To examine the role of the Lactobacillus reuteri 100-23C frc gene product in oxalate metabolism, host colonization and the acid stress response.
To examine the role of the Lactobacillus reuteri 100-23C frc gene product in oxalate metabolism, host colonization and the acid stress response.
Genes encoding putative formyl-CoA transferase (frc) and oxalyl-CoA decarboxylase (oxc) enzymes are present in the genome sequences of Lact. reuteri strains. Two strains isolated from humans harboured an IS200 insertion sequence in the frc ORF and a group 2 intron-associated transposase downstream of the frc gene, both of which were lacking in two strains of animal origin, which contained intact frc and oxc genes. An frc− insertional mutant of Lact. reuteri 100-23C was compared with the parent strain with respect to oxalate degradation, colonization of an RLF-mouse host model and growth in the presence of acids. Neither parent nor mutant degraded oxalate in vitro or in vivo. However, the parent outcompeted the frc− mutant in the mouse intestine during co-colonization and the frc− mutant showed a reduced growth rate in the presence of hydrochloric acid.
Intact oxc and frc genes do not ensure oxalate degradation under the conditions tested. The frc gene product is important during host colonization and survival of acid stress by Lact. reuteri 100-23C.
Oxalate metabolism by oxalate-degrading intestinal bacterial strains may be important in preventing urolithiasis and might lead to the derivation of probiotic products. To produce safe and efficacious probiotics, however, an understanding of the genetic characteristics of potential oxalate degraders must be obtained, together with knowledge of their functional ramifications.
Kidney stone disease or urolithiasis refers to the formation of insoluble stone-like structures in the urinary tract of humans. While the prevalence of this disease is related to a diverse range of risk factors, there is evidence to suggest that it is rising worldwide, leading to an increased burden on the health system (Romero et al. 2010). A large percentage of clinically significant kidney stones are composed of calcium oxalate, suggesting a central role for urinary oxalate in their formation (Liebman and Costa 2000). As a significant proportion of oxalate excreted in the urine is of dietary origin (Holmes et al. 2001), the presence of oxalate-degrading bacteria in the human gut may be an important factor in lowering the amount of oxalate available for stone formation.
Two genes appear to play a central role in the degradation of oxalate by bacterial strains present in the gastrointestinal tract. The first, frc, encodes a formyl-CoA transferase, which catalyses the transfer of a CoA moiety from formyl-CoA to oxalate to produce oxalyl-CoA and formate (Baetz and Allison 1990; Gruez et al. 2003). The second, oxc, encodes an oxalyl-CoA decarboxylase, which decarboxylates the oxalyl-CoA to produce formyl-CoA and CO2 (Baetz and Allison 1989; Werther et al. 2010). In species such as Oxalobacter formigenes that metabolize oxalate to produce energy, a third gene, oxlT, encodes an oxalate:formate antiporter, which generates the proton motive force necessary for ATP production through the exchange of intracellular formate and extracellular oxalate (Anantharam et al. 1989; Ruan et al. 1992).
For oxalate-degrading Bifidobacterium and Lactobacillus spp., both the frc and oxc genes have been shown to be important in the utilization of the substrate. In 2004, Federici et al. cloned and characterized an oxc ortholog in Bifidobacterium animalis subsp. lactis. An frc gene, which was up-regulated in the presence of oxalate under acidic conditions, was later identified in the same strain by Turroni et al. (2010). In the lactobacilli, Azcarate-Peril et al. (2006) identified frc and oxc orthologs in Lactobacillus acidophilus NCFM and they were able to show that the frc product was associated in vitro with both the degradation of oxalate by this strain and its survival in the presence of oxalic acid. Furthermore, Lewanika et al. (2007) observed up-regulated transcription of both the oxc and frc genes during growth of Lactobacillus gasseri ATCC 33323T in the presence of oxalate and showed that the two genes were co-expressed on a single transcript.
