agent. P2 medium [25] with 20 g/l glucose was used to screen for auxotrophic mutants.

C. sporogenes NCIMB 10956 and its mutant derivatives were grown in TYG medium (2% trypticase, 0.5% peptone, 0.1% glucose, 0.5% yeast extract and 0.1% cysteine-HCl) [26] for routine manipulations, or P2Y medium for fermentation assays. P2Y medium is P2 medium with some modifications, in which glucose was added at a final concentration of 50 g/l, 5 g/l calcium carbonate was added as a buffering agent and 2 g/l yeast extract was added for extra amino acids. MACC medium [27] with 20 g/l glucose was used to screen for auxotrophic mutants.

Plasmids, primers, DNA techniques

Plasmids and primers used in this study are listed in Tables 1 and 2, respectively. Chromosomal DNA preparation, plasmid isolation and purification of DNA fragments from agarose gels were carried out using the DNeasy Tissue kit, the QIAprep Miniprep kit and the QIAquick Gel Extraction kit, respectively (Qiagen, UK). Restriction enzymes were supplied by New England Biolabs and were used according to the manufacturer’s instructions. E. coli strains were transformed by electroporation using a Gene-Pulser (Bio-Rad), as recommended by the manufacturer. PCR amplifications were carried out using the KOD Hot Start Master Mix (Merck). Oligonucleotides used in this study are detailed in Table 2, which were synthesised by Eurofins MWG Operon, Germany.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Oligonucleotide primers used in this study.

https://doi.org/10.1371/journal.pone.0122411.t002

To establish whether transposition had occurred, inverse PCR (INV) was performed according to the procedure of Cartman and Minton. [15]. All DNA sequencing was carried out by Source BioScience, United Kingdom.

To identify the genomic location of transposon insertions, sequence data were analyzed using DNASTAR (www.dnastar.com) and compared to the genome sequences using Artemis (www.sanger.ac.uk/resources/software/artemis/). The sequences of C. acetobutylicum ATCC 824 genome and megaplasmid pSOL1 are published (Refseq number NC_003030.1 and NC_001988.2; GenBank accession number AE001437 and AE001438) [28]. The GenBank accession number of C. sporogenes NCIMB 10956 is CP009225.

Plasmid transfer in Clostridium species

C. acetobutylicum and C. sporogenes were transformed as described previously [29]. For C. acetobutylicum, prior to transformation, plasmid DNA was purified from E. coli TOP10 cells containing plasmid pAN2. This plasmid contains the ϕ3TI methyltransferase gene of B. subtilis phage ϕ3tI, which protects DNA from C. acetobutylicum Cac824I DNA restriction activity [30].

Construction of plasmids

Plasmid pMTL5401Fcat is a previously constructed plasmid [30] which mediates the IPTG inducible expression of the catP gene. To simplify and construct an IPTG inducible promoter cassette, a 982 bp DNA intervening region encompassing oriT/traJ was deleted from plasmid pMTL5401Fcat using the QuikChange Site-Directed Mutagenesis Kit (Stratagene) with primers Del-traJ and Del-traJ-antisense. Having deleted the traJ region, the 2728 bp lac inducible promoter cassette containing lacI, P_ptb, P_fac and cat was then PCR amplified using primers YZ4 and YZ5 and cloned into pMTL83251 via NotI/HindIII, resulting in plasmid pMTL83251-YZ2 (Fig 1A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Schematic diagram of the lac-based, IPTG inducible expression cassette in pMTL-YZ2 (A), and the demonstration of its function in C.

acetobutylicum (B) and C. sporogenes (C). A: Schematic diagram of the lac-based, IPTG inducible expression cassette. Key: LacI is the LacI repressor protein gene. LacI binds to the indicated lacO region, blocking transcription from the P_fac promoter. The P_ptb promoter (derived from the C. beijerinckii gene encoding phosphotransbutyrylase) directs the transcription of the lacI gene. B and C: IPTG induction of CAT production in of C. acetobutylicum (B) and C. sporogenes (C) harbouring pMTL-YZ2. Circles equate to cells which received no IPTG, squares represents samples from cells that were induced with IPTG. The arrow indicates the time of adding IPTG. Activity is expressed as units of CAT activity per mg of soluble protein.

https://doi.org/10.1371/journal.pone.0122411.g001

Having established the functionality of the P_fac inducible promoter, the cat gene was deleted from plasmid pMTL83251-YZ2, in order to bring the pCB102 replicon [31] under the transcriptional control of P_fac. Two plasmids were created. In the one (pMTL83251-lacI-T see Fig 2A), pMTL83251-YZ2 was cleaved with BamHI and HindIII, the sticky-ends created blunt-ended by treatment with T4 polymerase, and the resultant linear fragment subjected to self-ligation. In a second plasmid (pMTL83251-lacI, see Fig 2B), pMTL83251-YZ2 was digested with BamHI and AscI, the sticky-ends created blunt-ended by treatment with T4 polymerase, and the resultant linear fragment subjected to self-ligation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Plasmid maps and the conditionality analysis of the conditional replicon.

