Immunogenetics of Human American Cutaneous Leishmaniasis Study of HLA Haplotypes in 24 Families from Venezuela M. L. Lara, Z. Layrisse, J. V. Scorza, E. Garcia, Z. Stoikow, J. Granados, and W. Bias

ABSTRACT: Twenty-four families with one or multiple cases of localized cutaneous leishmaniasis (LCL) from an endemic region with the highest incidence of LCL in Venezuela were typed from HLA-ABC, DR, DQ antigens and complement factors. The parental HLA haplo* types segregated at random among healthy and affected siblings but in backcross families significantly higher frequencies of HLA-A28 (p = 0.0018), -Bw22 (p = 0.0122), or -DQw8 (p = 0.0364) were present in affected compared to healthy siblings. HLA-B15 showed a higher frequency (p = 0.0062) among the latter group. Haplotypes Bw22CF31 (p = 0.0076) and Bw22DRwl 1DQw7 (p = 0.0163) were also significantly

more frequent in affected compared to healthy siblings and A2Cw_ (p = 0.0445) among the latter. No HLA genetic linkage with a putative LCL susceptibility germ(s) could be demonstrated in this study. A case/control comparison of 26 unrelated LCL patients (one proband from each family) and healthy individuals of the same ethnic origin confirmed the association of HLA-Bw22 (p~ = 0.048) and -DQw3 (Pc = 0.036) with LCL The relative risk reached 12.5 for Bw22 and 4.25 for DQw3 with ethiologic factors of 0.17 and 0.64, resepectively. HLADQw3 apparently makes the major contribution as a genetic risk factor for LCL at the population level. Human Immunology 30, 129-135 (1991)

ABBREVIATIONS ACL American cutaneous leishmaniasis DHR delayed hypersensitivity reaction DCL diffuse cutaneous leishmaniasis EF ethiologic factor

LCL MCL PBMC RR

localized cutaneous leishmaniasis mucocutaneous leishmaniasis peripheral blood mononuclear cells relative risk

INTRODUCTION American cutaneous leishmaniasis (ACL) is endemic in some areas o f Central and South America where the ethiologic species belong to the complex Ldshmania mexicana and Leishmania braziliensis [1, 2]. The disease presents three different clinical manifestations caused From the Centro de Medicina Experimental, lnstituto Venezolanode lnvestigaciones Cientificas, Caracas, Venezuela. Lab.de Leishmaniasis. Centro de lnvestigaciones Trujillanas, U.L.A., Venezuela,Dopartamento de Inmunologia. Instituto Nacional de la Nutricion Salvador Zubiran. Mexico D.F., Mexico; and lmmunogenetics Laboratories, The Johns Hopkins University Schoolof Medicine, Baltimore, Maryland. Addressreprintsrequestto Dr. Z. Layrisse,Centrod¢ Medicina Experimental, lnstituto Venezolanode lnvestigacionesCientificas (IVIC). Apartade 21827, San Martin, Caracas 1020 A, Venezuda. ReceivedJune 4, 1990; acceptedSeptember4,1990,

HumanImmunology30,129-135 (1991) © AmericanSocietyfor Histocompatibifityand Immunogenedcs,1991

by different leishman/a species: localized cutaneous leishmaniasis (LCL), mucocutaneous le/shmaniasis (MCL), and diffuse cutaneous lelshmanlasls (DCL). Up to 90% o f the ACL are LCL [3]. Albeit the parasite species has been established as the factor with more influence [4-6], there is evidence to support the fact that the genetic background o f the host and the immune response play a major role in the outcome o f the disease. Thus in the routine model, genes outside the H-2 system are involved in the innate susceptibility to experimental leishmaniasis infection but genes within the H-2 system play a major role in the regulation o f the acquired immune response modulating the outcome o f the disease [7-10]. 129 0198-8859/91I$3.50

M. L. Lara et al.

130

TABLE 1 Trujillan families included in the study. Distribution of 101 individuals in 24 sibships Number of siblings in each sibship Parents tested

Two One None

1

2

3

3

7 2

9 1

sp. recognized as the ethiologic agent in the region: Leishmania braziliensis braziliensis and Leishmania braziliensis guyanensis. A delayed hypersensitivity reaction (DHR) test was done on all individuals participating in the study.

