If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Faculdade de Farmácia, Centro de Patogénese Molecular, URIA, Universidade de LisboaPublic Health Department, Public Health Laboratory: Mycobacteriology/Tuberculosis, Administração Regional de Saúde de Lisboa e Vale do Tejo, I.P.
Grupo de Micobactérias, Unidade de Microbiologia Médica, Instituto de Higiene e Medicina Tropical, Universidade Nova de Lisboa (IHMT/UNL), LisboaFaculdade de Ciências e Tecnologia, Centro de Recursos Microbiológicos (CREM), Universidade Nova de Lisboa, Caparica
Development of streptomycin resistance in Mycobacterium tuberculosis is usually associated with mutations in rpsL and rrs genes, although up to 50% of clinical streptomycin-resistant isolates may present no mutation in either of these genes. In the present report we investigate the role of gidB gene mutations in streptomycin resistance. We have analyzed 52 streptomycin-resistant and 30 streptomycin-susceptible Mycobacterium tuberculosis clinical isolates by sequencing and endonuclease analysis of the gidB and rpsL genes. All clinical isolates were genotyped by 12-loci MIRU-VNTR. The gidB gene of 18 streptomycin-resistant isolates was sequenced and four missense mutations were found: F12L (1/18), L16R (18/18), A80P (4/18) and S100F (18/18). The remaining isolates were screened by endonuclease analysis for mutations A80P in the gidB gene and K43R in the rpsL gene. Overall, mutation A80P in the gidB gene was found in eight streptomycin-resistant isolates and 11 streptomycin-susceptible multidrug-resistant isolates. Also noteworthy, is the fact that gidB mutations were only present in isolates without rpsL and rrs mutations, all from genetic cluster Q1. Streptomycin quantitative drug susceptibility testing showed that isolates carrying the gidB A80P mutation were streptomycin intermediate-level resistant and that standard drug susceptibility testing yielded inconsistent results, probably due to borderline resistance. We conclude that gidB mutations may explain the high number of streptomycin-resistant strains with no mutation in rpsL or rrs. These mutations might occasionally confer low-level streptomycin resistance that will go undetected in standard susceptibility testing.
Streptomycin is an aminoglycoside antibiotic, the first that was introduced in the treatment of tuberculosis (TB), revolutionizing the treatment of this disease. Although mostly regarded as a tuberculostatic agent, streptomycin is no longer a first option in standardized TB treatments for new cases. It is still incorporated in first-line therapeutic regimens for TB in patients who have previously been treated for TB (former World Health Organization, WHO, Category II “retreatment” regimen). In some countries it is incorporated in second-line regimens for drug-resistant TB, although it's not the first choice for injectable aminoglycosides given the high rates of streptomycin resistance in drug-resistant TB [
In spite of this, due to the global emergence of resistant Mycobacterium tuberculosis (M. tuberculosis) strains, such as multidrug-resistant (MDR) and extensively drug-resistant (XDR) strains, streptomycin may regain an important role in the management of TB.
Portugal has an intermediate incidence rate when compared with other countries of the European Union, with 2398 new cases (22.5 new cases per 100 000 habitants) in 2010, but a worrying situation regarding M/XDR-TB [
]. The latest TB surveillance data available and published by the European Centre for Disease Prevention and Control for streptomycin resistance reported, in 2008, a 9.5% prevalence of streptomycin resistance (156 cases out of 1641 cases with DST results), which exceeded other drug resistance prevalences, such as that of isoniazid (7.4%) and rifampicin (1.8%) [
Streptomycin inhibits protein synthesis through irreversible binding to the A-site of the 30S ribosomal subunit and to the ribosomal protein S12. Ribosomes bound to streptomycin are unable to initiate and elongate during the process of gene translation. Streptomycin also prevents dissociation of ribosomal 50S and 30S subunits [
]. Mutations conferring streptomycin resistance in rrs have also been described, usually in the 530 and 912 loop regions, and because M. tuberculosis only possesses one copy of the rRNA operon, one mutation in this gene is enough to cause resistance [
]. The gene product of gidB is involved in the methylation of 16S rRNA, more specifically at position G527. As binding of streptomycin to 16S rRNA is crucial to its action mechanism, mutations in gidB may well be responsible and explain streptomycin resistance when no mutation is found on rpsL or rrs [
]. Notably, in the streptomycin-resistant strains isolated in our laboratory, rrs mutations that confer streptomycin resistance were never detected in the streptomycin-resistant strains lacking rpsL mutations [
The gidB gene is highly conserved in all eubacteria, including Mycoplasma genitalium, and loss of function has been implicated in high-level streptomycin resistance in Salmonella but with low-level resistance in M. tuberculosis [
In this study, we screened 52 streptomycin-resistant and 30 streptomycin-susceptible isolates circulating in the Lisbon Health Region for mutations in the gidB gene and its association with streptomycin resistance.