Lactobacillus reuteri 100-23C is able to stably colonize the gastrointestinal tract of mice. The genome sequence of Lact. reuteri 100-23C has been determined, and the strain has been used in conjunction with a colony of reconstituted Lactobacillus-free (RLF) mice to examine the effect of various gene products on strain metabolism and persistence in the mouse host (Tannock et al. 1988, 2012; Tannock 2004). Analysis of the genome sequence of Lact. reuteri 100-23C revealed the presence of ORFs encoding putative full-length frc and oxc genes, neither of which have been characterized previously in this strain. We aimed to examine whether the frc gene product played a role in the metabolism of oxalate in Lact. reuteri 100-23C and whether it affected ecological fitness of the strain.
All strains and plasmids used in this study are listed in Table 1. Lactobacillus strains were propagated anaerobically in de Man, Rogosa, Sharpe (MRS) broth (Biolab, Modderfontein, Gauteng, South Africa) or MRS agar (1·5% w/v) containing erythromycin (Em; 5 μg ml−1) and/or chloramphenicol (Cm; 7·5 μg ml−1) where necessary. Lactobacillus strains were incubated in an anaerobic chamber under an atmosphere of 5% H2, 10% CO2 and 85% N2 (Forma Scientific, Model 1024, Marietta, OH). Lactococcus lactis harbouring the pVE6007 plasmid was grown aerobically in M17 medium containing Cm (7·5 μg ml−1). Escherichia coli strains were grown in Luria-Bertani (LB) medium containing Em (100 μg ml−1) when expressing the pOri28-based vectors.
|Strains, plasmids or primers||Relevant characteristic(s)a||Sourceb|
|Escherichia coli EC1000||Host for pOri28-based plasmids; RepA+ derivative of E. coli MC1000, Kmr, carries a single copy of the pWV01 repA gene in the glgB gene||Law et al. 1995|
|Lactococcus lactis pVE6007||Lact. lactis MG1363, Cmr, carries the pVE6007 plasmid||Maguin et al. 1992|
|Lactobacillus reuteri DSM 20016T||Type strain of Lact. reuteri||DSMZ|
|Lact. reuteri 100-23C||Plasmid-free, rodent gastrointestinal isolate||Wesney and Tannock 1979|
|Lact. reuteri 100-23C frc−||Lact. reuteri 100-23C, Emr, carries two copies of the pOri-Reut-Frc plasmid inserted into the frc gene||This Study|
|Lact. gasseri ATCC 33323T||Type strain of Lact. gasseri||ATCC|
|Lact. gasseri B72||Human isolate of Lact. gasseri||Magwira 2008|
|pVE6007||Cmr, 1·175 kb; temperature-sensitive ScaI deletion of pVE6006; provides RepA in trans||Maguin et al. 1992|
|pOri28||Emr, ori (pWV01), replicates only with repA provided in trans||Law et al. 1995|
|pOri-Reut-Frc||2·102 kb; pOri28 with 0·436 kb internal Lact. reuteri 100-23C frc fragment||This Study|
|LrIF-F||TGA CTG GAT CCT AAG CAG CGA CTA GGG CAG G||This Study|
|LrIF-R||TGA CTG AAT TCT CGG ACC TCG GGA TTG C||This Study|
|LrTF-F1||AAG AGC CGT TGA TTC AGG ACG||This Study|
|LrTF-R1||AGA TCT ATC GAT GCA TGC CAT GG||This Study|
|LrTF-F2||GCC AAC GAA TCG CCA ACG||This Study|
|LrTF-R2||GCC GTT CTC TGA CAA ATG AAA GG||This Study|
|LrIS-F||TGG GCA AGT GAT TGG CTG G||This Study|
|LrIS-R||CCA TTC AGG ATG ACC CAT CG||This Study|
Genomic sequence data for Lact. reuteri DSM 20016T (Accession No. NC_009513) and Lact. reuteri 100-23C (Accession No. NZ_AAPZ00000000) were obtained from the NCBI database. Sequence analysis, comparison and alignments were carried out using the BLAST algorithm (Altschul et al. 1997) and MEGA software (ver. 4; Tamura et al. 2007). RNA secondary structure was determined using the RNAdraw program (ver. 1.1; Matzura and Wennborg 1996) and visualized using RnaViz (ver. 2.0; de Rijk et al. 2003). Potential promoter consensus sequences were identified based on homology to those previously described for Lactobacillus spp. (McCracken et al. 2000; Chen and Steele 2005).
