INTERNATIONAL. JOURNAL OF LEPROSY^

volume 56. Number 4

Printed in the U.S. ...I.

Inhibition of Phenolic Glycolipid-I Synthesis inExtracellular Mycobacterium leprae as an

Indicator of Antimicrobial Activity 1 ' 2Eugene B. Harris, Scott G. Franzblau, and Robert C. Hastings 3

Previous aims to study the effect of drugsOn the metabolism ofillycobacterium lepraehave been hindered by the inability to cul-ture the organisms on artificial media. Themouse foot pad (M FP) technique is the mostcommon method for evaluating drug sus-ceptibility in M. leprae ("). This techniqueinvolves measuring the ability of a drug ad-ministered to a mouse to inhibit the limitedmultiplication of .I/. leprae inoculated intothe mouse's foot pad. The principal disad-vantages of this technique are its cost, thetime required (6-9 months) for evaluatinga drug's effectiveness, and the necessity forthe drug to have favorable pharmacokinet-ics in the mouse in order to show activity.This has prompted the recent developmentof a number of in vitro systems (") that relyon inhibition of bacterial metabolic func-tions as indices of drug activity. We havepreviously demonstrated the potential ofadenosine-5-triphosphate (ATP) measure-ment ( 3) and the radio-respirometric deter-mination of palmitate oxidation (') to rap-idly assess drug susceptibility in M. leprae.However, due to the possible presence ofother bacteria in infected armadillo tissuesand human leprosy tissues, it would' be ofvalue to develop, in addition, an assay basedon an M. leprae'-specific metabolite.

Within the past few years, an Al. leprae-specific glycolipid, phenolic glycolipid-I(PGL-I), has been identified and character-ized ( 13). Ramasesh, et al. ( 16) have recently

' Wayne M. Meyers, M.D., Ph.D., kindly served asEditor in regard to the submission, review, revision,and acceptance of this manuscript.

= Received for publication on 22 February 1988; ac-cepted for publication in revised form on 11 August1988.

E. B. Harris, B.S., Research Chemist, BiochemistryResearch Department; S. G. Franzblau, Ph.D., Micro-biologist, Pharmacology Research Department; R. C.Hastings, M.D., Ph.D., Chief, Laboratory ResearchBranch, GWL Hansen's Disease Center, Carville, Lou-isiana 70721, U.S.A.

reported the incorporation of [U 14C] pal-mitic acid ([U ' 4C] PA) into the PGL-1 frac-tion of IL leprae residing in cultured mouseperitoneal macrophages. This procedure hassubsequently been adapted for drug sensi-tivity studies (Ramasesh, et al., unpub-lished data).

We have observed a linear incorporationof [U ' .1C] PA into the PGL-I of extracellular

leprae as well (s). This simplified incu-bation system was sensitive to biophysicalperturbations ( 2) and to the anti leprosyagents dapsone and rifampin. With inhibi-tion of PGL-I synthesis as the criterion ofdrug susceptibility, an expanded series ofestablished antimicrobials were tested foractivity against M. leprae.

MATERIALS AND METHODSM. leprae suspension. Freshly harvested

organisms from experimentally infected footpads of nude mice (HSD athymic nu/nu/AF; Harlan Sprague-Dawley, Inc., India-napolis, Indiana, U.S.A.) were prepared aspreviously described (') in Dubos broth basewithout polysorbate 80 (Gibco Diagnostics,Madison, Wisconsin, U.S.A.) without pHadjustment and containing 20% (v/v) Du-bos medium albumin (Difco Laboratories,Detroit, Michigan, U.S.A.). The medium isreferred to as DA. Briefly, the procedureinvolves initial surface decontamination ofthe foot pads by ultraviolet irradiation, anethyl ether wash, Acudyne (Acme UnitedCorp., Bridgeport, Connecticut, U.S.A.)treatment, and a 70% ethanol rinse. Thefoot pads were then minced and homoge-nized in DA. A slow-speed centrifugation(100 x g for 5 min) removed most of thetissue debris. The organisms were then pel-leted from the supernatant by centrifugationat 2700 x g for 45 min. The final bacterialpellet was suspended in the desired volumeof medium and enumerated. Ampicillin (50µg/ml) and amphotcricin B (2.5 µg/ml) were

588

56, 4^Harris, et al.: Inhibition of PGL-I Synthesis in Vitro^589

added to the medium as protection againstany low-level contamination that might bepresent. Any contaminant in the suspensionmight rapidly overgrow Al. leprae and cornpete for the 14C-labeled substrate. Previousexperimental work has shown that thesecompounds have no effect on Al. leprac (', 18).

