|
A correction has been published: N Engl J Med 1999;340(10):824.
|
| |||||||||||||||||||||||||||||||||||||||||||||||
Background Neisseria meningitidis is nearly always susceptible to the penicillins, the cephalosporins, and chloramphenicol. Between 1987 and 1996, however, chloramphenicol-resistant strains were isolated from 11 patients in Vietnam and 1 in France.
Methods The minimal inhibitory concentration of chloramphenicol was determined for the 12 isolates. The isolates were analyzed by monoclonal-antibody–based serotyping and subtyping, pulsed-field gel electrophoresis, and multilocus enzyme electrophoresis. Bacterial DNA was analyzed by hybridization, the polymerase chain reaction, and sequencing to identify the resistance gene and determine the origin of the resistance.
Results The isolates were resistant to chloramphenicol (minimal inhibitory concentration,
Conclusions The high-level chloramphenicol resistance that we describe in N. meningitidis isolates is of great concern, since in developing countries, chloramphenicol given intramuscularly is the standard therapy for meningococcal meningitis. The resistance to chloramphenicol is due to the presence of the catP gene on a truncated transposon that has lost mobility because of internal deletions, and the transformation of genetic material between strains of N. meningitidis probably played an important part in the dissemination of the gene.
64 mg per liter) and produced an active chloramphenicol acetyltransferase. All 12 strains belonged to serogroup B but had a high degree of diversity, and 10 could not be typed with the use of monoclonal antibodies. The nucleotide sequence of the resistance gene and the flanking regions was identical to that of an internal portion of transposon Tn4451 that carries the catP gene in Clostridium perfringens. Moreover, this gene was located in the same genomic site in the chloramphenicol-resistant isolates.
N. meningitidis is differentiated into serogroups on the basis of the composition of the polysaccharide capsule. Serogroup A causes widespread epidemics in sub-Saharan Africa, where a 600-km-wide meningitis belt extends from the Gambia in the west to Ethiopia in the east. Outbreaks of serogroup A meningitis have occurred in the Middle East and in southern and eastern Asia. Serogroup B meningococci are the cause of most cases of meningococcal disease in developed countries, but the attack rates are substantially lower than those for serogroup A. During the late 1970s, a serogroup B clone (serotype 15, electrophoretic type 5) emerged in northwestern Europe. Intercontinental spread of this clone has been documented, with outbreaks in Cuba (in 1980), Chile (in 1985), and Brazil (in 1987). Strains of this type were also isolated from patients in the United States in 1988. Cases caused by serogroup B organisms began to decline slowly after 1988, whereas those caused by serogroup C organisms increased in several European countries and in the United States.1 There are vaccines against serogroups A, C, W-135, and Y, but no effective vaccine against group B meningococci is available.
Among the bacteria that cause serious infections, N. meningitidis is one of the least problematic in terms of antibiotic resistance.2 Resistance to penicillin G, sulfonamides, rifampin, and tetracyclines has been reported, but high-level resistance to penicillins through 
-lactamase production is still extremely rare. Low-level resistance to penicillins due to an alteration in penicillin-binding protein–2 is widespread.2 Sulfonamides have been used extensively, and resistance caused by a mutation in the chromosomal gene for the dihydropteroate synthase is common.3 High-level resistance to rifampin is due to mutations in the gene for the subunit of RNA polymerase (rpoB)4 and low-level resistance to a decrease in membrane permeability.5 The plasmid-borne tet(M) gene confers resistance to tetracyclines in genital and anorectal isolates of N. meningitidis.6,7 In addition, the development of drug resistance in isolates of N. gonorrhoeae, a closely related organism, has led to concern about the almost inevitable progression of resistance in N. meningitidis isolates.
We report high-level resistance to chloramphenicol in epidemiologically unrelated clinical isolates of N. meningitidis through the production of a chloramphenicol acetyltransferase. This observation is clinically important because intramuscular administration of chloramphenicol is standard therapy for meningococcal meningitis in developing countries.
