Genetic and Biochemical Studies on Drug-Resistant Mutants in Mycobacterium smegmatis

Transcription

1 Japan. J. Microbiol. Vol. 18(6), , 1974 Genetic and Biochemical Studies on Drug-Resistant Mutants in Mycobacterium smegmatis Yasuo MIZUGUCHI, Kiyoko SUGA, Kunitsugu MASUDA, and Takeshi YAMADA Department of Tuberculosis, National Institute of Health, Tokyo, and Research Institute for Microbial Diseases, Osaka University, Suita-shi (Received for publication, June 20, 1974) ABSTRACT Genetic analyses of the loci conferring resistance to antibiotics known to affect protein synthesis were made employing the conjugation system of Mycobacterium smegmatis. By the complementation tests, vic (viomycin-capreomycin-resistant) mutants were classified into two different groups; cistrons A and B. Neomycin-kanamyein-resistant (nek) mutants, on the contrary, fall into a single cistron. Erythromycin-resistance (ery) locus was linked to vic-nek region. The map order deduced was argb, arga, (vica, vicb), nek, ery. Biochemical studies showed that strains resistant to streptomycin, neomycinkanamycin or viomycin-capreomycin had altered ribosomes. In the previous report, we analyzed the linkage relationships among the markers for resistance to streptomycin (str), neomycinkanamycin (nek), and viomycin-capreomycin (vic) in Mycobacterium smegmatis [4]. Since these markers were presumed to define genes for ribosomal components, and known to form a cluster in a region of chromosome in Escherichia coli [1, 2] and Bacillus subtilis [3], linkage relationships among these loci in mycobacteria were of particular interest from the phylogenetical point of view. The results showed that nek and vic loci were closely linked with each other and also linked to the genes for the arginine biosynthesis (arga and argb). No linkage was observed, however, between str and nek or vic. We also showed that in certain recombinant types, nek and vic alleles formed stable heterogenotes. Accordingly, it is expected that complementation tests are possible between the different vic or nek mutations, because vicr or nekr is recessive to vics or neks alleles [4]. Requests for reprints should be addressed to Dr. Yasuo Mizuguchi, Department of Tuberculosis, National Institute of Health, Kamiosaki, Shinagawaku, Tokyo 141, Japan. In this paper, we report the results of the complementation tests of vic and nek genes, and the mapping of erythromycin resistance locus (ery), We also report the results of biochemical analysis of ribosomes derived from these mutants. MATERIALS AND METHODS Media. Nutrient broth and nutrient agar used for the experiment were described previously [5]. As a minimal medium, modified Karlsson's medium with following composition was used: (NH4)2SO4, 0.1%; MgSO4, 0.01%; CaCl2, 0.001%; KH2PO4, 0.15%; K2HPO4, 0.4%; glycerol, 2%; and agar 1.5%. Bacterial strains. M. smegmatis strain Rabinowitchi (R) and strain PM5 (P) were employed. Mutant substrains employed are shown in Table 1. Isolation of drug-resistant mutants. Isolation methods for vic and nek mutants were reported previously [4]. Spontaneous one-step erythromycin-resistant mutants were isolated by inoculating bacterial cells on complete agar plates containing erythromycin (500 Đg/ml). Determination of minimum inhibitory concentration (MIC). The method for the determina- 457

