Download Endemic Infection with Human T Cell Leukemia/Lymphoma Virus
Document related concepts
Transcript
944 Endemic Infection with Human T Cell Leukemia/Lymphoma Virus Type lIB in Argentinean and Paraguayan Indians: Epidemiology and Molecular Characterization Jorge F. Ferrer, Eduardo Esteban, Syamalima Dube, Miquel A. Basombrio, Amalia Segovia, Monica Peralta-Ramos, Dipak K. Dube, Keith Sayre, Nicolas Aguayo, James Hengst, and Bernard J. Poiesz Comparative Leukemia and Retrovirus Unit, New Bolton Center, University of Pennsylvania, Kennett Square, Pennsylvania; Area de Virologia, Facultad de Ciencias Veterinarias, Universidad Nacional del Centro de la Provincia de Buenos Aires, Tandil, and Laboratorio de Patologia Experimental, Facultad de Ciencias de la Salud, Universidad Nacional de Salta, Salta, Argentina; Departments of Medicine and Microbiology, State University of New York Health Science Center, Syracuse, and Cellular Products Inc., Buffalo, New York; AIDS/ Hepatitis Unit, Ortho Diagnostic Systems, Inc., Raritan, New Jersey; Programa Nacional de SIDA, Asuncion, Paraguay Human T cell leukemia/lymphoma virus (HTLV-II) type II infection was detected by polymerase chain reaction or serologic analyses (or both) in 22% of 697 Indians of six different ethnic backgrounds inhabiting the Argentinean and Paraguayan Gran Chaco. None was infected with HTLVI. The prevalence of HTLV-II increased with age (14% in those <13 years and 23% in those ~13 years). HTLV-II infection was found in all 20 Gran Chaco communities studied, but marked differences (44%-4%) in the rate of infection were observed even in communities separated by only a few miles. These variations correlated closely with ethnicity. In the high-incidence communities, infection clustered within families, with evidence for both sexual and perinatal transmission, primarily via breast-feeding. By contrast, only 2% of 94 Mapuche Indians from southern Argentina were positive for HTLV-II. Analyses of pol and long terminal repeat sequences from 15 Gran Chaco HTLV-II strains indicated that they constitute a highly conserved branch of the HTLV-lIB substrain. The primate T cell leukemia/lymphoma viruses (PTLVs), which include human T cell leukemia/lymphoma virus (HTLV) types I and II and simian T cell leukemia/lymphoma virus type I (STLV-I), plus bovine leukemia virus (BLV), make up a unique genus of oncogenic retroviruses (reviewed in [1]). The PTtVsand BLV encode for regulatory proteins that transactivate both viral and cellular gene transcription, a phenomenon believed to be involved in disease pathogenesis [1]. HTLV-I is regarded as the etiologic agent of adult T cell leukemia/lymphoma, HTLV-I-associated myelopathy/tropical spastic paraparesis, polymyositis, uveitis, and arthritis (reviewed in [1, 2]). The etiologic roles of STLV-I [3] and BLV [4] in neoplastic diseases of their respective species of origin have been well documented. Although HTLV-II is yet to be definitively associated with any disorder, suggestive evidence for its etiologic involvement in certain lymphomas/leukemias [5-8] and neurodegenerative diseases [9-11] has been re- Received 3 March 1996; revised 27 June 1996. Verbal informed consent was obtained from individuals in accordance to US government guidelines in place at the time of this field work. Grant support: Fundacion Hermanos Agustin y Enrique Rocca and Fundacion Antorchas, Buenos Aires; Secretaria de Ciencia y Tecnica de la Universidad Nacional de Centro de la Provincia de Buenos Aires, Tandil, Argentina; NIH (HB-67021). Reprints or correspondence: Dr. Jorge F. Ferrer, New Bolton Center, 382 W. Street Rd., University of Pennsylvania, Kennett Square, PA 19348. The Journal of Infectious Diseases 1996; 174:944-53 © 1996 by The University of Chicago. All rights reserved. 0022-1899/96/7405-0007$01.00 ported. This evidence and the results of seroepidemiologic surveys showing that infection with HTLV-II is on the rise, particularly among intravenous drug users (IVDUs) of the United State and Europe [12-18], emphasize the importance of thoroughly studying its natural history and pathogenicity. The availability of human populations with a high incidence of infection is essential for these studies, because if HTLV-II causes disease, it probably does so in a small percentage of infected hosts and after a long latency period. The two major subtypes of HTLV-II, designated A and B, differ genetically by ,....,7% [19-21]. While both subtypes have been demonstrated in the IVDU population ofthe United States, subtype A is the most prevalent [1,2]. We have reported serologic and molecular data showing HTL V-II infection in '" 13% of adult Toba and Wichi Indians living in isolated communities of northem Argentina [22]. Sequence analyses of DNA isolated from peripheral blood leukocyte (PBL) samples demonstrated that the virus present in these Indians belongs to the HTLVIIB subtype [20, 22], which is the predominant subtype found among other Paleo-Amerindian populations [20, 22-25]. The reliability of epidemiologic data requires the use of detection assays of proven sensitivity and specificity in the particular populations studied. Current commercially available ELISAs or Western blot (WB) assays for