Journal of Antimicrobial Chemotherapy (2003) 52, 5660DOI: 10.1093/jac/dkg287Advance Access publication 29 May 2003

56. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2003 The British Society for Antimicrobial Chemotherapy

Susceptibility of Mycobacterium tuberculosis to weak acids

Ying Zhang*, Hao Zhang and Zhonghe Sun

Department of Molecular Microbiology and Immunology, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21205, USA

Received 21 January 2003; returned 20 March 2003; revised 11 April 2003; accepted 16 April 2003

The susceptibility of Mycobacterium tuberculosis and Mycobacterium smegmatis to a range of weak acidsand acid pH was investigated. M. tuberculosis was found to be more susceptible to acid pH and weak acidsthan M. smegmatis. The weak acids were more active against M. tuberculosis at acid pH than at neutral pH.M. tuberculosis was found to be less able to maintain its internal pH and membrane potential at acid pHthan M. smegmatis. The antituberculous activity of weak acids correlated with their ability to disrupt themembrane potential but not the internal pH. The significance of these findings is discussed in relation toM. tuberculosis physiology and development of new antituberculous agents.

Keywords: weak acid, antimicrobial susceptibility, pH homeostasis, membrane potential, Mycobacterium tuberculosis

IntroductionTuberculosis (TB) still remains a major infectious cause of morbidityand mortality world-wide, especially in the developing countries.1Drug-resistant TB is becoming an increasing public health concern inrecent years and poses a potential threat to the control of the disease.1There is growing awareness that the current TB therapy is too long,taking a period of 6 months. Failure to adhere to the lengthy therapy isa frequent cause of drug-resistant TB. There is currently a great dealof interest in developing new drugs that are not only active againstdrug-resistant TB but can also shorten the duration of the therapy.2

During our study of the mode of action of the frontline TB drugpyrazinamide, a drug that has shortened TB therapy from 912months previously to 6 months, we have shown that Mycobacteriumtuberculosis seems to be uniquely susceptible to the weak acidpyrazinoic acid (pKa = 2.9), the active form of pyrazinamide; whereasother mycobacteria (e.g. Mycobacterium smegmatis) or bacteria (e.g.Escherichia coli) are more resistant to pyrazinoic acid.3,4 In addition,it is well known that during pyrazinamide susceptibility testing,which requires acid pH for activity, the growth of M. tuberculosis isinhibited if the medium pH is below 5.5.5 M. tuberculosis appears tobe quite susceptible to acid pH compared with other mycobacteria.68In Sautons simple salt medium, the growth of M. tuberculosis wasrestricted at pH 6.0, whereas other mycobacterial species grew quitewell.8 In this study, we tested whether M. tuberculosis is also suscep-tible to other weak acids in addition to pyrazinoic acid and comparedthe susceptibility of M. tuberculosis to acidic pH and a range of weakacids with that of M. smegmatis. We have shown that M. tuberculosisis significantly more susceptible to acidic pH and weak acids in gen-eral than M. smegmatis. The antimycobacterial activity of the weak

acids is enhanced at acid pH. The basis of the susceptibility ofM. tuberculosis to acid pH and weak acids is investigated.

Materials and methods

Mycobacterial growth and susceptibility to acid pHM. tuberculosis strain H37Ra was grown in 7H9 liquid medium (DIFCO)supplemented with 0.05% Tween 80 and 10% bovine serum albumin-dextrose-catalase enrichment (DIFCO) at 37C for 3 weeks with occa-sional shaking. M. smegmatis mc26 (MC2) was similarly cultivated in the7H9 medium at 37C for 4 days. To test the susceptibility of mycobacteriato different pH values, M. tuberculosis H37Ra or M. smegmatis cellswere resuspended in sodium phosphate buffer adjusted to different pHlevels (pH 3.0, 4.0, 5.0, 6.0, 7.0) in 1 mL to a cell density of 1.30 at OD600and incubated at 37C. At 1, 3, 5 and 7 days, aliquots of the cell suspen-sion were removed, washed and diluted before plating on 7H11 plates.The plates were then incubated at 37C for 4 weeks for M. tuberculosisand for 5 days for M. smegmatis to determine the number of survivingbacteria.

