Content uploaded by Donald J Hamel
Author content
All content in this area was uploaded by Donald J Hamel on Aug 25, 2023
Content may be subject to copyright.
Content uploaded by Christopher Mullins
Author content
All content in this area was uploaded by Christopher Mullins
Content may be subject to copyright.
AIDS RESEARCH AND HUMAN RETROVIRUSES
Volume 23, Number 10, 2007, pp. 1189–1196
© Mary Ann Liebert, Inc.
DOI: 10.1089/aid.2007.0037
Twenty Years of Prospective Molecular Epidemiology in
Senegal: Changes in HIV Diversity
DONALD J. HAMEL,
1
JEAN-LOUIS SANKALÉ,
1
GEOFFREY EISEN,
1
SEEMA THAKORE MELONI,
1
CHRISTOPHER MULLINS,
1
AISSATOU GUEYE-NDIAYE,
2
SOULEYMANE MBOUP,
2
and PHYLLIS J. KANKI
1
ABSTRACT
Over a 20-year period we have observed the dynamics of HIV-1 subtypes and HIV-2 infection in a prospec-
tive cohort of registered female sex workers (FSW) in Dakar, Senegal. Prevalence and incidence rates for
HIV-1 and HIV-2 are described from 290 seroprevalent and 193 seroincident subjects who were among the
3910 women enrolled between 1985 and 2004. We report a significant decrease of HIV-2 prevalence in the co-
hort, parallel to the introduction and rise of HIV-1 infection. In 328 HIV-1-infected women, a 385-bp C2–V3
fragment of the envelope gene was sequenced and classified into the following subtypes or recombinant forms:
239 (72%) were subtype A [of which 180 (55%) were CRF02_AG and 53 (16%) were A3], 10 (3%) were B,
12 (4%) were C, 11 (4%) were D, 18 (6%) were G, 24 (7%) were CRF06_cpx, and 7 (2%) were CRF09_cpx.
We found an increasing proportion of CRF02_AG over many years, but recently subsubtype A3 has over-
taken CRF02_AG, with the largest proportion of new infections. The predominance of existing HIV-1 sub-
types did not preclude the emergence and increase of other closely related subtypes or recombinant forms.
This 20-year prospective serological and sequence analysis of HIV viruses reveals a complex and changing
HIV epidemic in Senegal.
1189
INTRODUCTION
H
ISTORICALLY
,
THE FIRST
HIV-1
VIRUSES
isolated and char-
acterized were from the United States and Europe. In the
early 1990s, it was discovered that multiple HIV-1 subtypes ex-
isted, but surprisingly the U.S. and European epidemics resulted
from a single HIV-1 subtype, subtype B. It is now recognized
that HIV-1 can be classified into three groups (M, N, O) and
that the M group consists of at least nine different subtypes (A,
B, C, D, F, G, H, J, and K) and circulating recombinant forms
(CRFs). Most of the diverse HIV-1 subtypes and HIV-2 have
been described in sub-Saharan Africa where they are associated
with the largest proportion of HIV infections worldwide. It is
still poorly understood why these distinct subtypes and their
epidemics have emerged or how they will evolve over time.
Differences in transmission or disease potential between HIV-
1 subtypes have been described, but their association with the
emergence or contribution to the dynamics of the HIV epi-
demics has not been well studied.
1–3
Since 1985 we have conducted a prospective study of regis-
tered female sex workers (FSW) in Dakar, Senegal. HIV-2 was
first described in this cohort of FSWs and the initial prevalence
rate for HIV-2 in infection in 1985 was 8%.
4
By contrast, HIV-
1 was first detected in Senegal in 1985, and since then has re-
mained at a constant prevalence rate of about 1% in the gen-
eral population,
5,6
although HIV-1 infection in FSWs had
increased to over 13% by 2003.
Globally, most areas have just one or two HIV-1 subtypes
geographically. We, along with others, have shown that the pre-
dominant HIV-1 subtypes circulating in West Africa are sub-
type A and AG recombinants, including the circulating recom-
binant form CRF02_AG.
7–13
New subtypes, subsubtypes, and
CRFs continue to be identified, and the issue of recombination
has become an important consideration in tracking the global
spread of HIV. Most studies that have analyzed subtype distri-
bution are based on a cross-sectional design. Over a 20-year
period, we enrolled nearly 4000 women and sequenced all avail-
able HIV-1 samples to study the dynamics of HIV-1 group
1
Department of Immunology and Infectious Disease, Harvard School of Public Health, Boston, Massachusetts 02115.
