• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of aacPermissionsJournals.ASM.orgJournalAAC ArticleJournal InfoAuthorsReviewers
Antimicrob Agents Chemother. Oct 2009; 53(10): 4138–4146.
Published online Jul 13, 2009. doi:  10.1128/AAC.00162-09
PMCID: PMC2764220

The Arabinosyltransferase EmbC Is Inhibited by Ethambutol in Mycobacterium tuberculosis[down-pointing small open triangle]


Ethambutol (EMB) is an antimycobacterial drug used extensively for the treatment of tuberculosis caused by Mycobacterium tuberculosis. EMB targets the biosynthesis of the cell wall, inhibiting the synthesis of both arabinogalactan and lipoarabinomannan (LAM), and is assumed to act via inhibition of three arabinosyltransferases: EmbA, EmbB, and EmbC. EmbA and EmbB are required for the synthesis of arabinogalactan, and at least one enzyme (M. tuberculosis EmbA [EmbAMt]) is essential in M. tuberculosis. EmbCMt is also essential for the viability of M. tuberculosis but is involved in the synthesis of LAM. We show that mutations in EmbCMt that reduce its arabinosyltransferase activity result in increased sensitivity to EMB and the production of smaller LAM species in M. tuberculosis. Overexpression of EmbCMt was not tolerated in M. tuberculosis, but overexpression of Mycobacterium smegmatis EmbC (EmbCMs) led to EMB resistance and the production of larger LAM species in M. tuberculosis. Treatment of wild-type M. tuberculosis strains with EMB led to inhibition of LAM synthesis, resulting in the production of smaller species of LAM. In contrast, no change in LAM production was seen in EMB-resistant strains. Overexpression of EmbBMs in M. tuberculosis also resulted in EMB resistance, but at a lower level than that caused by EmbCMs. Overexpression of EmbAMt in M. tuberculosis had no effect on EMB resistance. Thus, there is a direct correlation between EmbC activity and EMB resistance, as well as between EmbC activity and the size of the LAM species produced, confirming that EmbC is one of the cellular targets of EMB action.

Tuberculosis (TB), one of the oldest diseases known to humans, remains a major public health threat. It is estimated that one-third of the world's population are infected with the causative agent, Mycobacterium tuberculosis. The scale of the problem is increasing, and the disease is becoming deadlier as it intersects with the spread of human immunodeficiency virus. The emergence of multidrug-resistant strains establishes the urgent need to fully understand the mechanisms of drug resistance. Ethambutol (EMB) is a bacteriostatic, antimycobacterial drug first described in 1961 (31) and has been prescribed for TB treatment since 1966. EMB is used worldwide for TB therapy in combination with isoniazid, pyrazinamide, and rifampin (rifampicin). It is also effective in the treatment of opportunistic mycobacterial infections of patients with human immunodeficiency virus. Unfortunately, resistance to EMB has been reported in as many as 4% of clinical isolates of M. tuberculosis and is prevalent among multidrug-resistant strains (5).

The effects of EMB are highly pleiotropic, and over the past 40 years, many efforts have been made to understand its intracellular target(s). Kilburn and Greenberg (12) were the first to indicate that EMB treatment of Mycobacterium smegmatis caused rapid bacterial declumping, suggesting cell wall changes. It was subsequently demonstrated that EMB inhibited the transfer of mycolic acids to the cell wall (29), resulting in the accumulation of trehalose monomycolate, trehalose dimycolate, and free mycolic acids in the medium (13). This was later related to inhibition of the biosynthesis of the mycobacterial cell wall core polymer, arabinogalactan (AG), leading to a lack of arabinan receptors for the mycolic acids (29). The mycolyl-AG-peptidoglycan complex (in which AG is linked to peptidoglycan and the mycolic acids) is a major structural component of the cell wall, providing a hydrophobic permeability barrier. It was noted subsequently that AG is not the only cell wall arabinan affected by EMB and that EMB also inhibits the synthesis of the arabinan core of lipoarabinomannan (LAM) (6, 17), a key surface molecule in the host-pathogen interaction (16). In M. smegmatis, EMB exposure leads to the accumulation of decaprenol phosphoarabinose (DPA), the arabinosyl donor for arabinan biosynthesis, confirming the effect of EMB on arabinosyltransferases involved in AG and LAM production (32).

The genetic basis of resistance to EMB in mycobacteria has been the subject of much scrutiny. Initial work with M. smegmatis led to the identification of a cluster of genes from the related species Mycobacterium avium that conferred EMB resistance when overexpressed (3). The locus has been characterized; it consists of a regulator, EmbR, and two arabinosyltransferases, EmbA and EmbB (3). In M. tuberculosis and M. smegmatis, the emb locus encodes three arabinosyltransferases (EmbC, EmbA, and EmbB); the EmbR regulator is located elsewhere on the chromosome. The identification of point mutations in codons 289 and 292 of M. smegmatis embB that conferred EMB resistance (1, 15) led to a search for similar mutations in codons 303 and 306 of embB in EMB-resistant clinical isolates of M. tuberculosis (15, 30). While mutations have been associated with EMB resistance in several studies (1, 15, 26, 30), there is still some controversy about the role of these mutations, particularly those at codon 306, in mediating EMB resistance, since the most common mutations have been found in fully sensitive isolates (9, 14, 18, 25). Most recently, work using isogenic strains has once again suggested that these mutations are linked to resistance (23). While much attention has been focused on the role of EmbB in mediating resistance, studies have suggested that multiple molecular pathways to an EMB resistance phenotype exist and that these may be associated with mutations elsewhere in the genome (22, 27).

The precise mode of action of EMB is still unclear, although given that EmbB has been implicated in EMB resistance, it has been considered to be the primary target of EMB. However, several lines of evidence suggest that EmbB is not the only target and that the target(s) may differ between mycobacterial species. In particular, inhibition of EmbB in M. smegmatis cannot be the sole antibacterial effect, since embB deletion mutants are viable, indicating that the gene product is not essential in culture (7). Thus, in the fast-growing nonpathogenic species, there must be another target of inhibition, and it seems likely that EMB must inhibit two or more of the Emb arabinosyltransferase activities to effect growth inhibition. In contrast, EmbA (2) and EmbC (8) are independently essential in M. tuberculosis, and the available data strongly suggest that EmbB is essential (24). Thus, in theory, inhibition of any one of these could lead to stasis.

