526

TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL MEDICINE AND HYGIENE (1990) 84, 526-529

Biochemical characterization of Leishmania (Viannia) braziliensis and Leishmania (Viannia) peruviana by isoenzyme electrophoresis

M Araua1S233, D. A. Evans’, A. Zolessi’, A. Llanos Cuentas”’ and J. Arevalo’p2F4’ ‘Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, P.O. Box 5045, Lima 100, Peru; %entro de Investigacio ‘Hugo Lumberas Cruz’, Instituto National de Salud, Lima, Peru; 3Department of Biology and 4Department of Biochemistry, Universidad Peruana Cayetano Heredia; SDepartment of Medical Parasitology, London School of Hygiene and Tropical Medicine, Keppel Street, London, WCIE 7HT, UK

Abstract Leishmanial organisms isolated from 24 patients

with Andean cutaneous leishmaniais (uta) and from 7 with sylvatic leishmaniasis in both cutaneous and mucosal forms were characterized on the basis of their isoenzyme profiles for 13 enzymes using both cellu- lose acetate (CA) and thin-layer starch gel (TLS) electrophoretic techniques. Malate dehydrogenase (MDH) after electroohoresis on CA or TLS and mannose phosphate i’somerase (MPI) on TLS were the only enzymes of 13 examined which discriminated between the organisms from patients with uta (L. (V.) r~~viana) and those with svlvatic leishmaniasis (L. IV., lna.&ensis). Mannose phosphate isomerase gave more clear-cut and reproducible discrimination than did MDH on either TLS or CA. and it is sueeested that MPI is a reliable enzyme marker that can 6: used in routine TLS electrophoresis to distinguish between L. (V.) peruviana and L. (V). braziliensis.

Introduction The Leishmania (Viunnia) braziliensis complex of

organisms includes parasites that are responsible for a varietv of different clinical manifestations of leish- maniasis in Central and South America. The course of the disease has been correlated with the species of Leishmania involved (LAINSON & SHAW, 1987); thus pian bois is associated with L. (V.j guykensis infection, uta with L. (V.) peruviana, both mucosal leishmaniasis (MCL) and simnle cutaneous leish- maniasis (CL) -with L. (V.) bra&ensis, and Panama- nian CL with L. (V.) panamensis.

When epidemiological studies are carried out in countries where the presence of more than one species of Leishmania is suspected, it is essential that charac- terization studies are carried out in order to establish the geographical distribution of the different organ- isms, and to forewarn the clinicians about the presence of aggressive organisms such as L. (V.) braziliensis in the lesions of their patients.

Two forms of leishmaniasis, Andean cutaneous leishmaniasis (uta) and sylvatic leishmaniasis (SL), are prevalent in Peru. Uta is found in the western Andean and inter-Andean valleys, whereas SL is widespread in the Amazonean jungle. Limited areas of overlap exist between both diseases, but these are increasingly complicated by migration of people back and forth between areas endemic for each disease.

ESCOMEL (1913) was the first to suggest that the

‘Author for correspondence and offprint requests.

aetiological agents were different for uta and espun- dia. His conclusion was based on differences in their clinical manifestations and geographical distribution. Recent studies using a variety of characterization techniques, including monoclonal antibodies specific for L. IV.) braziliensis. zvmodeme analvsis. and deoxyribonucleic acid (DNA) hybridization studies (ROMERO et al., 1987), which have looked for differences between Andean and sylvatic stocks, all failed to discriminate between them. The degree of similarity found was in marked contrast to the remarkable differences in the clinical manifestations of the Andean and sylvatic diseases. Thus uta lesions are rather benign, with rapid self-healing and good response to treatment with Glucantime@. SL, on the other hand, is more difficult to treat, the disease is more aggressive? and secondary metastatic spread can lead to the lughly destructive mucosal form of leishmaniasis known as espundia.

Here we report the use of the enzymes mannose phosphate isomerase (MPI) and malate dehyd- rogenase (MDH) to distinguish between the causative agents of uta and espundla.

Materials and Methods Parasites

Leishmanial organisms were isolated into culture from biopsy matesal taken from patients presenting with either Andean or svlvatic leishmaniasis. The organisms together with their geographical locations are listed in the Table. The parasites were cultured and cryopreserved as reported previously (ROMERO et al., 1987).