Plasmid maps of the non-conditional control plasmid pMTL83251-lacI-T (A) and conditional plasmid pMTL83251-lacI (B). Key: CD0164 terminator, a transcriptional terminator isolated from downstream of the Clostridium difficile strain 630 CD0164 gene; lacI, the E. coli gene encoding LacI repressor; P_ptb, the promoter of the C. beijerinckii gene encoding phosphotransbutyrylase; P_fac, the promoter of the C. pasteurianum ferredoxin gene derivatised to include an E. coli lac operator; repH, replication region of the Clostridium butyricum plasmid pCB102; ermB, the macrolide-lincosamide-streptogramin B antibiotic resistance gene of plasmid pAMß1; ColE1, the replication origin of plasmid ColE1, and; traJ, transfer function of the RP4 oriT region. Cpa fdx terminator, transcriptional terminator of the ferredoxin gene of C. pasteurianum (this feature is underlined due to its existence only in A: pMTL83251-lacI-T, but not in B: pMTL83251-lacI). C and D: the ability to replicate of pMTL83251-lacI-T and pMTL83251-lacI in of C. acetobutylicum (C) and C. sporogenes (D) in the absence (CFU account in black) and presence (CFU account in grey) of IPTG.

https://doi.org/10.1371/journal.pone.0122411.g002

To confine the conditional replicon to a cassette compatible with the pMTL80000 modular vectors series [32], the DNA fragment encompassing the P_fac/P_ptb::lacI cassette, plus the pCB102 replicon, was localised to a portable AscI-FseI fragment. This was accomplished by the following steps. In the first instance, the NotI site preceding the P_fac/P_ptb::lacI cassette of pMTL83251-lacI was changed to an AscI site using a QuikChange Site-Directed Mutagenesis Kit (Stratagene) and primers NotI/AscI and NotI/AscI-antisense, to yield the plasmid pMTL87250. This now allowed the excision of the P_fac/P_ptb::lacI cassette, together with the pCB102 replicon, as a 3537 bp AscI and FseI fragment. This was substituted in place of the equivalent restriction fragment encoding the pBP1 replicon in plasmid pMTL-SC1. The plasmid obtained was designated pMTL-YZ14. In parallel, pMTL-YZ13 was made replacing the pBP1 replicon in pMTL-SC1 with replicon pCB102.

Plasmid pMTL82254-P_tcdB as shown in Fig 3A, was constructed by excising a ca. 334 bp NotI/NdeI fragment from the plasmid pMTL-SC1 [15] and inserting it between the NotI and NdeI sites of plasmid pMTL82254 [32]. For comparative purposes, a second plasmid was constructed identical to pMTL82254-P_tcdB, but in which the fragment encompassing the P_tcdB promoter was replaced with an equivalent NotI/NdeI fragment encompassing the P_fdx promoter. This plasmid was designated pMTL82254-P_fdx as illustrated in Fig 3B.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Analysis of the functionality of the exogenous TcdR in two clostridial strains (CRG3011 and CRG3817).

Plasmid maps of the pMTL82254-P_tcdB (A) and pMTL82254-P_fdx (B). Key: CD0164 terminator, a transcriptional terminator isolated from downstream of the C. difficile strain 630 CD0164 gene; catP, a C. perfringens-derived gene encoding chloramphenicol acetyltransferase; Cpa fdx terminator, transcriptional terminator of the ferredoxin gene of C. pasteurianum; repA and orf2, replication region of the C. botulinum plasmid pBP1; ermB, the macrolide-lincosamide-streptogramin B antibiotic resistance gene of plasmid pAMß1; ColE1, the replication origin of plasmid ColE1, and; traJ, transfer function of the RP4 oriT region. CD tcdB promoter, the promoter region of the C. difficile tcdB gene; Csp fdx promoter: the promoter region of the C. sporogenes fdx gene. (C): CAT activity of either C. acetobutylicum ATCC 824 wild type or CRG3011 (tcdR containing C. acetobutylicum ATCC 824) carrying plasmids pMTL82254-P_tcdB and pMTL82254-P_fdx. (D): CAT activity of either C. sporogenes NCIMB 10969 wild type or CRG3817 (tcdR containing C. sporogenes NCIMB 10969) carrying plasmids pMTL82254-P_tcdB and pMTL82254-P_fdx. Black circles ●, wild type with pMTL82254-P_tcdB; black squares ■, wild type with pMTL82254-P_fdx; black triangles ▲, CRG3011/CRG3817 with pMTL82254-P_tcdB; black triangles ▼, CRG3011/CRG3817 with pMTL82254-P_fdx.