We investigated the role of EmbC as a target of EMB action. We constructed strains of M. tuberculosis carrying mutated alleles of embC with reduced arabinosyltransferase activity or strains overexpressing EmbC. Phenotypic analyses demonstrated a direct correlation between EmbC activity, the level of resistance to EMB, and the size of the LAM species produced.



Mycobacterium smegmatis was grown either in Lemco liquid medium (5 g liter−1 Lemco powder, 10 g liter−1 peptone, 5 g liter−1 NaCl) and 0.05% (wt/vol) Tween 80 where stated or on Lemco agar (5 g liter−1 Lemco powder, 10 g liter−1 peptone, 5 g liter−1 NaCl, 15 g liter−1 agar). M. tuberculosis was grown either in Middlebrook 7H9 liquid medium (4.7 g liter−1 Middlebrook 7H9) plus 10% (vol/vol) OADC (oleic acid, albumin, dextrose, catalase) supplement (Becton Dickinson) and 0.05% (wt/vol) Tween 80 where stated or on Middlebrook 7H10 solid medium (19 g liter−1 Middlebrook 7H10) plus 10% (vol/vol) OADC supplement. Kanamycin was used at 20 μg ml−1, hygromycin at 100 μg ml−1, and gentamicin at 10 μg ml−1 as required.

Cloning of emb genes into the pVV16 expression vector.

Primers were designed to amplify each emb gene. Forward primers had NdeI restriction sites, and reverse primers had HindIII restriction sites, at their 5′ ends (underlined in sequences below). Amplicons were directionally cloned into the pVV16 expression vector (hsp60 promoter) (10). The M. tuberculosis embA (embAMt) gene was amplified using TAF1 (CCC CAT ATG GTG CCC CAC GAC GGT AAT) and TAR1 (CCC AAG CTT TCA TGG CAG CGC CCT GAT). The embBMt gene was amplified using TBF1 (CCC CAT ATG ATG ACA CAG TGC GCG AGC) and TBR1 (CCC AAG CTT CTA TGG ACC AAT TCG GAT). The embCMt gene was amplified using TCF1 (CCC CAT ATG GCT ACC GAA GCC GCC) and TCR1 (CCC AAG CTT CTA GCC GCG CAA CGG). The M. smegmatis embB and embC (embBMs and embCMs) genes were amplified using primer pairs SBF1 (CCC CAT ATG GTG AGC GGC AAC ATG GAT) plus SBR1 (CCC AAG CTT TCA AGG TCG GAT GCG CAA) and SCF1 (CCC CAT ATG GTG ACC GGT CCG CAT GCA) plus SCR1 (CCC AAG CTT TCA ACT CAG CCG CAG GGG), respectively.

Mutagenesis of the embA, embB, and embC genes.

For site-directed mutagenesis, amplification reactions were carried out in a 50-μl total volume containing 1× Pfu Ultra reaction buffer, 0.5 mM deoxynucleosoide triphosphates, 10 pmol each primer, 10% dimethyl sulfoxide, 10 ng template, and 2.5 U Pfu Ultra. The thermocycling program used was 95°C for 1 min, followed by 18 cycles of 95°C for 1 min, 60°C for 1 min, and 68°C for 12 min, with a final extension cycle of 68°C for 20 min. Template DNA was degraded using 10 U DpnI at 37°C for 1 h. A 5-μl volume of each reaction product was used to transform competent Escherichia coli. Recombinant plasmids were isolated and sequences verified. Primers used were as follows: for embAMt(V280M), MutTAF (GCT ACC TTC TGA CCA TGG CCC GGG TCG CCC CGA) and MutTAR (TCG GGG CGA CCC GGG CCA TGG TCA GAA GGT AGC); for embBMt(M306V), MutTBFv (GGC TAC ATC CTG GGC GTG GCC CGA GTC GCC) and MutTBRv (GGC GAC TCG GGC CAC GCC CAG GAT GTA GCC); for embBMt(M306I), MutTBFi (GGC TAC ATC CTG GGC ATC GCC CGA GTC GCC) and MutTBRi (GGC GAC TCG GGC GAT GCC CAG GAT GTA GCC); for embBMt(M306L), MutTBFl (GGC TAC ATC CTG GGC CTG GCC CGA GTC GCC) and MutTBRl (GGC GAC TCG GGC CAG GCC CAG GAT GTA GCC); for embCMt(M300V), MutTCFv (GGC TAC ATC CTG ACC GTG GCC CGG GTG TCC GAG) and MutTCRv (CTC GGA CAC CCG GGC CAC GGT CAG GAT GTA GCC); for embCMt(M300I), MutTCFi (GGC TAC ATC CTG ACC ATC GCC CGG GTG TCC GAG) and MutTCRi (CTC GGA CAC CCG GGC GAT GGT CAG GAT GTA GCC); for embCMt(M300L), MutTCFl (GGC TAC ATC CTG ACC CTG GCC CGG GTG TCC GAG) and MutTCRl (CTC GGA CAC CCG GGC CAG GGT CAG GAT GTA GCC); for embCMt(D294G), MDCf (GTC GGG GCC AAC ACC TCC GAC GGA GGC TAC ATC CTG ACC ATG GCC) and MDCr (GGC CAT GGT CAG GAT GTA GCC TCC GTC GGA GGT GTT GGC CCC GAC); for embCMt(T270I), MTCf (CC GCG CGC TGG TGG TCG ATC GGC GGT CTG GAC ACC CTG GTT ATC G) and MTCr (C GAT AAC CAG GGT GTC CAG ACC GAT CGA CCA CCA GCG CGC GG); for embBMs(M292V), MutSBFv (GC TAC ATC CTG CAG GTG GCG CGC ACG GCC G) and MutSBRv (C GGC CGT GCG CGG CAC CTG CAG GAT GTA GC); for embBMs(M292I), MutSBFi (GC TAC ATC CTG CAG ATC GCG CGC ACG GCC G) and MutSBRi (C GGC CGT GCG CGG GAT CTG CAG GAT GTA GC); for embBMs(M292L), MutSBFl (GC TAC ATC CTG CAG CTG GCG CGC ACG GCC G) and MutSBRl (C GGC CGT GCG CGG CAG CTG CAG GAT GTA GC); for embCMs(M286V), MutSCFv (GGC TAC ATC CTG ACC GTG GCC CGT GTG TCC G) and MutSCRv (C GGA CAC ACG GGC CAC GGT CAG GAT GTA GCC); for embCMs(M286I), MutSCFi (GGC TAC ATC CTG ACC ATC GCC CGT GTG TCC G) and MutSCRi (C GGA CAC ACG GGC GAT GGT CAG GAT GTA GCC); and for embCMs(M286L), MutSCFl (GGC TAC ATC CTG ACC CTG GCC CGT GTG TCC G) and MutSCFl (C GGA CAC ACG GGC CAG GGT CAG GAT GTA GCC).

