Content uploaded by Anne-Mieke Vandamme
Author content
All content in this area was uploaded by Anne-Mieke Vandamme on Oct 27, 2014
Content may be subject to copyright.
AIDS RESEARCH AND HUMAN RETROVIRUSES
Volume 23, Number 4, 2007, pp. 503–507
© Mary Ann Liebert, Inc.
DOI: 10.1089/aid.2006.0203
The Close Relationship between South African and
Latin American HTLV Type 1 Strains Corroborated in a
Molecular Epidemiological Study of the HTLV Type 1
Isolates from a Blood Donor Cohort
AUGUSTO CESAR ANDRADE MOTA,
1
SONIA VAN DOOREN,
2
FLORA MARIA DE CAMPOS FERNANDES,
1,3
SERGIO ARAUJO PEREIRA,
3
ARTUR TRANCOSO LOPO QUEIROZ,
1
VIVIANA OLAVARRIA GALLAZZI,
3
ANNE-MIEKE VANDAMME,
2
BERNARDO GALVÃO-CASTRO,
1,3
and LUIZ CARLOS JUNIOR ALCANTARA
1,3
ABSTRACT
It has been difficult to explain why all HTLV-1 sequences in Salvador, a city in the northeast of Brazil, be-
long to the Transcontinental (A) subgroup of the Cosmopolitan (a) subtype, since according to historical data
the vast majority of slaves brought to Brazil (through Salvador) came from west Africa, where only the west-
ern African subgroup (C) has been found. To shed more light on this subject we conducted a phylogenetic
analysis of 23 isolates from blood donors of Salvador. DNA was extracted and submitted to a nested PCR for
amplification of the entire LTR region. The PCR products were purified and sequenced on an automated se-
quencer. Neighbor-joining and maximum likelihood phylogenetic analyses were performed. None of the new
sequences from Salvador clustered within the West-African subgroup C. Confirming previous results, all se-
quences belonged to the Transcontinental subgroup (A) of the Cosmopolitan subtype, and clustered in two
Latin American clusters. In addition we showed sequences from southern Africa clustering in both Latin
American clusters. One of the new sequences is ancestral to the larger Latin American cluster beta due to a
duplication of a 12-bp long fragment, a finding that has not been previously described. These findings sup-
port the hypothesis that HTLV-1 isolates circulating in Latin America have a closer relationship to South
African compared to West-African HTLV-1 strains. The 12-bp-long duplications in one of the sequences has
no obvious clinical or biological implications yet.
503
INTRODUCTION
M
OLECULAR EPIDEMIOLOGICAL STUDIES
have demonstrated
that the HTLV-1 genome has a high genetic stability.
1
This
stability can be explained mainly by the fact that proviral repli-
cation depends on the clonal expansion of HTLV-1-infected
cells.
2
This low HTLV-1 variability has been used to give sup-
port to theories about the origin, dissemination, and evolution of
this retrovirus, and also about the migrations of the ancestor hu-
man population across the world.
1
Seven different genetic sub-
types of HTLV-1 have been proposed based on phylogenetic
analyses: a, or Cosmopolitan, which is distributed worldwide;
3
b, from Central Africa;
4
c, highly divergent Melanesian strain
from Papua New Guinea and Australia;
5
d, isolated from Cen-
tral African Republic (CAR) pygmies, and from two patients in
Cameroon and Gabon;
6,7
e, isolated in a sample from an Efe
pygmy in Democratic Republic of Congo (DRC) and f, from a
Gabonese individual;
8
and g, recently described as a new sub-
type in Cameroon.
9
The Cosmopolitan subtype is divided into
five subgroups based on geographical distribution: Transconti-
1
Bahia School of Medicine and Public Health/Foundation for Scientific Development, Salvador, Bahia, Brazil.
2
Rega Institute for Medical Research, Katholieke Universiteit Leuven, Leven, Belgium.
3
Advanced Public Health Laboratory, Gonçalo Moniz Research Center, Oswaldo Cruz Foundation (LASP/CPqGM/FIOCRUZ), Salvador,
Bahia, Brazil.
nental (A), Japanese (B), West African/Caribbean(C), North
African (D), and Black Peruvian (E).