Genomic DNA was prepared from Lactobacillus strains using the Genomic DNA Purification Kit (Fermentas, Burlington, Ontario, Canada) according to the manufacturer's instructions. Where necessary, cells were pre-incubated in prelysis buffer (20 mmol l−1 Tris-HCL pH 8·5, 2 mmol l−1 EDTA, 20 mg ml−1 lysozyme and 1% v/v Triton X-100) for 90 min to increase DNA yields. Plasmid DNA was routinely extracted using a plasmid DNA midi-prep kit (Qiagen, Dusseldorf, Germany) following the manufacturer's instructions. A ‘prelysis’ step, similar to that employed in the genomic DNA extraction protocol, was added when isolating plasmid DNA from Lactobacillus spp. and Lact. lactis.
The presence of the IS200 insertion sequence in the frc gene of Lact. reuteri DSM 20016T was confirmed by PCR using the LrIS-F/R primer set (Table 1), which flank the insertion site and sequencing of the resulting 0·902-kb amplicon. PCR cycling parameters were as follows: 95°C for 5 min, followed by 25 cycles of 95°C for 30 s, 53°C for 30 s and 72°C for 1 min and finally 72°C for 7 min.
Strains were cultured and analysed for oxalate degradation according to the method of Azcarate-Peril et al. (2006) with slight modifications. Oxalate-degrading activity was induced by serial passage of the cells in medium containing oxalate where the medium pH was allowed to fall naturally over the incubation period. Briefly, 50 μl of each 16 h culture was inoculated into 5 ml of MRS broth supplemented with 3 mmol l−1 sodium oxalate (MRS-OX, pH 6·8) and incubated anaerobically at 37°C for 48 h. Thereafter, 100 μl of the culture was transferred to fresh broth with the same composition and incubated anaerobically at 37°C for a further 48 h. By the end of each subculture period, the broth pH measured approximately pH 4·5. The resulting cultures were harvested by centrifugation and resuspended to prepare a 0·54 OD600 suspension in phosphate-buffered saline ((PBS), pH 7·2). An aliquot (100 μl) of the suspension was inoculated into 8 ml of MRS broth (pH 6·5) supplemented with 10 mmol l−1 sodium oxalate and incubated anaerobically at 37°C for 120 h. The pH of the medium was allowed to decrease naturally over the incubation period reaching an end point of approximately pH 4·5. A 1 ml aliquot of each culture was removed at 48 and 120 h, centrifuged, and the supernatant used for the oxalate utilization assay. The oxalate present in culture supernatants was measured using an oxalate enzymatic assay kit (Trinity Biotech, Bray, County Wicklow, Ireland) according to the manufacturer's instructions with the addition of activated charcoal to eliminate phenolic compounds (Federici et al. 2004). Data from three biological replicates, each assayed in duplicate, were averaged and the final values reported as the amount of oxalate remaining in the supernatant. Uninoculated MRS-OX, incubated in parallel with the other cultures, was used as a negative control.
A mutant of Lact. reuteri 100-23C, in which the frc gene was insertionally inactivated, was constructed using the method of Russell and Klaenhammer (2001) with the modification that the temperature-sensitive pVE6007 helper plasmid was used as described by Walter et al. (2005). Briefly, a DNA fragment internal to the frc gene of Lact. reuteri 100-23C was PCR-amplified from genomic DNA using the LrIF-F/R primer set (Table 1, PCR cycling parameters as follows: 95°C for 5 min, followed by 25 cycles of 95°C for 30 s, 53°C for 30 s and 72°C for 30 s and finally 72°C for 7 min). The resulting 0·436-kb fragment was ligated into the pOri28 vector via the BamHI/EcoRI sites to form the pOri-Reut-Frc construct and transformed into E. coli EC1000 using standard protocols (Sambrook et al. 1989). Purified plasmid was electroporated into Lact. reuteri 100-23C harbouring the pVE6007 helper plasmid (Connell 1990). Subsequent loss of the temperature-sensitive pVE6007 plasmid and integration of the pOri-Reut-Frc construct were carried out using previously described protocols (Walter et al. 2005). Correct insertion of the pOri-Reut-Frc plasmid construct was confirmed by PCR using primers flanking the target region (LrTF1/R1, LrTF2/R2, PCR cycling parameters as follows: 95°C for 5 min, followed by 25 cycles of 95°C for 30 s, 52°C for 30 s and 72°C for 30 s and finally 72°C for 7 min). These junction fragments were sequenced. The insertion was further confirmed by Southern hybridization (Sambrook et al. 1989) using a probe corresponding to the internal fragment targeted by the integration vector that was nonradioactively PCR-labelled using digoxygenin (Roche, Johannesburg, Gauteng, South Africa) and the LrIF-F/R primer pair, according to the manufacturers' instructions.
