Abstract
Trimethoprim with sulfamethoxazole is a therapeutic agent combination used to treat infections caused by the facultative intracellular foodborne pathogen Listeria monocytogenes. The aim of this study was to assess the frequency of resistance of L. monocytogenes arising due to exposure to trimethoprim and subsequently investigate the molecular mechanisms of resistance. After exposure of a culture of L. monocytogenes ATCC 13932 to trimethoprim at 10-fold the minimal inhibitory concentration spontaneous resistant mutants were recovered, giving a frequency of resistance development of 6.85 ± 0.92 × 10−8. The isolates exhibited a 32–64-fold decrease in susceptibility compared with the parental strain. These results indicate the capacity of L. monocytogenes to develop low-level resistance toward trimethoprim after exposure to the drug. The trimethoprim resistance genes (dhfr) and their promoter regions from all trimethoprim-resistant isolates were amplified and sequenced, leading to the identification of four single amino acid substitutions (Met20-Val, Pro21-Leu, Thr46-Asn, Val95-Leu) and two double substitutions (Met20-Ile+Thr46-Asn and Thr46-Asn+Leu85-Phe) in DHFR. Of the identified mutations, the Thr46-Asn substitution has not been previously reported as the mechanism of resistance to trimethoprim in other bacteria; thus this substitution seems to be unique to L. monocytogenes. The expression of the mutated L. monocytogenes dhfr genes in Escherichia coli led to decreased susceptibility of the heterological host, therefore proving that the identified point mutations in dhfr serve as the molecular mechanism of acquired resistance of L. monocytogenes to trimethoprim.
Introduction
T
Since the 1980s, trimethoprim in combination with sulfamethoxazole (called co-trimoxazole) has been used successfully in the treatment of infections with the facultative intracellular foodborne pathogen Listeria monocytogenes. This pathogen can cause severe human infections, such as septicemia and meningitis, primarily in neonates and immunocompromised adults and perinatal infections that may result e.g., in abortions (Hof, 2004). Co-trimoxazole is a therapeutic alternative in treating listeriosis in the case of first-line treatment failure or intolerance to beta-lactams. Case reports and a case series demonstrated the efficacy of co-trimoxazole both in combination with other drugs (such as gentamicin, amoxicillin, and rifampicin) and as monotherapy (Günther and Philipson, 1988; Winslow and Pankey, 1982; Spitzer et al., 1986; Merle-Melet et al., 1996; Fernandez-Sabé et al., 2009; Grant et al., 2010).
Listeriosis represents a grave public health problem since it was fatal in around 20% of cases reported in the last two decades (Muňoz et al., 2012) and is among the most important causes of death from foodborne infections in industrialized countries (EFSA and ECDC, 2016).
Generally, bacterial resistance to trimethoprim is mediated by the following five main mechanisms: (i) a permeability barrier, (ii) a naturally insensitive intrinsic DHFR, (iii) spontaneous mutations in the intrinsic DHFR, (iv) increased production of the sensitive target enzyme by upregulation of gene expression or gene duplication, and (v) horizontal acquisition of dfr genes that encode resistant DHFRs (Huovinen et al., 1995; Sköld, 2001; Grape, 2006).
Nothing is known about the emergence of development and the genetic basis of chromosomal resistance to trimethoprim in L. monocytogenes. It is therefore important to assess the rate of resistance development, determine the level of resistance, and identify the mutation sites. In this study we selected trimethoprim-resistant mutants of L. monocytogenes in vitro and determined their genetic characteristics.
Materials and Methods
Selection of trimethoprim-resistant mutants and determination of resistance development frequency
L. monocytogenes ATCC 13932 (wild type) was used. Cultures were grown in brain-heart infusion broth (Oxoid, Basingstoke, United Kingdom) to a cell density of approximately 108/mL and then were concentrated by centrifugation to ∼1010 cells/mL. Then 100 μL was spread onto brain heart infusion agar plates containing 1.25 μg of trimethoprim (10 × MIC) (Sigma, Poole, United Kingdom) per mL and incubated at 37°C for 24 h. Mutation frequency determinations were performed in triplicate, and the results expressed as the number of antibiotic-resistant mutants recovered as a fraction of the viable count (Vikkers et al., 2007). The single colonies growing on agar plate containing 1.25 μg/mL of trimethoprim were selected for further study.