3,10–12
The HTLV-1 infection is endemic in Japan, the Caribbean
basin, and some regions of Africa and South America.
13
Sal-
vador, a city located in the northeast of Brazil, has the highest
seroprevalence of HTLV-1 in the country.
14
The origin and mo-
lecular epidemiology of the HTLV-1 infection in Salvador have
been enigmatic and difficult to trace. According to historical
data, most of the Africans brought to this city came from west-
ern Africa,
15,16
where only the C subgroup of the HTLV-1a has
been found.
3
In contrast, HTLV-1 isolates found in Salvador
are exclusively of the A subgroup.
17,18
Previous studies, comparing both the molecular phylogenies
of the HTLV-1 sequences and the genetic background of the
infected hosts from Salvador with those from South Africa, have
hypothesized that the virus was introduced in Salvador by in-
fected individuals from the southern region of Africa, during
the post-Columbian slave trade.
17,18
The phylogenetic analysis
revealed that South African sequences cluster inside the Latin
American group of the Transcontinental subgroup.
18
In a further attempt to understand the possible origin of
HTLV-1 infection in Salvador, we conducted a phylogenetic
analysis with sequences from infected blood donors from Sal-
vador who have participated in a case–control study exploring
the risk factors for this infection.
19
In this paper we report the
results of this new phylogenetic analysis, indicating a closer re-
lation between Latin American and South African compared to
West African HTLV-1 strains.
MATERIALS AND METHODS
DNA was extracted from peripheral blood mononuclear cells
of 23 positive samples using the QIAamp Blood kit (Qiagen
Inc., Chatsworth, CA). Nested polymerase chain reaction (PCR)
was performed and long terminal repeat (LTR) regions were
amplified as two overlapping fragments: an LTR-gag segment
of 473 bp and a tax-LTR segment of 479 bp, as previously de-
scribed.
12,17
All amplification products were electrophoresed on
1% agarose gel stained with ethidium bromide, and visualized
with ultraviolet light.
Ninety-one HTLV-1-infected blood donors residing in the city
of Salvador were enrolled in a case–control study of risk factors
for HTLV-1 infection, as previously described.
19
In brief, blood
donors were routinely screened for bloodborne infections (includ-
ing HLTV-1) as recommended by the Brazilian Ministry of Health.
Those who had a positive screening test by the ELISA method
had new blood samples reanalyzed in duplicate, and if persistently
positive were subsequently confirmed by Western Blot analysis.
DNA was extracted from peripheral blood mononuclear cells of
23 participants from whom blood samples were available, using
the QIAamp Blood kit (Qiagen Inc., Chatsworth, CA). Nested PCR
was performed and LTR regions were amplified as two overlap-
ping fragments: an LTR-gag segment of 473 bp and a tax-LTR
segment of 479 bp, as previously described.
18
All amplification
products were electrophoresed on 1% agarose gel stained with
ethidium bromide, and visualized with ultraviolet light.
The LTR amplification products were purified using the
QIAquick Gel Extraction kit (Qiagen) and sequenced directly on
an automated 3100 genetic analyzer (Applied Biosystems Inc.,
Foster City, CA). Sequencing reactions were carried out using the
inner PCR primers, and a Taq FS Dye terminator cycle sequenc-
ing kit (Applied Biosystems). The new LTR sequences and se-
lected references strains (including sequences from different eth-
nic groups and different geographic regions, and all African LTR
sequences of the Transcontinental subgroup) were downloaded
from the GenBank/EMBL database, aligned with ClustalX and
Dambe,
20
and manually edited using the GeneDoc programs.
21
Phylogenetic trees were generated using the neighbor-joining (NJ)
and maximum-likelihood (ML) methods of PAUP*, version
4.0b10.