2
Laboratory of Bacteriology-Virology, Cheikh Anta Diop University, and Institut d’Hygiene Sociale, Dakar, Senegal.
M subtypes. Our prospective design has allowed us to docu-
ment the introduction and dominance of specific subtypes and
CRFs through the analysis of incident infections.
MATERIALS AND METHODS
The study focuses on an open cohort of female commercial
sex workers who were able to enroll or leave the cohort through-
out the study period. Samples were obtained during regular
biannual visits to the Institut d’Hygiene Sociale clinic in Dakar,
Senegal, which are a requirement for legal registration of the
sex workers. At these visits, women were provided with a clin-
ical examination as well as treatment for any sexually trans-
mitted diseases. Women who gave informed consent were en-
rolled into our study group. Serology was performed using
immunoblot, as well as synthetic peptide assays, to diagnose
infection with HIV-1, HIV-2, or HIV dual infection (HIV-D).
14
The detailed protocols for study recruitment and the cohort de-
sign have been described previously.
5,15,16
Seroincident HIV-1
cases were defined as those who were HIV negative at enroll-
ment and subsequently seroconverted to HIV-1. Time of in-
fection, or seroconversion date, was calculated as the midpoint
between the last seronegative sample date and the first seropos-
itive sample date. HIV-1-infected individuals who were also
HIV-2 infected, or HIV dually infected, were included in this
study.
Blood samples were collected from the women upon enroll-
ment and at scheduled visits, and separated into plasma and pe-
ripheral blood mononuclear cell (PBMC) fractions. For HIV-
1-positive subjects, proviral DNA was extracted from PBMCs
(Qiagen Blood DNA Kit; Qiagen Inc., Chatsworth, CA). We
performed a nested polymerase chain reaction (PCR) of a 385-
bp fragment of the envelope gene (C2–V3) using previously
described primers and conditions.
1,17
The PCR product was then
direct sequenced using the second round PCR primers, KK30
and KK40. Products were cloned when necessary, using the
pCR2.1 vector (T/A Cloning; Invitrogen, San Diego, CA). Se-
quence reactions were performed by either dye terminator cy-
cle sequencing using Taq polymerase, or by the Big Dye Ter-
minator v1.1 system (both Applied Biosystems Inc., Foster City,
CA). For incident cases, DNA samples were selected as the first
available after seroconversion, typically from the same bleed
date as the first seropositive result.
Alignments of sequences were performed using the Clustal
multiple alignment software package (Clustal W 1.6).
18
Minor
manual adjustments were made where necessary to accommo-
date reading frames. Phylogenetic analysis was performed us-
ing the neighbor-joining method, and reliability was estimated
from 1000 bootstrap resamplings. Representative sequences
from the HIV database were included in the analysis.
19
The nucleotide sequences were submitted to GenBank (ac-
cession numbers AF085284–085327, AF020819–020827,
AF526650–526876, and DQ323178–323400).
RESULTS
Annual HIV prevalence and incidence
From 1985 to 2004, we enrolled 3910 women into the Dakar
FSW cohort; of these, 290 were HIV-1 seropositive (sero-
prevalent) on entry and 193 seroconverted to HIV-1 (seroinci-
dent) during observation in the cohort. Based on the serology
screening results, we calculated the annual prevalence and in-
cidence rates (IR) for HIV-1, HIV-2, and HIV-D. HIV-1 preva-
lence has climbed steadily since its introduction in 1985 to just
below 10% in 1993 (Fig. 1A). HIV-1 prevalence moderated
from 1993 to 1998, and then continued to rise again to 13.8%
in 2003. HIV-2 decreased from the range of 8–11% between
1985 and 1995, to 5.5% in 2003. HIV-D increased from 0.3%
in 1985 to 3.2% in 1995, and then decreased to 1.8% in 2003.
The annual incidence rate for HIV-1 increased from 0 in 1986
to 2.5/100 person years of observation (PYO) in 1992, while
HIV-2 and HIV-D have decreased to less than 0.3/100 PYO
each since 1999 (Fig. 1B). The incidence rate for HIV-1 has
also remained higher than that of HIV-2 and HIV-D since 1990,
in the range of 0.9–2.8/100PYO.