Methods
Patients and Strains
The properties of the chloramphenicol-resistant N. meningitidis strains isolated from cerebrospinal fluid are listed in Table 1. N. meningitidis LNP11890 was isolated from a two-year-old girl at the Hôpital de Metz (Metz, France) in July 1993. The patient, who had never been to Southeast Asia, presented with fever, vomiting, meningeal signs, and cloudy cerebrospinal fluid. She was treated with cefotaxime for eight days and recovered.
|
Strain LNP14610 was isolated on May 10, 1996, from a nine-year-old boy living in Ho Chi Minh City. The patient presented with high-grade fever, vomiting, headache, meningeal signs, petechial lesions, and cloudy cerebrospinal fluid. He recovered after treatment with three daily intravenous injections of cefotaxime (6 g per day) for eight days, followed by penicillin G (1,250,000 units per day) for five days.
Strain LNP14608 was isolated on May 31, 1996, from a two-year-old girl living in the same district of Ho Chi Minh City where the nine-year-old boy lived. The strain was isolated at the same hospital. The patient presented with fever, vomiting, meningeal signs, and cloudy cerebrospinal fluid. Treatment included intramuscular injections of gentamicin (0.08 g per day) twice a day for 7 days and intravenous injections of cephradine (4 g per day) four times a day for 10 days. The patient recovered after two weeks.
Strain LNP14609 was isolated in June 1996 from a five-year-old girl living outside Ho Chi Minh City; the strain was isolated at a different hospital from that where strains LNP14610 and LNP14608 were isolated. The patient presented with fever, headache, vomiting, and cloudy cerebrospinal fluid. She was treated with intravenous injections of penicillin G (400,000 units per kilogram of body weight per day) for 10 days and recovered.
Nine chloramphenicol-susceptible strains (minimal inhibitory concentration, 1 mg per liter) with various antigenic characteristics were used as controls in polymerase-chain-reaction (PCR) studies: LNP14299, LNP14308, and LNP14371 in serogroup A; LNP15339, LNP15340, and LNP15341 in serogroup B; and LNP15345, LNP15346, and LNP15347 in serogroup C.
Mediums and Resistance Studies
The strains were grown on GC-base agar (Difco) containing supplements, as described by Kellogg et al.,11 and were incubated at 37°C with 5 percent carbon dioxide. The minimal inhibitory concentrations of antibiotics were determined by the E test on supplement G–containing agar medium (Sanofi Diagnostics Pasteur). Chloramphenicol acetyltransferase and aminoglycoside-modifying enzymes were assayed in supernatants (centrifuged at 100,000xg) after ultrasonic disintegration.12,13 With Escherichia coli JM83, the following antibiotics were used: chloramphenicol (10 mg per liter) for cloning the resistance gene and kanamycin (20 mg per liter) for cloning the PCR product.14 With neisseriae, the following antibiotics were used: chloramphenicol (10 mg per liter), nalidixic acid (20 mg per liter), and tetracycline (2 mg per liter). The transformation of N. meningitidis BM4376 (a nalidixic acid–resistant derivative of N. meningitidis 8013,15 serogroup C, kindly provided by X. Nassif) and BM4377 (a recA::tet(M) tetracycline-resistant derivative of N. meningitidis 8013) with total DNA from chloramphenicol-resistant N. meningitidis LNP13947 was performed as described elsewhere.15
Antigenic Classification of N. meningitidis Isolates
The strains were serogrouped with the use of monoclonal antibodies and the agglutination technique, as described elsewhere.8 Isolates were typed and subtyped with the use of the whole-cell enzyme-linked immunosorbent assay of Abdillahi and Poolman.9
Multilocus Enzyme Electrophoresis
Protein extraction and selective enzyme staining were carried out as described by Caugant et al.10 Every isolate was characterized according to the combination of its alleles at the 13 enzyme loci, and distinctive multilocus genotypes were designated as electrophoretic types.16
Nucleic-Acid Techniques
Isolation of DNA, cleavage by restriction endonucleases, and purification of DNA fragments from low-temperature-gelling agarose type VII (Sigma Chemical) were performed as described elsewhere.17 Large restriction fragments were separated by pulsed-field gel electrophoresis in 0.5x TRIS–borate buffer in a CHEF-DRIII system (Bio-Rad). Purified DNA fragments to be used as probes were labeled with [
-32P]deoxycytidine triphosphate by nick translation. Hybridization was carried out under highly stringent conditions.17 The PCR assay was performed with a DNA thermal cycler (model 2400, Perkin–Elmer Cetus). Sequencing of double-stranded DNA was performed by the dideoxynucleotide chain-termination method18 with a modified T7 DNA polymerase and [-35S]deoxyadenosine triphosphate.