Materials and Methods
Two sets of isolates were analyzed in the present study, the first comprised a total of 82 M. tuberculosis isolates, 52 streptomycin-resistant isolates and 30 that were streptomycin susceptible (14 multidrug resistant and 16 susceptible to all first-line antibiotics), recovered from the same number of patients between 2004 and 2006. For this set of isolates, only DNA was available. All isolates were collected from hospitals and public health laboratories in the Lisbon Health Region and first-line drug susceptibility testing was carried out as described previously [
]. The subset of 18 isolates initially screened in this study comprised the following isolates: BK26, BK30, BK81, BK87, BK103, BK105, BK120, BK125, BK127, BK128, BK130, BK178, BK186, BK232, BK272, BK293, BK341 and BK353.
The second set of isolates comprised six isolates, three streptomycin-resistant isolates and three that were streptomycin susceptible, also from hospitals and public health laboratories in the Lisbon Health Region. This set was included in the study to provide data on streptomycin semi-quantitative drug susceptibility testing (qDST), in particular for Q1 strains.
Streptomycin susceptibility testing
All isolates were tested for streptomycin susceptibility by the fluorimetric BACTECtm MGITtm 960 system (Becton Dickinson Diagnostic Systems, Sparks, MD, USA) using a critical streptomycin concentration of 1.0 mg/L, according to the manufacturer's instructions.
Streptomycin qDST was carried out for the selected isolates using three streptomycin concentrations (1.0, 4.0 and 20.0 mg/L) with the MGIT 960 system and the Epicenter V5.80A software equipped with the TB eXIST module (Becton Dickinson Diagnostic Systems®) as previously described [
The entire open reading frame (ORF) of gidB, plus 96 nucleotides upstream and 56 nucleotides downstream, was amplified by PCR, resulting in an 828 bp fragment using forward oligonucleotide primer gidBF1 and reverse oligonucleotide primer gidbR1 (Table 1). Cycling conditions for this amplification consisted of an initial denaturation step at 94°C for 10 min, 40 cycles of a denaturation step at 94°C for 1 min, a primer annealing step at 67.5°C for 35 s and an extension step at 72°C for 1 min; a final extension step at 72°C for 10 min was performed.
TABLE 1Oligonucleotides used for amplification and sequencing of the rpsL and gidB genes
The entire rpsL ORF, plus 46 nucleotides upstream and 86 nucleotides downstream, was also amplified by PCR, resulting in a 504 bp fragment using forward oligonucleotide primer RPSL-1 and reverse oligonucleotide primer RPSL-2 (Table 1) [
]. Cycling conditions consisted of 40 cycles of a denaturation step at 94°C for 1 min, a primer annealing step at 60°C for 2 min and an extension step at 72°C for 2 min.
All PCR reactions were carried out using approximately 50 ng of template DNA, 10 pmol of each oligonucleotide, a final concentration of 1.5 mM of MgCl2, 0.2 mM of dNTPs and 0.2 U of AmpliTaq™ DNA Polymerase (Applied Biosystems, Foster City, CA, USA) in a T3 Thermocycler (Biometra, Goettingen, Germany). Newly described primers were designed based on the M. tuberculosis H37Rv genome (GenBank accession no. AL123456.3).
Amplicons were purified with Wizard® SV Gel and the PCR Clean-Up System (Promega, Madison, WI, USA) prior to sequencing. Sequencing reactions were performed with the BigDye Terminator Cycle Sequencing Kit with AmpliTaqTM DNA polymerase (Applied Biosystems) using oligonucleotide primers: RPSL-1 for rpsL amplicons, RRS-1 and RRS-2 for rrs amplicons and gidBF1 and gidBR1 for gidB amplicons (Table 1). Cycle conditions consisted of an initial denaturation step of 30 s at 96°C followed by 25 cycles of 10 s at 96°C, 5 s at 50°C and 4 min at 60°C.
Sequence data were aligned with the respective wild-type alleles from M. tuberculosis H37Rv (GenBank accession no. AL123456.2), using the CLC Sequence Viewer (CLC Bio, Aarhus, Denmark) and analysed using BioEdit (v.18.104.22.168, T.A. Hall).
Mutation A128G in rpsL was detected by MboII (New England Biolabs, Ipswich, MA, USA) digestion of the rpsL amplicon. Two micrograms of PCR product were digested with 5 U of MboII according to the manufacturer's instructions for 1 h at 37°C. Absence of restriction indicates the presence of mutation A128G.
Mutation G238C in gidB was detected as above using the BsrDI enzyme (New England Biolabs). Two micrograms of PCR product were digested with 1U of BsrDI according to the manufacturer's instructions for 1 h at 37°C. The occurrence of restriction indicates the presence of mutation G238C, yielding two fragments (501 + 329 bp).