Determination of trimethoprim susceptibility
The MICs of trimethoprim for all selected isolates and wild-type strain of L. monocytogenes were determined using a broth microdilution method on cation-adjusted Muller-Hinton Broth supplemented with 5% lysed horse broth according to guidelines of the Clinical and Laboratory Standards Institute (CLSI, 2016).
The MICs of trimethoprim for Escherichia coli transformants were determined using a broth microdilution method according to CLSI (2012) on cation-adjusted Muller-Hinton broth supplemented with 1% arabinose.
To ensure reproducibility, MIC determinations were repeated thrice.
Genetic characterization of trimethoprim-resistant mutants
To identify potential trimethoprim resistance mutations, the complete copy of the dhfr gene (483 bp) together with upstream (52 bp) and downstream (116 bp) regions was amplified from all phenotypically resistant mutants and L. monocytogenes ATCC 13932 strain using oligonucleotide primers: Lmo1873-F1 (5′ GACAAGCAAGAAGCGCCATA 3′) and Lmo1873-R2 (5′ TCCAGATCCATCGATTAAAGCA 3′). Primers were designed based on sequence of lmo1873 gene coding DHFR (position 1947416–1947898) and flanking regions from the genome sequence of L. monocytogenes EGD-e (GenBank accession no. AL591974).
To identify potential mutations in the promoter region of the operon 333 (lmo1875, lmo1874, lmo1873) (Toledo-Arana et al., 2009), the region located upstream of lmo1875 was amplified, using starters poFA (5′CATTAAATATGGATGTCGATGA3′) and poRD (5′CATTTCCTGTTCGTCCTTATCTA3′).
Chromosomal DNA of wild-type strain and resistant mutants was extracted using GeneMATRIX Bacterial & Yeast Genomic DNA Purification Kit (EURx, Gdańsk, Poland). Amplification was carried out in a 25 μL reaction mixture containing 0.2 μM of each primer, 0.2 mM dNTPs, 1 U Pfu polymerase (Fermentas, Lithuania), and 50 ng chromosomal DNA.
The conditions were heating either for 1 cycle at 94°C for 5 min; 30 cycles at 94°C for 30 s, annealing for 30 s at 53°C and at 72°C for 1 min, followed by 1 cycle at 72°C for 10 min (dhfr gene and flanking region) or for 1 cycle at 94°C for 2 min; 25 cycles at 94°C for 10 s, annealing for 15 s at 51°C and at 72°C for 15 s, followed by 1 cycle at 72°C for 5 min (the promoter region of the operon 333).
All PCR products were purified using QIAquick PCR Purification Kit (QIAGEN, Germany). Sequencing was performed using DYEnamic ET Terminator Cycle Sequencing Kit (GE Healthcare, Chalfont St. Giles, United Kingdom) on an Applied Biosystems 3730 automated sequencer (Applied Biosystems, Perkin-Elmer, Foster City, CA).
To confirm the role of the identified DHFR mutations in trimethoprim resistance, the dhfr gene preceded by its ribosome binding site was PCR-amplified from trimethoprim resistant strains encoding DHFRMet20Val, DHFRPro21Leu, DHFRThr46Asn, DHFRVal95Leu, DHFRMet20Val+Thr46Asn, DHFRThr46Asn+Leu85Phe and from parental ATCC 13932 strain, using primers lmo1873F(5′GTG
The amplified fragments were digested with KpnI and XbaI and ligated using the corresponding restriction sites into the vector pBAD18 downstream from the arabinose-inducible PBAD promoter (Guzman et al., 1995) to generate the constructs pBAD-DHFRwt (wild-type DHFR), pBAD-DHFRMet20Val, pBAD-DHFRPro21Leu, pBAD-DHFRThr46Asn, pBAD-DHFRVal95Leu, pBAD-DHFRMet20Ile+Thr46Asn, and pBAD-DHFRThr46Asn+Leu85Phe. Constructs were introduced into E. coli Top10F’ by standard heat-shock transformation. The correctness of constructs was verified by sequencing, followed by performing trimethoprim susceptibility testing for the transformants.
Results
Isolation and susceptibility testing of trimethoprim-resistant derivatives of L. monocytogenes ATCC 13932
Spontaneous trimethoprim-resistant mutants of L. monocytogenes ATCC 13932 were recovered on media containing 1.25 μg/mL trimethoprim. The mutants arose at frequencies of 6.85 ± 0.92 × 10−8.