https://doi.org/10.1371/journal.pone.0122411.g003

Integration of the tcdR gene into the clostridial host genome using ACE

A DNA fragment encompassing the tcdR structural gene encoding TcdR and its ribosome binding (RBS) site was PCR-amplified from the chromosome of the C. difficile strain 630 with primers tcdR-F and tcdR-R, and cloned into the vector pGEM-T. This localised the tcdR gene and RBS to a NotI-BamHI fragment which was excised and inserted between the equivalent sites of the plasmid pMTL-ME6C to yield the plasmid pMTL-ME6C-tcdR. Plasmid pMTL-ME6C is essentially plasmid pMTL-JH14, but which lacks the lacZ’- encoding region residing between the shorter Left Homology Arm (LHA, encoding a pyrE allele) and the longer Right Homology Arm (RHA, encoding CAC0028) and in which a region of DNA encompassing the transcriptional terminator of the Clostridium pasteurianum ferredoxin gene has been inserted 3’ to the pyrE allele, and preceding CAC0028.

To introduce the tcdR gene into the chromosome of C. acetobutylicum, Allele-Coupled Exchange (ACE) was employed as described in Heap et al. [10]. In vivo methylated pMTL-ME6C-tcdR plasmid DNA was transformed into the ACE-generated C. acetobutylicum pyrE mutant essentially as described in Heap JT et al., 2012. Transformed cells were plated onto CGM agar supplemented with 15 μg/ml Tm and 20 μg/ml uracil. After 24 h, fast growing single colonies were picked and re-streaked twice onto CGM agar containing 15 μg/ml Tm and 20 μg/ml uracil. These cells represented those in which the plasmid had integrated into the genome by single cross-over recombination between the RHA. Thereafter, cells were streaked onto CBM agar to select for cells able to grow in the absence of exogenous uracil as a consequence of plasmid excision (through recombination between the duplicated LHA) and restoration of a functional pyrE allele. The final construct has the tcdR gene, together with its RBS, inserted immediately downstream of the pyrE gene, residing upstream of the Clostridium pasteurianum transcriptional terminator. The tcdR gene is transcribed from the promoter upstream of CAC0025 (as demonstrated in Fig 4A). The C. acetobutylicum strain generated was designated CRG3011.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Schematic representations of the genome of tcdR strain of C. acetobutylicum (A) and C.

sporogenes (B). A promoter-less copy of the tcdR gene (including its ribosome binding site, RBS) of C. difficile strain 630 has been inserted into the genome using ACE technology immediately downstream of the pyrE gene (A: CAC0027, B: CS3413). Illustrated are the surrounding genes, and the position of the promoter responsible for expression of tcdR, and the position of the tcdR RBS. The illustrated terminator is the T1 transcriptional terminator of the C. pasteurianum ferredoxin gene.

https://doi.org/10.1371/journal.pone.0122411.g004

To generate a strain producing functional TcdR protein in C. sporogenes NCIMB 10696, the following steps were undertaken. A DNA fragment encoding TcdR and its ribosome binding (RBS) site was excised as a NotI-BamHI fragment from plasmid pMTL-ME6C-tcdR, and inserted between the equivalent sites of the plasmid pMTL-YZ29 to yield the plasmid pMTL-YZ29-tcdR. ACE vector pMTL-YZ29 is as same as pMTL-JH29 with the Gram-positive replicon being changed to pCD6 instead of pIM13. Plasmid pMTL-YZ29-tcdR plasmid DNA was transformed into the ACE-generated C. sporogenes pyrE mutant described in Heap et al. [10]. Transformed cells were plated onto TYG agar supplemented with 15 μg/ml Tm. After 24 h, fast growing single colonies were picked and re-streaked twice onto TYG agar containing 15 μg/ml Tm. These cells represented those in which the plasmid had integrated into the genome by single cross-over recombination between the RHA. Thereafter, cells were streaked onto MACC minimal medium agar to select for cells able to grow in the absence of exogenous uracil as a consequence of plasmid excision (through recombination between the duplicated LHA) and restoration of a functional pyrE allele. The final construct has the tcdR gene, together with its RBS, inserted immediately downstream of the pyrE gene from where it is transcribed from the promoter upstream of CS3415 (Fig 4B). The strain generated was designated CRG3817.