Subtype summary and phylogenetic analysis
We obtained subtype data for 151/193 (78%) seroincident
individuals identified, and similarly subtyped 177/290 (61%)
seroprevalent individuals. The subtyped seroincident and sero-
prevalent groups together comprise a total of 328 subjects
HAMEL ET AL.
1190
A
0
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
Year
HIV-1
Percent
B
HIV-2
HIV-D
0
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Ye a r
HIV-1
Incidence Rate/100 PYO
HIV-2
HIV-D
FIG. 1. (A) Annual prevalence for HIV. (B) Annual incidence rates for HIV.
(Table 1). Of the 328 subjects for whom we present subtype
data in this paper, 223 represent novel sequences that have not
been published or analyzed previously.
Assignment of specific subtypes was determined by phylo-
genetic analysis of all sample sequences with reference subtype
sequences
20
using the neighbor-joining method. Due to the large
number of samples analyzed here, we have presented repre-
sentative data in two phylogenetic trees, one for the prevalent
group (Fig. 2A) and one for the incident group (Fig. 2B). All
sequences, with the exception of three unclassifiable subtype A
viruses, clustered definitively with reference subtypes with
bootstrap values between 53 and 100%. With the small size of
the fragment sequenced, it can be difficult to achieve high res-
olution between closely related subtypes such as subtype G and
CRF06_cpx or between CRF02_AG and sub-subtype A3. Sub-
subtype A3 has only recently been identified,
21
and is not yet
included in subtype reference alignments published by the Los
Alamos National Laboratories.
22
A majority of the women, 72% (239/328), were infected with
subtype A viruses, and of those most 55% (180/328) were
CRF02_AG. CRF06_cpx contributed 7% (24/328), subtype G
6% (18/328), and all other subtypes and recombinant forms con-
tributed less than 5% (15/328) each to the overall total in the
cohort.
In the 151 individuals with documented new infection
(seroincident), the contributions from each subtype were deter-
mined for each calendar year. From 1986 to 1990 incident cases
included subtypes D, C, A1, and CRF02_AG. Subsubtype A3
was introduced into the cohort in 1991, CRF06_cpx in 1992,
subtype G in 1993, and CRF09_cpx in 1994. In 1999, all inci-
dent cases were subtype A, with 71% (12/17) from CRF02_AG
and 24% (4/17) from subsubtype A3.
Likewise, the distribution of each subtype was determined
for every year among the seroprevalent cases. We documented
the first prevalent case of subsubtype A3 in 1988, 3 years be-
fore the first incident case of subsubtype A3 was observed. The
first detection of CRF06_cpx was in 1990, 2 years before the
first incident case was observed. Also, the first evidence of sub-
type G among the prevalent was in 1992, 1 year before the first
incident case.
Subtype trends
A timeline was created for the 151 incident cases for which
we have subtype data (Fig. 3). The timeline graphically illus-
trates the introduction of subsubtype A3, CRF06_cpx, and sub-
type G into the incident cohort. Upon the introduction of sub-
subtype A3 in 1991, it has increased to represent 23% of new
incident infections over the 20-year study period. Within this
incident group we documented the diagnosis of AIDS in 13
women, also graphically illustrated on the timeline.
To better resolve trends of specific subtypes over time, and
to minimize the impact of small annual fluctuations, we created
tables collapsing data into 5-year intervals for both the HIV-1
seroincident cases (Fig. 4A) and for the HIV-1 seroprevalent
cases (Fig. 4B).
Among the HIV-1 seroincident cases, there was no evidence
of CRF06_cpx during the period from 1985 to 1989, but this
CRF represented 6% of infections from 2000 to 2004 (Fig. 4A).
We also documented the increase in new HIV-1 infections by
subsubtype A3 from 0% in 1985–1989 to 41% in 2000–2004.
For the seroprevalent group, subtype G was not observed
during the 1985–1991 period, but this subtype contributed 22%
of HIV-1 cases entering the cohort from 2000 to 2004 (Fig.
4B). The dominant subtype remained CRF02_AG throughout
the study period, representing 50% from 1985 to 1989 and 44%
from 2000 to 2004.
