Contribution of dfrA and inhA Mutations to the Detection of Isoniazid-Resistant Mycobacterium tuberculosis Isolates

Abstract

Screening of 127 isoniazid (INH)-resistant Mycobacterium tuberculosis isolates from Singapore for mutations within the dfrA and inhA genes revealed mutations in 0 and 5 (3.9%) isolates respectively, implying that mutations in dfrA do not contribute to the detection of INH-resistant M. tuberculosis and that mutations within inhA are rare. Thirty-seven (29%) of the 127 isolates had no mutations in any of the genes implicated in INH resistance (katG, kasA, and ndh; inhA and ahpC promoters), suggesting that there are new INH targets yet to be discovered.

Mycobacterium tuberculosis remains a major health concern worldwide, with approximately 2 billion people currently infected worldwide and 9.2 million new cases in 2006 (13, 27). The emergence of multidrug-resistant tuberculosis has increased global efforts to understand the molecular mechanisms of drug resistance in M. tuberculosis.

Isoniazid (INH) for M. tuberculosis chemotherapy has always been used as first-line treatment (25) since it was found to exhibit powerful bactericidal activity against the disease by effectively causing loss of acid fastness in M. tuberculosis through inhibition of mycolic acid synthesis (9). However, its widespread use, justified by its high specificity, low cost, and considerably low toxicity, has seen treatment failures due to increasing resistance to the drug.

Clinical resistance to INH is widely known to be caused by mutations within the katG and inhA genes and the promoter region of inhA (2, 13, 18, 19, 25, 29). Other genes, such as kasA and ndh, have been implicated in INH resistance, but mutations in these genes are rare and have been observed in INH-susceptible isolates and/or in association with katG mutations (8, 11, 12, 14, 15, 18, 25, 28). The ahpC gene is involved in the cellular regulation of oxidative stress (18, 22). Mutations within the oxyR-ahpC intergenic region that result in increased expression of alkyl hydroperoxidase are considered to compensate for peroxide sensitivity due to loss of KatG function found in clinical resistant strains (10, 22, 28). Nevertheless, screening of INH-resistant clinical isolates of M. tuberculosis for mutations within kasA, ndh, and ahpC is often performed.

We have previously extensively characterized INH-resistant isolates for mutations within all of these regions; however, 57 (36%) of 160 INH-resistant isolates from Singapore had no detectable alterations (11, 12). This finding suggests that mutations in other genes that confer resistance to INH may exist.

Recently, dihydrofolate reductase, encoded by the dfrA gene, was proposed as a new target for INH (1) as it has been reported that NADP-bound INH had been shown to inhibit dihydrofolate reductase, an enzyme essential for nucleic acid synthesis. Also, the importance of mutations within the structural region of enoyl reductase (InhA), encoded by the inhA gene, have been reported where a single point mutation allele (S94A) transferred by using specialized linkage transduction was able to confer clinically relevant levels of resistance and inhibit mycolic acid synthesis (18, 26). Most studies investigating mutations in the inhA gene analyzed the promoter region of inhA and not the region within the gene (6-8, 16, 17). Thus, we aimed to screen for mutations within both the inhA and dfrA genes of INH-resistant M. tuberculosis isolates to determine whether mutations within these genes contribute to INH resistance.

A total of 127 INH-resistant clinical isolates and 15 INH-sensitive isolates from Singapore were included in this study in order to gain molecular insight into INH resistance. DNA extracted from the isolates was analyzed by amplifying two overlapping fragments for the entire inhA gene and one fragment for the entire dfrA gene (Table 1). Amplification of the genes through PCR was done by using specific oligonucleotide primers for the respective gene fragments (Table 1). PCR products were purified (NucleoSpin Extract Column II; Macherey-Nagel) and directly sequenced with the BigDye Terminator sequencing kit V3.1 (Applied Biosystems), followed by analysis on the 3130XL Genetic Analyzer (Applied Biosystems). Confirmation of mutations was done by reamplification and resequencing. The nucleotide sequences obtained were aligned against the reference sequences of the respective genes of M. tuberculosis reference strain H37Rv by using the DNASTAR SeqMan II software (Lasergene).