All isolates were genotyped by 12-loci mycobacterial interspersed repetitive unit – variable number of tandem repeats (MIRU-VNTR) as described by Supply et al. [
]. A dendrogram was constructed with the MIRU-VNTRplus web application (available at http://www.miru-vntrplus.org) using the Dsw measure of genetic distance and the unweighted group method with arithmetic average (UPGMA). A cluster was defined as a group of more than one isolate sharing the same MIRU-VNTR profile.
Screening a set of 18 streptomycin-resistant isolates for gidB mutations we have found four different missense mutations: F12L, L16R, A80P and S100F. Mutations L16R and S100F were present in all isolates tested. When analysing the gidB homologous genes from both M. tuberculosis CDC1551 and F11 we found mutation S100F to be present on both strains and that M. tuberculosis F11 also had the L16R mutation. These two mutations were therefore considered natural polymorphisms and were not considered as mutations associated with streptomycin resistance. GidB A80P and F12L mutations were detected in four isolates and one isolate, respectively.
The A80P mutation was therefore considered the most prevalent mutation occurring in the gidB gene and a method to rapidly screen for G238C (A80P) mutation in gidB through PCR-RFLP was devised through amplicon digestion with BsrDI. Thirty-four additional streptomycin-resistant isolates and 30 streptomycin-susceptible isolates were characterized in this manner. The results obtained for each isolate were compared with the characterization of the rpsL gene.
Overall, the GidB A80P mutation was detected in 19 out of 82 isolates, eight streptomycin-resistant isolates and 11 that were streptomycin susceptible. Only MDR isolates were found to bear the A80P mutation, whereas the 16 pansusceptible isolates included in the study did not harbour this mutation (Fig. 1) . Also noteworthy, we have sequenced the rpsL and rrs genes of all strains containing the GidB A80P and all lacked mutations in both genes.
To examine the distribution of this prevalent mutation and in an attempt to associate this specific mutation with any genetic cluster, family or clade, all isolates were genotyped by 12-loci MIRU-VNTR. We found that all isolates carrying the GidB A80P mutation belonged to genetic cluster Q1 (Fig. 1). The occurrence of the GidB A80P mutation in all Q1 isolates and the fact that this mutation occurs among both streptomycin-resistant and susceptible isolates may suggest that this mutation has no contribution to streptomycin resistance but may rather be, from a resistance standpoint, a neutral phylogenetic polymorphism characteristic of Q1 isolates. Alternatively, this mutation might constitute an adaptation mechanism acquired by an ancestral Q1 isolate that yielded clinically significant but borderline streptomycin resistance on the standard MGIT drug susceptibility test.
To further elucidate the contribution of the GidB A80P mutation to streptomycin resistance, we decided to study the streptomycin qDST profiles in a second set of six M. tuberculosis clinical isolates composed of four Q1 isolates carrying the GidB A80P mutation, one Beijing strain carrying the rpsL K43R mutation and a pansusceptible isolate with no mutation in gidB, rpsL or rrs. The results obtained show that the four Q1 isolates were all resistant to 1 mg/L, and three of them were resistant to 4 mg/L, while the remaining isolate displayed an intermediate resistance to 4 mg/L (Table 2). All Q1 isolates were susceptible to 20 mg/L. The Beijing strain with the K43R mutation on the rpsL gene was resistant to the three tested streptomycin concentrations whereas the pansusceptible clinical strain and the control strain H37Rv were found to be susceptible to the three tested concentrations (Table 2). The results obtained show that the Q1 isolates carrying the GidB A80P mutation display an increased streptomycin resistance-level towards an intermediate-level resistance (Table 2). Three Q1 isolates with the GidB A80P mutation (two streptomycin susceptible and one resistant) were also blindly retested in duplicate, using standard DST methodology on BACTEC 960 MGIT with a streptomycin breakpoint of 1 mg/L. Surprisingly, the two isolates (IHMT308 and 361) initially considered as streptomycin susceptible were considered resistant upon retesting (Table 2).
TABLE 2Streptomycin qDST for selected strains carrying the GidB A80P mutation
In the present study, we have analysed the gidB gene, encoding an rRNA methyltransferase, in streptomycin-resistant and susceptible isolates. Methyltransferases have demonstrated a growing importance in the action mechanism of and resistance pathways to several antimicrobial agents in M. tuberculosis. Intrinsic resistance to macrolides is thought to be associated with 23S rRNA methyltransferases encoded by the erm genes whereas loss of TlyA methyltransferase confers resistance to cyclic peptides (e.g. capreomycin). However, most rRNA methyltransferases are still unknown in M. tuberculosis [
The screening of gidB mutations revealed that besides two naturally occurring polymorphisms (L16R and S100F), two different missense mutations were found, F12L and A80P, the latter being the most prevalent. Both mutations had never been described before and were only detected in isolates without rpsL and rrs mutations. As a result, gidB mutations may account for streptomycin resistance in isolates with both wild-type rpsL and rrs. Nevertheless, the A80P mutation was detected in MDR isolates susceptible to streptomycin but not in pansusceptible isolates.