Susceptibility to weak acids and isolation of weak acid resistant mutantsVarious weak acids were obtained from Sigma Chemical Co., and weredissolved in DMSO at appropriate concentrations. The weak acids wereincorporated into 7H11 agar at various concentrations. Three-week-oldstationary phase M. tuberculosis H37Ra culture or 4-day-old M. smeg-matis mc26 culture were tested for susceptibility to weak acids on 7H11plates at pH 6.8 and pH 5.5 as described.9 For the isolation of weak acidmutants, about 108 colony forming units (cfu) of M. tuberculosis H37Ra

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

*Corresponding author. Tel: +1-410-614-2975; Fax: +1-410-955-0105; E-mail: yzhang@jhsph.edu

by guest on April 30, 2014

http://jac.oxfordjournals.org/D

ownloaded from

ToshibaUnderline

Susceptibility of M. tuberculosis to weak acids

57

were plated on acidic 7H11 agar plates (pH 5.5) containing variousconcentrations of weak acids such as salicylate, benzoic acid, nonyloxy-benzoic acid and mefenamic acid. The plates were incubated at 37C for4 weeks before being examined for the emergence of spontaneousmutants.

Measurement of intracellular pH and membrane potentialThe internal pH of mycobacteria was measured as described previously.3Membrane potential was measured with [3H]tetraphenylphosphoniumbromide (TPP+) using the method as described.10 Briefly, 3-week-oldH37Ra or 4-day-old M. smegmatis cells were resuspended in Sautonsmedium at different pH values to measure the change in the membranepotential in response to changes in external pH after incubating the cellsat room temperature for 50 min. [3H]TPP+ (380 mCi/mmol) at 10 Mfinal concentration was then added to the cell suspension and the mixturewas fully mixed before silicone oil was added and the mixture incubatedfor another 10 min. The mixture was spun at 12 000 rpm for 3 min, and100 L supernatant was taken for scintillation counting. The cell pelletswere then snap-frozen in an alcohol/dry ice bath. The bottom of the tubescontaining the cell pellets were cut off for scintillation counting. Todetermine the effect of weak acids on membrane potential and internalpH, various weak acids were incubated with mycobacterial cells resus-pended in pH 5.5 Sautons medium for 1 h when the measurements weremade as described above. Valinomycin (10 M) and nigericin (10 M)were used as controls for the membrane potential and internal pH meas-urements.

Results

Susceptibility of M. tuberculosis and M. smegmatis to acid pHThe acid sensitivity of M. tuberculosis and M. smegmatis was deter-mined by exposing the bacilli to various acidic pH conditions usingpH 7.0 as a control, and plated for survivors after exposure for differ-ent times. The relative sensitivity of the two mycobacterial species toacidic pH was expressed as the percentage of bacterial survival bydividing the cfu obtained after exposure to acid pH by that at neutralpH. At pH 3.0, there was relatively little difference between M. tuber-culosis and M. smegmatis in terms of survival due to extreme acidity(Table 1). However, at pH 4.0 and 5.0, M. tuberculosis was signifi-cantly more sensitive to acid pH than M. smegmatis (Table 1).

Susceptibility of M. tuberculosis to weak acidsAs shown in Table 2, M. tuberculosis was more susceptible thanM. smegmatis to a range of weak acids. The antimycobacterial activ-ity of the weak acids was more pronounced at acid pH than at close to

neutral pH for both mycobacterial species. In addition, the activity ofthe weak acids appeared to correlate with their pKa values, i.e. thelower the pKa, the higher the antimycobacterial activity (Table 2). It isnoteworthy that M. tuberculosis was susceptible to linoleic acid(MIC 37 mg/L at pH 5.5) but not to linoleic acid ethyl ester (MIC >1000 mg/L at pH 5.5), indicating that the acid form COOH is activeand that M. tuberculosis does not have an appropriate esterase toconvert linoleic acid ethyl ester to the active acid form.

Inability to isolate weak acid resistant mutants of M. tuberculosisWe have shown previously that no pyrazinoic acid-resistant mutantsof M. tuberculosis could be isolated.11 To determine whether this is amore generalized phenomenon, we attempted to isolate M. tuber-culosis H37Ra mutants resistant to a range of weak acids such assalicylic acid, benzoic acid and 4-nonyloxybenzoic acid. However,we were unable to isolate any mutants resistant to the weak acidseven at very high density of cells (109 cfu/mL) on 7H11 plates (datanot shown).