Cell separation. Blood samples were collected in hepa4

Sibships

1

20 3 1

1

The whole sample tested includes 105 individuals. 101 organized in sibships: 42 parents and 59 siblings.Two sibships share the same mother. Four nonrelated individuals tested are not included in this table (three LCL patients and one control).

This article presents HLA-A, B, C, DR, D Q and complotype data obtained from families with one or more cases of localized cutaneous leishmaniasis. HLA haplotypes, gametic association, antigen segregation, and linkage analysis are presented for the first time in the literature. MATERIALS AND METHODS

Families. All families were living in Trujillo, a city in the Andean region of Venezuela, with the highest incidence of LCL in the country. They were randomly selected from those followed up at the Leishmaniasis Laboratory of the Los Andes University Research Center if there was at least one proband and one or two siblings available for testing. Twenty-four family groups including 42 parents and 59 siblings were tested (Table 1). There were 11 families with both parents healthy, 10 families with one affected parent, and 3 families with both parents affected with LCL. Families 4 and 5 shared the same mother. The Andean background was checked for all families up to two generations.

rinized vacutainers in Trujillo and transported immediately to the laboratory at Instituto Venezolano de Inves tigaciones Cientificas (IVIC) where they were processed the same day. Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation [12] and cryopreserved until needed for HLA typing.

HLA serology. Typing was performed by the microcytotoxicity method [13] using enriched T- or B-lymphocyte suspensions isolated [14] from the cryopreserved PBMC and sera from the Fifth Latin American Histocompatibility Workshop. HLA haplotypes were ascertained from individual phenotypes after analysis of the nuclear family relationship.

Statistical analysis. Gene frequencies were determined by direct gene counting. Significance was obtained from 2 x 2 table analysis and chi-square or Fisher's exact test as appropriate. Haplotype segregation analysis was performed using two approaches: families were grouped (1) according to the LCL status of the parents or (2) independently of the LCL status of the parents. The analysis was done following Nihenhuis' method [15] for families with two affected siblings, or by applying the chi-square test to haplotypes observed in sibling pairs (one affected, one healthy) compared to the expected proportions in random segregation patterns [16].

Linkage analysis. Linkage analysis was done at the Johns Hopkins University Immunogenetics Laboratories by the standard maximum likelihood method using the computer program LIPED [17].

Patients. Patients filled three diagnostic criteria [11]:

Complotypes. Complotypes were determined at the Im-

(1) an ulcer or scar lesion with parasites in a smear, (2) a positive leishmanin test (papule > 5 ram), and (3) identification of the ethiologic agent (leishmania species) in a reference center.

munology Department, Nutrition Institute in Mexico, as part of the collaborative studies carried out for the Fifth Latin American Histocompatibility Workshop.

Unrelated patients. Twenty-six unrelated patients were tested; 23 were probands from each family, three were unrelated individuals.

RESULTS

Leishmanin. Leishmanin was elaborated at the Leishmaniasis Laboratory of the Los Andes University Research Center (Lot no. NURR88) with antigens of Leishmania

HLA haplotype segregation in families. Tables 2, 3, and 4 show the segregation of HLA haplotypes in families grouped according to the LCL status of the parents. The inheritance of HLA haplotypes in all the families was as expected according to random segregation: 1 : 2 : 1 in healthy or affected siblings (p > 0.5). There was no

131

TABLE

2

P a r e n t a l H L A h a p l o t y p e s e g r e g a t i o n a m o n g h e a l t h y a n d a f f e c t e d s i b l i n g s in f a m i l i e s w i t h o n e o r more cases of cutaneous leishmaniasis: Families with one affected parent Number of sib|ing~ inheriting each haplotype pa/r Affected