22
The Tamura–Nei evolutionary model (which takes into
account different substitution rates for transversions and transi-
tions, as well as intersite substitution rate heterogeneity, using a
-distribution) was selected as the best model for the phyloge-
netic analysis (alpha parameter 0.5124). The likelihood ratio
test was employed to select the model as well as to test the mo-
lecular clock hypothesis. The NJ tree was constructed with an op-
timized nucleotide substitution rate matrix and a -shaped pa-
rameter, using empirical base frequencies. The reliability of the
NJ trees was assessed by analyzing 1000 bootstrap replicates. For
ML trees, a heuristic search was performed with a subtree-prun-
ing-regrafting branch swapping algorithm using the NJ tree as the
starting material, including its optimized parameters. The likeli-
hood ratio test (RT) method was used to calculate statistical sup-
port (expressed in p-values) for the branches. Trees were drawn
with TreeView, version 1.4.
Accession numbers: the GenBank accession numbers of the
new HTLV-1 fragments included in our phylogenetic study
were as follows: HB3167, DQ471187; HB3104, DQ471188;
HB3133, DQ471189; HB2966, DQ471190; HB3168, DQ471191;
HB3171, DQ471192; HB3170, DQ471193; HB3311, DQ471194;
HB3229, DQ471195; HB3114, DQ471196; HB2562, DQ471197;
HB3205, DQ471198; HB3135, DQ471199; HB3166, DQ471200;
HB3160, DQ471201; HB3119, DQ471202; HB3230, DQ471203;
HB3102, DQ471204; HB3203, DQ471205; HB3120, DQ471206;
HB3134, DQ471207; HB3202, DQ471208; HB3194, DQ471209.
The other HTLV-1 strains were: pyg19, L76310; ITIS, Z32527;
MEL5, L02534; HS35; D00294; GH78, D23693; CH26,
D23690; Bo, U12804; OD, U12805; Pr52, U12806; MOMJ,
Z31659 Pr144, U12807; GB233, D23692; RKI4, AF054627;
MT4, Z31661; HTLV24, DQ005565; HTLV04, DQ005557;
FNN100, DQ005547; Ni1.Peru, Y16484; Bl1.Peru, Y16481;
Me1.Peru, Y16478; Bl3.Peru, Y16483; Bl2.Peru, Y16482;
Qu1.Peru, Y16475; Me3.Peru, Y16480; Qu3.Peru, Y16477;
Me2.Peru, Y16479 and Ni2.Peru, Y16487; ATM, J02030;
ATK1, J02029; MT2, L03562; CR1, K02722; TBH2 to -4,
L76025, L76034 and L76028; TBH6 and -7, L76030 and
L76029; BOI, L36905; AMA, CMC, FCR and MAQS,
X88871–X88873 and X88876; MWMG, Z31662; StDen,
L76306; PH236, L76307; GAB7, L76311; 12503A, L76309;
H24, L76308; Ni3.Peru, Y16485; HKN, X88874; TBH5,
L76027; Cam, AF 063819; Nar, AF063820; H5, M37299;
AINU, D23694; Qu2.Peru, Y16476.
RESULTS
The phylogenetic analysis of the total LTR region showed
that all 23 sequences belonged to the Transcontinental (A) sub-
group of the Cosmopolitan (a) subtype, a finding that was sup-
ported by a bootstrap value of 78% (p0.001 for ML analy-
ses) (Fig. 1). The majority of our new sequences (n18)
MOTA ET AL.
504
clustered within the larger Latin American cluster (named clus-
ter beta in Fig. 1). One of these sequences formed a mono-
phyletic group with two Peruvian sequences (Me1 and Me2),
yet this cluster was supported by neither bootstrap nor ML. Four
of the other new sequences clustered monophyletically with
other sequences from South America, forming a smaller Latin
American cluster (named cluster alpha in Fig. 1), supported by
a bootstrap value of 93% and by ML analysis.