anti-PTLV antibodies are based on HTLV-I antigens or antigens of the HTLV-IIA substrain, or both. However, the sensitivities and specificities of these ELISAs and WB assays for the detection of the HTLVlIB subtype have not been adequately evaluated. Polymerase chain reaction (Pf.R) assays for HTLV-I1II pol DNA have Jll) 1996; 174 (November) HTL V-II in South American Indians been shown to be the most sensitive and specific procedures available for the diagnosis of PTL V infections [I, 12, 26 - 29]. Thus, as a first step for studying the epidemiology of HTL VII in Gran Chaco Indian populations, we compared the results of PCR analyses with those of several serologic assays among 246 Indians. These comparisons enabled us to establish more definitive criteria for identification of Indians infected with HTL V-IIB [29]. Applying these criteria, we have examined the distribution ofHTLV-II infection in Indians residing in the Argentinean and Paraguayan sectors of the Gran Chaco as well as in southern Argentina. In addition, HTLV-II strains from 15 Indians were subtyped by sequencing and oligomer restriction analyses and compared phylogenetically to other PTLV strains. Materials and Methods Study populations. The Indian populations studied inhabit two areas of South America that are separated by ~ 1400 km (figure 1). The Gran Chaco is a partially forested tropical flat land occupying > 1 million km2 in northeastern Argentina, central and southern Paraguay, and southeastern Bolivia. It is believed that this plain began to be populated 6000 years B.P. by the Proto-MacroGuaycuruan Indians, who underwent divisions and subdivisions into several linguistic families, including the Mataco-Mataguayan, Mascoian, and Guaycuru proper. Other ethnic groups, such as the Areas in South America in which Indian communities studied are located. Figure 1. 945 Zamucoan, Chiriguanos, and Arawak, also migrated to the Gran Chaco from the north [30, 31]. The Indians studied in the Argentinean sector of the Gran Chaco belong to ethnic groups called Chorote, Wichi, and Chulupi, which are members of the Mataco-Mataguayan linguistic family, and Toba, Lengua, and Ayoreo, which are members ofthe Guaycuruan, Mascoian, and Zamucoan linguistic families, respectively. They live in relatively small and isolated communities located in the northeastern Argentinean province of Salta, along the Pilcomayo river, in the areas called Santa Victoria Este and Los Blancos. The other Indians studied live in communities adjacent to the farming Mennonite colonies of Femheim and Salve Yanga in central Paraguay. Many of these Indians are employed by the Mennonites. By and large the subjects studied still obtain their food by hunting, fishing, and gathering, but many of those working for the Mennonites have incorporated farm products into their diet. None of these Indians were known IVDUs, nor have they received blood transfusions. Intermarriages between tribes belonging to the same linguistic families, particularly between Chorote and Chulupi, have occurred. Admixtures have been less frequent between tribes of different linguistic families and are very rare between Indians and nonIndian people. The large majority of the non-Indian people living in the Gran Chaco are of Spanish (Argentina) or German (Paraguay) descent. There have been no black populations in these areas. The Alumine is a mountainous region adjacent to the Andes in the eastern part of Neuken, a province of southwestern Argentina. The subjects studied in this region are the Mapuche. These Indians, who migrated from southern Chile in the 17th century, raise sheep and cattle and practice some agriculture [31]. Admixtures between the Mapuche and non-Indian people occur with considerable frequency. Collection and processing of blood specimens. Specimens were collected during several trips to the Indian communities. The selection of the subjects was based primarily on their accessibility and cooperation. The name, sex, age, ethnicity, community of residence, and family relationships of each Indian sampled were recorded. Heparinized and clotted blood samples were centrifuged on site at 1500 g for 20 min. Plasma and sera were stored at 4°C. PBL were obtained as described [22] and stored in liquid nitrogen. Because of space limitations, we were unable to store PBL from all of the Indians sampled. Serology. Sera or plasma from the 791 Indians included in this study (70 < 13 years' and 721 ~ 13 years) were tested for HTLV-I and -II antibodies using either the Cambridge Biotech (rp21 enhanced) ELISA (distributed by Ortho Diagnostic Systems, Raritan, NJ) or the Vironostika (Organon Teknika, Durham, NC) ELISA [32] or both. These two ELISAs use an HTLV-I lysate as the antigen; the Cambridge Biotech ELISA also includes recombinant HTLV-I p21E in the antigen preparation. All of these samples were also tested with the differential Select HTLV ELISA kit (Biochem ImmunoSystems, Montreal) [33], which uses synthetic type-specific HTLV-I and HTLV-II antigens in separate wells. Samples from 30 I of the Indians, including many with negative ELISA results, were also tested in the HTLV blot 2.3 WB assay (Diagnostic Biotechnology; distributed by Cellular Products, Buffalo, NY) [34], in