2 458 Y. MIZUGUCHI, K. SUGA, K. MASUDA AND T. YAMADA Table 1. Mutant strains employed Table 2. Characteristics of vic mutants a) -, Changes in MIC from the parental strains were less than a factor 2. +, Elevation in MIC from the parental strains were 2 to 8 times. ++, Change in MIC was greater than 25-fold from the parental strain. met+argb+ recombinant colonies were transferred to the masterplates and incubated for 2 days at 37 C, and then replicated onto agar plates containing 20 ƒêg/ml of viomycin or 10 ƒêg/ml of kanamycin. Growth was checked every 24 hr, and when colonial a) Strains of Rabinowitchi are prefixed R and those of PM5 are prefixed P. b) Markers are designated by gene letter or allele number when known, and abbreviations used for nutritional requirements are: arg, arginine; met, methionine; leu, leucine; his, histidine. For resistance markers: str, streptomycin; nek, neomycin-kanamycin; vic, viomycin-capreomycin; erv, erythromycin. tion of MIC was described previously [4]. Mating procedure and segregation analysis. The indirect plating method [5] was used to obtain recombinants. Analysis of the recombinants was carried out by the replica method described previously [5]. Complementation test. Crosses were carried out between drug-resistant mutants of R (met) and drug-resistant mutants of P (leu, his, argb). In one case, a drug-resistant mutant of R was crossed with a drug-resistant recombinant obtained from a cross of R by P. Heterogenomic recombinants in vic or nek allele could be obtained among met+ argb+ recombinants with high frequency [4]. The growth was observed not after 24 hr but after 72 to 96 hr, it was regarded that the mutants employed for the cross belonged to different complementation groups. Ribosomes and cell-free system. The method for the isolation of ribosomes and supernatant fluid from mycobacteria and the method for polyphenylalanine synthesis directed by polyuridylic acid (poly U) in the presence or absence of antibiotics were reported previously [6]. RESULTS Characteristics of vic Mutants As reported previously [4], mutants resistant to viomycin and capreomycin were divided into two groups by the cross resistance relationship to kanamycin; those which showed no or very little coresistance (vic-1, vic-12, vic-15) and those associated with coresistance in significant levels (vic-3, vic-4, vic-5, vic-11, vic-13, vic-14). Mutants belonged to the former group were resistant to 300 ƒêg/ml of viomycin and grew rapidly on the medium containing 20ƒÊg/ml of viomycin. On the contrary, mutants belonged to the latter group showed slightly lower level of

3 DRUG RESISTANCE IN MYCOBACTERIA 459 viomycin resistance. They were resistant to ƒÊg/ml of viomycin and the growth of these mutants on the medium containing viomycin was rather slow. One mutant (vic-17) was isolated from UV-irradiated P-22, and this mutant was resistant to 300 ƒê g/ml of viomycin but showed highest cross resistance to kanamycin (MIC; 100 ƒêg/ml). Characteristics of the mutants are shown in Table 2. recombinants replicated on agar plates containing viomycin could grow within 24 hr, indicating that vic-1 and vic-12 belonged to the same cistron. Crosses between vic-3 and viv-11, and vic-3 and vic-17 also did not produce viomycin-sensitive heterogenotes. When argb+met+ recombinants obtained from a cross of vic-1 and vic-11 were replicated on the medium containing viomycin, growth of 254 recombinants out of 321 was not visible after Complementation of vic Mutants Viomycin-capreomycin-resistant mutants of R (met) and P (leu, his, argb) were crossed by the indirect plating method and met+argb+ recombinants were selected. As reported previously [4], about 60% of the recombinants became stable heterogenotes in vic allele. It was easy to detect drug-sensitive heterogenomic recombinants by their growth characteristics on inhibitor-containing medium. When recombinants were replicated on the medium containing viomycin, growth of vicr haploid recombinants could be visible after 24 hr of incubation. If the mutants used for the cross belonged to the same cistron, growth of the heterogenomic recombinants in vic allele was also visible after 24 hr, because they were resistant to the drug. The vics haploid recombinants did not grow even after 4 to 5 days of incubation. On the contrary, when mutants belonging to different cistrons were employed for the cross, they produced drug-sensitive heterogenotes. Since these heterogenotes produced drug-resistant segregants at low frequencies, colonial growth due to vicr segregants became visible not after 24 hr but after 72 to 96 hr of incubation. As shown in Table 3, viomycin-sensitive heterogenotes did not appear in a cross between vic-1 and vic-12; all of 247 argb+met+ Table 3. Complementation test between vic mutantsa) 24 hr, but became visible after 72 hr of incubation, suggesting that they were drugsensitive heterogenotes. We confirmed that the slower growth of these recombinants on the plates containing viomycin was not due to prolonged generation time but to growth of vicr haploid segregants which were produced at a low frequency. The results indicated that vic-1 and vic-11 belonged to different cistrons. Similarly, vic-3 belonged to different cistron from vic-12. From a cross between vic-1 and vic-17, both of them were resistant to 300 ƒêg/ml of viomycin, viomycinsensitive heterogenotes were also obtained. Thus, all of the mutants were divided into two groups as shown in Table 4; cistron A contains vic-1, vic-12 and vic-15, and cistron B, vic-3, vic-4, vic-5, vic-11, vic-13, vic-14 and vic-17. Characteristics of nek Mutants All of the 11 nek mutants except one (nek-6) were resistant to 1000 ƒêg/ml of kanamycin and 500 ƒêg/ml of neomycin. These mutants were coresistant to viomycin; MIC of these mutants against viomycin was 2 to 4 times higher than that of parental strain. One mutant (nek-6) showed low level resistance to kanamycin (10 ƒêg/ml). This strain also showed coresistance to viomycin. When 97 argb+met+ recombinants obtained from a cross between R-47 (met-5, ery-1, nek-6) and P-22 (leu-11, his-13, argb-1) were analyzed for the segregation of nek-6 locus, we found that 67 recombinants (about 70%) were heterogenomic in nek-6 allele. This indi- Table 4. Complementation group of vic mutants a) Experimental procedures were described in the text. b) Number of drug-sensitive heterogenote/ Number of met+ argb+ recombinant tested.