MOLECULAR STUDY OF HTLV-1 ISOLATES FROM BRAZIL 505
FIG. 1. Rooted neighbor-joining tree of 80 HTLV-1 strains based on a 718-bp fragment of the LTR region. The bootstrap val-
ues (60% and using 1000 bootstrap samples) on the branches represent the percentage of trees for which the sequences located
at the right end of the branch form a monophyletic group. Mel5 was used as the out-group. Geographic origin is given in paren-
theses. Newly sequenced LTRs included in this analysis are in bold. **The ML method was highly statistically significant (p
0.001). The rectangle represents part of the sequenced LTR region showing the 12-bp duplication of the isolate HB3166 (the up-
per line represents the reference sequence ATK1).
Two HTLV-1 prototypes from South Africa (HTLV24 and
HTLV04) belonged to each of the Latin American clusters. The
sequence HTLV24 clustered into the smaller Latin American
cluster (alpha) (93% bootstrap; p0.001 for ML), and
HTLV04 into the larger Latin American cluster (beta) (65%
bootstrap; p0.001 for ML). We have tested the molecular
clock hypothesis for all branches intending to date our data. In
all cases the null hypothesis was rejected.
One of our new sequences (HB3166) clusters ancestral to the
larger Latin American cluster (beta), yet with no bootstrap or
ML support. This HB3166 sequence has a 12-bp duplication
within its LTR region, corresponding to position 9018 to 9029
of the ATK1 prototype. This position does not overlap tax and
is distal from the functional regions in the LTR.
DISCUSSION
In line with previous reports,
17,18
our phylogenetic analysis
has shown that all sequences belonged to the Cosmopolitan sub-
type, Transcontinental subgroup. Moreover, this study demon-
strates that sequences from southern Africa cluster into both al-
pha and beta Latin American groups. These findings indicate a
closer relation between Latin American and South African
strains compared to West African HTLV-1 isolates. In addition,
with our previously published reports we present the hypothe-
sis that infected individuals from the southern region of Africa
might have introduced HTLV-1 into Latin America.
12,17,18
However, in the current phylogenetic analysis there is no con-
clusive evidence on the direction of migration, either from
Southern Africa to Latin America or from Latin America to
southern Africa.
The fact that sequences clustered into two different clusters
suggests that there were multiple introductions of HTLV-1 into
Salvador. Nonetheless, the observation that the sequences clus-
tered homogeneously within each Latin American group sug-
gests that the introduction of HTLV-1 among people in Sal-
vador was relatively recent, most likely occurring during the
slave trade that took place during the seventeenth, eighteenth,
and nineteenth centuries, as previously reported.
12,17,18,23,24
Since the molecular clock was rejected, we could not date the
interior nodes.
However, these findings appear to contradict historical data,
according to which the vast majority of slaves brought to this
region originally came from western Africa,
15,16
where HTLV-
1aC predominates.
2
A possible explanation for this discrepancy
is that individuals originally from the Central Africa continent
(and infected with HTLV-1aA) migrated to southern Africa as
suggested by the presence of the
A
-globin CAR haplotype in
southern Africa.
18
The CAR haplotype is prevalent in Central
Africa, South Africa, Mozambique, and other regions of south-
ern Africa. The source of the CAR haplotype in South Africa
is not known, but may reflect the migration of the Bantu pop-
ulation from the Central to South African continent during the
last 3000 years. A plausible alternative explanation is that rel-
atively few isolates from all these regions have been analyzed,
hampering definitive conclusions.
The only sequence that clustered ancestral to the Latin Amer-
ican cluster beta had a 12-bp duplication in its LTR region (se-
quencing was performed three times for confirmation purposes).
As far as we know this is the first time that such a duplication
has been reported. The isolate is from an asymptomatic 37-year-
old male blood donor. It is not known at this point whether this
insertion could have any clinical or biological implication. We
intend to follow him up for clinical symptoms and proviral load
and to investigate his relatives.
ACKNOWLEDGMENTS
This paper is part of Augusto Cesar Andrade Mota’s Ph.D.
thesis, Escola Bahiana de Medicina e Saúde Publica, Postgrad-
uate Course in Medicine and Human Health. This study was
partially supported by FAPESB (Grants 303/03) and by the
Brazilian Ministry of Health (306/04 and 307/04). The authors
are grateful to Mr. Noilson Lazaro and Mrs. Elisabeth Deliège,
respectively, for their technical and editing assistance.