which the antigen preparation includes, in addition to whole HTLV-I viral lysate, both type-specific (rgp46 946 Ferrer et al. HTLV-I env and rgp46 HTLV-II env) and shared (HTLV-I recombinant p21E) recombinant proteins. All of the above ELISAs were done following the manufacturer's instructions. Samples that were initially positive were tested in duplicate and those reacting in at least one of the duplicates were considered to be positive. Samples that in the WB assays reacted with HTLV p24, rgp21E, and rgp46-I were considered to be HTLV-I - positive and those that reacted with HTLV p24, rgp21E, and rgp46-II were considered to be HTLV-II- positive. All other reactivities were considered indeterminate or negative. peR and sequence analyses. DNA was organically extracted from PBL as described [20]. All samples were first determined to contain amplifiable DNA by PCR with the human ,8-g1obin primers PC03 and PC04 [35]. One microgram of each DNA sample was amplified with the HTLV-I/II pol primers SKIIO/SKlll for 45 cycles of PCR. Specific HTLV-I or HTLV-II sequences were detected using a commercial assay (Cellular Products) [36]. Three HTLV-II-positive Indian samples (W175, Ch61O, and W43) were also amplified with the HTLV-II-specific long terminal repeat (L TR) primers HTIIL(26-4 7) + /HTIIL(351-329) - and HTIIL(257-278)+/HTIIL(624-606)-, and the amplified products were detected by Southern blot hybridization using oligonucleotides HTIIL(257-278)+d and HTIIL9(351-329)-d, respectively [20]. In order to prevent false-positive reactions, all PCR assays were done with dUTP and were subjected to uracil DNA glycosylase sterilization [37]. Pre- and post-PCR operations were done by separate personnel in separate facilities. Further, all primers had additional 5' nonviral, nonhuman sequences added to facilitate sequencing, and all positive retroviral samples were reamplified with "signature primers" comprising these same nonviral, nonhuman sequences [38]. Failure to amplify with the signature primers indicates the absence of contaminating amplified DNA. Amplified pol and LTR DNAs were cloned and sequenced as described [20], and HTLV-II pol sequences were subtyped as either A or B by oligomer restriction analysis using the enzymes HinfI or MseI as described [39]. Sequences were aligned using a commercial software package [40]. The neighbor-joining method using the maximum likelihood technique to determine distance matrices [41] was used to compare new and previously published BLV, PTLV-I, and PTLV-II pol sequences (GenBank accession numbers K02120, D00647, U45264-U45284, M99067-M99088, L27561L27571, L20351-L20361, U49798-U49820, L08408-L08409, L1l456, MlO060, M76755, L02534, M92846, U12117, U12102, L20651, D00294, M76751, 102020, U12114, U12111, L20641, L20639, 20664, L20656, L20646, L20660, Z46900; also see [4244]) and HTLV-II LTR sequences (GenBank accession numbers L37129-L37146, U10252-U10266, U46555-U46558, M99091, M10060, L06856, L06860, U25135, Ll1456; also see [21, 4547]). Both distance matrix and bootstrap (l00 replications) trees were generated. Determination of HTL V-II positivity. In our previous report on a subset of the Indian population [29], the sensitivities and specificities of the screening HTLV ELISA used herein were 73%74% and 96%-99%, respectively. The sensitivity and specificity for the HTLV Select ELISA were 72% and 98%, respectively, and for the 2.3 HTLV Western blot were 70% and 100%. In contrast, the sensitivity and specificity of the HTLV-II PCR assay were 97% and 100%, respectively. Hence, in the current study, any sample that was positive for HTLV-II by WB or PCR was deemed JID 1996; 174 (November) to be positive. In the absence of WB or PCR analyses, any sample that was positive in the Cambridge Biotech or Vironostika ELISA (or both) and also positive for HTLV-II in the Select ELISA was deemed HTLV-II-positive. Results peR and phylogenetic analyses. PCR analyses were done on DNA samples from 246 Indians. All were positive for human ,8-globin. None was positive for HTLV-I. Ninety-four samples were positive for HTLV-II pol DNA. When these HTLV-II - positive samples were reamplified with signature primers, all were negative, indicating that they were not contaminated with previously amplified DNA. Three of the 246 samples were negative for HTLV-II by PCR, yet the subjects were positive for HTLV-II antibodies by both Select and WB assays. Hence, these were deemed false-negative PCR results. In contrast, among the 94 PCR-positive samples, only 67 (71%) were confirmed seropositive. Eleven of the HTLV-II -seronegative, PCR-positive Indians were retested 19 months later and 2 had seroconverted. Amplified HTL V-II pol DNA samples from 15 infected Indians (including 2 who were seronegative) belonging to all six ethnic groups and all but one of the communities studied in the Gran Chaco were subtyped via cloning and sequencing and liquid hybridization oligomer restriction (figure 2). Similarly, LTR sequences were obtained from 3 HTLV-IIB-infected Indians (figure 3). All of these samples were found to belong to the HTL V-IIB subtype. If the Gran Chaco HTLV-IIB (T-1) strain is used as a reference, then 0 changes in 2100 bases were observed among