IN BRAZIL: XI. OBSERVATIONS ON THE MORPHOLOGY OF OF TOE BRAZILIENSIS AND MEXICANA COMPLEXES

J. J. SHAW and R. LAINSON

The Wellcome Parasitology Unit, Instituto Evandro Chagas,Belém, Pará, Brazil

Reprinled from lhe JOURNALOF TROPICALMEDICINE ANOHVGIENE, January 1976, Vol. 79, pp. 9-13. LEISHMANIASIS IN BRAZIL: XI. OBSERVATIONS ON THE MORPHOLOGY OF LEISHMANIA OF THE BRAZILIENSIS AND MEXICANA COMPLEXES J. J. SHAW and R. LAINSON Wellcome Parasitology Unir, IrlStituto Evandro Chagas,Belém. Pará. Brazil

Introduction H6 L. mexicana amazonensis: isolated While studying strains of Leishmania isolated from a non-ulcerated lesion on the from man and animaIs in the Amazon we thigh of a man from the Paragominas noted that they relI into two natural groups. region of Pará. (2057'S -47°13'W). These we called "fast" and "slow" strains H17 L. mexicana amazonensis: isolated (Lainson and Shaw, 1970) because of their from a lesion of a woman with ex- behaviour in culture and hamsters. It must be tensive lesions on legs and back. emphasizedthat the terms " fast" and " slow " many of which were ulcerated. from do not reter to their behaviour in mano When the River Maracá region of Amapá. examining Giemsa-stainedsmears from infected (OO20'S-51 °35'W). hamsters we gradually became aware that we H9 L. braziliensis braziliensis: isolated could distinguish the "fast" and "slow" from a single ulcer infection below strains. At first we thought that it was possibly lhe knee of a man from the Itinga due to differences in the cellular picture and region of Pará. (4°36'S -47°43'W). number of parasites, but blind tests in fields Ml678 L. braziliensis braziliensis: isolated with similar numbers of parasites and cells from a small single ulcer on the enabled us to distinguish parasites of the two thigh of a man from the Serra Norte groups, which we termed the braziliensis com- region of Pará (Lainson et al., 1973). plex (slow) and the mexicana complex (fast) (Lainson and Shaw, 1972). Clearly we were Amastigotes of the different strains were recognizingbasic morphological differences but obtained from non-ulcerated cutaneous nasal we could not be certain what they were. lesions of golden hamsters (Mesocricetus To determine the morphology of the ama- auratus). stigotes more precisely we decided to apply Previous morphological studies on Sarcocystis the classicalprocedures used to study the mor- zooites (Lainson and Shaw, 1969) showed that phology of trypanosomes. It soon became the methods of fixation and collection signifi- apparent that camera lucida drawings of such cantly effect size. Preliminary observations on small organisms were not very reliable. We Leishmania amastigotes showed that direct therefore used pho~omicrographic techniques smears from lesions were not a suitable source to determine the various mensural characters of parasitesas there was considerable variation (Shaw and Lainson, 1972) of the amastigotes. in shape depending on the pre~nc.e or absence This paper reports the basic morphological of blood. In general amastigotestended to be characters of some strains of the braziliensis and mexicana complexes. torpedo shaped when associated with blood. Least variation in shapewas noted in impression smears made from a piece of tissue from the Materiais and Methods lesion whose surface had been gently blotted The following Leishmania strains were used:- on filter paper. Smearsmade in this way were PH7 L. mexicana amazonensis: isolated fixed in Bouin's fluid and stained with a modi- from a naturally infected Lutzomyia fied Giemsa technique (Lainson, 1959;Wenyon, flaviscutellata captured in the Utinga 1926). Photomicrographs of arnastigotes were forest, Belém. (Lainson and Shaw taken on Adox KB 17 or Agfa Isopan IF 35 1968). fim. film using a Zeiss WL research micro- M22 L. mexicana amazonensis: isolated scope with a Planapo 100/1'3 oil objective and from a naturally infected Oryzomys the Zeiss 11 carnera body mounted on a sup- capito captured in the Utinga forest, port armoEach fllm included a photomicrograph Belém (Lainson and Shaw, 1968). of the scale of a slide micrometer which was