Nucleotide-Sequence Accession Number
The nucleotide sequence of the catP gene and flanking regions from N. meningitidis LNP13947 has been deposited in the GenBank under accession number AF031037.
Results
Bacterial Identification
The minimal inhibitory concentrations of chloramphenicol against the 12 isolates, determined by the E test, ranged from 64 to 192 mg per liter (Table 1). Disk-agar diffusion tests showed that all 12 were also resistant to streptomycin and sulfonamides but remained susceptible to penicillins, cephalosporins, tetracyclines, macrolides, rifampin, and quinolones (data not shown). All strains belonged to serogroup B, and most could not be serotyped with the use of monoclonal antibodies. The antigenic characterization of each strain is shown in Table 1.
The strains were quite diverse. Pulsed-field gel electrophoresis after digestion with the BglII restriction endonuclease identified nine chromosomal profiles19 (Figure 1A), and multilocus enzyme electrophoresis identified eight electrophoretic types16 (Table 1). Taken together, these results indicate that the chloramphenicol-resistant isolates were not clonally related. In particular, strains LNP14608, LNP14609, and LNP14610, which were isolated in 1996 and for which clinical information was available, differed from one another on the basis of both multilocus enzyme electrophoresis (Table 1) and pulsed-field gel electrophoresis (Figure 1A, lanes 11, 12, and 13).
|
Total DNA from N. meningitidis strain LNP13947 and pUC18 DNA digested with Sau3A and BamHI, respectively, were mixed, ligated, and introduced into E. coli JM83 (minimal inhibitory concentration of chloramphenicol, 2 mg per liter). The resulting transformed isolates (transformants) were selected on the basis of their resistance to chloramphenicol, and the smallest hybrid plasmid, pAT447, was found to contain a 6.5-kb insert. Subcloning was accomplished by introducing the 4.6-kb EcoRI–HindIII fragment of the insert into pUC18, which generated pAT448. The latter construct conferred high levels of resistance to the new host (minimal inhibitory concentration, >256 mg per liter) by synthesis of a chloramphenicol acetyltransferase, as determined by an enzyme assay (data not shown).
The sequence of a 1.5-kb DNA segment in the cloned fragment from N. meningitidis LNP13947 was determined. The sequence of the resistance gene and of the flanking regions was identical to approximately 1 kb of DNA (nucleotides 257 to 1243; numbering according to GenBank accession number AF031037) present within the 6.3-kb plasmid-borne transposon Tn4451 from Clostridium perfringens (nucleotides 2721 to 3707; numbering according to GenBank accession number U15027), which carries the catP gene.20,21 Two primers (primer A, 5'ATTCAGAGTTTAGGACGG3', and primer B, 5'ATCAAATAATCAATCC3') designed from the sequence of catP allowed amplification of a fragment of total DNA with the predicted size of 300 bp from all the resistant isolates but not from chloramphenicol-susceptible strains LNP14299 (serogroup A), LNP15339 (serogroup B), and LNP15345 (serogroup C) (data not shown).