the 15 Gran Chaco pol sequences, while 15 changes in 1400 bases were observed among the 10 non-Gran Chaco HTLV-IIB strains (P < .001, Fisher's exact t test). Similarly, the 3 Gran Chaco HTLV-IIB LTR sequences were significantly more homologous to each other than to other HTLV-IIB strains (data not shown). Hence, the Gran Chaco HTLV-lIB strains cluster as a highly conserved subgroup that differs from most other North and Central American and European HTLV-IIB strains. Distribution of HTLV-Il according to area, community, and ethnic background. Using the criteria for positivity as outlined in Methods, 158 (20%) of 791 Indians examined were positive for HTLV-IIB. None was positive for HTLV-I. Samples from 1008 non-Indian, 19-year-old Argentinian military conscripts from the Gran Chaco were screened in the HTLVI Cambridge Biotech ELISA. Two were repeatedly reactive but were negative when tested in the Vironostika and Select ELISA and in the 2.3 WB assays. None of82 adult non-Indians, including 10 with leukemia, seen in health clinics in the Gran Chaco were positive for HTLV antibodies. In addition, 50 volunteer blood donors from Syracuse, New York, were serologically and PCR-negative for HTLV infection, while 20 each of North American HTLV-I -, HTLV-II -, or HIV-1- infected patient samples were positive in their respective serologic or PCR assays. lID 1996;174 (November) HTLV-II in South American Indians 947 PTLV-I Australasian Figure 2. Bootstrap consensus tree (l00 replicates) of pol sequences (140 bases) of bovine leukemia virus (BLV) and primate T cell lymphoma/leukemia viruses (PTLVs) analyzed by neighbor-joining method using maximum likelihood technique to calculate distance matrices. Branch lengths shown do not correlate with distance. Bootstrap confidence values> 50% are shown on respective branches. Major subtypes of PTLV-I and PTLV-II and the clustering of HTLVlIB strains identified in Gran Chaco Indians are indicated. High bootstrap values for PTLV-IIA and -lIB branches indicate that Gran Chaco strains clearly belong to HTLV-lIB subtype. However, bootstrap value separating them from other HTLVlIB strains is not significant. West Africa BLV West Africa Gran Chaco PTLV-II Tables 1 and 2 summarize the distribution of HTLV-II in Indians ~ 13 years old among the 20 communities and various ethnic groups examined. As can be seen, the prevalence of HTLV-II infection was highest in the Chorote and Chulupi, somewhat less in the Toba, and considerably less in the communities composed of Wichi, Lengua, Ayoreo, or Mapuche. Infection was demonstrated by serology and PCR in all of the communities, except one inhabited by Mapuches. Although La Merced and San Luis, inhabited exclusively by Chorote and Wichi, respectively, have marked differences in the prevalence ofHTLV-II (43.9% and 3.7%, respectively), they are separated by <3 km. One of the positive Mapuche had received a blood transfusion at a local hospital; however, the donor is unknown. Table 3 shows the prevalence of HTLV-II in the Gran Chaco Indians according to age and sex. As can be seen, there was increased infection with age (14.2% in the cohort < 13 years and 23.2% in those 13-89 years; P = .05 by Fisher's exact t test), particularly for subjects ~40. Also, in those 13-89 years old, there was an increased prevalence of HTLV-II in women compared with men (26.0% vs. 19.2%; P = .03 by Fisher's exact t test). These data suggested both mother-to-child and sexual modes of viral transmission. The fact that infection tended to aggregate in families supported these hypotheses. In the high-incidence families studied (figure 4), infection was demonstrated in the offspring of 9 of the 11 positive mothers, whereas the spouses of only 3 of 17 infected married members of these families were positive. The data in table 4 indicate that the maternal transmission rate of HTLV-II was ,. . ., 30%. It should be noted that 1 of the infected children was born to a noninfected mother but was nursed from birth by an infected foster mother. We could not obtain reliable information to rule out the possibility that the 3 infected children born to noninfected or nontested mothers had been nursed by an infected foster mother. Among age- and sex-matched Indians > 13 years old, HTLV-II infection was demonstrated in 30 (48%) of 62 born to infected mothers but in only 9 (11%) of78 born to uninfected mothers. Our study includes 19 HTLV-II - infected Indians whose spouses and mothers were tested. As shown in table 5, infection was significantly more frequent among the mothers (89%) than among the spouses (37%) of these subjects. There were 115 married couples of which both spouses were tested for HTLV-II. In 68 of these couples, both spouses were negative, and in 17, both spouses were positive. There were 30 couples with discordant HTLV-II status; in 19 of them, the wife was positive and the husband was negative, and in 11 the wife was negative and the husband positive. Thirteen of the 16 tested mothers of the positive spouses in the HTLV-IIdiscordant couples were infected. Data further supporting both maternal and sexual transmission of HTLV-II are presented in table 6, which compares prevalence rates of infection in spouses, mothers, and fathers of infected Indians. Discussion Our previous work indicated that PCR, WB, or a combination of screening ELISA followed by Select ELISA testing were highly specific (100%) testing formats for HTLV-II infec- Ferrer et aI. 948 lID 1996; 174 (November) FUC lib C23723 PENN7a W43 ,~~~...