A total of sixty four colonies were isolated. The MICs of trimethoprim for these isolates ranged from 2 to 8 μg/mL, while the MIC against the parental L. monocytogenes ATCC 13932 strain was 0.125 μg/mL. The resistant mutants were thus 32- to 64-fold less susceptible to trimethoprim than the wild-type strain.
Characterization of trimethoprim resistance genotypes
All L. monocytogenes isolates phenotypically resistant to trimethoprim, as well as the wild type, were further characterized by sequencing the DHFR gene (lmo1873) and the promoter region of operon 333.
Sixty one isolates carried a single base pair alteration that resulted in an amino acid substitution. The most common single substitution was threonine for asparagine at position 46 of DHFR (33 of 61 isolates), followed by methionine for valine at position 20 (21 of 61 isolates). Four isolates revealed a change at position 21 (proline to leucine) and three others at 95 (valine to leucine). Three resistant isolates had double base pair alterations. Interestingly, all these isolates had the substitution Thr46-Asn. Two of them additionally had the substitution Met20-Ile. Only one isolate apart from the Thr46-Asn change exhibited the substitution Leu85-Phe. The active site mutation Thr46-Asn was detected in 56.25% (36 of 64) resistant isolates while the alteration at position 20 (Met20-Val and Met20-Ile) in 35.94% (23 of 64) mutants.
Sequence analysis of the promoter regions of all trimethoprim-resistant isolates of L. monocytogenes showed no differences between mutants and wild-type strain.
All identified chromosomal mutations led to low trimethoprim resistance (Table 1).
Confirmation of the role of identified DHFR mutations in resistance to trimethoprim
To verify if the trimethoprim resistance phenotype of the analyzed isolates of L. monocytogenes results from point mutations in gene dhfr, the effect of the expression of the mutated dhfr genes in a heterological host was examined. For this purpose, all identified variants of dhfr genes of L. monocytogenes were cloned into the expression vector pBAD18, and the resulting recombinant vectors were introduced into E. coli strain. The susceptibility to trimethoprim of the obtained E. coli isolates was subsequently examined. The study showed that E. coli strains which expressed mutated variants of dhfr of L. monocytogenes conferred a 4- to 256-fold decrease in susceptibility to trimethoprim compared with E. coli carrying the wild-type dhfr sequence (Table 2).
Discussion
L. monocytogenes is naturally susceptible to trimethoprim (Troxler et al., 2000), but resistance mediated by the acquisition of dfr genes has been reported. The resistance of L. monocytogenes against trimethoprim was first detected in France in an environmental isolate containing the dfrD gene encoded on the 3.7 kb plasmid pIP823 (Charpentier and Courvalin, 1997). Two further isolates from France were reported to contain this resistance gene (Morvan et al., 2010; Granier et al., 2011). Recently, a novel trimethoprim resistance gene dfrG located within transposon Tn6198 was reported in a human isolate from Switzerland (Bertsch et al., 2013).
In this study, we report the results of the analysis of spontaneous trimethoprim resistant mutants of L. monocytogenes ATCC 13932.
The resistant clones were recovered with frequency similar to the mutation frequency of single-step resistance to, for example, rifampicin in L. monocytogenes (Morse et al., 1999). In contrast, this mutation frequency is relatively high compared with the mutation frequency of resistance to trimethoprim in Staphylococcus aureus (Vickers et al., 2007).
All resistant mutants exhibited up to a 64-fold reduction in susceptibility to trimethoprim. However, the level of resistance is relatively low compared with resistant strains of L. monocytogenes that carry the transferable genes, dfrD and dfrG, which mediate high-level resistance to trimethoprim with MICs >1000-fold greater than the normal values (Charpentier and Courvalin, 1997; Morvan et al., 2010; Granier et al., 2011; Bertsch et al., 2013). An amino acid substitution in the dhfr gene and altered chromosomally encoded DHFR that confer resistance appear to be rare, but examples have been reported for S. aureus (Dale et al., 1997; Vickers et al., 2007), Streptococcus pneumoniae (Adrian and Klugman, 1997; Pikis et al., 1998; Maskell et al., 2001), and E. coli (Watson et al., 2007). In strains of trimethoprim-resistant Haemophilus influenzae (de Groot et al., 1996) and E. coli (Flensburg and Skold, 1987; Huovinen, 1987; Toprak et al., 2012), changes in both the promoter and coding regions of the dhfr genes have been found.