The H6 H9 plastjposterior end to middle óf nucleus (I1

Results I I A summary of the various mensural characters I micron is given in Table I. Analysis by the F-test Fig. (i). Amastigotes of L. m. amazonensisshowing mean mensural characters of strains H6 and H9 in showed that there were highiy significant dif- diagramatic formo ferences betweenthe lengths of L. m. amazon- ensis strains. ranging from 3' 18 to 3' 72 p.m printed with each ftlm and used to determine with a mean of 3' 44 p.m.The degreeof overlap lhe final magnifications of the prints. The sizes betweenthe smalIer strain M22 and lhe larger of the organisms.organelles and their positions strain H17 was 23 per cento The overlap be- were measuredfrom the photomicrographs with tween the two strains of L. b. braziliensis was a pair of sliding callipers. The anterior end was 26 per cento the mean major diameter being taken as being the end of lhe organisms in 2. 4 p.m. The overlap between the smaller which the ftagellar vacuole occurred (Figure i). L. m. amazonensisstrain M22 and the larger Dividing forms with two nuclei or kinetoplasts L. b. braziliensis strain H9 was 12 per cento were not included in the calculations of the while the overlap of the major diameter for the mean biometrical characters.Wherever possible two groups was 13 per cento The mean area extracellular parasites were measured. The of the amastigotesof strain M22 was 5.34 p.m2 meansand standard deviations were determined and H9 was 3,78 p.m2with an overlap of area for lhe different mensural characters including of 24 per cento The PK/pN fatia was greater the major diarneter of the kinetoplast. Vari- than one for the strains of L. m. amazonensis ations within and betweenthe two groups were and approximately one for the L. b. braziliensis analysed using the F-test and differences be- strains indicating that the kinetoplast is more tween means were compared using the fatia anterior to the-nucleus in the former. The over- d (Bailey. 193:,9).The percentageof overlap of lap of the distance P- K betweenstrains M22 certain characters was estimated using the and H9 was 20 per cent and betweenM22 and M1678 15.9 per cento The PKjKA fatia in- formula C= 1/2 (xl -x2). where S is the dicated that the kinetoplast is more posteriorly S placed in the two L. b. braziliensis strains and common estimate of the standard deviation of the overlap of this character between strains the two popu1ations as ca1cu1atedfor a l-testo M22 and H9 was 28.4 per centoThe PNjNA The percentageof over1apis 1- P, where P is fatia showed that the nucleus is more centralIy the integral of the normal distribution from placed in the L. b. braziliensis strains. -C to infinity. Thus if the above gives a deviate The sizesof the nuclei of alI the strains were C= 3.24, tab1esof the normal distribution give similar ranging from 1.2 X 1.01 p.m to 1.35 P=0.9994 and the over1apis 1-0.9994=0.06 Xl' 05 p.m. The kinetoplasts of the L. b. per cent [in the original examp1e(Shaw, 1969) braziliensis amastigoteswere significantly larger ..C of 23.4" shou1dhave read ..C of 3.24 "]. than those of L. m. amazonensis.An F test of Ratios for the fo110wingcharacters were also the major diameter of the kinetoplast was signi- ca1cu1ated;posterior end to midd1e of kineto- ficant for the L. m. amazonensisstrains but not

1)1::':)7~1J TABLB I. Mensural data in microns of four strains of Leishmania mexicana amazonensis and two strains of L. bra- ziliensis braziliensis isolated froro man and wiId animaIs in the state of Pará, Brazil (Data expressed as means ::!: standard deviations and range).