Epidemiologic analyses
For the total 483 women who tested HIV positive during the
study period, we were unable to subtype 155 (32%) due to lack
of an available DNA sample, predominantly from the study
years of 1985–1986 and 2002–2004. The low numbers of sub-
type data from recent years should not be inferred to necessar-
ily represent a decline in incidence or prevalence. Analyses
were performed to examine sampling bias as a result of the open
MOLECULAR EPIDEMIOLOGY OF HIV IN SENEGAL 1191
T
ABLE
1. S
UMMARY OF
HIV-1 S
UBTYPE
R
ESULTS
Seroincident cases Seroprevalent cases Total cases
Subtype nnn(%)
A 115 124 239 (72)
A1 2 1 3 (1)
A3 32 21 53 (16)
A unclassified recombinant 1 2 3 (1)
CRF02_AG 80 100 180 (55)
B 3 7 10 (3)
C 9 3 12 (4)
D 7 5 12 (4)
F 1 2 3 (1)
G 3 16 19 (6)
J 0 2 2 (1)
CRF 06 7 16 23 (7)
CRF 09 6 2 8 (2)
Total 151 177 328
A
B
FIG. 2. (A) Representative neigh-
bor-joining radial phylogenetic tree
of HIV-1 prevalent cases. Reference
strains representing group M sub-
types and CRFs, designated by ,
are as follows: A1_SE.94.SE7253,
A1_UG.92UG037, B_FR.83.HXB2,
B_US.90.WEAU160, CRF02AG_
FR.91.DJ264, CRF02AG_GH.G829,
C_BW.96BW0502, C_BR.92BR025,
CRF09_TM7808, CRF06_CX.ML.
95ML84, CRF06_CX.SN.97SE1078,
D_CD.83.ELI, D_CD.84ZR085,
F1_BE.93.VI850, F1_FR.96.M411,
G_NG.92NG083, G_SE.93.SE6165.
(B) Representative neighbor-joining
radial phylogenetic tree of HIV-1 in-
cident cases. Reference strains rep-
resenting group M subtypes and
CRFs, designated by , are as fol-
lows: A1_SE.94.SE7253, A1_UG.
92UG037, B_FR.83.HXB2, B_US.90.
WEAU160, CRF02AG_FR.91.DJ264,
CRF02AG_GH.G829, C_BW.96BW
0502, C_BR.92BR025, CRF09_
TM7808, CRF06_CX.ML.95ML84,
CRF06_CX.SN.97SE1078,
D_CD.83.ELI, D_CD.84ZR085, F1_
BE.93.VI850, F1_FR.96.M411, G_NG.
92NG083, G_SE.93.SE6165.
MOLECULAR EPIDEMIOLOGY OF HIV IN SENEGAL 1193
FIG. 3. HIV-1 seroincident data timeline representing the distribution of infections among the FSWs in our cohort. The line
before the infection date represents an individuals negative person-years of observation, and the color line after the infection date
represents the positive person-years of observation. The start of each line indicates the first bleed date for each subject, and the
end of line indicates the last bleed date.
HAMEL ET AL.
1194
A
FIG. 4. (A) Incidence rates of subtyped HIV-1, with 95% confidence intervals. Incidence rates for HIV-1 and individual sub-
types are calculated annually before being compiled into 5-year groups. Each person contributes only once toward incidence. (B)
Prevalence of subtyped HIV-1, with 95% confidence intervals. Prevalence rates for HIV-1 and each subtype are calculated an-
nually before being compiled. Mean annual prevalence is the sum of the prevalence for each year divided by the number of years
for each group.
B
cohort design, which could affect our overall conclusions. Pear-
son’s chi-square tests revealed that a higher proportion of in-
dividuals of Senegalese nationality (p0.001) and HIV-D
(p0.001) were likely to be subtyped, compared to the non-
subtyped individuals. Similarly, previous analyses have shown
that Ghanaian women were more likely to drop out of the co-
hort and less likely to have been subtyped.
6,7
Therefore we must
exercise some caution in extending our conclusions about sub-
type dynamics for groups with nationality other than Sene-
galese. Our analysis found no significant difference comparing
the subtyped to the nonsubtyped with respect to religion, edu-
cation level, marital status, or age at which the individual was
diagnosed as HIV positive.
DISCUSSION
This study describes the molecular epidemiology of the HIV-
1 subtypes in Senegal over the past 20 years. We have docu-
mented new observations of HIV-1 subtype dynamics in this
region; our data are also consistent with cross-sectional preva-
lence studies.
