Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, June 2002, p. 1934-1939, Vol. 46, No. 6
0066-4804/02/$04.00+0 DOI: 10.1128/AAC.46.6.1934-1939.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Anna King,2 David M. Livermore,3 and Lucinda M. C. Hall1*
Department of Medical Microbiology, Barts and The London School of Medicine and Dentistry, Queen Mary, University of London, London E1 2AD,1 Department of Infection, KCL, St. Thomas' Hospital Campus, London SE1 7EH,2 Antibiotic Resistance Monitoring and Reference Laboratory, Central Public Health Laboratory, London NW9 5HT, United Kingdom3
Received 6 August 2001/ Returned for modification 23 November 2001/ Accepted 16 March 2002
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
Sulfonamide antimicrobial agents inhibit the formation of dihydropteroic acid by competing with p-amino benzoic acid for condensation with 7,8-pterin pyrophosphate, a reaction catalyzed by dihydropteroate synthase (DHPS). Inhibition results in the cells becoming depleted of tetrahydrofolate (3).
Sulfonamide resistance is commonly mediated by alternative, drug-resistant forms of DHPS. In enteric bacteria two plasmid-borne genes, sul1 (or sulI) (23) and sul2 (or sulII) (17), encode resistant enzymes. sul1 forms part of the conserved region at the 3' end of most class 1 integrons (18). sul2 was originally found to be carried predominantly on small nonconjugative plasmids (17, 18) but in recent United Kingdom isolates was mainly on larger plasmids (7); sul2 is usually linked to the streptomycin resistance genes strA and strB. By contrast, sulfonamide resistance in a number of other species, including Streptococcus pneumoniae and Neisseria meningitidis, is mediated by mutation of the chromosomal gene encoding DHPS (13, 16).
Here, we present evidence that either acquisition of sul2 or an insertion in the chromosomal gene encoding DHPS, folP, can mediate sulfonamide resistance in H. influenzae. Both mechanisms were found among recent clinical isolates from two continents.
| MATERIALS AND METHODS |
|---|
|
|
|---|
|
Susceptibility testing. Isolates were grown for 18 h at 37°C in 5% CO2 on brain heart infusion agar (Oxoid, Basingstoke, United Kingdom), supplemented with 5% horse blood heated until chocolate. Colonies were suspended in sterile saline to conform to the density of a McFarland standard 0.5. Suspensions were diluted to provide a final inoculum of 103 (for sulfonamide) or 104 (for streptomycin) CFU when spotted on Iso-Sensitest agar (Oxoid) supplemented with 5% lysed defibrinated horse blood and 15 µg of NAD (Sigma Aldrich, Poole, United Kingdom) per ml containing doubling dilutions of sulfamethoxazole (0.5 to 1,024 µg/ml; Sigma Aldrich) or streptomycin (0.5 to 128 µg/ml; Sigma Aldrich). Plates were incubated at 37°C for 18 h in 5% CO2. Escherichia coli NCTC 10418 was included in each set as an MIC control, and Enterococcus faecalis NCIB 127556 was used to check the suitability of each batch of medium for sulfamethoxazole susceptibility testing (2).
PCR amplification. Primers to amplify sul1, sul2, folP, strA, and strB were designed with the Oligo-4 program (National Biosciences Inc., Plymouth, Minn.) and are described in Table 2. Primers were supplied by Amersham Pharmacia (St. Albans, United Kingdom). All PCRs were performed in reaction mixtures containing 10 mM Tris-HCl, 50 mM KCl, 200 µM concentrations each of dATP, dCTP, dGTP, and dTTP, MgCl2 as specified in Table 2, 20 pmol of each primer, 2.5 U of Taq polymerase, and 1 µl of template per 100 µl of reaction mixture. All reagents were supplied by PE Biosystems (Warrington, United Kingdom). The template comprised total DNA prepared by a guanidium thiocyanate method as described previously (13). Amplification was carried out in a PE Biosystems 2400 thermal cycler and consisted of 3 min of initial denaturation at 95°C, 30 cycles of denaturation at 95°C for 1 min, annealing at the temperature specified in Table 2 for 1 min, and extension at 72°C for 1 min, followed by a final extension at 72°C for 4 min. Products of primers DHPS5F and DHPS5R were separated on 4% polyacrylamide gels to differentiate fragments with small length differences. Other products were separated on 1% agarose gels.
|
Determination of folP copy number. In order to determine whether isolates had one or two copies of folP, total DNA was digested with EcoRV or SnaBI (Promega) and separated on 0.9% agarose gels. Gels were stained in ethidium bromide, visualized under UV, and blotted using a Posiblotter (Stratagene, Amsterdam, The Netherlands) as recommended by the manufacturer. Blots were hybridized by standard methods (20) with digoxigenin-labeled probes made from a PCR amplification product of DHPS1F and DHPS2R, and hybridization was detected via antibody-conjugated alkaline phosphatase with a DNA labeling and detection kit (Roche Diagnostics, Lewes, United Kingdom) according to the manufacturer's instructions.