Gene switching.

We previously demonstrated that high-efficiency replacement of L5-based integrating vectors occurs in M. tuberculosis (19). We used this “switching” technique to test mutant alleles of embC for functionality in M. tuberculosis. The resident integrated vector in the M. tuberculosis embC del-int strain (8), which carried the only functional copy of embC together with a gentamicin resistance gene, was switched with a second integrating vector carrying alternative alleles and a hygromycin resistance gene. Transformants were plated on hygromycin to select for the incoming plasmid. To confirm that vector replacement rather than cointegration had occurred, 24 transformants from each transformation were patch tested for gentamicin resistance, and all were gentamicin sensitive, confirming that the original plasmid was lost. Switching was confirmed by PCR amplification and sequencing of the embC locus.

EMB resistance.

M. tuberculosis strains were plated onto 7H10 medium plus 10% OADC with increasing concentrations of EMB (0 to 128 μg ml−1). Twenty microliters of exponential-phase liquid cultures was spotted onto agar in 12-well tissue culture plates. Plates were incubated at 37°C, and after 10 to 30 days, growth was scored and compared to the lawn of cells on the positive control (no EMB). The MIC was determined as the highest concentration of EMB allowing growth. Similarly, M. smegmatis strains were plated onto Lemco medium containing EMB (0 to 128 μg ml−1), and growth was scored after 2 days of incubation. MICs were determined using the proportion method: briefly, 104 CFU was plated onto 7H10-OADC medium containing EMB or drug-free controls. CFU were scored after 4 weeks. The MIC99 was defined as the lowest concentration at which <1% of cells formed colonies.

Extraction and analysis of lipomannan and LAM and immunoblotting.

M. smegmatis was grown in Lemco broth containing 100 μg ml−1 hygromycin, 20 μg ml−1 kanamycin, and increasing concentrations of EMB (0 to 128 μg ml−1). After 24 h of culture, cells were harvested, resuspended in 400 μl phenol-water (1:1), and incubated at 80°C for 2 h. One hundred microliters of chloroform was added, and 10 μl of the aqueous phase was analyzed in denaturing, nonreducing 16% acrylamide gels, followed by periodic acid-Schiff staining.

M. tuberculosis was grown in 7H9 broth plus 10% (vol/vol) OADC supplement containing 100 μg ml−1 hygromycin, 20 μg ml−1 kanamycin, and increasing concentrations of EMB (0 to 128 μg ml−1). After 14 days of culture, cells were harvested, resuspended in 400 μl phenol-water (1/1), and incubated at 80°C for 2 h. One hundred microliters of chloroform was added, and 10 μl of the aqueous phase was analyzed in denaturing, nonreducing acrylamide gels, followed by periodic acid-Schiff staining or by immunoblotting using monoclonal antibody CS-35, as described previously (21).


Point mutations in the third loop alter EmbC activity but are tolerated on the chromosome.

We are interested in the role of EmbC in M. tuberculosis and the possibility that EmbC is a bona fide target of the antitubercular agent EMB. We recently demonstrated that the embC gene is essential in M. tuberculosis under normal culture conditions, and as such, it is not possible to investigate the role of EmbCMt by gene deletion (8). Previous work revealed a functional glycosyltransferase motif in EmbCMs and identified point mutations that affected the activity of EmbCMs (4). Therefore, we decided to determine whether similar mutations would affect the activity of EmbCMt and the sensitivity of M. tuberculosis to EMB, and if so, whether such a mutant allele would support LAM production and bacterial growth in M. tuberculosis.

Alignment of the primary amino acid sequences of EmbCMs and EmbCMt allowed us to identify the glycosyltransferase motif in the M. tuberculosis protein (Fig. (Fig.1).1). We introduced a point mutation at the conserved aspartate in the glycosyltransferase motif (D294G) that has previously been shown to affect the activity of EmbCMs (4), in order to determine whether a similar effect would be seen with the M. tuberculosis counterpart.

FIG. 1.
Alignment of the glycosyltransferase motif regions of the three Emb proteins of M. smegmatis (MS) and M. tuberculosis (MT). Conserved amino acids are asterisked; residues of high and low similarity are indicated by colons and periods, respectively. The ...

We constructed the D294G mutant allele in an L5-based integrating vector. To test if the allele retained functional arabinosyltransferase activity, we transformed this vector into the embC deletion strain of M. smegmatis (Fig. (Fig.2).2). This strain lacks EmbC and as a result fails to produce any LAM (33). Analysis of LAM isolated from the complemented strain confirmed that EmbC(D294G) did possess arabinosyltransferase activity, since LAM production was restored in the embC deletion strain. However, the activity of the mutant allele was lower than that of the wild-type allele, as evidenced by the fact that the LAM species produced in the strain complemented with embCMt(D294G) was significantly smaller than that in the strain complemented with the wild-type embCMt allele.

FIG. 2.
Analysis of LAM from M. smegmatis recombinant strains expressing mutant alleles of EmbC. Recombinant strains carried single-copy integrated plasmids expressing EmbC alleles from the Ag85a promoter. LAM was extracted from M. smegmatis and analyzed by sodium ...