Chloramphenicol acetyltransferase (CAT) assay

CAT activity was determined according to the method of Shaw. [33]. A quartz cuvette was prepared containing 540 μl of 100 mM Tris buffer (pH 7.8), 200 μl of 2.5 mM DTNB (5,5′-dithiobis-2-nitrobenzoic acid) solution in 100 mM Tris buffer (pH 7.8), 200 μl of 5.0 mM freshly prepared acetyl Coenzyme A solution in deionized water and 10 μl of cell lysate. The cuvette was pre-warmed to 25°C, and the reaction initiated by adding 10 μl of 0.3% w/v Cm solution in deionized water. The initial rate of increase of absorption at 412 nm was measured using an Analytik Jena SPECORD 250 PLUS spectrophotometer.

Fermentation

For C. acetobutylicum, batch fermentation experiments were carried out in 100 ml volumes of CBMS broth in 250 ml conical flasks. For C. sporogenes, batch fermentation experiments were carried out in 100 ml volumes of P2Y broth in 250 ml conical flasks. Cultures were inoculated to a starting OD₆₀₀ of approximately 0.05 with an overnight starter culture, and incubated in an anaerobic cabinet at 37°C for 96 h. At each sample point, 1 ml samples were removed, placed on ice and centrifuged at 16,000 x g for 1.5 minutes. Supernatants were stored at -80°C before analysis by high performance liquid chromatography. OD₆₀₀ was also measured at each time point. Samples subjected to heat shock (80°C, 10 min) were taken at time points 11, 24, 48, 72 and 96 h for sporulation assay.

Analytical techniques

The supernatant samples described above were subjected to HPLC analysis using a Varian Prostar HPLC (model 240) equipped with an Aminex HPX-87H Ion Exclusion Column Column (300 x 7.8 mm Biorad), a refractive index detector (Varian 356LC) and a photo diode array (UV) detector (Varian). The mobile phase used was 0.005 M H₂SO₄ and the column was operated at 30°C with a flow rate of 0.6 ml/min. The injection volume was 20 μl and elution was isocratic. Supernatants were diluted appropriately in Milli-Q-water and filtered through a 0.2 μm pore size membrane filter prior to analysis.

Isolation of transposon mutants

For C. acetobutylicum, plasmid pMTL-YZ14 was in vivo methylated and transformed into C. acetobutylicum CRG3011 as described previously [30]. The transformed cells were selected on CGM agar containing Em and were incubated for 48 to 72 h in order to expand the population of cells containing the transposon plasmid. All growth was harvested by flushing the whole plate with CGM broth with 10% Glycerol, and serial dilutions were made and plated onto CGM agars in duplicate: CGM agars without any antibiotics to calculate transposition frequency, CGM agars containing Tm (15 μg per ml) and IPTG (1 mM) to select for the transposon-based catP antibiotic marker and induce plasmid loss at the same time. Cell suspensions were stored at -80°C for future use. Individual colonies, visible after 12 to 16 h, were picked and patch plated onto CGM agars containing Tm only and CGM agars containing Em, respectively, to check for plasmid loss and for further analysis and/or phenotypic screening.

For C. sporogenes, plasmid pMTL-YZ14 was transformed into C. sporogenes CRG3817 strain as described previously without methylation. The transformed cells were selected on TYG agar containing Em and were incubated for 24 to 48 h. The plating and selections were performed essentially as described for C. acetobutylicum except for the growth medium employed.

Transcription can affect maintenance of pCB102 replicon-based plasmids

During the construction of the ClosTron plasmid pMTL007 [30], it was noted that one of the derivative plasmids, pMTL540F, was unable to transform C. acetobutylicum. In contrast, a previously constructed plasmid pMTL540FT, transformed the same clostridial host at normal frequencies [34]. The only difference between the two plasmids is the presence of a transcriptional terminator, positioned between the strong P_fac promoter and the pCB102 plasmid replicon. We hypothesised that transcription into the replication region was interfering with the ability of the plasmid to stably establish itself in a transformed cell. It follows, that were the transcription from P_fac to be placed under regulatory control, and repressed, then the plasmid should be successfully installed in the transformed cell. Moreover, by adding the requisite inducer to a transformed cell, then the activation of transcription into the replicon region should subsequently lead to plasmid loss. To test this system, an inducible promoter system was required.