Transformation. H. influenzae Rd RM118 and clinical isolate R162 (Table 1) were transformed with total DNA or purified PCR amplification products, using the M-IV method of Herriot et al. (12). Transformants were selected on triplicate plates of Iso-Sensitest agar supplemented with 5% lysed horse blood, 15 µg of NAD per ml, and 256 µg of sulfamethoxazole per ml. Two colonies were selected at random from each plate, and the MIC of sulfamethoxazole was determined. A parallel reaction with no added DNA was carried out for each transformation experiment to control for the selection of spontaneous resistant mutants.
Nucleotide sequence accession numbers. Sequences of folP for isolates A12, Z26, TOM36, A101, K57, B167, B98, O38, R162, and T194 have been given GenBank accession numbers AF378235 to AF378244, respectively.
| RESULTS |
|---|
|
|
|---|
1,024 µg/ml). Five of the eight highly sulfamethoxazole-resistant isolates from the United Kingdom were also resistant to streptomycin, as were 3 of the 16 resistant isolates from Kenya. Representative isolates with a range of susceptibilities were selected from among the United Kingdom, St Thomas' Hospital, and Kenya collections for further study (Table 1).
|
Detection of sul2 in sulfonamide-resistant isolates.
PCR amplification for the detection of sul1 and sul2 was carried out on all highly sulfonamide-resistant isolates (MIC
1,024 µg/ml) described in Table 1. The sul1 gene was not detected in any isolate but sul2 was present in eight isolates, which represented five different PFGE types (Table 1). There was complete correlation between carriage of sul2 and streptomycin resistance (streptomycin MIC,
16 µg/ml, compared to
4 µg/ml for other isolates), and both strA and strB genes were detected by PCR in all eight sul2-positive isolates.
Characterization of folP. It was postulated that in highly resistant isolates lacking both sul1 and sul2, resistance could be due to mutation in the chromosomal DHPS gene, folP. The seven highly resistant isolates investigated further comprised two isolates sharing PFGE pattern HI001 from the United Kingdom, four sharing pattern HI002 from Kenya, and the unique United Kingdom isolate TOM36 (Table 1).
In the complete genome sequence of sulfonamide-susceptible H. influenzae strain Rd (9), the folP gene is part of a 5.4-kb segment that is duplicated in the Rd genome, the two copies being separated by 135 kb. The number of folP copies present in clinical isolates was assessed by digestion of genomic DNA with EcoRV and SnaBI followed by Southern hybridization with a folP probe. These digestions were both predicted from the published sequence to yield differently sized fragments for the two folP copies in Rd.
Two folP-hybridizing EcoRV fragments were detected in strain Rd and in six of the seven sul2-negative highly resistant isolates, while only one fragment was detected in resistant isolate TOM36 and eight isolates for which MICs were in the range of 2 to 64 µg/ml. Only one SnaBI fragment was detected in all isolates except Rd, which yielded two fragments as predicted. These results suggest that all of the sul2-negative highly resistant isolates except TOM36 have two copies of folP but that polymorphisms between Rd and the clinical isolates resulted in comigration of both restriction fragments or colocation of two copies of folP on the same fragment in SnaBI digests.
Segments of 874 bp including the complete folP-coding region were sequenced from the seven sul2-negative isolates with high-level resistance, four isolates requiring MICs of 32 to 64 µg/ml, and four isolates requiring MICs of 2 to 16 µg/ml. In isolates with two chromosomal copies of folP, PCR would be predicted to amplify both copies equally; if the copies were not identical, this should be detected by the presence of two superimposed peaks at polymorphic positions, as is seen in heterozygote detection in analysis of diploid genomes. No such sequence polymorphism was seen in the clinical isolates, suggesting that when two copies were present they were identical.
Among the eight clinical isolates for which MICs were 2 to 64 µg/ml, A101 had the same folP sequence as K57, and B167 had the same as B176, while the folP sequences of the other isolates were all different. The folP sequences of these isolates diverged from folP of Rd by 3 to 9%, corresponding to 1.5 to 5.5% divergence in the encoded protein (see Fig. 3). Two of the least susceptible isolates among these eight, B98 and R162 (Table 1), had a 3-bp insertion resulting in an additional Asp residue after Pro-64. Sequences of folP from the seven highly resistant isolates were all identical and contained, relative to Rd, a 15-bp insertion followed by an A
G point mutation, corresponding to the addition of codons for Ser-Phe-Leu-Tyr-Asn after Pro-64 and replacement of Asn-65 by Asp (Fig. 2). (Alternatively, folP of highly resistant isolates could be described as having a 12-bp insertion in the sequence represented in intermediate isolates B98 and R162.) Outside the insertion, the sequence from the highly resistant isolates diverged from Rd folP by 7.9% at the nucleotide level and 4.8% at the amino acid level. Only one amino acid substitution, Ala-243
Thr, was found in the highly resistant isolates but not in any other isolates (Fig. 3).
|
|
Transformation of susceptible strains to resistance.