We next determined whether the D294G allele could complement a chromosomal deletion in M. tuberculosis. We used a strain of M. tuberculosis we had previously constructed, in which there is an unmarked, in-frame deletion of the chromosomal embC gene, together with a wild-type copy of embC expressed from the Ag85A promoter on an L5 mycobacteriophage-derived integrating vector (del-int strain; genotype, embCΔ [embCint gm L5-int]) (8). The plasmid carrying the D294G allele was introduced into the del-int strain using a gene-switching method (19) so that the wild-type allele was replaced with the mutated allele, i.e., strains would contain only a single copy of embC (see Materials and Methods). Replacement of the integrated wild-type gene with the D294G allele was achieved with high efficiency, indicating that this was indeed a functional allele that could support growth. A control switching plasmid (pRG603) with a wild-type allele was used to confirm that replacement of the resident wild-type copy occurred efficiently (Table (Table1).1). Switching efficiency was approximately 103 transformants per μg of plasmid DNA for all plasmids.

EmbCMt mutations lead to EMB resistance and changes in LAM in M. tuberculosisa

LAM exists as a heterogeneous population within the cell, with variations in the overall length of the arabinan chain and the degree of branching, as well as in acylation and mannose capping. The size and precise structure of LAM differ in different species and strains, although the size and amount synthesized are generally constant in the same strain grown under the same culture conditions. It is well established that the size of LAM is dependent on the activity of EmbC in M. smegmatis (4). Since EmbC is involved in LAM biosynthesis, we looked at the effect of the D294G mutation on LAM production in M. tuberculosis (Fig. (Fig.3).3). There was no difference in the size of the LAM species produced between the wild-type strain and the recombinant del-int strain (embCΔ [embCint gm L5-int]), indicating that expression of embC from the Ag85a promoter, instead of its native promoter, did not affect LAM production. In contrast, the D294G allele resulted in the production of a smaller LAM species, although there was no effect on viability or growth in liquid or on solid medium. Thus, we confirmed that the D294G mutation reduces, but does not completely abolish, the arabinosyltransferase activity of EmbC and that the glycosyltransferase motif identified by sequence homology is functional.

FIG. 3.
Analysis of LAM from recombinant M. tuberculosis strains expressing mutant alleles of EmbC. Recombinant strains carried single-copy integrated plasmids expressing EmbC alleles from the Ag85a promoter. LAM was extracted from M. tuberculosis and analyzed ...

Reduced EmbC activity leads to increased sensitivity to EMB.

EmbC is required for the biosynthesis of the arabinan portion of LAM, and LAM biosynthesis is inhibited by EMB exposure (6, 17). This suggests that EmbC is a direct target of EMB. We tested the EMB susceptibility of the D294G strain using two methods in order to determine whether reduced EmbC activity had any bearing on EMB resistance (Table (Table1).1). Using a rapid method whereby cultures were spotted onto plates and scored for growth, we saw an increase in EMB sensitivity: the wild-type strain and the del-int strain were both inhibited by 3 μg/ml EMB in the spot test on solid medium. The D294G allele conferred a marked increase in sensitivity to EMB: strains were inhibited by only 0.5 μg/ml (a sixfold increase in sensitivity). Since changes in the inoculum or the growth rate of the mutant strain could lead to apparent changes in the MIC by this method, we also determined the MIC99 values for these strains by the agar proportion method. The mutant strain again showed increased susceptibility (MIC99, <0.5 μg/ml), confirming that there is a direct link between the activity of EmbC and EMB sensitivity and strongly suggesting that EmbC is a direct target for inhibition.

Mutations that affect EmbC activity are associated with EMB sensitivity.

Since we had shown that a single point mutation could result in EMB sensitivity, we looked at other mutations that have been suggested to play a role in mediating EMB resistance. Mutations in codon 306 of the EmbB arabinosyltransferase (M300L, M300I, and M300V) have been associated with EMB resistance (30). The three arabinosyltransferases have conserved regions, including the region around codon 306 (Fig. (Fig.1).1). In addition, mutations in EmbC itself (T270I) have been seen in EMB-resistant clinical isolates (22, 27). We decided to determine whether mutations in these regions of EmbC would lead to changes in EMB sensitivity.

We introduced several independent mutations into the embCMt gene (M300L, M300I, M300V, T270I) by site-directed mutagenesis. As before, we used the M. smegmatis embC deletion strain to determine if functional arabinosyltransferase activity was seen. Western blotting using an anti-LAM antibody confirmed that all of the complemented strains produced LAM, although in all cases the LAM species produced was smaller than that in wild-type M. smegmatis (Fig. (Fig.2).2). Thus, we confirmed that all four mutant alleles of EmbC are functionally active. Interestingly, the sizes of LAM species produced by strains carrying different embCMt alleles were different, suggesting that the mutations might affect EmbC activity. Since the wild-type M. tuberculosis allele already resulted in a LAM species smaller than that of wild-type M. smegmatis, it was difficult to be precise on sizing, but it appeared that the three mutations in codon 300, but not T270I, further reduced the size of LAM (and, by extension, EmbC activity).

We introduced each mutant allele of embCMt (M300L, M300I, M300V, T270I) into the M. tuberculosis chromosome by gene switching as before. All of the alleles were functional, in that they supported bacterial growth (Table (Table1).1). Analysis of the LAM produced by the recombinant M. tuberculosis strains demonstrated that the M300L and M300V alleles produced smaller LAM species, while the M300I allele had no effect (Fig. (Fig.3).3). In contrast, the T270I mutation produced a slightly larger LAM species, suggesting that this mutation increases the activity of EmbC. Sensitivity to EMB by both the spot test and the MIC99 assay paralleled these results, in that the strains with reduced EmbC activity and smaller LAM species (M300L, M300V) were more sensitive to EMB (MIC, 0.5 μg/ml; MIC99, <0.5 μg/ml).