Preparation of lvsates About 1 x 10” promastigotes were harvested from

culture. washed in nhosnhate-buffered saline (PBS) at DH 7.4, then resuspended in an enzyme stabilizing solution (2 mM ethylene diaminetetraacetic acid, 2 mM l -aminocaoroic acid. DH 7-O). The oreanisms were lvsed bv 3& cvcles of ireezini and tha;fring in liquid htrogen and stored either ai crude homogen- ates at -79°C. or after hleh sneed centrifueration as soluble fractions under liquids nitrogen. -

Isoenzyme analysis Cellulose acetate electrophoresis. The following en-

zymes were analysed: aspartate aminotransferase (E.C.2.6.1.1 ASAT), alanine aminotransferase (E.C.2.6.1.2 ALAT), glucose-6-phosphate dehyd- rogenase (E.C. 1.2.1.49 G6PD), glucose phosphate isomerase (E.C.5.3.1.9 GPI), 6-phosphogluconate de-

527

Table. Stocks of Leishmania used in this study

Code Lesiona Department Region Iosenzyme identification

MHOM/PE/83/LCOl MHOMIPE1841LC03

,OMIPE184/LC48 .OiWPE/84/LC53 :OM/PE/84/LC58 :OM/PE/84/LHlOS :OM/PE/86/LH282 :OM/PE/84/LC26 ;OMIPE/84iLC27 ;OM/PEi84/LC32 .OMIPEW4lLC37 :OMIPEl84lLC39 :OM/PEi85/LC106 :OM/PEI87ILC157 :OM/PE/87/LC180 :OMPE/87/LC187 :OMIPE/87/LC196 :OM/PE/87/LC202 :OM/PE/84/LC30 :OM/PE/84/LH78 :OM/PE/SWLH 107 :OMPE/84/LH108 :OM/PE/87/LH387 :OM/PE/87/LH379 !OM/PE/87/LH389 :OMIPE/87iLH440 :OMIPE/87/LH452 :OM/PE/87/LH460 :OM/PE/87/LH469 :OMIPE/88/LH483 :OMIPE/88/LH488

MCL MCL MCL MCL CL

MCL

:“L

MH MH MH

El

iti: MH

is:: MCL MCL

Madre de Dios Madre de Dios cuzco Madre de Dios Madre de Dios Madre de Dios Huanuco Ancash Ancash Ancash Ancash Ancash Lima Ancash Ancash Ancash Ancash Ancash Ancash Ancash Huanuco Lima Ancash Lima Lima Lima Lima Lima Lima Lima Lima

Jungle Jungle Jungle Jungle Jungle Jungle Jungle Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean Andean

L . braziliensis L. braziliensis L. brazilimsis L. braziliensis L . braziliensis L braziliensis L . brazilimsis L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruvia?la L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L. peruviana L . peruviana L peruviana L. peruviana L. peruviana L. peruviana

YZL=cutaneous leishmaniasis; MCL=mucosal leishmaniasis.

hydrogenase (E.C. 1.1.1.44 6PGD), phosphogluco- mutase (E.C.2.7.5.1 PGM), mannose phosphate isomerase (E.C.5.3.1.8 MPI), malic enzyme (E.C.1.1.1.40 ME), and malic dehydrogenase (E.C.l. 1.1.37 MDH). The electrophoretic conditions were as described by KREUTZE~ & CHRISTENSEN (1980). exceot for MDH where the conditions were as given’& MILES et al. (1980). Enzyme staining was bv Fhe meihod of LANH~M et ‘al. (1981). Crude-extracts for MDH isozvme analvsis were incubated with Nonidet@ P-40 -(a non-ionic detergent) at 3% (v/v) final concentration for 15 min at room temperature.

Thin-layer starch gel electrophoresis. In addition to ALAT, ASAT, GPI, MDH, MPI, 6PGD and PGM described above, the following enzymes were analysed by the methods described by EVANS et al. (1984): esterase (E.C.3.1.1.1 ES), nucleoside hydrolase (E.C.3.2.2.1 NH), proline iminopeptidase (E.C.3.4.11.5PEPD),pyruvatekinase(E.C.2.7.1.40 PK), and superoxide dismutase (E.C.1.15.1.1 SOD).

Isoelectric focusing of munnose phosphate isomerase. Crude homogenates of leishmanial promastigotes prepared as described above were electrofocused on agarose gels. The isoelectric focusing was carried out in an LKB horizontal electrophoretic apparatus with constant cooling (2°C). The agarose gels were made as follows: 0.8% agarose IEF (Pharmacia); 12.5%

glycerol; and 1.1 ml, 0.5 ml, and 0.5 ml respectively of the following ampholines: pH 2-11, pH 3-5, pH 5-7 (Servalyta) for a gel of 40 ml final volume. Flat gels (22.5 cmx 11.5 cmx0.1 cm) were run at 1000 V and 2” for 1 h. Electrode wicks soaked in 0.5 M NaOH and 0.5 M phosphoric acid were used as cathode and anode respectively. Enzyme activity was vizualized by overlaying the electrofocusing gel with an agar gel containing 0.1 M Tris/HCl pH 7.4, 0.66 M mannose-6-phosphate, 0.13 mM nicotina- mide adenine dinucleotide phosphate, 10 mM M&l,, 4 units of GPI, 2.5 units of G6PD, 0.32 mM methylthiazole tetrazolium, 0.22 mM phenozine methosulnhate and 0.6% (w/v) Noble aear. Incuba- tion was-for 15 mitt at 37C.’