-SPAN130 CAM--~ SPAN129 ITA47a Figure 3. Neighbor-joining distance tree of long terminal repeat sequences (576 bases) from HTLV-IIA and HTLV-lIB isolates worldwide. Branch lengths correlate with distance between strains. Bootstrap confidence values (100 replicates) >50% are shown. High bootstrap value (88%) on branch separating HTLV-IIA and -liB strains is significant. Gran Chaco Indian strains W43, W175, and CH610 belong to branch of HTLV-liB strains that also includes Seminole strains SEM1050 and SEM1051. GI2 is strain from Guaymi Indian from Panama. WYU I and WYU2 are strains from Wayu Indians living in Colombia. Gran Chaco strains group together as clade, but bootstrap value (38%) is not significant given amount of sequence analyzed. Kayapo Indian strains Braz A21 and Kayapo form distinct subgroup of HTLVIIA sequences. 88% KAYAPO FLN Iia HTLV-II in South American Indians lID 1996; 174 (November) Table 1. 949 Prevalence of HTLV-II infection by community among Gran Chaco Indians ~ 13 years old. Area Community La Merced La Paz, Kilometro I, and Kilometro 2 La Esperanza, Pozo del Tigre, and La Gracia Monte Carmelo and La Curb ita Santa Maria San Luis La Vertiente San Patricio Chulupi aldeas Chulupi aldeas Lengua colonies Lengua colonies Ayoreo colonies Pupenhuen Ruccahroi Ethnic composition No. HTLVII-positivel total (%) No. positive by PCRlno. tested Santa Victoria Este Chorote 36/82 (43.9) 11/27 Santa Victoria Este Chorote, Wichi 47/144 (32.6) 31/59 Santa Victoria Este Chorote, Wi chi 13/39 (33.2) 6/25 Santa Victoria Este Santa Victoria Este Santa Victoria Este Santa Victoria Este Los Blancos Paraguay-Femheim Paraguay-Salve Yanga Paraguay-Femheim Paraguay-Salve Yanga Paraguay-Salve Yanga Alumine Alumine Toba, Wichi Wichi Wichi Wichi Wichi Chulupi Chulupi Lengua Lengua Ayoreo Mapuche Mapuche 5/21 (23.8) 2/48 (4.1) 2/53 (3.7) 4/33 (12.1) 2/17 (11.7) 4112 (33.3) 24178 (30.7) 4/37 (10.8) 1/12 (8.3) 2/51 (3.9) 2/33 (6.0) 0/61 5/20 2111 2/22 4/31 2/3 1/1 13113 2/2 III 1/2 1/1 NOTE. Closely related communities of common ethnicity and comparable incidences of HTLV-II infection have been combined. PCR, polymerase chain reaction. tion in Indians of the Gran Chaco [29]. PCR testing, however, was more sensitive than a purely serologic approach (97% vs. ~ 72%). In the present study, the diagnosis of HTLV-II infection was made in 43 Indians by ELISA testing alone, whereas 115 were diagnosed as infected by additional PCR or WB analyses (or both). Hence, we believe that only a minority of false-negative results were obtained and that the overwhelming majority of HTLV-II-positive samples were true-positives. The extensive precautions and controls that were used should preclude the possibility that the PCR results were due to contamination. This assumption was verified by the fact that analyses of the HTLV-II sequences obtained from 15 of the HTLV- Table 2. Prevalence ofHTLV-II in Indians ~ 13 years old according to ethnic groups. Ethnic group Present study Chorote Toba Wichi Chulupi Chorote/Wichi' Chorote/Chulupi" Lengua Ayoreo Mapuche 611171 (35.6) 5115 (25.0) 26/204 (12.7) 32/94 (34.0) 4/14 (28.5) 8110 (80) 5/49 (10.2) 2/51 (3.9) 2/94 (2.1) NOTE. Data are no. positive/total (%). * [32]. , Ethnicity of mother/ethnicity of father. Previous study* 181121 (14.8) 2/40 (5.0) Both studies 61/171 (35.6) 23/136 (16.9) 28/244 (11.5) 32/94 (34.0) 4/14 (28.5) 8110 (80.0) 5/49 (10.2) 2/51 (3.9) 2/94 (2.1) II-positive Indian samples (including 2 from seronegative subjects) typed them as unique and highly related HTLV-lIB substrains. The observation that 19 months later, only 2 of 11 HTLV-II PCR-positive, serologically negative Indians had seroconverted indicates that the seroconversion "window" following HTLV-II infection may be quite prolonged or that some subjects may harbor a defective provirus. The data allow for several conclusions regarding the epidemiology ofHTLV-Il infection among Paleo-Amerindians. Our observations show that the prevalence rates of HTLV-II infection vary markedly among the various ethnic communities, even those in close proximity to each other. This is unlikely to be due to genetic factors in the hosts, because when communities that are cohabited by high-prevalence (e.g., the Chorotes) and low-prevalence (e.g., the Wichi) ethnic groups are tested, the frequency of infection in the two ethnic groups is virtually Table 3. Prevalence ofHTLV-II in Gran Chaco Indians according to age and sex. No. postive/total (%) Age (years) <13 13-19 20-33 34-47 48-61 ;;.62 13-89 Female Male All 3/32 (9.3) 12/69 (17.3) 221127 (17.3) 31/97 (31.9) 19/55 (34.5) 12/26 (46.1) 97/373 (26.0) 7/38 (18.4) 3/41 (7.3) 15/87 (17.2) 7/47 (14.8) 13/52 (25.0) 11126 (42.3) 49/254 (19.2) 10170 (14.2) 151110 (13.6) 37/214 (17.2) 38/144 (26.3) 32/107 (29.9) 23/52 (44.2) 146/627 (23.2) 950 Ferrer et al. JID 1996; 174 (November) Family 1 Family 2 Family 3 Family 4 Figure 4. Pedigrees of 4 families in which presence