Antimicrobials. All antimicrobials wereobtained from Sigma Chemical Co., St.Louis, Missouri, U.S.A., except: ciprollox-acin (Miles Laboratories, West Haven,Connecticut, U.S.A.); clofazimine (CIBA-GEIGY, Summit, New Jersey, U.S.A.) andrifabutin (Farmitalia Erba, Milan, Italy).

In vitro susceptibility testing. Suspen-sions of Al. leprae were diluted in DA con-taining ampicillin (50 pg/m1) and ampho-tericin 13 (2.5 kig/m1) to a concentration of1 x 10 8 bacilli per ml, and 6 ml aliquotswere distributed into 12-m1 screw-cap tubes.Drugs were added at the appropriate con-centrations, and the cultures were subdivid-ed into three 2-ml aliquots. Medium con-taining bacilli but devoid of antimicrobialagents served as controls. The tubes weretightly capped and incubated at 33°C for 4days, after which 0.5 piCi of [U ' 4C] PA (850mCi/mmol; New England Nuclear, Boston,Massachusetts, U.S.A.) was added to eachtube. The reaction tubes were then tightlycapped and incubated at 33°C for an addi-tional 8 days.

PGL-I extraction and quantitation. PGL-Iwas recovered from the lyophilized samplesby a procedure adapted from Hunter andBrennan ( 13) and slightly modified by Franz-blau, et al. ( 5). In brief, the total lipid ex-tracts from the lyophilized samples weresubjected to a biphasic wash, and the gly-colipid fractions in the lower organic layerswere collected by methanolic chloroformelution from silicic acid : celite columns. Af-ter further separation by thin-layer chro-matography (TLC), the areas correspondingto the PGL-I were identified, removed byscraping from the TLC plates, and their ra-dioactivity measured.

RESULTSThe results of the in vitro drug-sensitivity

testing are presented in The Table. Controlvalues in four individual experiments rangedfrom 2.2-8.2 pico moles of [U 14C] PA in-corporated per 2 x 10 8 M. leprae. As de-

THE TABLE. E.Iket of antimicrobial agentson incorporation of [U140 palmitate intoPGL-I fraction of M. leprae.

Antimicrobials% Inhibition' at drug

concentration of

2 pg/mI 20 pg/m1

Clinical antileprosyDapsone 87"^(6)g N D'Rifampin 65" (10) NDclofazimine 48' (16) NDEthionamide + I ND

Protein synthesis inhibitorsErythromycin 77g^(4) 88"^(5)Chloramphenicol 66' (22) 80g^(9)Minocycline 72"^(6)TetracyclineStreptomycin 26'^(7) 42'^(8)Gentamicin 7^(6) 33^(15)Clindamycin 32^(12) 60'^(7)

Cell wall synthesis inhibitorsCephalothin 43'^(II) 56"^(3)Cloxacillin 13^(14) 24^(9)Cycloserine 3^(I) 4^(3)Bacitracin 40^(15) 10^(9)

Nucleic acid synthesis inhibitorsCiprolloxacin 46g^(8) 71"^(1)Nalidixic acid 14^(8) 66g^(2)RillAbutin (LM 427) 43^(5)Griseofulvin 52'^(11) 45^(17)

MiscellaneousCerulenin 76g^(7) 92"^(8)Isoniazid 35^(19) 77"^(5)Polymyxin B 0 0

Controls and antibiotic activities were determinedin triplicate and the results averaged.

p < 0.01 compared to control values, t test.Numbers in parentheses represent standard devia-

tions.' 1 ND = not done.

p < 0.05 as compared to control values, t test.I.2-1.9-fold increase in PGL-1 incorporation as

compared to control values.p < 0.02 compared to control values, t test.

tcrmined by [U 14C] PA incorporation intothe PGL-I fraction, there was decreasedsynthesis of the glycolipid in the presenceof the most frequently used antileprosy drugsdapsone, rifampin, and clofazimine. On theother hand, the antileprosy drug ethion-amide at 2 kig/m1 caused an increase inPGL-I synthesis.

Among the protein synthesis inhibitors,erythromycin and chloramphenicol werehighly active and minocycline, streptomy-cin, and clindamycin also showed signifi-cant inhibition. Cephalothin alone was ac-

590^ International Journal of Leprosy^ 1988

tive among the inhibitors of cell wallsynthesis. The quinolones ciprolloxacin andnalidixic acid demonstrated activity, but theresults with griseofulvin were inconclusive.Cerulenin was highly active as was isoniazidat the higher concentration.

DISCUSSIONIn general, those compounds inhibiting

PGL-I synthesis have also been active inthe in vitro ATP system ( 3) and/or in thetraditional MFP system (6-8. I. 12. 17-19) . I nall three systems minocycline is active whiletetracycline and cycloserine are inactive orweakly active. Streptomycin and clinda-mycin are bacteriostatic in the MCP andweakly active in the PGL-1 system, whilegentamicin had questionable activity in bothsystems.