Modularisation of an IPTG inducible promoter cassette

We have previously demonstrated that P_fac controlled expression of a catP gene in C. acetobutylicum and C. sporogenes could be repressed by the incorporation of an E.coli lacI gene onto a plasmid backbone in which expression of the repressor gene was placed under the transcriptional control of the P_ptb promoter of the C. beijerinckii ptb (phosphotransbutyrylase) gene. Moreover, repression could be relieved by the addition of exogenous IPTG, resulting in production of cat-encoded chloramphenicol acetyltransferase (CAT). For convenience, we elected to localise the P_ptb::lacI element and P_fac promoter to a single portable restriction fragment. As the region encompassing these elements on the previously made plasmid, pMTL5401Fcat [30], were separated by a region of DNA encompassing oriT, we elected to delete this element during this process—oriT is superfluous as it is not needed for clostridia that can be transformed, and is in any case already part of the modular vector series (31) that we later planned to use. Accordingly, we used QuikChange to delete oriT, converted the P_ptb::lacI + P_fac + cat region to a NotI/HindIII restriction fragment (Materials and Methods), and inserted it between the equivalent sites of plasmid pMTL83251 to yield plasmid pMTL-YZ2 (Fig 1A).

To confirm that expression of the cat gene remained under IPTG-inducible control, we transformed pMTL-YZ2 into C. acetobutylicum or C. sporogenes, and then grew the transformants in the presence or absence of added inducer. IPTG-mediated induction of CAT activity was observed in both organisms, but a significant basal level of expression was observed in C. acetobutylicum (Fig 1B) in the absence of IPTG, indicating the repression by LacI is less stringent than in C. sporogenes, where tight repression was evident (Fig 1C). This is consistent with a decrease in expression of lacI due to a reduced activity of the P_ptb promoter as cells entered the solventogenesis phase in C. acetobutylicum [35].

Development of a conditional replicon

In order to test our hypothesis, that the establishment of a pCB102 replicon-based plasmid could be controlled by the IPTG inducible promoters system, we constructed two new plasmids, pMTL83251-lacI-T and pMTL83251-lacI (Fig 2A and 2B). The essential difference between the two plasmids is that pMTL83251-lacI-T carries a transcriptional terminator (that of the ferredoxin gene of Clostridium pasteurianum) between the P_fac promoter and the repH replication gene of the pCB102 replicon. Plasmid pMTL83251-lacI is broadly equivalent to pMTL-YZ2, except it lacks the cat gene.

Transformed cells of C. acetobutylicum ATCC 824 carrying either plasmid were cultivated and the effect of adding IPTG on plasmid retention estimated. Specifically, C. acetobutylicum transformants were cultured for 12 h in CGM broth supplemented with Em to select for the plasmid. The culture was then washed twice in PBS to remove the antibiotic and then used to inoculate fresh, CGM broth with and without IPTG (1mM) at 1% (vol/vol). After 12 h of growth, the culture was plated to enumerate Em^R Colony Forming Units (CFU). As predicted, plasmid pMTL83251-lacI was found to only be stably maintained in the absence of IPTG (Fig 2C). In the presence of IPTG, the plasmid was rapidly lost from the cell population (Fig 2C), as evidenced by an almost complete loss of CFU on plates supplemented with Em. A similar loss was not evident in cells harbouring pMTL83251-lacI-T (Fig 2C). It was concluded that transcriptional read-through into the pCB102 replication region was interfering with plasmid replication/ maintenance. One explanation for this phenomenon might have been that the RepH protein was being overexpressed, and that over production of RepH was detrimental to the cell. However, no difference in growth rate was observed between cells carrying pMTL83251-lacI cultivated in medium in the presence or absence of IPTG (data not shown). To test whether a similar phenomenon was observed in a different clostridial species, the equivalent experiment was replicated in C. sporogenes NCIMB 10696, in this case using TYG broth. A similar result to that seen in C. acetobutylicum was obtained. Thus, in the presence of IPTG, plasmid pMTL83251-lacI was rapidly lost from the population (Fig 2D), whereas retention of plasmid pMTL83251-lacI-T was unaffected by addition of the inducer (Fig 2D). These data indicate that the system has broader application than just a single Clostridium species.

Endowment of the genome with tcdR

Having derived a prototype conditional delivery vehicle for use in C. acetobutylicum and C. sporogenes we sought to implement our mariner transposon system in these clostridia by endowing them with a genomic copy of the C. difficile tcdR gene. The presence of TcdR in either clostridia would direct the expression of the transposase gene under the control of the P_tcdB promoter when introduced on a plasmid. Accordingly, a promoter-less copy of the tcdR gene was cloned into the requisite ACE vectors needed to both convert the pyrE allele of the clostridial pyrE mutant hosts used back to wild type using the ACE protocol (Materials and Methods), while simultaneously inserting the tcdR gene into the genome immediately downstream of the restored pyrE gene (Fig 4A). This placed the tcdR gene under the control of the promoter responsible for pyrE transcription. The strain of C. acetobutylicum carrying tcdR in its genome was designed CRG3011, whereas the equivalent strain of C. sporogenes was designated CRG3817 (Fig 4B).