Inefficient maintenance of intracellular pH in M. tuberculosisWe compared the intracellular pH of M. tuberculosis and M. smeg-matis in response to changes in external pH (Figure 1a). Between pH5 and pH 7, the two organisms behaved similarly in terms of changesin internal pH. However, under more acidic conditions (pH 35), theinternal pH of M. smegmatis remained fairly stable at values of 5.75.9; in contrast, the internal pH of M. tuberculosis became moreacidic, reaching 5.2 at an external pH of 3.2 (Figure 1a). This indi-cates that M. tuberculosis is less efficient at maintaining the internalpH than M. smegmatis. In addition, valinomycin and nigericin had amore pronounced effect on lowering the internal pH of M. smegmatis

Table 1. Comparison of relative acid sensitivity of M. tuberculosis and M. smegmatis

M. tuberculosis H37Ra and M. smegmatis MC2 were exposedto different acid pH conditions and pH 7.0. The percentage ofsurviving bacteria on day 7 was calculated by dividing the cfuunder acidic pH conditions pH 3, 4, and 5 by that of the controlat pH 7.

Bacterial species pH 3.0 pH 4.0 pH 5.0

H37Ra 0.009 0.073 9.1M. smegmatis MC2 0.67 6.7 33

Table 2. MICs (mg/L) of weak acids for M. tuberculosis H37Ra and M. smegmatis

BA, benzoic acid; SA, salicylic acid; ASP, aspirin (acetyl-salicylic acid);MFA, mefenamic acid; NA, nicotinic acid; INA, isonicotinic acid; NBA, 4-nonyloxybenzoic acid; OBA, 4-octylbenzoic acid; BFCA, benzofurancarboxy-lic acid; DBA, 4-dodecyloxylbenzoic acid; PDA, 11-phenoxyundecanoic acid;LOA, linoleic acid; LAE, linoleic acid ethyl ester. NA, not available.apKa values are from The Merck Index.14

H37Ra M. smegmatis

Weak acids pH 6.8 pH 5.5 pH 6.8 pH 5.5 pKa valuesa

BA 111 37 >333 >111 4.2SA 50100 1020 1000 333 3.0ASP 111 37 1000 333 3.49MFA 33 11 1000 333 4.2NA >500 200 >1000 >500 4.85INA >1000 500 >1000 1000 4.96NBA 33 11 >1000 333 NAOBA 33 3.7 333 111 NABFCA 33 11 1000 111 NADBA 33 11 >1000 >1000 NAPDA 33 11 333 111 NALOA 111 37 >1000 1000 NALAE >1000 >1000 >1000 >1000 NA

by guest on April 30, 2014

http://jac.oxfordjournals.org/D

ownloaded from

Y. Zhang et al.

58

but had little effect on M. tuberculosis (Figure 1b and c). This findinglends further support to the idea that M. smegmatis has a more activeapparatus to maintain its internal pH at acid pH conditions (pH 35)than M. tuberculosis.

Inefficient maintenance of membrane potential in M. tuberculosisWe compared the membrane potential of M. tuberculosis andM. smegmatis in response to changes in external pH. The membranepotential of M. tuberculosis was generally higher than that ofM. smegmatis except at the very acidic pH of 3.5, at which there waslittle difference in the membrane potential between the two organ-isms. However, the membrane potential of M. tuberculosis was moresensitive to changes in external pH than M. smegmatis between pH 4and pH 8.5 (Figure 2). The more responsive change in the membranepotential of M. tuberculosis compared with M. smegmatis is mostlikely due to a poor ability of M. tuberculosis to maintain its mem-brane potential under different external pH conditions.

Correlation between activity of weak acids and their ability to disrupt membrane potential or lower internal pHThe susceptibility of M. tuberculosis and M. smegmatis to weak acidswas examined in the context of membrane potential and internal pH.It was found that the susceptibility of M. tuberculosis to weak acidsappeared to correlate with their ability to disrupt membrane potential(Figure 3a). In contrast, weak acids had little effect on the disruptionof membrane potential in the non-susceptible species M. smegmatis(Figure 3a). The antimycobacterial activity of the weak acids did notcorrelate well with their ability to decrease the internal pH (Figure3b).