4 460 Y. MIZUGUCHI, K. SUGA, K. MASUDA AND T. YAMADA Table 5. Genetic constitution of recombinants obtained from a cross between R-46 (met-5, ery-1) and P-31 (leu-11, his-13, argb-1, nek-12, str-23) cated that nek-6 was also linked to argb like other nek mutations [4]. were resistant to 1000 Đg/ml of erythromycin Complementation Test between nek Mutants Neomycin-kanamycin-resistant mutants and did not show cross resistance to streptomycin, kanamycin and viomycin. To investigate whether or not ery locus was linked to nek or vic, erythromycin-resistant strain were also subjected to complementation tests. No kanamycin-sensitive heterogenotes were found among argb+met+ recombinants obtained from crosses between nek mutants of strain R and strain P. Crosses between nek-6 and nek-12, nek-6 and nek-3 also did not produce kanamycin-sensitive heterogenotes. The results indicated that all nek mutations analyzed did not complement with each other. Mapping of ery Gene R-46 (met-5, ery-1) was crossed with P-3l (leu-11, his-13, argb-1, nek-12, str-23) and recombinants were analyzed for their unselected markers. Table 5 shows the results of the analysis. No linkage was observed among argb, leu, his and str loci. Percentage of linked transfer of ery-1 with argb+ in met+argb+ recombinants was 54 which was less frequent than linked transfer of nek allele (67%), but this value indicated close linkage of ery to argb locus. Erythromycin-sensitive heterogenotes were not detected among In E. coli [1], and B. subtilis [3], a locus met+argb+ recombinants. conferring resistance to erythromycin due to 50 S ribosomal alteration has been mapped within ribosomal gene cluster. To test whether or not this is the case in M. smegmatis, erythromycin-resistant mutants were isolated by plating large amount of cells on plates containing 500 Đg/ml of erythromycin. Frequencies of ery mutation were about 8.5 ~10-7 in the case of strain R and 5 ~10-8 in the case of strain P. Mutants isolated Three point crosses were carried out to determine the location of ery-1. When argb+nekr recombinants were selected, 129 out of 137 were sensitive to erythromycin. The results suggested that Model I shown in Table 6 was more likely than Model II. Similarly, most of eryrnekr recombinants (58 out of 65) obtained from the same cross were argb-. The results also satisfied the gene order of Model I. From the data together Table 6. Segregation of ery and argb loci in recombinants obtained from a cross between R-46 (met-5, ery-1) and P-31 (leu-11, his-13, argb-1, nek-12, str-23)