To test the functionality of TcdR in the two clostridial strains (CRG3011 and CRG3817), the P_tcdB or the P_fdx promoter were inserted upstream of the promoter-less catP gene of the reporter plasmid pMTL82254, to yield the respective plasmids pMTL82254-P_tcdB and pMTL82254-P_fdx (Fig 3A and 3B). Each plasmid was introduced into both the wild type strains of C. acetobutylicum and C. sporogenes, and their respective derivatives (CRG3011 and CRG3817) carrying tcdR. Subsequently, the four cell lines were cultivated in CGM or TYG broth, cells harvested at designated time points and lysates prepared for assays of CAT activity (Fig 3C and 3D). In both organisms it was evident from the low CAT activity observed in the wild type cell lines that the P_tcdB promoter was essentially inactive in the absence of TcdR. In contrast the level of CAT expression seen in CRG3011 and CRG3817 cells carrying pMTL82254-P_tcdB was broadly equivalent to that seen in cells carrying plasmid pMTL82254-P_fdx, where expression was due to the strong P_fdx promoter.

It was of interest to determine whether the presence of tcdR in the genome of either organism had any discernible effect. It was evident that in either case, there were no detectable alterations to either clostridial species in terms of appearance under the microscope, colony morphology, sensitivity to heat shock, growth rate in liquid media or the solvent/acid supernatant profiles observed after four days growth of the cells (S1, S2 and S3 Figs).

Use of the conditional vector for transposon delivery

By endowing the two clostridial hosts with tcdR, effective transposition of the mini-transposon from plasmid pMTL-SC1 should now occur. Before testing this assumption, the plasmid was first modified to incorporate the pCB102 conditional replicon in place of the pBP1 replication region. This was accomplished by substituting the modular pBP1 replicon, excised as an AscI and FseI restriction fragment, with an equivalent fragment encompassing the modular, conditional pCB102 replicon of pMTL87250. The plasmid generated was designated pMTL-YZ14 (Fig 5). To test its effectiveness as a transposon delivery system, the plasmid was transformed into the tcdR-containing strains of C. acetobutylicum and C. sporogenes, CRG3011 and CRG3817. Transfer of the plasmid was initially selected on solidified media supplemented with Em and left on plates for 24 (CRG3817) to 72 (CRG3011) hours, before being plated on media containing Tm and IPTG to select for transposon events and promote plasmid loss, respectively. After 12 to 16 h of incubation, Tm ^R colonies were visible at a frequency of 2.6 (±0.6) ×10^–4 and 3.2 (±0.5) ×10^–4 (calculated as the ratio of Tm ^R CFU to total CFU) in C. acetobutylicum and C. sporogenes, respectively. A total of 100 Tm ^R colonies of each Clostridium species were picked and patch plated onto appropriate solidified rich media containing either Tm or Em to estimate the percentage of cells that had lost the plasmid. In total, 80% of C. acetobutylicum and 100% of C. sporogenes colonies were Tm resistant (^R) and Em sensitive (^S), indicative of successful insertion of the catP mini-transposon into chromosome and substantive plasmid loss.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Vector map of plasmid pMTL-YZ14.

Expression of the hyperactive mariner transposase gene Himar1 C9 was driven by the C. difficile toxin B promoter, P_tcdB. The plasmid backbone consisted of the conditional replicon between restriction sites AscI and FseI, the macrolide-lincosamide-streptogramin B antibiotic resistance gene ermB, and the Gram-negative replicon, ColE1. The whole mariner element (i.e., transposase gene and catP mini-transposon) can be excised as a SbfI fragment. The control plasmid pMTL-YZ13 was identical, except that the Gram-positive replicon is the pCB102 replicon from C. butyricum. This plasmid conforms to the pMTL80000 modular system for Clostridium shuttle plasmids [32].

https://doi.org/10.1371/journal.pone.0122411.g005

To establish whether transposition had indeed occurred, a more thorough molecular characterisation was undertaken. Genomic DNA of 60 randomly selected Tm ^R and Em ^S clones was prepared, digested with HindIII and subjected to Inverse PCR after self-ligation. The gel-purified inverse PCR products were subjected to nucleotide sequence analysis using the primer catP-INV-R2 to determine the sites of insertion. The data demonstrated that each transposon insertion had taken place at a different position around the genome as illustrated in Fig 6.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Genetic map of mariner transposon insertions in C.