These results confirmed and extended our previous observation with the D294G allele that a reduction in the activity of EmbC confers EMB sensitivity, while an increase confers resistance. Sensitivity to other antibiotics (ampicillin, isoniazid, rifampin) remained unchanged (results not shown), excluding the possibility that this is a general permeability effect. Thus, this is evidence that EmbC is a direct target of EMB in M. tuberculosis.

Overexpression of EmbCMt.

Our data strongly suggested that EmbC is a target of EMB, since a reduction of EmbC activity resulted in increased EMB sensitivity. From this we would predict that overexpression of EmbC should confer EMB resistance. We attempted to confirm this by overexpressing the M. tuberculosis allele from the hsp60 promoter in a multicopy vector, pVV16 (10). Multiple transformations of M. tuberculosis with this plasmid gave very low efficiencies, in the range of 102 transformants per μg of DNA compared to 106 transformants per μg for the control plasmid pVV16 (Table (Table2),2), suggesting that the plasmid was not tolerated by the cells. Plasmids were recovered from transformants and analyzed by sequencing. The recovered plasmids all had different deletions, with at least 1 kb of the embC gene missing at either the 5′ or the 3′ end; in addition, most deletions resulted in a downstream frameshift. We also attempted to overexpress embCMt alleles M300L, M300I, M300V, T270I, and D294G. Again, the transformation efficiencies were low, and recovered plasmids had deletions in the embC gene. Thus, it seems that overexpression of embCMt, even that of alleles with reduced activity, is extremely inhibitory to growth.

Phenotypic effects of overexpression of EmbCMt and EmbCMs in M. tuberculosisa

Since overexpression of EmbC from M. tuberculosis was toxic, we tried an alternative strategy, overexpressing the homolog from M. smegmatis. We used the same multicopy plasmid, containing the M. smegmatis allele under the control of the hsp60 promoter (pVV16), as we had used for the M. tuberculosis alleles. In addition to testing the wild-type allele, we introduced similar mutations into EmbCMs, namely, D280G, M286L, M286I, and M286V. We attempted to transform these plasmids into M. tuberculosis. Transformation with the wild-type allele resulted in a low frequency of transformants (none of which expressed the protein, as determined by Western blotting [data not shown]). In contrast, transformation of the mutant alleles was successful in all cases, with a high frequency of transformation (>1 × 106 transformants per μg of DNA). M. tuberculosis strains overexpressing mutant embCMs alleles were tested for their susceptibilities to EMB by the rapid method (spot test) only. All strains were highly resistant to EMB by this method and were able to grow in the presence of EMB concentrations as high as 32 μg/ml. The wild-type strain had a reproducible cutoff point of 3 μg/ml in this assay using different inocula. The two strains overexpressing EmbCMs(D280G) or EmbCMs(M286I) could grow with 64 μg/ml of EMB (Table (Table2).2). These findings further support the hypothesis that EmbC is a direct target of EMB in M. tuberculosis.

The arabinosyltransferase activity of each embCMs allele was assessed by analyzing LAM production in the two Mycobacterium species (Fig. (Fig.4).4). All alleles had functional activity, but all three mutations in codon 286 resulted in reduced activity, as demonstrated by the production of smaller LAM species in M. smegmatis. In contrast, expression of any of these three alleles in M. tuberculosis resulted in an increase in the size of the LAM species produced, which correlates with the increased level of EMB resistance (Fig. (Fig.5).5). Taken together with our previous data, these findings confirm that the M. smegmatis allele is intrinsically more active than the M. tuberculosis allele and could, at least partly, account for the difference in size between the LAM molecules isolated from these two species.

FIG. 4.
Analysis of LAM from recombinant M. smegmatis strains overexpressing mutant alleles of embC. LAM was extracted from recombinant M. smegmatis strains and was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (on a denaturing, nonreducing ...
FIG. 5.
Analysis of LAM from recombinant M. tuberculosis strains overexpressing mutant alleles of embCMs. LAM was extracted from M. tuberculosis strains and was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (on a denaturing, nonreducing ...

Overexpression of embB in M. tuberculosis leads to low-level EMB resistance.

We have shown that increased EmbCMs activity results in EMB resistance and that decreased EmbCMt activity results in sensitivity; normally, these would be sufficient indications to confirm that EmbC is the target of EMB action. However, previously published data indicate that EMB inhibits all of the Emb arabinosyltransferases, although it is not clear to what extent each is inhibited by a particular concentration of EMB. In addition, our data indicate that the EmbC enzymes of the two species most studied in this area have different intrinsic activities, and this may have a significant impact on the mode of action of EMB in the two organisms.

In order to address the relative roles of EmbB and EmbC in EMB resistance, we attempted to overexpress EmbB from M. tuberculosis. Despite many attempts, we were unable to clone embB from M. tuberculosis into an expression vector, either in Escherichia coli or M. smegmatis. This strongly suggests that the overexpression of M. tuberculosis embB is toxic in both E. coli and M. smegmatis. In contrast, we were able to clone embBMs into the pVV16 expression vector and to transform this vector successfully into M. tuberculosis, but EMB resistance was unchanged. We also introduced expression vectors carrying mutated alleles of embBMs (M292L, M292I, and M292V) into M. tuberculosis (Table (Table3).3). Overexpression of these alleles resulted in low-level EMB resistance (MICs, 4 to 8 μg/ml). Thus, although resistance was seen, its level was not as high as that of the resistance engendered by overexpression of embC.

Phenotypic effects of overexpression of EmbA and EmbB in M. tuberculosisa

Overexpression of embAMt in M. tuberculosis does not lead to EMB resistance.

We also looked at the effect of EmbAMt overexpression on EMB resistance. In contrast to EmbC and EmbB, we were able to overexpress EmbAMt in M. tuberculosis, but it had no effect on EMB resistance (Table (Table3).3). Interestingly, M. tuberculosis EmbA does not share the conserved methionine in the second periplasmic loop with EmbB and EmbC from M. tuberculosis. At this position, M. tuberculosis EmbA has a valine, an amino acid suspected to increase EMB resistance in EmbB. In order to look at this effect, we mutated this codon to methionine in EmbAMt, but it had no impact on resistance to EMB (Table (Table3).3). Thus, it appears that EmbA does not mediate EMB resistance in M. tuberculosis.