of HTLV-II was examined in members of 3 generations. Black circles and black squares denote infected females and males, respectively. Numbers within circles and squares indicate ages of subjects. NT, not tested. HTLV-II in South American Indians JID 1996; 174 (November) Table 4. HTLV-II infection in children < 13 years old born to or breast-fed (or both) by infected women. HTLV-II infection of mother or foster mother Positive Negative or unknown 951 Table 6. HTLV-II infection in spouses, mothers, and fathers of infected Indians of whom at least one of these relatives was tested. No. of infected children/total (%) Ages of infected children (years) Sex of infected subject Spouse Mother Father 7*/24 (29.1)t 3/46 (6.5)t 6, 7, 8, 8, 8, 9, 9 8,9, 12 Female Male Female + male 17/36 (47.2) 17/28 (60.7) 34/64 (53.1)* 22/29 (75.8) 14/19 (73.6) 36/48 (75.0)t 7/20 (35.0) 4/10 (40.0) 11/30 (36.6)t * 5 were PeR-positive. t p < .01 (x 2 ) . the same. Rather, the high prevalence of HTLV-II in certain communities would seem to be due to a founder effect or a result of certain disparate behavioral customs. Previously reported data on the natural mode of transmission of HTL V-II in Indian populations have been obtained in studies in which the subjects were examined only by serology [4951]. However, as shown by our present and previous [29] studies, in high-incidence populations, serologic tests fail to identify up to 25% of the HTLV-II-infected persons who are PCR-positive; thus, it is evident that extensive PCR testing is required to obtain more definitive information on the natural mode of transmission of the virus. Therefore, in the present study a large number of serologically negative and serologically positive subjects were examined by PCR. Our data indicate a very high rate (""""30%) of mother-to-child transmission, presumably by breast feeding. Results supporting significant levels of perinatal HTLV-II transmission are evident in all age groups studied, particularly in the high-incidence families. Likewise, a reasonable case can be made for sexual transmission, as the prevalence of infection increases after puberty. By age 34, the prevalence rate was almost double that observed in children < 13 years old. The fact that after puberty, the prevalence of BTLV-II is significantly greater in women than in men suggests that transmission from male to female via intercourse is more Table 5. HTLV-II infection in spouses and mothers of 19 infected Indians of whom both were tested. HTLV-II status Sex ofHTLV-IIpositive subject Female Female Female Female Male Male Male Male Female Female Female Female + male + male + male + male Mother Spouse No. Positive Positive Negative Negative Positive Positive Negative Negative Positive Positive Negative Negative Positive Negative Positive Negative Positive Negative Positive Negative Positive Negative Positive Negative 3 6 * 6 of 10 seronegative spouses were PCR-negative. o 1 4 4 o 1 7 10* o 2 NOTE. Data are no. positive/total (%). P < .02, Fisher's exact t test * vs. t; P < .001, Fisher's exact t test, * vs. t and t vs. i. ' * 14 of 30 seronegative spouses were tested by PCR and all were found negative. t 7 of 12 seronegative mothers were tested by PCR and all were found negative. i 11 of 19 seronegative fathers were tested by PCR and all were found negative. efficient than vice versa. These observations are all consistent with those previously reported for the PTLV (reviewed in [1, 2, 48-51]). The data herein also strengthen the hypothesis that HTLVlIB is endemic throughout Paleo-Amerindian groups and may have been either brought by early migrants from Asia or transmitted to them from an animal source after they crossed Beringia 15,000-30,000 years ago. Because 1% of divergence within the SKllO/SKlll pol sequence of the PTLV is thought to represent """"500-1000 years of separation of the host populations [1, 2, 19, 25, 28], our genetic data would suggest that the Gran Chaco Indians have been separated from the Guaymi (Panama), Wayu (Colombia), and Seminole Indians (North America) for 500-3000 years and from the Kayapo Indians (Brazil) and other HTLV-IIA-infected groups (e.g., the Navajo and Pueblo in North America) by 3500 to > 7000 years. Whether these theoretical time frames represent points of genetic divergence or cessation of contacts that could lead to retroviral infections among Indian groups is uncertain. There has been some debate as to the origin of HTLV-II strains among modem people. The above data, plus the fact that there is a higher prevalence of both HTLV-IIA and HTLV-lIB in humans of European or African descent living in the New World compared with the prevalence of these infections in current populations from their ancestral continents, suggest that both HTLV IIA and HTLV-IIB were originally endemic in the Americas and only relatively recently were disseminated throughout the world [1, 2, 12-19]. Additional molecular comparisons ofPTLV strains from different native populations from around the world will be crucial to gain new insights into the origin and evolutionary relationships of HTLV-II variants. Clinical and epidemiologic studies of these infected populations would also be warranted to discern if the various