Sequence analysis revealed that resistant mutants of L. monocytogenes had either single base pair changes or double ones in dhfr gene coding DHFR that resulted in an amino acid substitution. In addition, the expression of the mutated dhfr genes in a heterological host, E. coli, clearly indicates that the identified point mutations in dhfr of L. monocytogenes indeed are responsible for the acquired trimethoprim resistance phenotype.
Certain mutations were frequently found: the substitution threonine for asparagine at position 46 in 36 isolates and methionine for valine at position 20 in 21 and for isoleucine in two, respectively. Four isolates revealed a change at position 21 (proline to leucine) and three others at 95 (valine to leucine).
Substitutions at the same position have also been found in other bacteria. The amino acid substitution of methionine for valine at position 20 in L. monocytogenes was identified in E. coli (corresponding position 20). Another substitution, Pro21-Leu, observed in L. monocytogenes isolates, was also found in trimethoprim-resistant E. coli strains (Watson et al., 2007; Toprak et al., 2012). These mutations fall on the highly studied M20 loop (residues 10–24) of DHFR from E. coli (Sawaya and Kraut, 1997). The Met20 loop lies directly over the active site (Asp27) and is primarily responsible for determining the active site architecture (Schnell et al., 2004).
The Val95 of the L. monocytogenes DHFR is homologous to Ile94 of E. coli DHFR, which is one of few residues that are in van der Waals contact with the pyrimidine ring of trimethoprim (Matthews et al., 1985). By replacing isoleucine or valine with a similar amino acid, leucine, the affinity of trimethoprim for the active site may probably be reduced without affecting dihydrofolate binding. The corresponding mutation in S. pneumoniae (Ile100-Leu) is critical for the generation of trimethoprim resistance in this bacterium (Adrian and Klugman, 1997; Pikis et al., 1998). An identical Ile-Leu substitution was also identified in a trimethoprim-resistant clinical isolate of H. influenzae (de Groot et al., 1996). Thus, the aforementioned mutations observed in L. monocytogenes DHFR are also sites for mutations seen in pathogenic bacteria resistant to trimethoprim.
Interestingly, the mutation leading to replacing Thr to Asn at position 46 which has been detected at high frequency among the resistant isolates of L. monocytogenes has not been described in isolates of E. coli (Flensburg and Skold, 1987; Watson et al., 2007; Toprak et al., 2012), S. aureus (Dale et al., 1997; Vickers et al., 2007), S. pneumoniae (Adrian and Klugman, 1997; Pikis et al., 1998; Maskell et al., 2001), or H. influenzae (de Groot et al., 1996). This strongly indicates that the Thr to Asn substitution is unique to the mechanism of resistance of L. monocytogenes to trimethoprim. Threonine at position 46 and at position corresponding in DHFRs from other species is a conserved residue (Liu et al., 2013). As has been shown for E. coli DHFR, Thr46 is one of few amino acids involved in NADPH binding (Oyeyemi, 2008).
As shown by others, the promoter mutation may lead to DHFR overproduction and resistance to trimethoprim (Huovinen, 1987). In this study the DNA sequence of the promoter regions of all trimethoprim-resistant isolates of L. monocytogenes was also analyzed. There were no differences in sequences between the mutants and wild-type strain.
These data suggest that only an amino acid substitution in DHFR is responsible for the trimethoprim resistance phenotype of these isolates.
Conclusion
In summary, in this study for the first time the emergence and genetic basis of spontaneous resistance of L. monocytogenes toward trimethoprim, used in therapy of infections caused by this foodborne pathogen, were investigated.
The results indicate the capacity of L. monocytogenes to select resistance toward trimethoprim after exposure to the drug. The molecular mechanism of resistance results from mutations leading to substitutions in key amino acids in DHFR. While most of the identified substitutions are common in pathogenic bacteria, the most frequently identified Thr46-Asn substitution has not been previously reported, which makes our finding seem unique to L. monocytogenes.
Footnotes
Acknowledgments
The authors are greatly thankful to Mariola Siwek for technical assistance and Prof Zdzislaw Markiewicz for critical review of the article.
Disclosure Statement
No competing financial interests exist.