Effect of EMB inhibition on LAM biosynthesis in EMB-resistant strains.

EMB inhibits the formation of LAM via inhibition of EmbC (6, 17). We looked at the effects of EMB on LAM production in the wild-type strain of M. tuberculosis and in a strain that was resistant to EMB. In the wild type, inhibition of LAM synthesis was clearly seen at concentrations below the MIC and as low as 1 μg/ml (Fig. (Fig.6),6), and much smaller LAM species were produced, confirming that EmbC activity was being inhibited directly. In contrast, when the embCMs(D280G) allele was overexpressed in M. tuberculosis, LAM production remained unaffected by EMB concentrations as high as 32 μg/ml. Thus, the inhibition of EmbC activity correlated with the EMB MIC.

FIG. 6.
Effect of EMB exposure on LAM production, measured in liquid cultures of M. tuberculosis overexpressing embCMs alleles from the hsp60 promoter. C, control vector (pVV16); D280G, embCMs(D280G) allele.

The fact that the effect of EMB on LAM production was directly correlated with the level of resistance further reinforces the hypothesis that EmbC is a target of EMB action in M. tuberculosis. These results also suggest that the action of EMB is mediated by an interaction between EMB and the catalytic domain of EmbC (the second loop, containing the D280 and M286 residues) (4).


Our data show for the first time that EmbC is a direct cellular target of EMB in M. tuberculosis. Since EmbC is essential in M. tuberculosis, its inhibition would prevent growth independently of any effect on the other arabinosyltransferases. We demonstrated that embCMt alleles with reduced enzymatic activity conferred increased EMB sensitivity and that EMB treatment resulted in inhibition of LAM synthesis in wild-type strains at concentrations just below the MIC. Taken together, the data on the effect of EMB on LAM production in sensitive and resistant strains and the correlation between EmbC activity and EMB resistance demonstrate that EmbC is one of the cellular targets of EMB action. Overexpression of embCMs alleles could confer high-level resistance in the absence of any changes in EmbB. This has implications for the detection of clinical resistance and suggests that studies to look at the role of embC mutations in mediating clinically relevant phenotypes should be extended.

We used two methods for determining the susceptibilities of strains to EMB-mediated growth inhibition: a rapid method for scoring changes in resistance and, for a subset of strains, MIC99 determination. MIC99s are routinely determined for clinical isolates by using the agar proportion method, and they provide a robust determination of the concentration of a drug required to prevent colony formation by 99% of cells. Since inocula are adjusted and viable bacilli counted, the proportion method corrects for small deviations in the inoculum size and growth rate that could be masked in growth as spots or lawns. However, MIC99 determinations are resource- and time-intensive. We used a rapid method for determining whether EmbC mutations resulted in changes in resistance, relying on spotting log-phase cultures onto drug plates and scoring for growth as a lawn. A comparison of the two methods using a subset of strains (Table (Table1)1) confirmed that the trends in the results were the same, i.e., mutant strains showed increased sensitivity to EMB in both assays, although the absolute values for MICs were different (more EMB was required to prevent the growth of spots than that of individual colonies). For resistant strains, small differences in inocula or growth rates would not be expected to lead to an ability to grow on the higher concentrations of EMB observed, and results for the wild-type strain with several different inocula were reproducible. Thus, the spot method can be used as an indication of resistance, although rigorous determination of MIC99s using the agar proportion method would likely reveal more-subtle changes in resistance.

Overexpression of mutant alleles of embCMs, but not embCMt, was tolerated in M. tuberculosis, although neither wild-type allele could be overexpressed. Overexpression of one allele, embC(D280G), gave rise to LAM species larger than the wild type in M. tuberculosis. To our knowledge, this is the first report of the modulation of LAM size in isogenic strains, and this finding paves the way toward a greater understanding both of the role of EmbC in controlling the length of the arabinan chains and of the role of LAM in host-pathogen interactions.

EmbC is an integral membrane protein with a large number of transmembrane domains, making it extremely challenging to purify the active enzyme from whole cells. Thus, there is no direct enzyme assay for EmbC. We used the M. smegmatis embC deletion strain to assess the activity of embC alleles. It is well documented that the level of EmbC activity directly affects the composition of LAM and that increased activity results in larger LAM species, whereas decreased activity results in smaller LAM species. Analysis of LAM from such strains allowed the identification of a number of mutant alleles with altered activity.

Our data showing that mutations in EmbBMs confer low-level EMB resistance in M. tuberculosis are in agreement with a previous study that established the role of embBMt codon 306 mutations in mediating low-level EMB resistance by using isogenic strains of M. tuberculosis (23). There is accumulating evidence that high-level EMB resistance is a multistage process, requiring two or more independent mutations (20, 23, 28). EmbB mutations are often associated with high-level resistance in clinical isolates, but mutations in codon 306 are not sufficient on their own to generate high-level resistance (23), so presumably there are secondary mutations (which may be dependent on strains already having EmbB mutations). In support of this, a recent study showed that a combination of mutations in EmbB codon 306 and EmbC codon 270 resulted in higher levels of resistance (11). In this light, it is interesting that construction of an isogenic strain carrying EmbC(T270I) alone demonstrated that this mutation played no role in mediating EMB resistance. Overexpression of wild-type EmbCMt or EmbCMs was not possible in M. tuberculosis; all transformants with overexpression plasmids had large deletions in the embC gene, and this seems to be the most likely explanation for the lack of embC mutations in clinical isolates. Overexpression of EmbC is likely to lead to higher competition for the substrate DPA and to the depletion of DPA for AG synthesis by EmbB and EmbA. Since AG is also essential, forming part of the structural integrity of the cell wall, this would lead to a severe growth defect. Given the probable competition of EmbB and EmbC for the donor DPA, it is possible that EmbC “up” mutations are tolerated only in the presence of EmbB mutations. In the future, it would be interesting to determine whether particular combinations of EmbB and EmbC mutations can give rise to high-level resistance.