HTLV-II subtypes in their particular hosts manifest different biologic properties and diseases. Acknowledgments We thank N. Auza and P. Soto (Facultad de Ciencias Veterinarias, Universidad del Centro de la Provincia de Buenos Aires, 952 Ferrer et a1. Argentina) for their invaluable support; C. M. Julia (Ministerio de Salud y Accion Social, Buenos Aires), J. C. Gomez Alvarenga, 1. Nunez Burgos, C. Marquez, and A. Aleman (Ministerio de Salud Publica De Salta, Argentina), D. Lucero (Ministerio de Salud Publica de Neuquen, Argentina), W. Kaethler and A. Stahl (Programa Pro-Salud Indigena ASCIM, Filadelfia, Paraguay), and L. C. Estigaribia (Ministerio de Asistencia Social y Salud Publica of Paraguay) for facilitating our access to the Indian communities and the transportation of equipment; D. Campos for assistance in several logistic aspects of our trips to the communities; and R. Boston for suggestions regarding the analysis of data. References 1. Poiesz Bl, Shennan MP, Saksena NK, et a1. The biology and epidemiology of human T cell lymphoma/leukemia viruses. In: Neu HC, Levy lA, Weiss RA, eds. Frontiers of infectious diseases: focus on HIV. 1st ed. London: Churchill Livingstone, 1993:189-205. 2. Poiesz Bl. Etiology of acute leukemia: molecular genetics and viral oncology. In: Wiernik PH, Canellos GP, Dutcher lP, Kyle RA, eds. Neoplastic diseases of the blood. 3rd ed. New York: Churchill Livingstone, 1995: 159-75. 3. Homma T, Kanki PJ, King NW Jr, et a1. Lymphoma in macaques: association with a virus of the human T-Iymphotropic family. Science 1984; 225:716-8. 4. Ferrer IF. Bovine lymphosarcoma. Adv Vet Sci Comp Med 1980;24: 1-68. 5. Kalyanaraman VS, Sarngadharan MG, Robert-Guroff M, et a1. A new subtype of human T-cell leukemia virus (HTLV-II) associated with a T-cell variant of hairy cell leukemia. Science 1982;218:571-5. 6. Rosenblatt JD, Giorgi lV, Golde DW, et a1. Integrated human I-cell leukemia virus II genome in CD8+ T-cells from a patient with "atypical" hairy cell leukemia: evidence for distinct T- and B-celllymphoproliferative disorders. Blood 1988; 71:363-9. 7. Loughran TP, Coyle T, Shennan MP, et a1. Detection of human I-cell leukemia/lymphoma virus type II in a patient with large granular lymphocyte leukemia. Blood 1992;80:1116-9. 8. Zucker-Franklin D, Hooper WC, Evatt BL. Human lymphotropic viruses associated with mycosis fungoides and evidence that human I-cell lymphotropic virus type II (HTLV-II) as well as HTLV-I may playa role in the disease. Blood 1992; 80:1437-45. 9. Hjelle B, Apenzeller 0, Mills R, et a1. Chronic neurodegenerative disease associated with HTLV-II infection. Lancet 1992;339:645--6. 10. Harrington Wl Jr, Sheremata W, Hjelle B, et aI. Spastic ataxia associated with human T-cell lymphotropic virus type II infection. Ann Neurol 1993; 33:411-4. 11. Sheremata WA, Harrington Wl Jr, Bradshaw PA, et a1. Association of "(tropical) ataxic neuropathy" with HTLV-II. Virus Res 1993;29: 71-7. 12. Ehrlich GD, Glaser JB, LaVigne K, et a1. Prevalence of human 'l-cell leukemia/lymphoma virus (HTLV) type II infection among high risk individuals; type-specific identification ofHTLVs by polymerase chain reaction. Blood 1989;74:1658-64. 13. Manns A, Blattner WA. The epidemiology of the human T-cell Iymphotropic-virus type I and II: etiological role in human disease. Transfusion 1990; 67 -75. 14. Kaplan JE, Khablaz R. The epidemiology of human T-Iymphotropic virus type I and II. Rev Med Virol 1993;3:137-48. 15. Calabro ML, Luparello M, Grottola A, et a1. Detection of human T-lymphotropic virus type lIb in human immunodeficiency virus type I-coinfected persons in southeastern Italy. 1 Infect Dis 1993; 168: 1273-7. 16. Soriano V, Calderon E, Esparza B, et a1. HTLV-I/II infection in Spain. Int 1 EpidemioI1993;22:716-9. lID 1996; 174 (November) 17. Vignoli C, Zandotti C, de Lamballerie X, et al. Prevalence of HTLV-II in Hl'V-Lc-infecteddrug addicts in Marseille [letter]. Eur 1 EpidemioI1993; 1:351-2. 18. Flo RW, Samdal HH, Kalland KH, et a1. Diagnosis of infection with human T-Iymphotropic virus type II (HTLV-II) in Norwegian HIVinfected individuals. Clin Diagn Virol 1993;1:143-52. 19. Hall WW, Takahashi H, Lui C, et a1. Multiple isolates and characteristics of human T-cell leukemia virus type II. 1 Virol 1992; 64:2456-63. 20. Dube DK, Shennan MP, Saksena NK, et a1. Genetic heterogeneity in human T-cell leukemia/lymphoma virus type II. 1 Virol 1993;67: 1175-84. 21. Lee H, Idler KB, Swanson P, et a1. Complete nucleotide sequence of HTLV-II isolate NRA; comparison of envelope sequence variation of HTLV-II isolates from US blood donors and US and Italian iv drug users. Virology 1993; 196:57-69. 22. Ferrer lF, del Pino N, Esteban E, et a1. High rate of infection with the human T-cell leukemia retrovirus type II in four Indian populations of Argentina. Virology 1993; 197:576-84. 23. Hjelle B, Zhu SW, Takahassi H, et a1. Endemic human 'l-cell leukemia virus type II infection in southwestern US Indians involves two prototype variants of virus. 1 Infect Dis 1993; 168:737-40. 24. Pardi D, Switzer WM, Hadlock KG, Kaplan JE, Lal RB, Folks TM. Complete nucleotide sequence of an Amerindian human T-ce1llymphotropic virus type II (HTLV-II) isolate: identification of a variant HTLVII subtype b from a Guaymi Indian. 