acetobutylicum (A) and in C. sporogenes (B). Sixty independent transposon mutants were identified and the insertions were sequenced. Insertions in the plus orientation are marked on the circle exterior. Insertions in the minus orientation are marked on the circle interior. Numbers indicate the precise point of insertion according to genome sequence data for (A) C. acetobutylicum ATCC 824 genome and megaplasmid pSOL1 (Refseq number NC_003030.1 and NC_001988.2; GenBank accession number AE001437 and AE001438) [28] and (B) C. sporogenes NCIMB 10956 (GenBank accession number CP009225).

https://doi.org/10.1371/journal.pone.0122411.g006

For C. acetobutylicum (Fig 6A), all of the Tm ^R clones sequenced had a single transposon insertion, with the exception of one which had a double insertion revealed by Inverse PCR (S4 Fig). In one clone the mini-transposon had an insertion into the megaplasmid pSOL1. The remainder had all inserted at random around the chromosome. In total, there were 34 insertions in the plus strand and 27 in the minus strand, with 51 of the insertions located within open reading frames. In the case of C. sporogenes (Fig 6B), two out of the sixty clones contained double insertions of the mini-transposon. The remaining 58 clones all contained single insertions scattered at random around the genome. In total, there were 30 insertions in the plus strand and 32 in the minus strand. The number of insertions within open reading frames was 49, representing a slightly higher ratio of insertions into intergenic regions compared to C. acetobutylicum.

As a control, both clostridial strains were also transformed with pMTL-YZ13, which is identical to pMTL-YZ14 except that the pCB102 replicon is not preceded by the IPTG inducible cassette (P_ptb::lacI::P_fac). In this case, following growth on plates supplemented with Em, cells were plated on CGM agar containing Tm, but no IPTG. A total of 24 Tm ^R colonies of each strain were picked and re-streaked 3 times, one more passage than described in Cartman and Minton. [15] Colonies were then patch plated onto solidified media with and without Em to estimate plasmid loss. Only 4 of clones had become Em ^S in the case of C. acetobutylicum and just 3 out of 24 in the case of C. sporogenes. These data underlined the advantages of using the conditional pCB102 replicon.

Isolation of phenotypic mutants affected in sporulation/ spore germination

To demonstrate the utility of the developed system for forward genetic studies, we screened representative transposon libraries of the two clostridial species for mutants with defects in the ability to form colony forming units (CFUs) after heat shock. These were presumed to be defective in either the process of sporulation or spore germination, ie., spo/ger¯ mutants. The number of Tm ^R clones screened for C. acetobutylicum and C. sporogenes were approximately 3000 and 6000, respectively. Clones were inoculated into fresh CBM or YTG liquid broth (C. acetobutylicum and C. sporogenes, respectively) in replica 96-well microtitre plates and anaerobically incubated for one week to allow sporulation to take place. The microtitre plates were subjected to heat treatment (80°C for 10 min), while the other was held at room temperature. The contents of each microtitre well were individually plated onto either CGM (C. acetobutylicum) or YTG (C. sporogenes) agar media supplemented with Tm. Those cells unable to form colonies after heating 80°C for 10 min, but still able to produce colonies when not subjected to heat shock, were deemed possible sporulation or germination mutants.

In C. acetobutylicum a single transposon mutant was isolated with the desired spo/ger¯ phenotype, whereas in C. sporogenes, two such mutants were obtained. Inverse PCR was carried out on the isolated genomic DNA of the three mutants to identify the genes which had been inactivated (Table 3). In the case of C. acetobutylicum, the mini-transposon was found to have inserted in a gene (CAC2631) encoding a hypothetical protein that exhibited homology (70% similarity in DNA sequence) to the Bacillus subtilis spore protein YkuD reported to be involved in assembly of proteins on the forespore [36]. The two genes affecting the ability of C. sporogenes to form CFUs after heat shock were CLSPO_c23320, a flavodoxin oxidoreductase, and CLSPO_c31790, annotated as encoding the stage V sporulation protein SpoVAD. The spoVA operon in B. subtilis encodes six proteins, each of which has several putative membrane-spanning domains, and together they have been suggested to be involved in dipicolinic acid DPA transport into the developing forespore and the release of Ca²⁺-DPA and other small molecules during spore germination [37,38]. The defect in this C. sporogenes gene is therefore consistent with the observed phenotype.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Transposon insertion sites.

https://doi.org/10.1371/journal.pone.0122411.t003

The phenotype associated with the mutation in CLSPO_c23320 is less clear. Oxidoreductase is commonly related to aerotolerance in anaerobic organisms, providing protection against oxidative stress [39]. Since spores are important in the dissemination of the clostridial species it is likely that proteins with ability to scavenge the oxidative radicals are present in the outermost layers of the spores, and play a dual role—one in structural built-up of spores coat layers by mediating dityrosine cross-linking among proteins and/or second in resistance against oxidative stress [40,41] or they may have some other hitherto unknown function.