Discussion

In this study, we have shown that M. tuberculosis is more susceptibleto acidic pH than the fast growing M. smegmatis (Table 1). Thehigher susceptibility of M. tuberculosis to acid pH is presumably areflection of its poor ability to maintain pH homeostasis than the lesssusceptible M. smegmatis. Indeed, comparison of internal pH inresponse to changes in external pH indicated that M. tuberculosis hasa lower ability to maintain internal pH at acid pH between pH 3 andpH 5 than M. smegmatis (Figure 1a). This is further strengthened bythe finding that valinomycin and nigericin had a significant effect onlowering the internal pH in M. smegmatis but not in M. tuberculosis(Figure 1b and c). The deficiency of M. tuberculosis in maintainingthe internal pH towards neutrality at very acidic pH conditions

Figure 1. Changes in internal pH of M. tuberculosis H37Ra (Ra) and M. smeg-matis mc26 (MC2) in response to external pH and valinomycin plus nigericin.Comparison of internal pH changes in response to external pH is shown in (a).The changes in internal pH of M. tuberculosis and M. smegmatis in response tovalinomycin (V) plus nigericin (N) are shown in (b) and (c), respectively.

Figure 2. Comparison of the membrane potential of M. tuberculosis H37Ra andM. smegmatis MC2 in response to changes in external pH.

by guest on April 30, 2014

http://jac.oxfordjournals.org/D

ownloaded from

Susceptibility of M. tuberculosis to weak acids

59

(pH 35) could result from an increased proton permeability of theM. tuberculosis membrane or a decreased proton extrusion by themembrane-embedded ATPase compared with M. smegmatis. Furtherstudies are needed to distinguish the two possibilities.

An important observation of this study is that M. tuberculosisis uniquely susceptible to a range of weak acids compared withM. smegmatis (Table 2) and indeed other bacteria such as E. coli (datanot shown). The antituberculous activity of the weak acids appearedto inversely correlate with the pKa of the weak acid (Table 2), i.e. thelower the pKa (the stronger the weak acid), the stronger the anti-tuberculous activity. For example, salicylic acid and nicotinic acidhave pKa values of 3 and 4.8, respectively, and their MICs forM. tuberculosis were 1020 and 200 mg/L at pH 5.5, respectively. Inaddition, the antimycobacterial activity of the weak acids wasenhanced at acid pH (Table 2). This is consistent with the fact that,at acidic pH, weak acids become protonated and form unchargedspecies that permeates through the membrane easily compared withcharged anion species.10 Enhanced activity of weak acids at acid pHis consistent with the observation that uptake and accumulation ofweak acids are increased at acidic pH, as shown for pyrazinoic acid.3The consequence of weak acid accumulation and recycling couldlead to disruption of the proton motive force that is required for thetransport of many nutrient substances into bacterial cells as amechanism of action of weak acids.12

The finding that M. tuberculosis is susceptible to weak acids ofdiverse structures suggests that these weak acids do not have aspecific cellular target besides their general effect on disrupting themembrane function. Failure to isolate M. tuberculosis mutants resist-ant to various weak acids is also in keeping with this proposition. Thesusceptibility of M. tuberculosis to weak acids may be a result ofits inefficient ability to maintain membrane potential compared withM. smegmatis. The observation that various weak acids appeared topreferentially disrupt the membrane potential of M. tuberculosis overthat of M. smegmatis (Figure 3a) supports this notion. This differ-ential disruption of membrane potential in M. tuberculosis by theweak acids could result from the slow metabolism and consequentlyslow energy production in the slow growing M. tuberculosis and adefective efflux mechanism as shown for pyrazinoic acid.3

Whereas there is no difference in membrane potential betweenthe two organisms at very acidic pH (pH 3), it is surprising that themembrane potential of M. smegmatis is generally lower than that ofM. tuberculosis (Figure 2). This could indicate that the probe TPP+used to measure the membrane potential is actively extruded by M.smegmatis but not by M. tuberculosis. The observation that valino-mycin and nigericin did not affect the membrane potential inM. smegmatis but did so in M. tuberculosis (not shown) could bedue to valinomycin and nigericin not getting into M. smegmatis cellsor an active efflux mechanism for the membrane potential probeTPP+. Because valinomycin and nigericin were shown to affect theinternal pH of M. smegmatis (Figure 1c), the first possibility of theseagents not getting into the cells can be ruled out. Therefore, it is likelythat M. smegmatis has an active efflux for TPP+, which is responsiblefor the measured lower membrane potential in this organism com-pared with M. tuberculosis.