Erythromycin and ciprolloxacin are ac-tive in vitro but inactive in the MFP. It islikely that the poor pharmacokinetic prop-erties of these compounds in mice are re-sponsible for their lack of in vivo efficacy.Erythromycin could not be detected in theserum of mice receiving erythromycinethylsuccinate at 0.1% w/w in the diet ( 4 ).Pefloxacin, which reaches and maintainshigher serum concentrations than ciproflox-acin in mice, is active in suppressing thegrowth of ..11. leprae in the MFP (9.

In several instances the presence of thedrug at a concentration of 2 µg/ml (mino-cycline, tetracycline, and ethionamide) re-sulted in the stimulation of PGL-I synthe-sis. In the case of minocycline, a 10-foldincrease in drug concentration resulted ininhibition. Low concentrations of variousdrugs have been reported to cause morpho-logical changes in cell walls, including in-creased synthesis of exopolysaccharides(9, 14) .

The antibiotic cerulenin acts upon thecondensation of acyl-acyl carrier protein andmalonyl-acyl carrier protein (ACP), a vitalstep in the biosynthesis of fatty acids (' 5).The almost complete inhibition of PGL-Isynthesis in the presence of this compoundindicates that Al. leprae is actively synthe-sizing PGL-I in vitro. Cerulenin probablyacts in this system by inhibiting a specificACP-dependent elongation system.

It is recognized that the use of albuminin the medium may result in some degree

of drug-protein binding, thereby reducingdrug availability and activity. In this re-spect, the effective in vitro concentrationsof the various drugs tested may not be a truereflection of free drug concentrations underin vivo conditions. Additionally, there is apossibility that amphotericin B and ampi-cillin, present in the medium throughout thetesting procedure, may exert some degree ofantimicrobial synergism. However, their usedoes not negate the usefulness of this pro-cedure, rather it emphasizes the need to ap-proach the screening of antileprosy drugswith caution.

From the data provided by this study, itappears that the synthesis of the Al. leprae-specific phenolic glycolipid can be exploitedas a relatively rapid in vitro assay for drugsusceptibility.

SUMMARY'The effects of 22 antimicrobial agents on

the incorporation of [U' 4C] palmitic acid([U' 4C] PA) into the unique phenolic gly-colipid-1 (PG L-I) antigen of Mycobacteriumleprae were studied. Nude-mouse-propa-gated Al. leprae were incubated in a modi-fied Dubos medium in the presence of anti-microbial agents for 4 days. [U' 4C] PA wasthen added and incubation was continuedfor 8 days. The antileprosy agents dapsone,rifampin, and clofazimine (2 pg/m1 each)caused a significant reduction in [u14c] PAincorporation into PGL-I. Among otheragents, the most active were erythromycin,chloramphenicol, and cerulenin. Low con-centrations of ethionamicle, tetracycline, andminocycline stimulated label incorpora-tion. This system may prove useful in theevaluation of antileprosy agents.

RES UM ENSe estudiaron los efectos de 22 agentes antimicro-

bianos sobre la incorporaciOn del ãcido palmitico mar-cado con "C [(UNC)PA] en el glicolipido fenOlico-1(GLF-I) del Mycobacterium leprae. El M. leprae pro-pagado en ratones desnudos se incubO en un medio deDubos modificado en presencia de los agcntes anti-microbianos durante 4 dias. Enseguida se adicionO(U"C)PA y la incubaciOn sc continuo por 8 dias. Losagentes antileprosos dapsona, rifampina, y clofazimina(2 pg/ml de cada uno) causaron una reducciOn signi-ficante en la incorporaciOn de (U"C)PA en el GLF-I.Entre otros agentes, los más activos fucron la eritro-micina, el cloranfenicol, y la cerulenina. La etiona-

56, 4^Harris, et al.: Inhibition^PGL-I Synthesis in Vitro^591

mida, la tetraciclina, y Ia minociclina estimularon laincorporaciOn de la marca. Este sistema podria resultarout en Ia evaluaciOn de agentes antileprosos.

RESUMEOn a êtudie les effets de 22 agents antimicrobiens

sur l'incorporation de l'acide palmitique marque parl'U"C (U "C PA) dans l'antigene specifique PGL-I deAlycobacu'rium leprae. Des bacilles de la lepre pro-pages chez la souris glabre ont etc incubes pendant 4jours dans un milieu de Dubos modifié, en presencede ces agents antimicrobiens. L'U"C PA a alors etcajoute, et on a poursuivi ('incubation pendant 8 jours.La dapsone, la rifampicine, et la clofazimine (A la dosede 2 Ng/ml pour chacun de ces medicaments antile-preux), a entrains une reduction significative dans l'in-corporation de l'U"C PA dans Ic PGL-I. Parmi lesautres composes etudies, les plus actifs se sont revelesetre l'erythromycine, Ic chloramphênicol, et la ceru-lenine. Des concentrations faibles d'ethionamide, detetracycline, de minocycline, ont stimule ('incorpora-tion du produit marque. Cc systeme pourrait etre utilepour evaluer les medicaments antilepreux.