Isolation of auxotrophic phenotypic mutants

To screen for auxotrophic mutants, the two libraries of Tm ^R colonies were simply replica plated onto either solidified rich media (CGM for C. acetobutylicum and YTG for C. sporogenes) or an appropriate minimal medium agar medium (P2 for C. acetobutylicum and MACC for C. sporogenes), in all cases supplemented with Tm.

A single auxotrophic mutant of C. acetobutylicum was isolated that failed to grow on P2 agar medium after 7 days of incubation at 37°C. Nucleotide sequencing of isolated genomic DNA demonstrated that the min-transposon insertion had inserted into CAC0971 (CitB), which is annotated as encoding aconitase (aconitate hydratase). In aerobic bacteria this enzyme is responsible for the conversion of citrate to iso-citrate as part of the tricarboxylic acid (TCA) cycle. Iso-citrate is further converted into the TCA cycle intermediates oxalosuccinate and α-ketoglutarate, and the later into the non-TCA cycle metabolite glutamate, which can subsequently be converted into arginine and glutamine. Whilst strict anaerobes are not generally regarded as being in the possession of a TCA cycle, two independent research groups have recently suggested that C. acetobutylicum possesses a complete, albeit bifurcated TCA cycle in which oxaloacetate flows to succinate both through citrate/α-ketoglutarate and via malate/ fumarate [42,43].

The gene coding for citrate synthase has been tentatively identified as CAC0970 and is the first gene of a putative tricistronic operon, comprising citrate synthase (CAC0970), aconitase (CAC0971) and isocitrate dehydrogenase (CAC0972), which is expressed at higher levels at stationary phase [43]. Accordingly, the mutant, together with C. acetobutylicum CRG3011, was plated onto P2 minimal agar medium, individually supplemented (1 mM) with the tricarboxylic and amino acids, citrate, iso-citrate, α-ketoglutarate, glutamate, glutamine, proline and arginine (Table 4). Qualitative growth of the mutant was only fully restored to wild type by exogenous glutamate, whereas growth was partially restored by glutamine. Intriguingly, the mutant remained unable to grow on the minimal media when supplemented with α-ketoglutarate, which is suggested to be the product of the conversion of iso-citrate by isocitrate dehydrogenase (CAC0971) and the substrate for the synthesis of glutamate [42]. Our results clearly implicate CAC0971 in glutamate biosynthesis, but may suggest that glutamate is derived from a source other than α-ketoglutarate, assuming that this ketone can be taken up by C. acetobutylicum. Whilst glutamine is most likely derived from glutamate, the fact that it could not entirely restore the growth of the citB mutant to wild type levels suggest that the reversal of the glutamate to glutamine reaction is not thermodynamically favourable under normal laboratory conditions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Characterization of the found auxotroph mutant with different additives in P2 minimal medium.

https://doi.org/10.1371/journal.pone.0122411.t004

As a proteolytic clostridia, C. sporogenes is not able to undertake the biosynthesis of as many amino acids as the saccharolytic C. acetobutylicum. It was for this reason that a larger library was screened. In total, three mutants were isolate which were compromised in their ability to grow on minimal media compared to the wild type control. The inactivated genes were CLSPO_c16750, encoding a collagenase (annotated as ColA), CLSPO_c21170, coding for a hypothetical protein and CLSPO_c29860, annotated as an ABC-2 family transporter protein. A BLAST search of the protein sequence of CLSPO_c21170 revealed no putative conserved domains. The role of this protein in supporting growth on the minimal media employed, therefore, remains unknown. Similarly, the function of the protein encoded by CLSPO_c29860 is unclear, although the annotated name of the ABC-2 family of transporter proteins indicates it is likely involved in the translocation of an as yet unknown substrate(s)/ nutrient(s) into the cell. In the case of ColA, whilst there is no direct evidence to suggest bacterial colA genes are essential for survival and growth in defined medium, it is interesting to note that some of the organisms shown to elaborate collagenolytic enzymes are asaccharolytic [44]. Since their metabolism is dependent on the uptake of small peptides and amino acids, the production of collagenolytic enzymes by these organisms may be essential for survival and growth. Closely related with C. botulinum, C. sporogenes is proteolytic and asaccharolytic, and extra yeast extract is known to be essential for growth in fermentation compared to other saccharolytic clostridia. Hence, ColA could play an important role for the uptake of small peptides and amino acids in C. sporogenes.