1. Introduction
One of the greatest challenges in the fight against tuberculosis (TB) has been the emergence and spread of drug-resistant (DR) and multidrug-resistant (MDR) strains of Mycobacterium tuberculosis. The development of new molecular techniques targeting specific molecular mutations associated with drug resistance creates a valuable adjunct to conventional drug susceptibility testing (DST) for M. tuberculosis. These techniques can be performed directly on clinical samples without a culturing step and thus allowing a reliable diagnosis of drug-resistant TB to be achieved as fast as within a 24-hour period.
Ethambutol (EMB), an arabinose analogue, is a bacteriostatic, antimycobacterial drug, which has been used for the treatment of TB since the mid-1960s. The drug is routinely recommended for the intensive phase of TB therapy, as part of a four-drug regimen, including isoniazid (INH), rifampicin (RMP), and pyrazinamide (PZA) [1]. Disturbingly, almost 4% of all M. tuberculosis clinical isolates have been shown to display resistance to EMB [2].
Ethambutol appears to target the cell wall of tubercle bacilli through interfering with arabinosyl transferases, encoded by the embCAB operon, comprised of three homologous genes, designated embC, embA, and embB, and involved in the biosynthesis of arabinogalactan and lipoarabinomannan, the key structural components of the mycobacterial cell wall. The proposed scenario of EMB action on M. tuberculosis is that upon interaction with the EmbCAB proteins EMB inhibits the arabinan synthesis leading to a lack of arabinan receptors for mycolic acids and accumulation of mycolic acids results in cell death [3]. Resistance to EMB has repeatedly been associated with alterations in the embB gene, particularly in embB codon 306, referred to as EMB resistance determining region (ERDR). Sequence analysis of the ERDR has been considered a rapid screening tool for detection of resistance to EMB [4–6]. Several allelic exchange studies have demonstrated that mutations in codons embB306, embB406, and embB497 are responsible for low and moderate levels of EMB resistance [7, 8]. However, this correlation is uncertain because all these codons have also been found mutated in isolates susceptible to EMB [9–12].
The aim of this study was to examine mutational “hot spots” in the embB gene, including the ERDR region, among MDR M. tuberculosis clinical isolates from Poland and to find a possible association between embB mutations and resistance to EMB.
Part of the results of this study was presented as a poster (A-527-0001-03736) at the 5th Congress of European Microbiologists (FEMS 2013), Leipzig, Germany, July 21–25, 2013.
4. Discussion
Although EMB has been used for the treatment of TB for over 40 years, the molecular mechanisms of EMB resistance still remain poorly understood. Previous studies have correlated the EMB resistance phenotype with mutations in genes of the embCAB operon, most notably in the embB gene. Mutations at codon position 306 of the embB gene have been found to occur most frequently. The high detection rates of mutations at codon embB306 among EMB-resistant M. tuberculosis isolates were reported from Korea (47%) [16], China (55%) [17], Cuba and the Dominican Republic (70%) [12] and also from countries neighboring to Poland, such as Russia (48%) [18] and Germany (68%) [5]. The role of the embB306 alterations in creating resistance of tubercle bacilli to EMB was confirmed by allelic exchange mutagenesis [19]. Additionally, strains with the Met306Ile substitution were found to have lower MICs of EMB (20 μg/mL) than strains with Met306Val or Met306Leu replacements (40 μg/mL) [20]. However, a number of reports have indicated that only less than 35% of EMB-resistant M. tuberculosis isolates harbored mutations in the embB codon 306 [21, 22]. Moreover, mutations in codon embB306 have been found also in M. tuberculosis strains susceptible to EMB, and the frequencies of those mutations in EMB-susceptible strains approached those among EMB-resistant strains [9, 11].
The results of this study are in line with previous findings, showing mutations in embB306 codon to predominate (40% of all MDR-TB isolates) and to occur at higher frequency in EMB-resistant than EMB-susceptible isolates (53% versus 33%).
Studies on EMB resistance showed that mutations in the embBAC gene cluster, outside embB codon 306, do occur but are quite rare. Only two other substitutions, found in embB codons 406 and 497, have been consistently associated with EMB resistance. The percentage of embB406 mutations among EMB-resistant isolates is rather low, usually not exceeding 10%, whereas mutations in codon embB497 are twice as frequent [12, 17, 21].
Allelic exchange experiments performed at codons 406 and 497 of the embB gene have concluded that point mutations at these codons only slightly increase resistance to EMB [7]. Interestingly, mutations in codons 406 and 497 have—similarly to mutations in codon 306—been identified also in M. tuberculosis strains susceptible to EMB [17, 23]. In our study mutations at embB406 were found exclusively in 4 (23%) EMB-resistant strains, whereas a single mutation at embB497 was found only in an EMB-susceptible isolate. Mutations at codons other than 306, 406, and 497 were identified only in two EMB-susceptible and one EMB-resistant isolates.
Three novel mutations in the examined fragment of the embB gene were observed in this study. The sequence variations in codons 504 and 507 have already been described before in EMB-resistant isolates, yet the amino acid replacements were different from those observed here [24]. The substitution at codon 413 is reported for the first time. Of the three new mutations described in this work, only that in codon 507 may have an impact on EMB resistance, since it was found in an EMB-resistant isolate. Yet, the extent of this impact was masked by the cooccurrence of the emb306 change in that isolate.
More than a half (54%) of MDR-TB clinical isolates tested had mutations in the examined region of the embB gene. These mutations occurred nearly twice as frequently in EMB-resistant than EMB-susceptible isolates (76.5% versus 42%).
The high frequency of embB mutations with no association between the presence of mutation and EMB-resistant phenotype can be explained by the fact that mutations in the embB gene occur significantly more frequently in MDR than EMB-monoresistant strains [9, 25, 26]. Several studies have demonstrated a strong association between embB306 mutations and resistance to INH or RMP, or MDR phenotype [19, 25, 26]. It has been suggested that embB306 mutations may have selective advantage upon treatment with multiple drugs. In other words, these mutations inhibit the synergistic effect of anti-TB drugs when used in combination [19]. The molecular mechanism behind this phenomenon can only be speculated and may involve changes in the cell wall permeability as a result of embB306 mutations [19].
Another possible explanation for the lack of correlation between the embB gene alterations and EMB resistance may relate to a cumulative effect of multiple mutations on the development of EMB resistance. Acquisition of resistance to EMB is thought to be a gradual process that may involve numerous genes [3, 27]. Strains bearing embB mutations are susceptible to EMB because these mutations alone are not sufficient to generate EMB resistance unless accompanied by alterations in other genetic loci. Recently, Safi et al. have shown that mutations in the embB, embC, Rv3806c, and Rv3792 genes, involved in the decaprenylphosphoryl-β-D-arabinose (DPA) biosynthetic and utilization pathway, produce a wide range of ethambutol MICs by interacting in different ways and that the acquisition of EMB resistance does not occur in a single step but requires a multistep process [28].
Finally, conclusions concerning EMB resistance can be inaccurate because of the false-negative DST results, and thus importance of mutations in the embB gene can be underestimated. Quite often, the MIC values for EMB have varied depending upon culture medium, strain condition, or the DST method used [29]. The results of previous studies have shown that EMB resistance can indeed be phenotypically missed by routine laboratory procedures [30, 31].