Family ld no. #5 #6 #8 #9 #10 #11 #13 #15 #18 #23

Father's haplotypes a/b

Mother's haplotypes c/d

a¢

(a) A3 Cw-B8 CS31 DRIDQw(b) A30 Cw-B44 CF31 DR7 DQw2 "A2 Cwl Bw22 CF31 DRwl 1 DQw7 "A24 Cw-B35 CS31 DR4 DQw7 *A28 Cw-B39 CS31 DR-DQw*A24 Cw3 B40 CF32 DR2 D Q w l A2 Cwl Bw51 CS32 DRwl I DQw8 A24 Cwl B35 CS31 DRwl4 DQw7 °A2 Cw-Bw63 CF31 D R w l l DQwl ° A l l Cw-BI3 CS31 D R w l l D Q w l *A32 Cw-BI4 CS42 D g w 6 D Q w l eA2 Cw5 B I 8 CFI31 DR4 DQw8 "A24 Cw3 B40 CF31 DR4 DQw8 *A32 Cw-Bw57 CF30 DR- DQw4 "A2 Cw3 B35 CS31 DRwI3 DQw7 "A24 Cw3 B40 CS31 DR4 OQw7 A28 Cw- Bw22 CF31 D R w l l DQw7+'~ A2 Cw3 B40 D R w l l D Q w 7 + ~ A2 Cw3 B40 °" A9 Cw- B51 "®

(c) °A2 Cw-Bw51 CS31 DR2 DQwl (d) "A26 Cw3 Bw22 CF31 DR2 DQwl A28 Cw-B39 CF31 DRwl2 DQwA23 Cw7 B7 CS21 DR2 DQwl A24 Cw-B35 CS31 DRw6 DQwl A24 Cw-B45 CS02 DR4 DQw"A2 Cw3 B40 CS32 DRwI4 DQw7 "A2 Cw3 B40 CS01 DR4 DQw7 A24 Cw-B44 CF31 DRwI2 DQw8 A24 CwoB44-CS02 DR4 DQw7 A2 Cw8 Bw51 CS31 DR2 DQwl A2 Cw8 Bw51 CF142 DR2 DQwl A1 Cw8 BI4 CF31 DRwl I DQw7 A24 Cw3 B44 CF30 DR7 DQw2 A2 Cw- BwS1 DR- DQwl ~ A3 Cw- B7 DR2 DQwl ~ "A29 Cw- B44 CF33 DR7 DQw2¢ "A32 Cw- B- D R w l l DQw7 = A2 Cw4 B35 [']= A24 Cw- BSI [°]*

1

ad

ix

Healthy bd

ac

ad

ix

bd

1

1

1

1

2 1

1

1

I

l

1

I 1

I

I

"Leishmanin-lmsitive individual with leishmania scar and referred illness. ~'] Leishmanin-posidveindividual without scar or referred illness. "Complotype not determined for this haplotylm. Familyno. 15: only father's complotype information. Parents l~plotypes (a/b) (c/d) are given as shown in familyno. 5. Complotype gene order is C2-Bf-C4A-C4B. + Deduced haplotylms. t Deduced ¢omplotype.

TABLE

3

P a r e n t a l H L A h a p l o t y p e s e g r e g a t i o n a m o n g h e a l t h y a n d a f f e c t e d s i b l i n g s in f a m i l i e s w i t h m u l t i p l e x cases of cutaneous leishmaniasis: Families with two affected parents Number of siblings inber/¢/ng each haplotype pair Affected

Family Id no. #2 #4 #7

Father's haplotypes a/b

Mother's haplotypes c/d

ac

"A24 Cwl Bw42 DR4 DQw -= "A30 Cw5 B44 DR7 DQW2 °A2 Cw3 B40 CS31 DR- DQW1 Aw30 Cw4 B35 CF31 D R W I I DQW7 *A28 Cwl B62 CF31 DR2 DQW7 *A24 CW3 B35 CS21 DR4 DQW7

"A28 Cw- BW35 DR4 DQW8 = "A24 Cw- B35 DR4 DQw8 °A2 Cw- BwS1 CS31 DR2 DQwl "A26 Cw3 Bw22 CF31 DR2 DQwl °A2 Cw; B44 CF31 DR7 DQw2 "A24 Cw3 B35 CS31 DR4 DQw8

1

,Samelegend as in Table 2.

ad

ix

Healthy bd

ac

ad

ix

bd 1

l 1

132

M.L. Lara et al.