The A80P mutation has not, to our knowledge, been previously described and bioinformatic analysis revealed that this mutation occurs in the alpha-helix D close to the S-adenosyl-L-methionine binding consensus sequence, according to homology analysis with E. coli GidB and respective crystal structure [
]. Moreover, this substitution does not appear to be chemically conservative and may eventually hamper the functionality of GidB in methylating the 16S rRNA, leading to an increase in the streptomycin resistance level [
]. Streptomycin qDST revealed that the majority of tested strains carrying the A80P mutation exhibit an intermediate streptomycin resistance level (i.e. either resistant or intermediate resistance to 4 mg/L but susceptible to 20 mg/L), which may correlate with the streptomycin low-level resistance reported by others relative to other gidB mutations in M. tuberculosis clinical isolates [
]. It is possible that strains carrying this mutation may occasionally yield inconsistent DST results as demonstrated by the streptomycin retesting of selected isolates. In the present study the use of the novel qDST methodology takes into account the growth rate in both drug-containing and drug-free media to establish intermediate resistance levels relative to a specific concentration, rather than just resistant or susceptible as in the traditional MGIT methodology [
]. Although the present study could benefit from having a larger sample of isolates subjected to the qDST assay, the fact that all four Q1 isolates showed a consistent resistance at 4 mg/L demonstrates that these results are representative of Q1 strains.
Our results indicate that a mutation in gidB may be enough to confer clinically relevant streptomycin resistance, although we do not discard a possible contribution of efflux mechanisms to the overall streptomycin resistance level, as proposed by Spies et al. [
] when studying the synergistic effect between gidB mutations and efflux pump inhibitors.
The association of the GidB A80P mutation with the genetic cluster Q1, a highly prevalent cluster of M/XDR-TB isolates in Lisbon, Portugal, is important because it allows the use of this mutation as a surrogate marker for Q1 isolates and can be useful for its detection. The occurrence of the A80P mutation in MDR isolates is therefore only linked with the high association of the Q1 cluster with MDR, as fully susceptible Q1 isolates have never been found and described. The detection of these isolates may be useful in managing M/XDR-TB in the region because early detection of Q1 isolates would prove valuable for the design of adequate treatment regimens and patient isolation [
Other mutations may be useful for assignment to M. tuberculosis lineages. For example, the L16R substitution is associated with isolates of Latin American Mediterranean (LAM) lineage, which include the M. tuberculosis F11 strain [
]. In fact, approximately 50% of the strains circulating in the Lisbon Health Region belong to the LAM lineage, of which 1.67% represent Q1 isolates belonging to the LAM4 sub-lineage (J. Perdigão, I. Portugal, unpublished data) [
] and from the updated version of M. tuberculosis H37Rv (Genbank accession no. AL123456.3) confirm that this mutation is due to a sequencing error present in the previous release (Genbank accession no. AL123456.2).
Given the high resistance rates in this and other settings, the correct assessment of streptomycin resistance level and eventual association with a molecular marker can be useful in the decision of whether streptomycin should be used in some therapeutic regimens.
Nevertheless, it is also important to stress that the addition of streptomycin alone to a treatment regimen, such as the former WHO-recommended Category II regimen, is counterproductive as it may lead to the acquisition of streptomycin resistance and amplification of resistance to other drugs in use [
] showed that the cure rate was lower for the WHO Category II regimen compared with the Category I regimen. According to the authors, this should support the phasing-out of the WHO Category II regimen. Additional data from other countries seem to also support this notion [
In conclusion, our results suggest that the cluster-specific gidB polymorphism A80P is responsible for an intermediate level of streptomycin resistance, which has implications for routine streptomycin DST testing as currently performed using a single critical concentration, streptomycin prescription and M/XDR-TB detection. Our data further contribute to global awareness of the role of gidB mutations in the process of streptomycin resistance development.
This work was partially supported by Project Ref. SDH49: ‘Early Molecular Detection of M/XDRTB in the Great Lisbon Healthcare Region’ from Fundação Calouste Gulbenkian (FCG, Portugal). J. Perdigão, D. Machado and C. Silva were supported by FCT grants SFRH/BD/45388/2008, SFRH/BD/65060/2009 and SFRH/BD/73579/2010, respectively.
The authors declare no conflicts of interest.
Standardized treatment of active tuberculosis in patients with previous treatment and/or with mono-resistance to isoniazid: a systematic review and meta-analysis.