That M. tuberculosis appears to be uniquely susceptible to weakacids may have implications for the design of new antituberculosisdrugs. However, weak acids may not be easily absorbed through thegastrointestinal tract or bind to serum proteins. To circumvent thispotential problem, it may be necessary to make precursors of weakacids such as ester or amide of weak acids for in vivo use. To showactivity the weak acid precursors will have to be hydrolysed byenzymes present in M. tuberculosis, which is known to contain arange of esterases and amidases in the genome.13 Future studies areneeded to determine whether weak acid precursors can be developedinto antituberculosis agents useful for the treatment of TB.

AcknowledgementsWe thank Peter Maloney for helpful discussions. The researchsupport from NIH (AI-44063) and the Potts Memorial Foundation toYZ is gratefully acknowledged.

References1. WHO. (2002). Global Tuberculosis Control. World Health Organ-

ization [Online.] http://www.who.int/gtb/ (December 2002, date lastaccessed).

2. OBrien, R. J. & Nunn, P. P. (2001). The need for new drugsagainst tuberculosis. Obstacles, opportunities, and next steps. AmericanJournal of Respiratory and Critical Care Medicine 163, 10558.

3. Zhang, Y., Scorpio, A., Nikaido, H. et al. (1999). Role of acid pHand deficient efflux of pyrazinoic acid in the unique susceptibility of Myco-bacterium tuberculosis to pyrazinamide. Journal of Bacteriology 181,20449.

4. Schaller, A., Guo, M., Gisanrin, O. A. et al. (2002). Escherichia coligenes involved in resistance to pyrazinoic acid, the active component of

Figure 3. Relationship between weak acid susceptibility of M. tuberculosisH37Ra and M. smegmatis MC2 and disruption of membrane potential (a) andinternal pH (b). Membrane potentials (mV) along the y axis are negative values.The membrane potential and internal pH values represent the average in tripli-cate. BA, benzoic acid; SA, salicylic acid; ASP, aspirin (acetyl-salicylic acid);NBA, 4-nonyloxybenzoic acid; LOA, linoleic acid. The concentration of theweak acids used in the experiments was 4 mM. by guest on A

pril 30, 2014http://jac.oxfordjournals.org/

Dow

nloaded from

Y. Zhang et al.

60

the tuberculosis drug pyrazinamide. FEMS Microbiology Letters 211,26570.

5. Heifets, L. B. & Iseman, M. D. (1985). Radiometric method for test-ing susceptibility of mycobacteria to pyrazinamide in 7H12 broth. Journalof Clinical Microbiology 21, 2004.

6. Chapman, J. S. & Bernard, J. S. (1962). The pH tolerance ofunclassified mycobacteria. American Review of Respiratory Diseases86, 5823.

7. Portaels, F. & Pattyn, S. R. (1982). Growth of mycobacteria in rela-tion to the pH of the medium. Annales de Microbiologie 133, 21321.

8. Piddington, D. L., Kashkouli, A. & Buchmeier, N. A. (2000). Growthof Mycobacterium tuberculosis in a defined medium is very restricted byacid pH and Mg(2+) levels. Infection and Immunity 68, 451822.

9. Sun, Z. H. & Zhang, Y. (1999). Reduced pyrazinamidase and thenatural resistance of Mycobacterium kansasii to the antituberculosis drugpyrazinamide. Antimicrobial Agents and Chemotherapy 43, 53742.

10. Rottenberg, H. (1979). The measurement of membrane potentialand pH in cells, organelles, and vesicles. Methods in Enzymology 55,54769.

11. Scorpio, A., Lindholm-Levy, P., Heifets, L. et al. (1997). Character-ization of pncA mutations in pyrazinamide-resistant Mycobacteriumtuberculosis. Antimicrobial Agents and Chemotherapy 41, 5403.

12. Zhang, Y. & Telenti, A. (2000). Genetics of drug resistance inMycobacterium tuberculosis. In Molecular Genetics of Mycobacteria(Hatfull, G. F. & Jacobs, W. R., Eds), Chapter 15, pp. 23554. ASMPress, Washington, DC, USA.

13. Cole, S. T. Brosch, R., Parkhill, J. et al. (1998). Deciphering thebiology of Mycobacterium tuberculosis from the complete genomesequence. Nature 393, 53744.

14. Budavari, S. (1989). The Merck Index, 11th edn. Merck & Co., Inc.,Rahway, NJ, USA.

by guest on April 30, 2014

http://jac.oxfordjournals.org/D

ownloaded from