for the two L. b. braziliensisstrains. The overlap major (L. major). Doubt as to the validity of of the major diameter of the kinetoplast be- this differentiation was expressed by some of tween strains PH7 and M1678 was nine per cento the Russian workers but Kellina (1962), in a Amastigotes of L. m. amazonensisin general detailed study of the sizesof the parasites from had larger flagellar vacuoles. The mean men- 13 human cases of , sural characters of strains H6 and H9 are concluded that there were significant differences shown in diagramatic forro in Figure (i). - in sizes of amastigotes associated with the urban and zoonotic forms of the disease.The DiscvssioD mean sizes of parasites of the urban fornl (L. tropica) were 3,33 X 1,99 p.mwhile those Such statements as "the leishmania are a of the zoonotic form (L. major) measured group of morphologically identical parasites" 4,48 X 3, 33 Jl.m. She emphasized that great (Schnur et al.. 1973) have often appeared in cale was needed in the making of the smears the literature. Muniz and Medina (1948), how- from which amastigoteswere measured. ever, described a species of Leishmania from the guin~a pig whose amastigotes measured Our results indicate that the amastigotesof 5.2 X 2.5 p.m, which is considerably larger L. m. amazonensisare larger than those o-fL. b. than speciesknown to infect manoAd1er (1963) braziliensis, thus, as in the Old World, there states that Parrot and bis colleagues appear to be two morphologically different showed that in dogs the amastigotes of L. groups of parasites. It was also noted that the tropica were considerably larger than those of major diameter of the kinetoplast of L. m. L. infantum. The Russian workers have studied amazonensiswas smaller than that of L. h. the gizes of the parasites causing urban and braziliensis. Apart from these two major rural cutaneousleishmaniasis (see Kellina, 1962 differences there were also diflerences in the for a review of these works). On the basis of position of the kinetoplast and nucleus. In L. size two subspeciesof L. tropica were created m. amazonensis amastigotes the kinetoplast L. tropica minar (L. tropica) and L. tropica was more central and the nucleus more pos- terior. The closer approximation of the nucleus to be quite reasonable especialIy in the light and kinetoplast in L b. -:braziliensis and-tlre .-of KelIina's biometrical studies on L. tropica. sma-ller size of this parasite often makes it It might be mentioned that it is rewarding to difficult to determine the boundaries of the in- see that some biochemical characters, such as dividual organelles in intracellular parasites nuclear and kinetoplast DNA buoyant densities and the amastigotes may appear as Histo- (Chance et ai., 1973), conform remarkabIy plasma-like bodies. welI with identifications based on bio10gica1 The statistical differences between mensural and epidemio10gicaldata. characters is clearly significant if the tech- It has been noted that Iarger strains of Dique is sufficiently standardized. Walker trypanosomes such as T. vivax (Fairbaim, (1963) noted that batches of Giemsa stain of 1953), T. evansi (Hoare, 1956) and T. congo- different ages caused notable differences in the lense (Godfrey, 1961) alI appear to be more total lengths of T. brucei trypomastigotes. We pathogenic for domestic animaIs than are the have measured the sizes of amastigotes of smalIer ones. In this respect it is interesting strain M22 from different hamsters ;;taitled to note that the 1arger parasite, L. mexicana, with different batches of Giemsa but noted is more viru1ent in hamsters than the smalIer no significant variation in their major dia- L. braziliensis,and that the larger 0Id World meters. We are thus led to conclude that varia- parasite L. major. is more pathogenic for tions in size between the different L. m. mice that the smalIer L. tropica (Moskovski amazonensisstrains is most probably a strain and Southgate, 1971). specific variation rather than an artifact due to technique. What then is the taxonomic im- Summary portance of statistically significant differences? The present work shows that the amastigotes Much has been written about this and it Ís of L. m. amazonensis are morpho10gicalIy generally taken that the degree of separatioll is the leveI for specific or subspecific ranking. distinguishab1efrom those of L. b. braziliensis. The authors consider that such differences Such levels are wel1 beyond statistically add further support to the use of the name significant differences and are clearly not braziliensis and mexicana at the specific leveI. equivalent in, for example, sexually and asex- ually reproducing organisms. Shaw (1969) sug- gested that a degree of overlap of 10 per cent Acknowledgement or less approximated the morphological differ- These studies were conducted under the ences used by most workers for species of auspices of the WelIcome