TABLE 4

Parental H L A haplotype segregation a m o n g healthy and affected siblings in families with o n e o r m o r e cases o f cutaneous leishmaniasis: Families with two healthy parents Number of siblings inheriting each haplotype pair Affected

Family Id no. #1 #3 #12 #14 #16 #17 #19 #20 #21 #22 #24

Father's haplotypes a/b

Mother's haplotypes c/d

A24 B45 Cw- CF31 DRw8 DQw7+ All Cw- B17 CS61 DRI1 DQw8+ A2 Cw- B- CS42 DRw13 DQwl+(e) A- Cw- B44 DR7 DQw2+ (f) A2 Cwl Bw51 CS30 DR4 DQw8 A3 Cw7 B7 CF31 DR2 DQw5 A24 Cw- B35 CS31 DR2 DQwl A24 Cw- B38 CS31 DR- DQwl A1 Cw3 Bw22 CF30 DRwll DQw7 A2 Cw- B44 CS31 DR4 DQw7 A24 Cw4 B35 DR10 DQwl ® A24 Cw8 B- DRwI0 DQwl A2 Cw- B8 CS01 DR4 DQw8 A3 Cw- B7 CS31 DR2 DQwl All Cw- Bw52 DR2 DQwI+ (x) A3 Cw7 B7 DR2 DQwl+ (y) A3 Cw- Bw52 DR2 DQwl+(z) A1 Cw- B8 CS20 DR3 DQw2 A28 Cw7 B39 CS31 DR2 DQw5 A29 Cw- Bw51 CS31 DR7 DQw2 A31 Cw- Bw51 CF30 DRwll DQw7 A3 Cw4 B35 CS31 DR2 DQwl A24 Cw- B14 CSO1 DR4 DQw7

A30 Cw6 B13 CS31 DR7 DQw2+ A28 Cwl Bw22 CF31 DRwll DQw7+ A24 Cwl B27 CS31 DR7 DQw2 A32 Cw7 B7 CF31 DR2 DQwl A2 Cwl B35 CS30 DR4 DQw8 A24 Cwl B63 CS33 DR- DQwA2 Cw3 BwSl CS31 DR4 DQw8 A24 Cw3 B40 CS31 DR4 DQw8 A2 Cw3 B62 CF31 DRwll DQw7 A24 Cw4 B35 CF30 DR4 DQw7 A24 Cw4 B35 DR4 DQw8® A24 Cw- B63 DRwll DQw7 A2 B44 Cw2 SC21 DR- DQwl A31 BS1 Cw- SC42 DRll DQwl A31 Cw4 B35 DRwll DQw7 ® A2 Cwl Bw62 DRwll DQw7

ac

ad

bc

1

Healthy bd

ac

1

1

ad

ix

yd

xc 1

1

1

1

1

2

§1

1

1

1

yc

A3 Cw- B- CF31 DR2 DQwl A24 Cw3 Bw62 CS21 DR2 DQw5 A28 Cw- Bw22 CF31 DR11 DQw7 A29 Cw- B44 CF31 DR7 DQw2 A3 Cw- B7 CF20 DR4 DQw8 A24 B39 Cw- CS31 DRwl2 DQwl

bd

1

1

xc

zd

1

1

1

Samelegendas in Table 2. § Complotypeof the affectedsibling(individualno. 75) CCFS3011;e-f-x-y-z,deduced haplotypesof differentfathers.

apparent influence o f t h e LCL status o f t h e parents (Table 5). W h e n t h e analysis was d o n e following N i h e n h u i s ' m e t h o d in families with two affected siblings or with o n e affected and two healthy siblings, again rand o m segregation o f paternal and maternal haplotypes in affected and healthy siblings was f o u n d (0.5 < p < 0.6) (data n o t shown).