REFERENCES
1. Gessain A, Gallo R, and Franchini G: Low degree of human T-cell
leukemia/lymphoma virus type I genetic drift in vivo as a means
of monitoring viral transmission and movement of ancient human
populations. J Virol 1992;66:2288–2295.
2. Wattel E, Vartanian J-P, Pannetier C, et al.: Clonal expansion of
human T-cells leukemia virus type I infected cells in asymptomatic
symptomatic carriers without malignancy. J Virol 1995;69:2863–
2868.
3. Miura T, Fukunaga T, Igarashi T, et al.: Phylogenetic subtypes of
human T-lymphotropic virus type I and their relations to the an-
thropological background. Proc Natl Acad Sci USA 1994;91:1124–
1127.
4. Hahn BH, Shaw GM, Popovic M, et al.: Molecular cloning and
analysis of a new variant of human T-cell leukemia virus (HTLV-
Ib) from an African patient with adult T-cell leukemia-lymphoma.
Int J Cancer 1984;34:613–618.
5. Gessain A, Boeri E, Yanagihara R, et al.: Complete nucleotide se-
quence of a highly divergent human T-cell leukemia (lympho-
tropic) virus type I (HTLV-I) variant from Melanesia: Genetic and
phylogenetic relationship to HTLV-I strains from other geograph-
ical regions. J Virol 1993;67:1015–1023.
6. Mahieux R, Chappey C, Georges-Courbot MC, et al.: Simian T-
cell lymphotropic virus type I from Mandrillus sphinx as a simian
counterpart of human T-cell lymphotropic virus type I subtype D.
J Virol 1998;72:10316–10322.
7. Mboudjeka I, Zekeng L, Yamashita M, et al.: Prevalence and phy-
logenetic analysis of HTLV-I isolates in Cameroon, including those
of the Baka Pygmy. Jpn J Cancer Res 1997;88:619–624.
8. Salemi M, Van Dooren S, Audenaert E, et al.: Two new human T-
lymphotropic virus type I phylogenetic subtypes in seroindetermi-
nates, a Mbuti pygmy and a Gabonese, have close relatives among
African STLV-I strains. Virology 1998;246:277–287.
9. Wolfe ND, Heneine W, Carr JK, et al.: Emergence of unique pri-
mate T-lymphotropic viruses among central African bush meat
hunters. Proc Natl Acad Sci USA 2005;102:7994–7999.
10. Gasmi M, Farouqi B, d’Incan M, et al.: Long terminal repeat se-
quence analysis of HTLV type I molecular variants identified in
four North African patients. AIDS Res Hum Retrovirol 1994;10:
1313–1315.
11. Vidal AU, Gessain A, Yoshida M, et al.: Phylogenetic classifica-
tion of human T cell leukaemia/lymphoma virus type I genotypes
in five major molecular and geographical subtypes. J Gen Virol
1994;75:3655–3666.
MOTA ET AL.
506
12. Van Dooren S, Gotuzzo E, Salemi M, et al.: Evidence for a post-
Columbian introduction of HTLV-I in Latin America. J Gen Virol
1998;79:2695–2708.
13. Proietti FA, Carneiro-Proietti ABF, Catalan-Soares BC, et al.:
Global epidemiology of HTLV-I infection and associated diseases.
Oncogene 2005;24:6058–6068.
14. Dourado I, Alcantara LCJ, Barreto ML, et al.: HTLV-I in the gen-
eral population of Salvador, Brazil: A city with African ethnic and
sociodemographic characteristics. J Acquir Immune Defic Syndr
2003;34:527–531.
15. Curtin PD: The Slave Atlantic Trade: A Census. The University of
Wisconsin Press, Milwaukee, 1969.
16. Verger P: Flux et reflux de la traite des nègres entre le Golfe de
Benin et Bahia de Todos os Santos. Mouton, Paris, 1968.