Since it is likely that high-level resistance to EMB in M. tuberculosis is multifactorial, this would support the hypothesis that there is more than one intracellular target; mutations in one target might cause low-level resistance, but mutations in both might be required for high-level resistance. In this scenario, overexpression of either target could lead to resistance, simply by virtue of by binding excess drug. If EMB binds to both EmbB and EmbC, mutations that affect the binding affinity or level of expression of one protein would also affect the level of binding to the other protein. Thus, low-level resistance could be mediated by a mutation in either target protein via a titration effect, i.e., if EMB is being bound and titrated out by an increased level of the target protein (which could be either EmbB or EmbC), then the effective concentration available to inhibit the other target is reduced.

We have not excluded the possibility that changes in EMB sensitivity are related to changes in the cell wall architecture resulting from altered LAM biosynthesis. However, we think this is unlikely, since resistance to the other antibiotics tested remained the same. This would also apply to EmbB mutations, since EmbB is also involved in cell wall (AG) synthesis. For antibiotics that target components of the cell wall, it is hard to distinguish between direct and indirect effects on resistance levels.

There are striking differences between the Emb proteins of M. tuberculosis and M. smegmatis. Most notably, any one of the three genes can be deleted in the latter, while it is highly likely that all three are independently essential in the former. This points toward a difference in the mechanism of action of EMB between the two species, in that inhibition of any Emb protein would be static for M. tuberculosis, while inhibition of two or more is required to affect the growth of M. smegmatis. Thus, M. smegmatis appears to be a poor model for studying the mode of action of EMB and EMB resistance.


This work was funded by Wellcome Trust grant 074612, awarded to T.P., and National Institutes of Health grant AI 37319, awarded to D.C.

We are grateful to Stefan Berg for valuable discussions.


[down-pointing small open triangle]Published ahead of print on 13 July 2009.