After susceptible strains of N. meningitidis had been independently transformed with chromosomal DNA from N. meningitidis LNP13947, four randomly selected chloramphenicol-resistant transformants were studied further: BM4378 and BM4379, obtained from BM4376, and BM4380 and BM4381, obtained from BM4377.
Genomic Region of the catP Gene
The probe made from the 300-bp fragment internal to catP from N. meningitidis LNP13947 hybridized to a BglII fragment of approximately 100 kb in the 12 chloramphenicol-resistant isolates (Figure 1B, lanes 2 through 13). The 45-kb difference in the size of the BglII fragment in transformants BM4378 and BM4380 (Figure 1B, lanes 15 and 16, respectively) was due to the presence of a BglII site within the tetracycline-resistance determinant15 in BM4377. To investigate the junctions between the DNA of the truncated transposon and that of the host, two primers were designed on the basis of the 5' and 3' meningococcal sequences flanking the Tn4451-like insertion (primer C, 5'CTAAATCAATAATAATATTCCC3', and primer D, 5'ACCCCAGGTAGAATACAATGAA3').
These primers, which allow amplification of a 1206-bp fragment with the insertion of pAT448 as a template (Figure 2, lane 6), gave rise to products of approximately 1200 bp in the chloramphenicol-resistant isolates (Figure 2, lanes 7 through 18) and to products of approximately 200 bp in the susceptible N. meningitidis strains belonging to serogroup A (three strains), B (three strains), and C (four strains) (Figure 2, lanes 3, 4, and 5). The predicted size of the PCR product corresponding to the target site before the acquisition of transposon sequences was 218 bp, on the assumption that there was no deletion in the adjacent meningococcal DNA. Moreover, with the use of the same primers, a 1200-bp product was also obtained from the four independent resistant transformants (Figure 2, lanes 20 and 21) as opposed to a 200-bp product from the chloramphenicol-susceptible corresponding strains (Figure 2, lane 19). Taken together, these results indicate that the catP gene was located in the same genomic region in all the resistant N. meningitidis strains.
|
The sequence of the catP gene and flanking regions from transposon Tn4451, from the pAT448 insert, and from the 1200-bp PCR products obtained from the transformants was compared with that of the 200-bp product obtained from susceptible N. meningitidis strains (Figure 3). Insertion of the catP gene occurred at the same site in a region with a large proportion of adenosine and thymidine (77 percent, as compared with 50 percent under normal conditions).22 Two thymidines are present in the meningococcal DNA and at the 3' end of the truncated transposon, indicating that there is a 1-bp or 2-bp deletion in the target DNA or no loss of meningococcal DNA in the transformants (Figure 3).
|
The entire gene for the dihydropteroate synthase from two sulfonamide-susceptible strains (LNP14299 and LNP15339) and two sulfonamide-resistant strains (LNP13943 and LNP14610) was amplified by PCR with the use of specific primers23 and sequenced. The resistant strains had a substitution of guanosine for thymidine at position 92 (N. meningitidis numbering), resulting in the replacement of the conserved phenylalanine by a leucine at position 31 and a 6-bp insertion (TCCGGC) resulting in two extra amino acids (serine at position 195 and glycine at position 196) in a highly conserved region of the enzyme.23 Each of these mutations has been shown to be associated with sulfonamide resistance in N. meningitidis strains.23
DNA from the 12 streptomycin-resistant isolates did not hybridize with probes specific for the streptomycin–adenylyltransferase ant(3")(9) gene and the streptomycin–phosphotransferase aph(3")-I and aph(6)-I genes. No streptomycin adenylyltransferase or phosphotransferase activity was detected in crude extracts from these strains. In other bacterial genera, streptomycin resistance can also result from mutations in the gene for ribosomal protein S12 (rpsL), leading to the substitution of lysine at positions 43 and 88 (E. coli numbering),24 or in the portions of the rrs gene that correspond to the highly conserved 530-loop and 912 region of 16S ribosomal RNA.25 Sequencing of the rpsL gene (from position 75 to position 300) and the rrs gene (from 368 to 979) in susceptible strains (LNP14299 and LNP15399) and resistant strains (LNP13943 and LNP14610) did not reveal any mutation.