17. Alcantara LC Jr, Van Dooren S, Gonçalves MS, et al.: Globin hap-
lotypes of human T-cell lymphotropic virus type I-infected indi-
viduals in Salvador, Bahia, Brazil, suggest a post-Columbian
African origin of this virus. J Acquir Immune Defic Syndr 2003;
33:536–542.
18. Alcantara LCJ, Oliveira T, Gordon M, et al.: Tracing the origin of
Brazilian HTLV-1 as determined by analysis of host and viral
genes. AIDS 2006;20:780–782.
19. Mota A, Nunes C, Mello A, et al.: A case-control study of HTLV-
1 infection among blood donors in Salvador, Bahia, Brazil—As-
sociated risk factors and trends towards declining prevalence.
Brazil J Hematol Hemother 2006;28(2):120–126.
20. Xia X: DAMBE: Data analysis in molecular biology and evolu-
tion. Department of Ecology and Biodiversity, University of Hong
Kong, Hong Kong, 2000.
21. Nicholas KB, Nicholas HBJ, and Deerfield DW: GeneDoc: Analysis
and visualization of genetic variation. EMBNEW NEWS 1997:30.
22. Swofford DL, Olsen GJ, Waddell PJ, et al.: Phylogenetic infer-
ence. In: Molecular Systematics, 2nd ed. (Hillis DM et al., eds.).
Sinauer, Sunderland, MA, 1996, pp. 407–514.
23. Yamashita M, Veronesi R, Menna-Barreto M, et al.: Molecular epi-
demiology of human T-cell leukemia virus type I (HTLV-1) Brazil:
The predominant HTLV-1s in South America differ from HTLV-
Is of Japan and Africa, as well as those of Japanese immigrants
and their relatives in Brazil. Virology 1999;261:59–69.
24. Van Dooren S, Pybus OG, Salemi M, et al.: The low evolutionary
rate of human T-cell lymphotropic virus type-1 confirmed by anal-
ysis of vertical transmission chains. Mol Biol Evol 2004;21:603–
611.
Address reprint requests to:
Luiz Carlos Junior Alcantara
LASP/CPqGM/FIOCRUZ
Rua Waldemar Falcão 121
Brotas, Salvador
Bahia, Brazil 40295-001
E-mail: lalcan@cpqgm.fiocruz.br
MOLECULAR STUDY OF HTLV-1 ISOLATES FROM BRAZIL 507
This article has been cited by:
1. Themístocles Magalhães, Aline Cristina Mota-Miranda, Luiz Carlos Junior Alcantara, Viviana Olavarria, Bernardo Galvão-Castro,
Maria Fernanda Rios-Grassi. 2008. Phylogenetic and molecular analysis of HTLV-1 isolates from a medium sized town in
Northern of Brazil: Tracing a common origin of the virus from the most endemic city in the country. Journal of Medical Virology
80:11, 2040-2045. [CrossRef]
2. René Gastaldello , Alena Mayo Iñiguez , Koko Otsuki , Gabriela Lamas , Marcos Balangero , María Gabriela Barbas , Andrea
Mangano , Luisa Sen , Eduardo Maturano , Carlos Remondegui , Ana Carolina Paulo Vicente , Sandra Gallego . 2008. HTLV
Type 1 Genetic Types among Native Descendants in ArgentinaHTLV Type 1 Genetic Types among Native Descendants in
Argentina. AIDS Research and Human Retroviruses 24:9, 1139-1146. [Abstract] [PDF] [PDF Plus]
3. Filipe Ferreira de Almeida Rego , Luiz Carlos Junior Alcantara , Jose Pereira Moura Neto , Aline Cristina Andrade Mota
Miranda , Osmario de Souza Pereira , Marilda de Souza Gonçalves , Bernardo Galvão-Castro . 2008. HTLV Type 1 Molecular
Study in Brazilian Villages with African Characteristics Giving Support to the Post-Columbian Introduction HypothesisHTLV
Type 1 Molecular Study in Brazilian Villages with African Characteristics Giving Support to the Post-Columbian Introduction
Hypothesis. AIDS Research and Human Retroviruses 24:5, 673-677. [Abstract] [PDF] [PDF Plus]