To determine whether the insertion of four or five amino acids in the DHPS of H. influenzae could confer resistance, transformation experiments were undertaken. Initially, H. influenzae Rd RM118 was used as a transformation recipient and was transformed with total DNA from all six representative isolates of PFGE types HI001 and HI002. DNA from each of the six isolates transformed the sulfamethoxazole MIC for Rd from 8 to
1,024 µg/ml. Transformation was then attempted with a PCR product amplified by DHPS1F and DHPS2R encompassing the complete folP gene, including the 15-bp insertion. Again, highly resistant transformants were obtained. It was not possible to determine the folP sequence from transformants, as the sequence profiles obtained were clearly the product of more than one version of the gene. Transformation efficiencies were approximately 2.3 x 10-3 for total DNA and 1.6 x 10-4 for PCR products. Resistant colonies were not detected in controls containing no added DNA, and it was estimated that spontaneous mutation to sulfonamide resistance must therefore occur at a frequency of <10-8.
These experiments demonstrated that sequence differences in folP between resistant isolates and Rd were responsible for resistance. However, in addition to the insertion, there were 12 amino acid differences between the proteins encoded by folP of Rd and the highly resistant isolates (Fig. 3). To test whether the insertion, rather than the other differences, was critical for high-level resistance, transformation experiments were performed with isolate R162 (MIC, 32 µg/ml) as a recipient. Apart from the insertion, the region of folP coding for amino acids 17 to 233 from isolate R162 was identical at the amino acid level to the folP of highly resistant isolates, despite seven nucleotide changes (Fig. 2 and Table 3). PCR primers DHPS3F and DHPS3R (Table 2) were designed to amplify a 599-bp fragment from this region. The resulting amplification product from highly resistant isolates was capable of transforming R162 to high-level resistance (MIC for transformants
1,024 µg/ml), confirming the role of the insertion (Table 3).
|
The DHPS4F/R primer pair was further used to test for the presence of the folP insertion in other highly sulfonamide-resistant isolates. The insertion was present in nine additional sulfonamide resistant isolates from the orphanage in Kenya but absent in the eight sul2-positive isolates tested (Table 1).
| DISCUSSION |
|---|
|
|
|---|
16 µg/ml. Whether any of these changes is a prerequisite for the acquisition of high-level resistance was not confirmed. Several highly resistant clinical isolates carried two copies of folP. It appeared that both copies of the gene were identical in all cases, as no size polymorphism was detected within PCR products and no evidence of polymorphism was found during sequence determination. Results from sequencing reactions with laboratory transformants of Rd suggested that polymorphism would have been readily detected.
Pro-64 is highly conserved among DHPSs (8), and changes around this position are involved in mutation to resistance in other species. Various duplications of one or two amino acids in this region of the protein can mediate resistance in S. pneumoniae (13). In N. meningitidis a change from Pro-68 (equivalent to Pro-64 in H. influenzae) to Ser or Leu has been found to influence the level of sulfonamide resistance, although other changes were also required for resistance (15). In a laboratory mutant of E. coli, a Pro-64 to Ser substitution resulted in the development of sulfathiazole resistance (24). The crystal structures of DHPSs from E. coli (1) and Staphylococcus aureus (11) have shown that the polypeptide is folded into an eight-stranded
/ß TIM barrel (i.e., a structure with the same fold configuration as triosephosphate isomerase). In E. coli, residues 58 to 71 form an interconnecting loop between ß-strand 5 and
-helix E. Within this loop, Thr-62 is involved in the binding site for 7,8-pterin pyrophosphate and Arg-63 is involved in the binding site for sulfonamide and p-amino benzoic acid (1). Clearly, a large insertion in this region, as observed here, could significantly alter binding specificity.
The insertion of extra residues in DHPS has been implicated as a resistance mechanism in a number of species (22). These insertions have mostly reflected short duplications in the DNA sequence (8, 13). However the H. influenzae insertion cannot be interpreted as a duplication and must have arisen by a different mechanism. In this context it is noted that the sequences of the folP genes from highly resistant (sul2-negative) isolates were identical, though they were collected from two different continents and some 8 years apart, whereas those from isolates without high-level resistance were variable. Nevertheless, PFGE typing showed that isolates with the folP insertion obtained from different locations were not related. It is postulated that the resistant form of the gene has arisen through a single event and then spread among the population by natural transformation.
In conclusion, two distinct mechanisms can mediate high-level sulfonamide resistance in H. influenzae: acquisition of a resistant DHPS encoded by sul2 and alteration of the chromosomal DHPS gene folP. Both mechanisms were found among isolates collected in Kenya in 1999 and among isolates collected in the United Kingdom in 1991 and 1999.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Present address: Department of Pathology and Microbiology, University of Bristol, Bristol BS8 1TD, United Kingdom. ![]()
| REFERENCES |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Clin. Vaccine Immunol. | Clin. Microbiol. Rev. |
|---|---|
| J. Clin. Microbiol. | ALL ASM JOURNALS |