Discussion
N. meningitidis is one of the few bacterial species that causes serious infections for which penicillin G and chloramphenicol are still routinely recommended in developing countries. Strains of N. meningitidis with high levels of resistance to chloramphenicol were isolated from cerebrospinal fluid in Vietnam and in France between 1987 and 1996. In Vietnam, chloramphenicol and penicillin G were used for the treatment of meningococcal meningitis in the 1980s, and penicillin G and cephalosporins were used in the 1990s, whereas cephalosporins were used in France during that period. The isolates were also resistant to streptomycin and sulfonamides but remained susceptible to penicillins, cephalosporins, tetracyclines, rifampin, and quinolones; quinolones are used for the treatment of meningococcal disease and also as prophylaxis to reduce nasopharyngeal carriage of meningococci.26 The resistance of the N. meningitidis isolates to sulfonamides was due to the production of an altered dihydropteroate synthase, but the mechanism of streptomycin resistance remains unknown.
The resistance of the N. meningitidis isolates to chloramphenicol was due to the presence, at a unique site, of the catP gene as part of a truncated copy of transposon Tn4451 from C. perfringens. Although it has been shown that excision of Tn4451 from C. perfringens and from the E. coli chromosome is precise,27 no intact copy of the transposon was found in N. meningitidis. Deletions within transposons have been reported, in particular in N. meningitidis, N. gonorrhoeae,28 and commensal neisseria species.29 In the chloramphenicol-resistant strains, there were large deletions in Tn4451, accounting for approximately 80 percent of the transposon, that included the genes responsible for excision and made the transposon immobile.
The initial acquisition of chloramphenicol resistance by N. meningitidis could have occurred by transformation or conjugation, raising the possibility of a transfer of genetic information from a strict anaerobe to a strict aerobe. For conjugation, a meeting point between C. perfringens and N. meningitidis is not mandatory, since — although it was first detected in the former species — a Tn4451-related element could also be present in an aerobic–anaerobic bacterium that acted as a donor. Unlike conjugation, transformation does not require physical contact between the donor and recipient cells and can therefore occur between microorganisms in different ecosystems.
Even if these two possible mechanisms cannot be inferred with certainty from the primary transfer event, the subsequent spread of chloramphenicol resistance in N. meningitidis strains can be envisioned, given the extreme transformability of this species. Transformation has been shown to be responsible for the dissemination of other genetic traits30,31 in this naturally competent species and can easily be reproduced under laboratory conditions. As mentioned above, because of deletions, the truncated transposon is a stable component of the N. meningitidis genome. Resistance can therefore easily be transferred by transformation and integration of the incoming DNA through homologous recombination between the flanking regions and the corresponding portion of the chromosome of the recipient and insertion of the heterologous catP gene. This process would account for the observation that an identical fragment of foreign DNA was inserted at the same site in all the transformants.
The emergence of high-level chloramphenicol resistance in N. meningitidis isolates is of great concern, since intramuscular administration of chloramphenicol in oil is the standard treatment for meningococcal meningitis in developing countries.
Supported in part by a Bristol-Myers Squibb Unrestricted Biomedical Research Grant in Infectious Diseases.
Source Information
From the National Reference Center for Antibiotics and the Unité des Agents Antibactériens (M. Galimand, G.G., P.C.) and the National Reference Center for Meningococci (M. Guibourdenche, J.-Y.R.), Institut Pasteur, Paris.
Address reprint requests to Dr. Galimand at the Unité des Agents Antibactériens, Institut Pasteur, 28 Rue du Dr. Roux, 75724 Paris CEDEX 15, France.
References
3 in the pathogenic Neisseriae: a genetic tool for mutagenesis. J Bacteriol 1991;173:2147-2154.
| |||||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