Results Characterization of the parasites

Twenty-four isolates of Leishmania obtained from patients suffering from uta and 7 isolates from patients with SL in cutaneous and mucosal forms were studied (Table).

Of the 12 enzymes examined by thin-layer starch gel electrophoresis, only MPI gave consistently diffe- rent patterns for uta and sylvatic isolates (Fig. la, b), with the enzyme from Andean isolates migrating more slowly than that from sylvatic isolates. MDH elec-

MPI MPI

- --.“l.-~-- 123456 78 12 34 5 6 7

MDH MDH

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7

Fig. 1. Photographs of thin layer starch gel zymograms of the enzymes MPI and MDH. (a) MPI: 1, L. br&Lkmis marker (M2903); 2, ,!,. po~mensis marker (LS94); 3, L. guyanensis marker (M4147); 4, L. brazilienris marker (LTB300); 5, sylvatic stock (LCOl); 6, sylvatic stock (LC03); 7, sylvatic stock (LC53); 8, Sylvatic stock (LHlOS). (b) MPI: 1, L. braailienris marker (M2903); 2, L. pmuzdr marker (LS94); 3, L. guyan& marker (M4147); 4, L. braziliensis marker (LTB300); 5, Andean stock (LC26); 6, Andean stock (LC39); 7, Andean stock (LH78). (c) MDH: tracks l-8 as in (a). (d) MDH: tracks l-7 as in (b).

PI

5.6

5.7

5.8

la9 LC53 Fig. 2. Diagram of the isoelectric focusing patterns of the enzyme MPI for Andean stock LC39 and the sylvatic stock LC53.

trophoretic patterns on both cellulose acetate (not shown) and thin-layer starch gels (Fig. lc, d) discri- minated in most cases between the organisms isolated from uta patients (L. (V.) penczkzna) and those from SL patients (~5. (V.) bruziliemis). On both matrices MDH from Andean isolates migrated faster than that from sylvatic isolates, but a degree of infra-specific isoenzyme variation was encountered that sometimes made the interpretation of the zymograms difficult (Fig. lc). The MDH isoenzyme patterns were distinct and reproducible on starch, but on cellulose acetate they depended on the conditions used to lyse the promastigotes. Incubation of the lysate for 15 min at room temperature with Nonidet@ P-40 (3% final concentration) was necessary in order to reduce smearing of the patterns on cellulose. The other 9 enzymes tested on cellulose, and 10 tested on starch, gave identical patterns for both Andean and sylvatic isolates.

Isoelectric point of mannose phosphate isomerase The MPI isoenzyme isoelectric points were deter-

mined from Andean and sylvatic isolates (Fig. 2). Two isoenzymes were present in organisms from both sources, but they showed differences in their respec- tive values of p1; the MPIs from L. (V.) peruviana (LC39) had pIi=S-77 and pIr=5.68, and those from

L. (V.) braziliensis (LC53) had pIi=5.72 and pI=5.64.

Discussion There has been some controversy in recent years as

to whether or not L. (V.) peruviana exists as a taxonomic entity separate from L. (V.) braziliensis. Clinical, epidemiological and geographical evidence all support the assumption that classical uta from the Peruvian Andes is caused by an agent different from that responsible for the cutaneous and mucosal leishmaniases from the forest areas of the country. Nevertheless DNA hybridization, monoclonal anti- bodies and, until now, isoenzyme techniques all failed to distinguish between the 2 organisms-(ROMERO et al.. 1987). This has been comnounded to some degree by’the habit of labelling most leishmanial isol%es from CL patients in Peru as ‘uta’ strains. The isoenzyme study reported here shows for the first time that there are useful biochemical markers for the organisms causing Andean cutaneous leishmaniasis (uta) and sylvatic leishmaniasis in Peru. The markers are the enzymes MPI and MDH. Using these markers we were able to demonstrate that, as predicted, the organism responsible for MCL and simple CL in the jungle regions of Peru we investigated (Madre de Dios, Cuzco and Huanuco) was L. (V.) braziliensis, while that from the Andean regions was L. (V.) peruviana. There were, however, 2 patients from Lima who were suffering from mucosal leishmaniasis and from whom we isolated organisms isoenzymically indistinguishable from L. (V.) peruviana (LH460 and LH469, Table). The mucosal lesions of these patients were very much less dramatic than those of patients infected with L. (V.) braziliensis and, in contrast to the true espundia patients, the lesions were confined to the nose and made excellent and rapid responses to antimony treatment. Most probably the mucosal lesions caused by L. (V.) peruviana were primary lesions as opposed to those of the true espundia patients, where one would expect the mucosal in- volvement to be secondary as a result of metastatic spread from a primary lesion elsewhere on the skin.