For the growth curves in the presence of sodium oxalate and acid, Lact. reuteri 100-23C wild-type and frc− mutant cells were grown initially for 16 h in MRS broth (pH 6·5). Cells were harvested anaerobically by centrifugation (6000 g, 10 min, 4°C), washed once in PBS (pH7·2) and resuspended to prepare a 0·05 OD600 nm suspension in prereduced MRS medium alone (pH 6·5), MRS medium containing sodium oxalate (10 mmol l−1 or 50 mmol l−1, pH 6·5) or MRS medium acidified to pH 4·5 using hydrochloric, lactic or oxalic acids. A 200 μl aliquot of each suspension was transferred in triplicate to separate wells of a 96-well microtitre plate (Nunc, Kamstrupvej, Roskilde, Denmark) and the entire plate sealed in optically clear plastic film. Incubations were carried out in a microplate reader (Multiskan FC model, Thermo Scientific, Pretoria, Gauteng, South Africa) at 37°C for 24 h without shaking and the growth monitored by measuring the absorbance at OD600 nm at 30-min intervals. Uninoculated MRS broth containing sodium oxalate (10 or 50 mmol l−1) or uninoculated MRS acidified to pH 4·5 using hydrochloric, lactic or oxalic acid was used as the culture blank in each case. Maximum growth rates were calculated by determining the slope of a linear regression line during exponential growth.
Examination of the ecological performance of Lact. reuteri 100-23C frc− mutants in reconstituted Lactobacillus-free (RLF) mice was carried out as described by Walter et al. (2005). Lactobacillus reuteri 100-23C wild-type and Lact. reuteri 100-23C frc− mutant cells were grown in MRS broth for 16 h, harvested by centrifugation (6500 g, 10 min, 4°C) and washed in NaCl (0·9% w/v). The resulting cell pellets were resuspended in fresh NaCl (0·9% w/v) solution and administered by intragastric gavage to 3- to 4-week-old anaesthetized RLF mice (six per group) in a 1 : 1 ratio of mutant to wild type (approximately 106 CFU per dose). Mice were maintained on a diet supplemented with 1% (w/w) ammonium oxalate for 2 weeks, after which they were euthanized and the percentage of the Lact. reuteri 100-23C frc− mutant of the total Lactobacillus population in the forestomach and caecum of the mice determined by inoculating homogenates prepared from both sites onto MRS medium with and without Em (Heng et al. 1999).
RLF mice and RLF mice colonized by wild-type Lact. reuteri 100-23C (six per group) were maintained on a diet containing 1% (w/w) ammonium oxalate for 2 weeks. Thereafter, urine was collected, acidified to pH 5·0 with 2 mol l−1 HCl and EDTA added to a final concentration of 1·34 mol l−1. Urinary oxalate was measured by HPLC. Urinary creatinine, used as an internal standard, was measured using a commercially available creatinine test kit (Wako, Wako Pure Chemical Industries, Osaka Japan) according to the manufacturer's instructions.