Multiplex affectedfamilies.

In four families (nos. 1, 5, 12, and 15) t h e r e were two siblings affected: t h r e e pairs were haploidentical and o n e was H L A different with t h e proband.

HLA antigen segregationin backcrossfamilies.

W e considered as backcross families t h o s e in which o n e p a r e n t was heterozygous for o n e antigen o r haplotype which was absent in t h e o t h e r parent. T w e n t y - t h r e e antigens could be analyzed. For only four H L A antigens we f o u n d remarkable differences f r o m t h e 1 : I proportion expected w h e n t h e r e is r a n d o m segregation. T h u s H L A - A 2 8 was p r e s e n t in 7 o f 7 affected siblings c o m p a r e d to 2 o f 10 healthy siblings (p = 0.0018); H L A - B w 2 2 was p r e s e n t

in 7 o f 8 affected siblings and in only 2 o f 9 healthy siblings (p = 0.0122); H L A - D Q w 8 was p r e s e n t in 6 o f 9 affected siblings and in 2 o f 10 healthy siblings (p = 0.0364). T h e f r e q u e n c y o f D Q w 7 was very high a m o n g t h e diseased families and only f o u r o f t h e s e were informative for t h e D Q w 7 segregation analysis. In t h e s e families t h e D Q w 7 segregation occurred at r a n d o m . Finally. H L A - B 1 5 (Bw62 and Bw63) was f o u n d m o r e freq u e n t l y in healthy t h a n in affected siblings: 7 / 1 0 versus 0/7 (p = 0.0062). W h e n t h e s a m e analysis was perf o r m e d with two H L A antigen pairs in informative families, statistically significant differences b e t w e e n affected and nonaffected siblings inheriting t h e antigen combinations was f o u n d for H L A - A 2 / C w _ (p = 0.0445), B w 2 2 / CF31 (p = 0.0076), and B w 2 2 / D R w 1 1 / D Q w 7 (p = 0.0163) (Table 6).

Linkage analysis.

B o t h d o m i n a n t and recessive m o d e s were tested at varying p e n e t r a n c e , various frequencies for t h e disease g e n e , and r e c o m b i n a t i o n fractions f r o m 0 to 0.5, with n o evidence for linkage b e t w e e n H L A and p r e s e n c e o f t h e disease. M a x i m u m lod scores obtained

133

TABLE

5

P a r e n t a l H L A h a p l o t y p e s e g r e g a t i o n a m o n g p a i r s o f h e a l t h y a n d a f f e c t e d siblings in F a m i l i e s ~ w i t h o n e o r m o r e LCL c a s e s a n d c o m p a r i s o n w i t h t h e e x p e c t e d proportion 1 : 2 : 1 Haplotype pair inherited by each sibling

Family id no.

Affected

Nonaffected

Number of pairs of affected/nonaffected siblings that share or not parental haplo~pes 2 haplotypes

1 haplotype

None

0 / 0 I 0

2 0 1 0 0

0 0 0 0 2

2 1 1 I :)

2 0 0

0 2 1

0 0 1

2 2 2

II

I

0

I

1 o

(l

~,

2 1

0

2

2

4

0 1

2 2

0 0

2 3

0

0

1

Families with one affected parent 5 ac bd ix 6 ac ac 8 ac bc 9 bc ix 10 ix ad ad 11 ac ac ac 13 15

ad

ac bd

ad bd bc 18 ac ix Families with two affected parents 2 . . . . bd 4 bd ix Families with two healthy parents 1 ad bd ac bc 14 ad ac bd 16 ac ac bc bc 17 ac bd 19 ac bc 21 ix ac ad Total observed Total expected X = 2.06 /, > 5

0 0 6 7.5

I I 16 14.5

Total pairs

I

0 I 8 7.5

1 2 30

"The familiesincluded in this study had at least one affectedand one healthysibling.The prolmrtionof the HLA haphnypesin all families tested irrespectiveof the LCLstatus of the parentswasas expected in randomsegregation,whenanalyzedseparatelyor by addingall of them together.