1. Alcaide, F., G. E. Pfyffer, and A. Telenti. 1997. Role of embB in natural and acquired resistance to ethambutol in mycobacteria. Antimicrob. Agents Chemother. 41:2270-2273. [PMC free article] [PubMed]
2. Amin, A. G., R. Goude, L. Shi, J. Zhang, D. Chatterjee, and T. Parish. 2008. EmbA is an essential arabinosyltransferase in Mycobacterium tuberculosis. Microbiology 154:240-248. [PMC free article] [PubMed]
3. Belanger, A. E., G. S. Besra, M. E. Ford, K. Mikusova, J. T. Belisle, P. J. Brennan, and J. M. Inamine. 1996. The embAB genes of Mycobacterium avium encode an arabinosyl transferase involved in cell wall arabinan biosynthesis that is the target for the antimycobacterial drug ethambutol. Proc. Natl. Acad. Sci. USA 93:11919-11924. [PMC free article] [PubMed]
4. Berg, S., J. Starbuck, J. B. Torrelles, V. D. Vissa, D. C. Crick, D. Chatterjee, and P. J. Brennan. 2005. Roles of conserved proline and glycosyltransferase motifs of EmbC in biosynthesis of lipoarabinomannan. J. Biol. Chem. 280:5651-5663. [PubMed]
5. Bloch, A. B., L. M. Onorato, and K. G. Castro. 1994. Multiple-antibiotic-resistant bacteria. N. Engl. J. Med. 331:678-679. [PubMed]
6. Deng, L., K. Mikusova, K. G. Robuck, M. Scherman, P. J. Brennan, and M. R. McNeil. 1995. Recognition of multiple effects of ethambutol on metabolism of mycobacterial cell envelope. Antimicrob. Agents Chemother. 39:694-701. [PMC free article] [PubMed]
7. Escuyer, V. E., M. A. Lety, J. B. Torrelles, K. H. Khoo, J. B. Tang, C. D. Rithner, C. Frehel, M. R. McNeil, P. J. Brennan, and D. Chatterjee. 2001. The role of the embA and embB gene products in the biosynthesis of the terminal hexaarabinofuranosyl motif of Mycobacterium smegmatis arabinogalactan. J. Biol. Chem. 276:48854-48862. [PubMed]
8. Goude, R., A. G. Amin, D. Chatterjee, and T. Parish. 2008. The critical role of embC in Mycobacterium tuberculosis. J. Bacteriol. 190:4335-4341. [PMC free article] [PubMed]
9. Hazbón, M. H., M. B. del Valle, M. I. Guerrero, M. Varma-Basil, I. Filliol, M. Cavatore, R. Colangeli, H. Safi, H. Billman-Jacobe, C. Lavender, J. Fyfe, L. Garcia-Garcia, A. Davidow, M. Brimacombe, C. I. Leon, T. Porras, M. Bose, F. Chaves, K. D. Eisenach, J. Sifuentes-Osornio, A. P. de Leon, M. D. Cave, and D. Alland. 2005. Role of embB codon 306 mutations in Mycobacterium tuberculosis revisited: a novel association with broad drug resistance and IS6110 clustering rather than ethambutol resistance. Antimicrob. Agents Chemother. 49:3794-3802. [PMC free article] [PubMed]
10. Jackson, M., D. C. Crick, and P. J. Brennan. 2000. Phosphatidylinositol is an essential phospholipid of mycobacteria. J. Biol. Chem. 275:30092-30099. [PubMed]
11. Jadaun, G. P., R. Das, P. Upadhyay, D. S. Chauhan, V. D. Sharma, and V. M. Katoch. 2009. Role of embCAB gene mutations in ethambutol resistance in Mycobacterium tuberculosis isolates from India. Int. J. Antimicrob. Agents 33:483-486. [PubMed]
12. Kilburn, J. O., and J. Greenberg. 1977. Effect of ethambutol on the viable cell count in Mycobacterium smegmatis. Antimicrob. Agents Chemother. 11:534-540. [PMC free article] [PubMed]
13. Kilburn, J. O., K. Takayama, E. L. Armstrong, and J. Greenberg. 1981. Effects of ethambutol on phospholipid metabolism in Mycobacterium smegmatis. Antimicrob. Agents Chemother. 19:346-348. [PMC free article] [PubMed]
14. Lee, A. S. G., S. N. K. Othman, Y. M. Ho, and S. Y. Wong. 2004. Novel mutations within the embB gene in ethambutol-susceptible clinical isolates of Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 48:4447-4449. [PMC free article] [PubMed]
15. Lety, M. A., S. Nair, P. Berche, and V. Escuyer. 1997. A single point mutation in the embB gene is responsible for resistance to ethambutol in Mycobacterium smegmatis. Antimicrob. Agents Chemother. 41:2629-2633. [PMC free article] [PubMed]
16. McNeil, M. R., and P. J. Brennan. 1991. Structure, function and biogenesis of the cell envelope of mycobacteria in relation to bacterial physiology, pathogenesis and drug resistance; some thoughts and possibilities arising from recent structural information. Res. Microbiol. 142:451-463. [PubMed]
17. Mikusová, K., R. A. Slayden, G. S. Besra, and P. J. Brennan. 1995. Biogenesis of the mycobacterial cell wall and the site of action of ethambutol. Antimicrob. Agents Chemother. 39:2484-2489. [PMC free article] [PubMed]
18. Mokrousov, I., T. Otten, B. Vyshnevskiy, and V. Escuyer. 2002. Detection of embB306 mutations in ethambutol-susceptible clinical isolates of Mycobacterium tuberculosis from Northwestern Russia: implications for genotypic resistance testing. J. Clin. Microbiol. 40:3810-3813. [PMC free article] [PubMed]
19. Pashley, C. A., and T. Parish. 2003. Efficient switching of mycobacteriophage L5-based integrating plasmids in Mycobacterium tuberculosis. FEMS Microbiol. Lett. 229:211-215. [PubMed]
20. Perdigão, J., R. Macedo, A. Ribeiro, L. Brum, and I. Portugal. 2009. Genetic characterisation of the ethambutol resistance-determining region in Mycobacterium tuberculosis: prevalence and significance of embB306 mutations. Int. J. Antimicrob. Agents 33:334-338. [PubMed]
21. Prinzis, S., D. Chatterjee, and P. J. Brennan. 1993. Structure and antigenicity of lipoarabinomannan from Mycobacterium bovis BCG. J. Gen. Microbiol. 139:2649-2658. [PubMed]
22. Ramaswamy, S. V., A. G. Amin, S. Göksel, C. E. Stager, S. J. Dou, H. El Sahly, S. L. Moghazeh, B. N. Kreiswirth, and J. M. Musser. 2000. Molecular genetic analysis of nucleotide polymorphisms associated with ethambutol resistance in human isolates of Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 44:326-336. [PMC free article] [PubMed]
23. Safi, H., B. Sayers, M. H. Hazbón, and D. Alland. 2008. Transfer of embB codon 306 mutations into clinical Mycobacterium tuberculosis strains alters susceptibility to ethambutol, isoniazid, and rifampin. Antimicrob. Agents Chemother. 52:2027-2034. [PMC free article] [PubMed]
24. Sassetti, C. M., D. H. Boyd, and E. J. Rubin. 2003. Genes required for mycobacterial growth defined by high density mutagenesis. Mol. Microbiol. 48:77-84. [PubMed]
25. Shen, X., G. M. Shen, J. Wu, X. H. Gui, X. Li, J. Mei, K. DeRiemer, and Q. Gao. 2007. Association between embB codon 306 mutations and drug resistance in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 51:2618-2620. [PMC free article] [PubMed]
26. Sreevatsan, S., K. E. Stockbauer, X. Pan, B. N. Kreiswirth, S. L. Moghazeh, W. R. Jacobs, Jr., and A. Telenti. 1997. Ethambutol resistance in Mycobacterium tuberculosis: critical role of embB mutations. Antimicrob. Agents Chemother. 41:1677-1681. [PMC free article] [PubMed]
27. Srivastava, S., A. Garg, A. Ayyagari, K. K. Nyati, T. N. Dhole, and S. K. Dwivedi. 2006. Nucleotide polymorphism associated with ethambutol resistance in clinical isolates of Mycobacterium tuberculosis. Curr. Microbiol. 53:401-405. [PubMed]
28. Starks, A. M., A. Gumusboga, B. B. Plikaytis, T. M. Shinnick, and J. E. Posey. 2009. Mutations at embB codon 306 are an important molecular indicator of ethambutol resistance in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 53:1061-1066. [PMC free article] [PubMed]
29. Takayama, K., and J. O. Kilburn. 1989. Inhibition of synthesis of arabinogalactan by ethambutol in Mycobacterium smegmatis. Antimicrob. Agents Chemother. 33:1493-1499. [PMC free article] [PubMed]
30. Telenti, A., W. J. Philipp, S. Sreevatsan, C. Bernasconi, K. E. Stockbauer, B. Wieles, J. M. Musser, and W. R. Jacobs, Jr. 1997. The emb operon, a gene cluster of Mycobacterium tuberculosis involved in resistance to ethambutol. Nat. Med. 3:567-570. [PubMed]
31. Thomas, J. P., C. O. Baughn, R. G. Wilkinson, and R. G. Shepherd. 1961. A new synthetic compound with antituberculous activity in mice: ethambutol (dextro-2,2′-(ethylenediimino)-di-l-butanol). Am. Rev. Respir. Dis. 83:891-893. [PubMed]
32. Wolucka, B. A., M. R. McNeil, E. de Hoffmann, T. Chojnacki, and P. J. Brennan. 1994. Recognition of the lipid intermediate for arabinogalactan/arabinomannan biosynthesis and its relation to the mode of action of ethambutol on mycobacteria. J. Biol. Chem. 269:23328-23335. [PubMed]
33. Zhang, N., J. Torrelles, M. McNeil, V. E. Escuyer, K. H. Khoo, P. J. Brennan, and D. Chatterjee. 2003. The Emb proteins of mycobacteria direct arabinosylation of lioparabinomannan and arabinogalactan via an N-terminal recognition region and a C-terminal synthetic region. Mol. Microbiol. 50:69-76. [PubMed]

Articles from Antimicrobial Agents and Chemotherapy are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...