An analysis of the published genome sequences for the type strain, human isolate, Lact. reuteri DSM 20016T, and the rat isolate, Lact. reuteri 100-23C, revealed putative oxc and frc genes for both strains (Fig. 1). In both strains, the frc and oxc ORFs were situated adjacent to one another; however, the frc gene in Lact. reuteri DSM 20016T was interrupted by an IS200 sequence carrying the tnpA gene, which encodes a transposase. This insertion resulted in the frc gene in Lact. reuteri DSM 20016T being divided into two ORFs (frc(a) and frc(b) in Fig. 1), with the frc(b) ORF beginning at an ATG start codon introduced at the site of the division. When translated, the predicted Frc(a) and Frc(b) products together matched the full-length Frc product in Lact. reuteri 100-23C, with an additional six amino acid residues present at the end of the Frc(a) product, possibly associated with the IS200 sequence insertion. The presence of this insert in the genome of Lact. reuteri DSM 20016T was confirmed by PCR amplification and sequencing of a DNA fragment containing the tnpA gene. A nucleotide sequence corresponding to a typical Lactobacillus spp. ribosomal binding site (AAGGAG) and a promoter consensus sequence were located upstream of the oxc gene, and a potential rho-independent transcriptional terminator was located downstream of the frc gene in each case. In the Lact. reuteri 100-23C genome sequence, the predicted Oxc and Frc products showed significant homology to oxalyl-CoA decarboxylase and formyl-CoA transferase proteins described in Lact. gasseri, Lact. acidophilus and Bif. animalis subsp. lactis. The translated putative Oxc product contained the cd07035 and cd02004 domains found in members of the class of thiamine pyrophosphate (TPP)-requiring enzymes, while the full-length Frc product contained the pfam02515 domain common to members of family III of the Co-A transferases. In addition to frc and oxc orthologs, both gene clusters contained an ORF encoding a putative transporter protein (Fig. 1). The translated product showed homology to hypothetical proteins predicted by the genome sequences of various Lactobacillus spp. and Pediococcus spp. and contained a conserved domain present in members of the diverse and poorly characterized family of xanthine/uracil/vitamin C permeases.
An unexpected feature of the gene cluster in the Lact. reuteri DSM 20016T genome sequence was the presence of two additional ORFs that were not present in the Lact. reuteri 100-23C genome (Fig. 1). The first of these was an IS200-like insertion sequence interrupting the frc gene, which contained a single ORF encoding a 152 amino acid protein that showed significant homology (>80% amino acid similarity) to uncharacterized transposases present in the genome sequences of Lactobacillus salivarius, Finegoldia magna and Selenomonas sputigena and 51% amino acid similarity to the 152 amino acid TnpA protein in Salmonella typhimurium (Biserčić and Ochman 1993). In addition, in silico analysis of the region upstream of the putative tnpA gene in the Lact. reuteri DSM 20016T gene cluster confirmed the presence of nucleotide sequences capable of forming significant stem-loop structures similar to those described for the IS200 element in Salmonella spp. (Beuzón et al. 1999). A comparative analysis of the oxc-frc gene clusters in several genome sequences currently available for Lact. reuteri strains in the NCBI database revealed that the IS200 element was present in strains of human origin (Lact. reuteri DSM 20016T and Lact. reuteri MM4-1A) whereas isolates of animal origin (Lact. reuteri 100-23C and Lact. reuteri ATCC 53608) harboured full-length, uninterrupted frc genes.
The second addition to the Lact. reuteri DSM 20016T gene cluster was an ORF (Lreu_0498) encoding a putative group II intron-associated reverse transcriptase inserted downstream of the frc gene (Fig. 1). The predicted protein product contained conserved domains similar to the reverse transcriptase- and maturase-specific domains present in the ORFs associated previously with bacterial group II introns. Analysis of the nucleotide sequences flanking the Lreu_0498 gene revealed the 5′ (GTGCG) consensus sequence and conserved ‘domain 5’ secondary structure architecture characteristic of transcribed bacterial group II introns. Further comparison with sequences present in the database of mobile group II introns maintained by the Zimmerly lab (http://www.fp.ucalgary.ca/group2introns/index.htm) allowed a putative secondary structure (structural class C) for the transcript to be determined (Figure S1). The six looped domains common to all bacterial group II introns were present in the proposed structure, with the Lreu_0498 ORF located in domain 4 as described previously (Toor et al. 2001).