TABLE

6

H a p l o t y p e s a n d H L A a n t i g e n s s e g r e g a t i o n in a f f e c t e d a n d n o n a f f e c t e d siblings in i n f o r m a t i v e families h e t e r o z y g o u s for the haplotypes A 2 CW-, B W 2 2 CF31, and B W 2 2 D R W 1 1 D Q W 7 or for antigens HLA- A28, HLA-BW22, HLA- B15, and HLA- E~W8 Affected ID no of family

Haplotypes A2, CwBw22, CF31 Bw22, DRwl 1, DQw7 Antigens A28 Bw22 BI5 (w62, w63) DQw8

Nonaffected

Number of families

+

-

+

-

Total siblings

p

4, 5, 10, 16, 19 1, 4, 5, 6, 18 1, 6, 16, 18

5 5 4

2 6 5

6 1 0

7 1 2

2 5 5

17 13 12

0.0445 0.0078 0.0163

1, 2, 6, 8, 18, 21 1, 4, 5, 6, 16, 18 16, 10, 12, 17, 20, 21 1,9, 10, 11, 13, 17, 19

6 6 6 7

7 7 0 6

0 1 7 3

2 2 7 2

8 7 3 8

17 17 17 19

0.0018 0.0062 0.0062 0.0364

Only familieswith pairs of affectedand healthysiblingswere included.This analysisincludesall the siblings(affectedand nonaffected)studied in each informative family.Significantdifferencein segregationwas mantalnedwhen only one affectedand nonaffectedsiblingpa/r was taken into account for the analysis. DR4DQw8 had a tendencyto be more frequentlypresent in affectedthan in nonaffectedsiblingsbut the significanceof the differencecould not be demonstratedsinceamongthe seven familiesinformativefor this haplotypethere were two families(uos. 7 and 24) without healthy siblings.

M. L. Lara et al.

134

TA~I~ 7

Antigens associated with LCL in Venezuelan mesitzos Antigen frequencies(%)

Antigens HLA-Bw22 HLA-DQW3

LCLpatients

Controls

(n = 26) 19.23 (n = 24) 83.32

(n = 160) 1.87 (n = 132) 54.00

p

R.R."

EFb

0.002' 12.5 0.17 0.012d

4.25 0.64

"RR = relativerisk. bl/F = ethiologicfactor. 'Pc = 0.048 alp, = 0.036.

for a recessive model were 0.01872 with 70% penetrance, a gene frequency of 0.2 for leishmaniasis, and a recombination fraction of 0.3 in 24 families; and of 0.01248 for a penetrance of 90%, a gene frequency of 0.05 for leishmaniasis, and a recombination fraction of 0.4 in 21 families. With the dominant: model all the lod scores were 0 or negative for the 24 families. Association analysis. When the HLA antigen frequencies observed in 26 unrelated LCL patients were compared to those of a control series of similar ethnic origin, increased frequencies of HLA A24 (53.8 versus 27.5), A28 (30.7 versus 17.5%), Bw22 (19.2 versus 1.9), DR4 (45.8 versus 24.2%), DRw12 (16.7 versus 4%), and DQw3 (83.3 versus 54%) were detected in the patients. The DQw7 and DQw8 splits of DQw3 were independently analyzed and showed higher frequencies in affected versus healthy controls (54.16% versus 32% and 29.16% versus 16%, respectively) but the difference was not significant. Only the differences corresponding to HLA-Bw22 (Pc = 0.048) and to DQw3 (Pc = 0.036) were statistically significant at the Pc level (Table 7). Similarly, HLA-Cw4 and -Cw7 showed significantly decreased frequencies among patients (Pc 0.032 and 0.046, respectively) (data not shown). DISCUSSION In contrast to the murine model in which there is extensive information about the genetic regulation of cutaneous leishmaniasis [10], only one HLA study among ACL patients has been reported in the literature [18]. The authors found a decrease of HLA-Cw7 in ACL patients and random segregation of HLA parental haplotypes in seven pairs and two trios of affected siblings. Nevertheless, evidence supporting the hypothesis of a different immune host response in the three types of r2. cutaneous disease [ 19, 20] suggests variability in the