Trust, Londol1, the haemoflagellates. If the technique is at fault Evandro Chagas Institute of the Fundação such distinctions may be accentuated but such Instituto Oswa1do Cruz of Brazi1, and the factors as the quality of stain (Walker, 1963) World Hea1thOrganization. or fixative (Lainson and Shaw, 1971) still We are most grateful to Roberto Naiff, only give rise to overlaps of 30 to 40 per cent SebastiãoOliveira, and Henrique Bul1a for their in identical material. ab1etechnica1 assistance. Amongst the haemoflagellates morphology has for long been the standard character for References the separation of speciesand at present much Adler, S. (1963). of the research on haemoflagellatesis devoted Dilferentiation of from L. to the determination of biochemical characters mexicana and L. tropica. Rev. Inst. Salubr. Enferm. that may be of taxonomic use. The present trop., v23, 139. studies, however, show that there are mor- Bray, R. S., Ashford, R. W. and Bray, M. A. (1973). The parasite causing cutaneous leishmaniasis in phological differences between parasites classi. Ethiopia. Trans. R. Soco trapo Med. Hyg., v67, 345. fied as belonging to the mexicana and bra-ziliensisChance, M. L., Peters, W. and Griffiths, H. W. (1973). complexes (Lainson and Shaw, 197:'). A comparative study of DNA in the genus Leish- Such differencesare within the range of specific mania. Trans. R. Soco trapo Med. Hyg., v67, 24. Fairbairn, H. (1953). differentiation of the haemoflagellatesand we Studies on Trypanosoma vivax. IX. Morphological therefore, seeno reason why such old 'specific' dilferences in strains and their relation to patho- names as L. mexicana and L. braziliensis genicity. Ann. trapo Med. Parasit., v47, 394. should be soft pedalled as was implied by Godfrey, D. G. (1961). Types of Trypanosoma congolense. ll. Dilferences Lumsden (1974). Similar~ythe raising of major in the courses in infection. Ann. trapo Med. Parasit., to specific status by Bray et ai. (1973) seems v55, 154. Hoare, C. A. (1956). Lainson, R. and Shaw,J. J. (1972). Morpho1ogical and taxonomic studies on mam- Leishmaniasis of the New World: Taxonomic malian trypanosomes VIII. Revision of Trypano- problems. Brit. medo Buli., v28, 44. soma evansi. Parasitology, v46, 130. Lainson, R., Shaw,J. J., Ward, R. D. and Fraiha, H. Kellina,O. L (1962). (1973). O razmerakh leishmanial'nykh forro leishmania Leishmaniasis in Brazil: IX. Considerations on the tropica major i minar. Med- Leishmania braziliensis complex:- importance of skaya Parazit., v3I, 716. sandflies of the genus Psychodopygus (Mangabeira) (A translation of this paper is available from the in the transmission of L. braziliensis braziliensis in British National Lending Library, Boston Spa, as north Brazil. Trans. R. Soco trop. Med. Hyg.. v67, paper RTS 7567 of the Translating Programme). 184. Lainson,R. (1959). Lumsden,W. H. R. (1974). Atoxoplasma Garnham, 1950, as a synonym of Biochemical Taxonomy of Leishmania. Trans. R. Lankesterella Labbé 1889. Its life cycle in the SOCotrapo Med. Hyg., v68, 74. English sparrow (Passer domesticus domesticus). Moskovskij, S. D. and Southgate, B. A. (1971). J. Protozool., v6, 360. Clinical Aspects of Leishmaniasis with Special Lainson, R. and Shaw,J. J. (1968). Reference to the U.SSR. Buli Wld Hlth Org., v44. Leishmaniasis in Brazil: I. Observations on enzootic 491. rodent leishmaniasis -incrimination of Lutz~myia Shaw,J. J. (1969). ftaviscutellata (Mangabeira) as the vector in the The Haemoflagellates of Sloths. London School Lower Amazonian Basin. Trans. R. Soco trop. Hyg. trop. Med., Memoir 13. H. K. Lewis &; Co. Med. Hyg., v62, 385. Ltd.. London. Lainson, R. and Shaw,J. J. (1970). Shaw,J. J. and Lainson, R. (1972). Leishmaniasis in Brazil: V. Studies on the epi- Trypanosoma vivax in Brazil. Ann. trapo Med. demiology of cutaneous leishmaniasis in Mato Parasit., v66, 25. Grosso State, and observations on two distinct Walker, P. J. (1963). strains of Leishmania isolated from man and forest Variation in the length of trypanosomes after animaIs. Trans. R. Soco trop. Med. Hyg., v64, 654. different methods of staining. Trans. R. SOCotrop. Lainson,R. and Shaw,J. J. (1971). Med. Hyg., v57, 237. Sarcocystis gracilis n. sp. from the Brazilian Tor- Wenyon,C. M. (1926). toise Kinosternon scorpioides. J. Protozool., vIS, Protozoology. London Bailliére. Tindali and Cox, 365. p1324.

CORRIGENDUM." Fig.(i)o Amastigotes of ~o ~o ama,zonensis showing .000 should read Fig.(i). Amastigotes of ~o T.° amazonensis & ~. ~. braziliensis showing .O OO O The following two references included in the text did not appear in the Referenceso Schnur, Lo F., Zuckerman, Ao, & Greenblatt, C. Lo (1973). The relationship between the clinical types and serotypes of Leishmaniao --J. Protozoolo, (Supplement), 20, 5340 Muniz, Jo, & Medina, H. (1948)0 Leishmaniose tegumentar do çobaio .Leishmania Enriettii no spo ~-.!!9l, Rio de Janeiro, ~, 7-25.