TABLE 1.

Oligonucleotide primers used for the amplification of the entire dfrA^a and inhA^b genes

Gene	Primer sequence	Annealing temp (°C)	Amplicon size (bp)	Nucleotide positions
dfrA (F)	GAC GAA GCG ATG AGG AGA AG	55	632	300,058-300,072
dfrA (R)	TCG TTG TGA AGA ACT ACG ATC C			300,699-300,678
inhA (F)1	CTA CAT CGA CAC CGA TAT GAC	55	700	290,948-290,968
inhA (R)1	GAC CGT CAT CCA GTT GTA G			291,647-291,629
inhA (F)2	GCA TCA ACC CGT TCT TCG AC	55	677	291,469-291,488
inhA (R)2	TAA TGC CAT TGA TCG GTG ATA C			292,145-292,124

^aThe M. tuberculosis sequence used to design the primers was obtained from GenBank (accession no. BX842580.1).

^bThe M. tuberculosis sequence used to design the primers was obtained from GenBank (accession no. BX842576.1).

Screening of the entire inhA gene revealed mutations in five (3.9%) of 127 INH-resistant isolates (synonymous mutation GGA→GGC at nucleotide position 9 or G3G [n = 2], I21T [n = 2], or I194T [n = 1]). Of the five isolates with mutations, four also had mutations in the ahpC promoter, the inhA promoter, and the kasA gene, which were detected in our previous study (Table 2) (11). Such an occurrence might suggest that some of the observed substitutions are not unique mutations that actually contribute to INH resistance. Also, the low mutational frequency within the inhA gene among INH-resistant isolates suggests this (11). Furthermore, a previous study reported that a previously known mutation, at position 491 of the rrs gene, was in fact a polymorphism within the F11 family and not actually involved in streptomycin resistance (24). Hence, mutations found, especially rarely occurring ones, should not be hastily perceived to confer resistance.

TABLE 2.

M. tuberculosis isolates with mutations in the inhA gene

Isolate	Mutation	Amino acid substitution	No. (%) of isolates	Mutation(s) at:
				katG	ahpC	inhA promoter	kasA	ndh
I59	GGA→GGC	G3G	1 (0.8)
I82	ATC→ACC	I194T	1 (0.8)				T6T (ACC→ACT), H307H (CAC→CAT)
IR27	GGA→GGC	G3G	1 (0.8)		−6 (G→A)a	−15 (C→T)
IRS7	ATC→ACC	I21T	1 (0.8)		27 (G→T)b	−15 (C→T)
MDR2	ATC→ACC	I21T	1 (0.8)			−15 (C→T)

^aPosition relative to the mRNA start site.

^bNucleotide substitution within the defective oxyR gene.

Screening of the entire dfrA gene did not reveal any mutations in any of the 127 INH-resistant isolates, suggesting that dfrA mutations may not contribute to INH resistance and that screening for dfrA mutations in INH-resistant isolates is unnecessary.

To determine if the mutations detected within the inhA gene are mutations and not polymorphisms, we screened 15 INH-susceptible isolates and found no mutations in inhA. These isolates also did not have mutations in the dfrA gene.

Mutations at both I21T (3, 5, 18) and I194T (5, 13) were previously reported to confer resistance to INH. Ile21 and Ile194 were shown to be 2 of the 10 most important amino acid residues making conserved hydrogen bonds with NADH cofactor in the wild-type InhA protein, which encourages INH sensitivity (21). Amino acid replacements will result in lowered NADH affinity and therefore confer resistance to INH since InhA inactivation is dependent on the presence of bound NADH (4, 13, 20, 23, 28).

Of the 127 INH-resistant samples we have screened, there remain 37 samples (29%) with no mutations in any of the genes reported to be involved in INH resistance. This possibly implies that there are new INH targets yet to be discovered. Our findings suggest that mutations in dfrA do not contribute to the detection of INH-resistant M. tuberculosis and that mutations within the inhA gene are rare.