immunogenetic regulation of the host antileishmanic response. The study reported here was carried out on patients affected with the most common type of cutaneous disease, localized cutaneous leishmaniasis or LCL. These patients live in an Andean region of Venezuela with long-standing high endemicity, presenting the best climatic and ecological conditions for the vecmr's (sand fly) development. Thus affected and nonaffected subjects had been exposed to comparable extrinsic factors but the data indicate the existence of an intrinsic risk factor in which the HLA antigens apparently play an important role. The definition of HLA class I and II antigens and complement factors in 24 families with one or more cases of LCL has provided preliminary negative evidence for linkage of the HLA system with primary susceptibility to LCL: HLA haplotype segregation occurs at random among affected and nonaffected siblings independently of the LCL status of the parents. However, a statistically significant difference on the frequency of one HLA class I antigen (Bw22) and one HLA class II antigen (DQw8) was found in affected compared to healthy siblings. The more frequent occurrence of HLA-Bw22 with either the complotype CF31 or with D R w l 1DQw7 in affected siblings of informative backcross families is due to linkage disequilibrium; these combinations of alleles are well-known among Oriental populations [21] and are present also among the 70 parental haplotypes reported here (data not shown). The HLA-Bw22 and -DQw3 (DQw7 and 8) association with LCL present in backcross families was confirmed when the frequencies of these antigens among 26 unrelated patients were compared to a large control series including subjects from all regions of the country. In order to verify that the difference was independent of regional frequencies, the HLA data from the large control series was compared to that of a local control group including 20 unrelated subjects (one healthy sibling of each family) from 19 of the 24 families plus one he',,lthy control of the same region. N o significant difference between the two control groups was seen (data not shown) except for HLA-Cw4 and -Cw7 which were also present in lower frequency among healthy subjects from Trujillo (p = 0.0095 and 0.024, respectively). The relative risk (RR) values for HLA-Bw22 and -DQw3 indicate that an individual with the former antigen has 12.5 times more chance of developing LCL in an endemic region than a person Bw22-negative; with DQw3, he has 4.25 times more chance than a DQw3negative person. According to the ethiologic factor (EF) values (EF = 0.64) though, the DQw3 antigen is more relevant at the population level as a genetic risk factor. This finding has been confirmed in a study of Mexican

HLA Haplotypes in Leishmaniasis

mestizos [22] reporting a DQw3 (including DQw7 and DQw8) association to LCL. Thus, it is possible that at the molecular level a common region in the DQw3 antigen, present in both splits, may be involved with presentation of the antigen and susceptibility to LCL. HLA-DQw3 could be not only an important factor for the susceptibility to LCL but also a protective antigen for the development of a more severe form of the illness. The ethiologic agents for the majority of the cases presented here were L. braziliensis guyanensis and L. braziliensis braziliensis; the latter is the species usually associated with the mucocutaneous (more severe) form of the disease in other regions of the country and the continent [2]. In vitro antileishmanic cell-response experiments using HLA-typed lymphncytes to search for high and low responder haplotypes are necessary to understand the HLA system association to LCL. Furthermore, the characterization of HLA antigen profiles in other clinical forms of ACL could contribute to defining the possible role of the major histocompatibiiity complex in the modulation of the disease manifestation.

ACKNOWLEDGMENTS The authors wish to thank the families for their cooperation; Elina Rojas, Julio Marques, and Juan B. Roias for help in the sample collection; Lubomira Rybak for secretarialassistance; and Alvaro Rodriguez L. for helpful discussion in the preparation of the manuscript. This work was partiallysupported by CONICIT's Grant 89002321 for the MSc. thesis of M. L. Lara.