To examine further the role of the frc gene product, a mutant strain of Lact. reuteri 100-23C was generated by gene disruption targeting the frc gene. PCR and Southern analysis of the mutants showed that the pOri28-based construct had inserted twice into the target site. To check the stability of the insertion, cells were cultured for approximately 90 generations in the absence of Em and then plated on medium with and without the antibiotic. Reversion events, as measured by the percentage of cells that had lost the antibiotic resistance phenotype, were less than 5%, indicating that integration of the construct was stably maintained.
To determine whether the presence of the integrated construct had any effect on the in vitro growth of the mutant strain in the presence of sodium oxalate, wild-type and frc− mutant cells were cultured in MRS medium alone or MRS medium containing sodium oxalate (Fig. 2a). No difference in growth was observed between wild-type Lact. reuteri 100-23C and the frc− mutant under any of the tested conditions. Addition of 50 mmol l−1 sodium oxalate to the growth medium did not affect the maximum growth rate nor the final OD600 nm reached for either strain.
To examine the effect of the frc mutation on acid tolerance of Lact. reuteri 100-23C, growth of the wild-type and mutant strains was monitored in MRS medium acidified to pH 4·5 using hydrochloric, lactic or oxalic acids (Fig. 2b–d). The mutant had a longer lag phase and showed a slight nonsignificant reduction in average maximum growth rate versus the wild type in medium containing oxalic acid (0·22 vs 0·30 h−1, respectively). When cells were grown in medium containing lactic acid, the frc− mutant had a longer lag phase yet showed a similar maximum growth rate to the wild type (0·28 vs 0·30 h−1, respectively), and both strains reached a similar final OD600 nm (1·3). During growth in medium acidified with hydrochloric acid, however, the maximum growth rate of the frc− mutant was significantly reduced compared with the wild type (0·16 vs 0·35 h−1, respectively; P = 0·05). The cells had a longer lag phase and reached a lower final OD600 nm (0·9 vs 1·3, respectively).
Oxalate degradation by various Lact. reuteri strains was examined in MRS medium containing sodium oxalate over a period of 120 h (Fig. 3). The Lact. gasseri ATCC 33323T strain and a human isolate Lact. gasseri B72, which have previously been shown to degrade oxalate (Magwira 2008), were included as positive controls. While the Lact. gasseri ATCC 33323T type strain and Lact. gasseri B72 strain degraded approximately 50 and 95%, respectively, of the supplemented oxalate after 120 h, no degradation was seen for any of the Lact. reuteri strains. The failure to degrade oxalate as measured by the enzymatic kit assay was confirmed for the Lact. reuteri 100-23C wild-type strain by both HPLC analysis of the spent culture medium and gas chromatography-mass spectrometry analysis to determine degradation of 13C-labelled oxalate added to the growth medium (results not shown).
To examine the in vivo functioning of the Lact. reuteri 100-23C frc gene, a colony of reconstituted Lactobacillus-free (RLF) mice was utilized. These mice have a gut microbiota that is otherwise equivalent to conventional mice and thus can be used to compare independently the ecological performance of individual Lactobacillus strains (Tannock et al. 1988). Administration of wild-type Lact. reuteri 100-23C did not alter the urinary excretion of oxalate by the experimental mice compared with their uninoculated littermates. However, when equal titres of wild-type Lact. reuteri 100-23C and the frc− mutant were co-administered to RLF mice, the relative abundance of the frc− mutant as a percentage of the total Lactobacillus population was significantly reduced at sample sites in both the forestomach (median 2·3%, range 0–61·1%) and the caecum (median 0·95%, range 0–52·3%).