Two kinds of supporting matrix were used in the electrophoretic separation of the isoenzymes, thin- layer starch and cellulose acetate. Thin-layer starch gels gave better resolution than cellulose acetate with MDH, but both methods showed uta strains to have a single MDH band which migrated faster than that of the sylvatic strains. Some problems were encountered with streaking of the MDH bands on cellulose acetate, but this was largely overcome by the incuba- tion of the lysed organisms with a non-ionic detergent (Nonidet@ P-40). Infra-specific isoenzyme variation of MDH also presented some problems in the interpreta- tion of the zymograms, in that some strains of L. (V.; braziliensis. for examnles M2903, had an MDH which migrated faster than that of the WHO reference strain ILTB300) and brought it verv close to that of the uta strains (Pig. lc, d): ’

MI’1 on thin-layer starch, however, proved to be a good, reliable enzyme marker and showed clear discrimination between uta and the svlvatic strains. This was further corroborated by the finding that the isoelectric wints of MPI from svlvatic strains were more acidic than those from uta. MPI was usually resolved into 2 bands by electrophoresis on starch

529

gels, but occasionally only a single rather diffuse band stained. due nrobablv to a degree of overloading of the sample’ on io the gel and ‘ihe resultant faihire to discriminate between two closely running bands. This explanation is supported by the finding that the isoelectric points of the 2 isoenzymes of MPI were very close, with values of 5.77 and 5.68 for uta, and 5.72 and 564 for svlvatic strains. Whv MPI on cellulose acetate failed to discriminate between the 2 organisms, when the difference was so clear-cut on starch, is. not immediately apparent, but it was nrobablv due to the different nhvsical and chemical properies of hydrolysed potato starch and cellulose acetate.

MPI is therefore an enzyme marker that can be used routinely in thin-layer starch gel electrophoresis to discriminate between L. (V.) braziliensis and L. (V.) peruviana. From an isoenzyme point of view these organisms are no more difficult to separate than 2 other members of the L. (V.) braziliensis complex, namely L. (V.) panamensis and L. (V.) guyanensis, which are consistently discriminated between by only one (6PGD) of the 12 enzymes routinely used for characterization purposes.

Acknowledgements This study received financial support from the UNDPi

World Bank/WHO Special Programme for Research and Training in Tropical Diseases, and from the Research Development Programme in the Field of Science and Technology for Development, Commission of the European Communities (Project TSD-312).

References Escomel, E. (1913). Premiere decouverte de Leishmania

tropica flagellee dans le corps humain. Bulletin de la Socitti de Patholoeie Exotiaue. 6. 237-238.

Evans, D. A., Lanha& S. M.,,‘Baldwin, C. I. & Peters, W. (1984). The isolation and isoenzyme characterization of Leishmania braziliensis subsp. from patients with cutaneous leishmaniasis acquired in Belize. Transactions of the Royal Society of Tropical Medicine and Hygiene, 78, 35-42.

Kreutzer, R. D. & Christensen, H. A. (1980). Characteriza- tion of Leishmania spp. by isozyme electrophoresis. American Journal of Tropical Medicine and Hygiene, 29. 199-208.

Lainson, R. & Shaw, J. J (1987). Evolution, classification and geographical distribution. In: The Leishmaniases in Biology and Medicine, Vol. l,, Biology and Epidemiology Peters, W. & Killick-Kendrick, R. (editors). London: Academic Press, pp. l-120.

Lanham, S. M.. Grendon. 1. M.. Miles. M. A.. Povoa. M. & de’Souza,A. A. (1981j. A comparison of electrophore- tic methods for isoenzyme characterization of trypanoso- matids. 1: standard stocks of Trypanosoma cruzi zymodemes from northeast Brazil. Transactions of the Royal Society of Tropical Medicine and Hygiene, 75, 742-750.

Miles, M. A., Lanham, S. M., de Souza, A. A. & Puvoa, M. (1980). Further enzymic characters of Ttypanosoma cruzi and their evaluation for strain identification. Transactions of the Royal Society of Tropical Medicine and Hygiene, 74, 221-237.

Romero, G. G., Arana, M., Lopez, I., Montoya, R., Bohl, R., Campos, M., Arevalo, J. & Llanos, A. (1987). Characterization of Leishmania species from Peru. Trans- actions of the Royal Society of Tropical Medicrne and Hygiene, 81, 14-24.

Received 31 August 1989; revised 22 Januap 1990; accepted for publication 23 January 1990