5 DRUG RESISTANCE IN MYCOBACTERIA 461 with the map order reported earlier [4], the gene order deduced is shown in Fig. 1. In Vitro Protein Synthesis To determine whether or not the strains used in our study had altered ribosomes, poly U-directed polyphenylalanine synthesis was tested in the presence or absence of antibiotics. As shown in Table 7, ribosomes obtained from R-16 (str-15) were resistant to the inhibition of polyphenylalanine synthesis by streptomycin, but they were sensitive to viomycin, kanamycin, and capreomycin. Ribosomes from R-33 (vic-3) were resistant to viomycin, capreomycin and showed coresistance to kanamycin, but the level of resistance to kanamycin was lower than that obtained from R-30 (nek-3). Thus, the cross resistance to kanamycin of vic-3 was also observed in in vitro system. Ribosomes of R-30 (nek-3) were resistant to the inhibition of polyphenylalanine synthesis by kanarnycin, Fig. 1. Linkage map of M. smegmatis. neomycin and showed coresistance to viomycin and capreomycin. DISCUSSION By the complementation tests between various mutants, we found that vic mutants were classified into two complementation groups; cistrons A and B. In general, the mutants belonging to cistron A showed higher level of resistance to viomycin and did not show cross resistance to kanamycin. All of the mutants belonging to cistron B showed cross resistance to kanamycin, and except vic-17, lower level of resistance to viomycin. Thus, correlation between complementation groups and cross resistance to kanamycin rather than the level of resistance to viomycin was observed among the vic mutants. Neomycin-kanamycin-resistant mutants, on the contrary, fall into a single cistron. This suggested that nek-6, although its resistance level to kanamycin was very low, might have altered ribosome like nek-3. We have reported that vic mutations occurred at a chromosomal site very closely linked to nek [4]. It is not yet known which cistron of the vic gene is closer to the nek. Further investigation is necessary to know the precise relationship among the mutations. Table 7. Antibiotic sensitivity of polyphenylalanine synthesis a) Experimental procedures were reported previously [6]. b) Not done.

6 462 Y. MIZUGUCHI, K. SUGA, K. MASUDA AND T. YAMADA Previous work in M. smegmatis strain ATCC has shown that mutants resistant to viomycin had either altered 30 S ribosomes or altered 50 S ribosomes [6]. Present studies in cell-free system showed that strain R-33 (vic-3) which belonged to the complementation group B, had altered ribosomes like viomycin-resistant mutants of M. smegmatis strain ATCC Experiments to elucidate the localization of resistance of R-33 on the ribosomal subunit and the biochemical changes occurred in a mutant of cistron A are in progress. Biochemical analysis revealed that the neomycin-kanamycin-resistant strain R-30 (nek-3) also had altered ribosomes. Although, further analysis has not yet been attempted, it seems probable that the resistance of R-30 to kanamycin localizes in 30 S subunit like nek mutants of E. coli [2]. It also seems certain that erythromycin-resistant mutant (ery-1) had altered ribosomes. The data obtained from the genetic analysis indicated that ery-1 was linked to vic-nek loci. This fact seems to support the above assumption, because in E. coli [1] and B. subtilis [3], ery due to 50 S ribosomal alteration has been mapped within ribosomal gene cluster. We reported that streptomycin-resistance marker (str-15) was not linked to the vic-nek segment [4]. Biochemical analysis showed, however, strain R-16 (str-15) also had altered ribosomes. Since the R-16 is resistant to more than 1000 Đg/ml of streptomycin, str-15 seems to correspond to the mutation of high level resistance to streptomycin of E. coli and B. subtilis (stra locus). If this is the case, unlike E. coli or B. subtilis, stra is not linked to neomycin-kanamycin locus in M. smegmatis. Attempts to isolate str mutants in which str locates close to vic-nek region have been unsuccessful. Cross resistance relationships among the drugs observed in in vivo, were also found in in vitro system. Further investigations on this subject together with the biochemical change occurred in an erythromycin-resistant mutant will be reported elsewhere. ACKNOWLEDGEMENT We thank Drs. T. Murohashi and M. Hori for their support throughout the study. REFERENCES [1] Apirion, D Three genes that affect Escherichia coli ribosomes. J. Mol. Biol. 30: [2] Apirion, D., and Schlessinger, D Mapping and complementation of three genes specifying 30 S ribosomal components in Escherichia coli. J. Bacteriol 96: [3] Harford, N., and Sueoka, N Chromosomal location of antibiotic resistance markers in Bacillusubtilis. J. Mol. Biol. 51: [4] Suga, K., and Mizuguchi, T Mapping of antibiotic resistance markers in Mycobacterium smegmatis. Japan. J. Microbiol. 18: [5] Tokunaga, T., Mizuguchi, Y., and Suga, K Genetic recombination in mycobacteria. J. Bacteriol. 113: [6] Yamada, T., Masuda, K., Shoji, K., and Hori, M Analysis of ribosomes from viomycinsensitive and -resistant strains of Mycobacterium smegmatis. J. Bacteriol. 112: 1-6.