Lactobacillus spp. that are able to degrade dietary oxalate may have an important role to play in the prevention and management of kidney stone disease. In silico analysis of the genome sequences for several Lact. reuteri strains revealed a gene cluster containing putative frc and oxc genes. However, IS200 and group II intron-associated genes were found only in the clusters present in human-derived Lact. reuteri strains. IS200 elements, originally identified in Salm. typhimurium, are amongst the most compact transposable elements, lacking terminal repeat sequences and containing a single copy of the transposase gene, tnpA (Lam and Roth 1983; Beuzón and Casadesús 1997). Expression of the tnpA gene is tightly controlled by the presence of internal inverted repeats, one located upstream of the gene that acts as a rho-independent terminator to prevent transcriptional read-through from any proceeding ORFs and another forming a stem-loop structure that occludes the ribosomal binding site of the gene (Beuzón et al. 1999). Therefore, the presence of the IS200-like sequence in the frc gene of human Lact. reuteri strains is likely to interfere substantially with transcription of the frc gene and, potentially, with oxalate metabolism in Lact. reuteri strains of human origin. The observation that this insert is not present in strains of animal origin possibly reflects the fact that Lact. reuteri strains have co-evolved with their respective hosts leading to significant genome differences as a result of niche specialization (Frese et al. 2011). The second addition to the human-derived Lact. reuteri DSM 20016T gene cluster was a group II intron-associated reverse transcriptase. Bacterial group II introns typically consist of a catalytic RNA component that carries out self-splicing and integration reactions and an associated ORF that encodes a multifunctional protein with reverse transcriptase activity (Cousineau et al. 2000; Lambowitz and Zimmerly 2004). Overall, the predicted topology of the Lact. reuteri DSM 20016T secondary structure was similar to that found in members of the structural class C of bacterial group II introns (Dai and Zimmerly 2002). This class of introns inserts specifically downstream of rho-independent terminator sequences, such as those following the frc gene in the Lact. reuteri DSM 20016T sequence, a behaviour which is thought to reduce their impact on host fitness, because transposition rarely results in the interruption of a pre-existing ORF. In the case of the Lact. reuteri DSM 20016T insertion, the insertion site is far enough downstream of the frc gene for it to be unlikely to interfere with the functioning of the potential terminator sequences. Mobile elements such as these are thought to be important factors in the evolution of bacterial genomes and in addition have been implicated in the horizontal transfer of genes between strains (Ley et al. 2006; Callanan et al. 2008; Kaleta et al. 2010). It has been speculated previously that frc and oxc genes in oxalate-degrading intestinal strains were acquired via horizontal gene transfer (Azcarate-Peril et al. 2006), although the origin of these genes in Lact. reuteri strains is not yet clear.
Wild-type Lact. reuteri 100-23C was not able to degrade oxalate under the in vitro and the in vivo conditions tested despite having full-length, uninterrupted versions of the frc and oxc genes, which encoded proteins showing significant homology to those characterized in other lactic acid bacteria. However, it is possible that the failure of Lact. reuteri 100-23C to degrade oxalate was due to the lack of a suitable oxalate transporter. Turroni et al. (2010) previously identified a putative permease-encoding ORF that may be involved in oxalate transport in Bif. animalis subsp. lactis. Expression of the permease was observed after pre-adaptation in the presence of 5 mmol l−1 oxalate followed by growth in pH-controlled batch cultures containing 50 mmol l−1 oxalate at pH 4·5. In addition, ORFs showing significant homology to the Bif. animalis subsp. lactis permease were identified in similar oxalate gene clusters in other potential oxalate-degrading Bifidobacterium spp. Analysis of publically available genome sequences in the current study revealed that homologues of the putative permease gene were also present in the vicinity of oxalate gene clusters in several Lactobacillus spp. as well as oxalate-degrading E. coli and Providencia rettgeri. However, this gene was absent from the genome sequences of both the human- and animal-derived Lact. reuteri strains included in this study. Little is known regarding the target substrates for the predicted transporter in any of the species, although the E. coli YfdV homologue has been shown to be important in host colonization (van Diemen et al. 2005). Nevertheless, the frequent association of this gene homologue with genes encoding proteins involved in oxalate metabolism does imply a role for the transporter in oxalate transport. Experiments are currently underway to determine whether the predicted Lact. gasseri ATCC 33323T transport protein, encoded by the LGAS_0249 gene, can function in Lact. reuteri 100-23C. An alternative explanation for the failure of Lact. reuteri 100-23C to degrade oxalate might be that the strain has lost this ability during growth under laboratory conditions and does not regain it even on re-introduction into an in vivo mouse model environment. The loss of degradative ability on subculture has been observed previously for P. rettgeri, which lost the ability to degrade oxalate over time, yet still maintained an frc gene homologue that hybridized to a probe prepared from the O. formigenes frc gene (Hokama et al. 2005). The presence of an oxc gene homologue in P. rettgeri was not confirmed by the authors. However, a protein that cross-reacted with antibodies prepared against the O. formigenes Oxc protein was observed in oxalate-degrading P. rettgeri. This protein was not expressed in cells that had lost the ability to degrade oxalate during subculture. The loss of oxalate-degrading activity along with the possible suppression of other beneficial probiotic characteristics during laboratory subculture is important to bear in mind when evaluating the efficacy of potential probiotic strains for use in kidney stone prophylaxis.