Acknowledgments. These studies were supported inpart by Public Health Service grant R-22-AI-22492from the National Institute of Allergy and InfectiousDiseases. The authors would like to thank Dr. NaliniRamasesh for helpful discussions, and Mrs. Baljit Ran-dhawa and Mr. Kenneth White for technical assistance.

REFERENCES1. FRANZBLAU, S. G. Oxidation of palmitic acid by

Mycobacterium leprae in an axenic medium. J.Clin. Microbiol. 26 (1988) 18-21.

2. FRANZBLAU, S. G. and HARRIS, E. B. Biophysicaloptima for metabolism of Mycobacterium leprae.J. Clin. Microbiol. 26 (1988) 1124-1129.

3. FRANZBLAU, S. G. and HASTINGS, R. C. Rapid invitro metabolic screen for antileprosy compounds.Antimicrob. Agents Chemother. 31 (1987) 780-783.

4. FRANZBLAU, S. G. and HASTINGS, R. C. In vitroand in vivo activities of macrolides against Mr-cobacterium leprae. Antimicrob. Agents Chemo-ther. 32 (1988) (in press).

5. FRANZBLAU, S. G., HARRIS, E. B. and HASTINGS,

R. C. Axenic incorporation of [U-"C] palmiticacid into the phenolic glycolipid-I of Mycobacte-rium leprae. FEMS Microbiol. Lett. 48 (1987) 407-411.

6. GAtiGns, M. M. Antimicrobial therapy of exper-imental human leprosy (Airco. leprac)infection inthe mouse foot pad. Lepr. Rev. 38 (1967) 225-230.

7. GELBER, R. H. Activity of cycloscrinc and struc-turally related compounds against Al. leprae-in-fected mice. Int. J. Lepr. 52 (1984) 536-537.

8. GELBER, R. H. Activity of minocycline in Myco-bacterium leprae infected mice. J. Infect. Dis. 156(1987) 236-239.

9. GOVAN, J. R. W., FYFE, J. A. M., and MCMILLAN,

C. Mucoid Pseuclomonas aeruginosa and cysticfibrosis: resistance of the mucoid form to carben-icillin, flucloxacillin and tobramycin and the iso-lation of mucoid variants in vitro. J. Antimicrob.Chemother. 4 (1978) 233-240.

10. Guelpa-Lauras, C.-C., Perani, E. G., Giroir, A. M.and GROSSET, J. H. Activities of pefloxacin andciprolloxacin against ..11ycobacterium leprae in themouse. Int. J. Lepr. 55 (1987) 70-77.

11. HASTINGS, R. C. and FRANZBLAU, S. G. Chemo-therapy of leprosy. Ann Rev. Pharmacol. Toxicol.28 (1988) 231-245.

12. HASTINGS, R. C. and JACOBSON, R. R. Activity ofansamycin against .Ilycobacierium leprae in mice.Lancet 2 (1983) 1079-1080.

13. HUNTER, S. W. and BRENNAN, P. J. A novel phe-nolic glycolipid in Mycobacterium leprae-possi-bly involved in immunogenicity and pathogenici-ty. J. Bacteriol. 147 (1981) 728-735.

14. LORIAN, V. Effect of low antibiotic concentrationson bacteria: effects on ultrastructure, their viru-lence and susceptibility to immuno-defense. In:Antibiotics in Laboratory Medicine. 2nd ed. Lor-ian, V., ed. Baltimore: Williams and Wilkins, 1986,pp. 596-668.

15. OMURA, S. Philosophy of new drug discovery. Mi-crobiol. Rev. 50 (1986) 259-279.

16. RAMASESH, N., HASTINGS, R. C. and KRAHENBUHL,

J. L. Metabolism of Mycobacterium leprae in mac-rophages. Infect. Immun. 55 (1987) 1203-1206.

17. SHEPARD, C. C. A kinetic method for the study ofactivity of drugs against Mycobacterium leprae inmice. Int. J. Lepr. 35 (1967) 429-435.

18. SHEPARD, C. C. A survey of the drugs with activityagainst M. leprae in mice. Int. J. Lepr. 39 (1971)340-348.

19. SHEPARD, C. C. and CHANG, Y. T. Effect of severalanti-leprosy drugs on multiplication of human lep-rosy bacilli in footpads of mice. Proc. Soc. Exp.Biol. Med. 109 (1962) 636-638.