REFERENCES 1. BlackweU J, McMahon-Pratt DY, ShawJJ: Molecular biology of Leishmania. Parasitol Today 2:45, 1986. 2. Laison R, Shaw JJ: Epidemiology and ecology ofleishmaniasis in Latin America. Nature 273:595, 1978. 3. Convit J, Pinardi M: Cutaneous leishmaniasis. The clinical and immunological spectrum in South America, CIBA Found Syrup 20:160, 1984. 4. Medina R, Romero J: Leishmaniapifanoi n. sp. El agente causal de la leishrnaniasis tegumentaria difusa. Arch Venez Med Trop 4:349, 1962. 5. Bryceson ADM: Pathogenesis of diffuse cutanous leishmaniasis. Trans Roy Soc Med Hyg 64:387, 1970. 6. Akuffo H, Schurr E, Anderson G, Yamaneberhan T, Britton S: Responsiveness in diffuse versus local cutaneous leishmaniasis is due to parasite difference. Scand J Immunol 26:717, 1987. 7. Perez H, Labrador F, TorrealbaJW: Variation in the re-

135

sponse of five strains of mice to Leishmania mexicana. Int J Parasiml 9:27, 1979. 8. Bradley DJ: Genetic control of natural resistance to L. donovanL Nature 250:353, 1974. 9. BlackweU J, Freeman J, Bradley D: Influence of H-2 complex on acquired resistance to Leis~mania donovani. Narare 283:72, 1980. 10. Roberts M, Alexander J, Blackwetl J: Influence of Lsl~ H-2 and an H-1t linked gene on viscem!ization and metastasis associated with Leishmania mexi~ana infection in mice. Infect Immua 57:875, t989. 11. Rodriguez M: Diagnostic criteria for American cutaneous leishmaniasis used in the Fifth L~in American Histocompatibility Workshop. Mexico, Dept, de Parasitolog~, Ins~ituto de Salubridad y Enfermedades Tropicales, 1988. 12. Boyum A: Isolation of mononuclear ceils and granulocytes from human blood. Scand J Clin Lab Invest 21:77, 1968. 13. Terasaki P|, Mc Clelland J, Park MS, McCurdy B: Microdroplet lymphocyte cytotoxicity test. In Ray JG, Hare D~, Pedermn PD, Kayb~oe Dt: (eds): Manual of tissue Typing Techniques, Bethesda, MD, NIH, 1974. 14. DanilovsJA, Ayoub G, Termaki Pt: B lymphocyte isolation by thrombin-nyion wool. in Termaki PI (ed): His¢ocompatibility Testing 1980. Los Angeles, UCLA Press, 1980, p 287. 15. de Vries RRP, Niienhuis LE, Lai RFM, Fat A, van Rood JJ: HLA-linked genetic control of host respon~ to Mjcobacterium leprae. Lancet Dec 18:1328, 1976. 16. Cavalli Sforza LL, Bodmer WF: The genetics of human populations. San Francisco, WM Freeman and Co, 1971. 17. OttJ: Estimation of the recombination fraction in human pedigrees: Effident computation of the likelihood for h~ man linkage studies A m J Hum Gen 26:588, 1974. 18. Barbier 13, Dennis F, Lefait JF, David B, Blanc M, HorsJ, Feingolon N: Susceptib/lity to human leishmanimis and HLA, Gin, Km markers. Tissue Antigens 30:63, 1987. 19. Castes M, Agnelli A, Rondon AJ: Mechanism associated with immunoregulation in human American curanous leishmaniasis. Clin Exp Parasitol 57:279, 1984. 20. Castes M, Cabrera M, Trujillo, Convit J: T-cell subpopulations, expression of interleukin-2 receptor, and production of interleukin-2, gamma interferon in human American cutanous leishmaniasis. J Clin Microbio126:1, 1988. 21. Baur MP, Nengebaner M, Albert ED: Reference tables of two-locus haplowpe frequencies for all MHC marker loci. in Albert ED, Baur MP, Mayr WR, (eds): Histocompatibility Testing 1984. New York, Sprirger-Verhg, 1984, p 677. 22. Gorodezky C, Guzm~n J, Granados G, Hobart O, [:lores J, Castro I,, Perez H: HLA linked susceptibility in Mexicans with localized cutaneous leishmaniasis. V Taller Larinoamericano de Histocompatibilidad. Abstract book, 1989, p 42.