1 ViroI1993;67:4659-64. 25. Ijuchi S, Tajima K, Zanimovic V, et aI. Identification of human I-cell leukemia virus type lIb infection in the Waju, an aboriginal population of Colombia. Jpn 1 Cancer Res 1993;84:1215-8. 26. Dube DK, Dube S, Ersenoy S, et a1. Serological and nucleic acid analyses for HIV and HTLV infection on archival human plasma samples from Zaire. Virology 1994;202:379-89. 27. Abbott MA, Poiesz Bl, Byrne BC, et a1. Enzymatic gene amplification: qualitative and quantitative methods for detecting proviral DNA amplified in vitro. 1 Infect Dis 1988; 158:1158-69. 28. Saksena NK, Herve V, Durand lP, et al. Seroepidemiologic, molecular, and phylogenetic analyses of simian T-cell leukemia viruses (STLV-I) from various naturally infected monkey species from Central and Western Africa. Virology 1994; 198:297-310. 29. Poiesz BJ, Dube S, Jones B, et aI. Comparative performances of ELISA, Western blot, and PCR assays for HTLV-II infection among Indians of the Gran Chaco. Transfusion (in press). 30. Alvarsson lA. The Mataco of the Gran Chaco: an ethnographic account of change and continuity in Mataco socio-economic organization. In: Uppsala studies in cultural anthropology. Vol II. Stockholm: Almquist and Wiksell, 1988. 31. Metraux A. Indians of the Gran Chaco. In: Steward Il-l, ed. Handbook of South American Indians. Vol I. Washington, DC: Smithsonian Institution, Bureau of American Entomology, 1946; bulletin 143. 32. Cossen C, Hagens S, Fukuchi R, et al. Comparison of six commercial human T-celllymphotropic virus type I (HTLV-I) enzyme immunoassay kits for detection of antibody to HTLV-I and -II. J Clin Microbiol 1992; 30:724-5. 33. Horal P, Hall WW, Svennerholm B, et al. Identification of type-specific linear epitopes in the glycoproteins gp46 and gp21 of human I-cell leukemia viruses type I and type II using synthetic peptides. Proc Nat! Acad Sci USA 1991;88:5754-8. 34. Lipka JJ, Miyoshi I, Hadlock KG, et al. Segregation of human T-cell Iymphotropic virus type I and II infections by antibody reactivity to unique viral epitopes. 1 Infect Dis 1992; 165:268-72. 35. Iannone R, Shennan MP, Rodgers-Johnson PEB, et al. HTLV-I DNA sequences in CNS tissue of a patient with tropical spastic paraparesis and HTLV-I associated myelopathy. 1 Acquir Immune Defic Syndr 1992; 5:810-6. 36. Dyster LM, Abbott L, Bryz-Gornia V, et al. Microplate-based DNA hybridization assays for detection of human retroviral gene sequences. 1 Clin MicrobioI1994;32:547-50. JID 1996; 174 (November) HTL V-II in South American Indians 37. Longo MC, Berninger MS, Hartley JL. Use of uracil DNA glycosylase to control carry-over contamination in polymerase chain reactions. Gene 1990;93:125-8. 38. Abbott LZ, Spicer T, Bryz-Gornia V, et al. Design and use of signature primers to detect carry-over of amplified material. J Virol Methods 1994;46:51-9. 39. Dube S, Spicer T, Bryz-Gornia V, et al. A rapid and sensitive method of identification ofHTLV-II subtypes. J Med ViroI1995;45:1-9. 40. Needleman SB, Wunsch CD. A general method applicable to the search for similarities in the amino acid sequences of two proteins. J Mol Bioi 1970;48:443-5. 41. Felsenstein 1. PHYLlP-phylogeny inference package. Cladistics 1989; 5:164-6. 42. Nerurkar VR, Babu PG, Song KJ, et al. Sequence analysis of human Tcelllymphotropic virus type I strains from southern India: gene amplification and direct sequencing from whole blood blotted onto filter paper. J Gen Virol 1993;74:2799-805. 43. Ratner L, Philpot T, Trogridge DV. Nucleotide sequence analysis of isolates of human T-lymphotropic virus type I of diverse geographic origins. AIDS Res Hum Retroviruses 1991;7:923-41. 44. Evangelista A, Maroushek S, Minnigan H, et al. Nucleotide sequence analysis of a provirus derived from an individual with tropical spastic paraparesis. Microb Pathog 1990; 8:259- 78. 953 45. Igarshi T, Yavashita N, Mura T, et al. Isolation and genomic analysis of human T-lymphotropic virus type II from Ghana. AIDS Res Hum Retroviruses 1993; 9:1039-42. 46. Zella D, Caviechini A, Saleni M, et al. Molecular characterization of two isolates of human T-cell leukemia virus type II from Italian drug abusers and comparison of genome structure with other isolates. J Gen Virol 1993; 74:437 -44. 47. Switzer WM, Owen SM, Pieniazek DA, et al. Molecular analysis of human T-celllymphotropic virus type II from Wayu Indians of Colombia demonstrates two subtypes ofHTLV-IIb. Virus Genes 1995; 10:153-62. 48. Murphy EL. The epidemiology of HTLV-I: modes of transmission and their relation to patterns of seroprevalence. In: Blattner WA, ed. Human retrovirology. New York: Raven, 1990:295-305. 49. Vitek CR, Gracia FI, Giusti R, et al. Evidence for sexual and mother-tochild transmission of human T Iymphotropic virus type II among Guaymi Indians, Panama. J Infect Dis 1995; 171:1022-6. 50. Black FL, Biggar RJ, Neel N, et al. Endemic transmission ofHTLV type II among Kayapo Indians of Brazil. AIDS Res Hum Retroviruses 1994; 10:1165-71. 51. Ishak R, Harrington WJ, Azevedo VA, et al. Identification of human Tcell Iymphotropic virus type IIA infection in the Kayapo, an indigenous population in Brazil. AIDS Res Hum Retroviruses 1995; 11:813-21.