An unexpected finding of the current study was that growth of the Lact. reuteri 100-23C frc− mutant in the presence of lactic acid and hydrochloric acid was diminished compared with that of the wild type. Acid tolerance is an important characteristic of bacterial strains that need to survive transit of the acidic environment of the stomach and persist in the gastrointestinal tract. A relationship between the acid stress response and the metabolism of oxalate by lactic acid bacteria has been suggested by previous studies. Mildly acidic conditions, such as those present in the gastrointestinal tract, are required for the expression of the frc and oxc genes and the induction of oxalate degradation in Lact. acidophilus NCFM, Lact. gasseri ATCC 33323T, Bifidobacterium dentium Bd1 and Bif. animalis subsp. lactis (Azcarate-Peril et al. 2006; Lewanika et al. 2007; Ventura et al. 2009; Turroni et al. 2010). Furthermore, E. coli strains harbouring wild-type frc and oxc homologues exhibit an improved tolerance to acid stress after pre-exposure to oxalate compared with frc− and oxc− mutants (Fontenot et al. 2013). It has been proposed that as the conversion of oxalate to formate has the effect of consuming one H+ ion per cycle, the degradation of oxalate may help to maintain the intracellular pH of cells that are under acid stress (Turroni et al. 2010). It is unclear, however, whether a similar strategy is employed by Lact. reuteri strains in the absence of observable oxalate degradation. The fact that none of the Lact. reuteri strains utilized oxalate under the experimental conditions suggests that this was not the case in the current study, although it does not rule out the possibility of Lact. reuteri 100-203C using this method under specific conditions that are as yet unknown. Alternatively, the Lact. reuteri 100-23C Frc protein may act on another dicarboxylic substrate such as succinate, the subsequent decarboxylation of which would also consume one H+ ion per cycle. The Frc protein from O. formigenes, for example, has been shown to catalyse the transfer of the CoA moiety from formyl-CoA to succinate (Toyota et al. 2008). However, as there have not yet been studies reporting the substrate specificities of the Lact. reuteri Frc protein, further experimentation is needed to support this hypothesis.
Experiments using RLF mice suggested that, while Lact. reuteri 100-23C did not degrade oxalate in vivo, the Frc gene product plays a role during colonization of the murine digestive tract because loss of the frc gene led to reduced levels of survival in the gut. Indeed, it has been shown that the Lact. acidophilus Frc protein is antigenic in humans, suggesting in vivo expression of the gene in the human host (Prangli et al. 2010). The diminished growth of the frc− strain compared with the wild-type strain during acid stress may have contributed towards this reduced ecological fitness. Given the possible roles of the Frc product in acid tolerance and host colonization, it will be interesting in future studies to assess whether the Lact. reuteri 100-23C frc and oxc genes are expressed under both these stress conditions.
In conclusion, while it is clear intact oxc and frc genes are necessary for the utilization of oxalate by intestinal bacterial strains, their presence alone is not sufficient to ensure degradation of the substrate under the conditions tested. Future studies will need to consider this when screening candidate oxalate-degrading intestinal strains, in particular those of human origin, based on the presence of these genes. Furthermore, the reduced fitness of the Lact. reuteri 100-23C frc− mutant during acid growth and host colonization points towards a broader role for the Frc product during stress responses as has been suggested by previous studies.
No conflict of interest declared.