6,7,23
Our observations on subtype introduction
and dynamics are further confirmed by the longevity and
prospective design of the study. An overview of the annual in-
cidence data for HIV (Fig. 1B) reveals a decreasing rate for
HIV-2 and HIV-D infections, but a fairly stable rate of new
HIV-1 infections. The incidence rate for HIV-1 has remained
approximately 1–1.5% since 1990. Our analysis of the annual
prevalence (Fig. 1A) demonstrates a decrease of HIV-2 and
HIV-D infections. However, we have seen an increase in HIV-
1 infection, after moderation between 1992 and 1999, reaching
a prevalence of 13.8% in 2003 and continuing to increase in
2004.
This cohort consists of women, self-identified as registered
sex workers, with higher rates of HIV infection than other pop-
ulations in Senegal, such as blood donors or pregnant women
visiting antenatal clinics. Previous analyses have demonstrated
that the women enrolled in this cohort are representative of the
registered sex worker population with regard to potential con-
founding risk factors including nationality, age, years of regis-
tered prostitution, and distributions of HIV serostatus. These
studies of the cohort have also shown the predominant nation-
ality of the women enrolled to be Senegalese (75.2%), or
Ghanaian (16.7%), with Guinea-Bissau and other nationalities
comprising the remainder (8.1%).
15
The initiation of this study early in the Senegal HIV-1 epi-
demic and the prospective study design carried out for a 20-
year period are unique, and provide important new information
on the epidemics of both HIV-1 and HIV-2, as well as the in-
troduction of distinct HIV-1 subtypes in this region. We have
described and characterized the entry and different rates of
spread of subtype G, subsubtype A3, and CRF06_cpx. The rapid
increase in subsubtype A3 infection suggests a “fitness” ad-
vantage that is worthy of further study. Sarr et al. have previ-
ously described a larger proportion of subsubtype A3 infections
in HIV/HIV-2 dually infected individuals in this cohort, which
led to the possibility that this variant might have a transmission
advantage.
24
Our subtype classification is somewhat limited by
the subgenomic sequence analysis, which does not allow a de-
finitive conclusion of subtype throughout the genome. For the
328 individuals with sequence data presented here, we have also
performed gag gene sequencing on a subset of 44 (data not
shown) that was classified as CRF02_AG or subsubtype A3
based on envelope region sequence. In phylogenetic analysis
42 of 44 (95%) of these gag sequences clustered with the same
subtype as the env sequences. In the two discordant samples, a
recombination event may have occurred, as one sample clus-
tered with CRF02_AG in env and subsubtype A3 in gag, and
the other sample clustered with subsubtype A3 in env and
CRF02_AG in gag. Therefore the majority of samples tested
revealed a similar sequence subtype in two regions, lending sup-
port for our subtype designations by env subgenomic sequenc-
ing. However, we note that the proportion of recombinant
viruses may be underestimated based on our methodology.
HIV-2 was the dominant virus in 1985, when HIV-1 was
first entering this population. Between 1985 and 2004, HIV-2
prevalence dropped from 8% to 4%, a 50% decrease over 20
years. Concurrently, HIV-1 prevalence increased from 0 to 13%
during the same time period. Anderson and May discussed pre-
dictions for the interaction of HIV-2 and HIV-1 on a popula-
tion level.
25
Their prediction that HIV-2 would decrease in the
face of an increasing HIV-1 prevalence appears to be supported
by our data. The attenuated rate of HIV-1 infection in a high
risk group of sex workers would also suggest that the interac-
tion of the HIV viruses in this population may inhibit explo-
sive increases in HIV-1 that have been described in similar HIV-
1 studies of sex workers.
26
In addition, the registration and
health care system for these women supported by a strong gov-
ernment HIV/AIDS control, prevention, and treatment program
may have contributed to the decline or stabilization of infec-
tion rates for both viruses.
Our prospective study design has reduced the risk of a sam-
pling bias that would, for example, be of concern in a cross-
sectional study focusing on hospitalized patients.
7
Examining
this cohort over time has allowed us to describe the rapid rise
of subsubtype A3 since the late 1990s. Our timeline of incident
HIV infection reveals the emergence of subsubtype A3 in 1991,
CRF06_cpx in 1992, and subtype G in 1993.
13
Similarly, anal-
ysis allowed the identification of recombinant CRF09_cpx,
which entered the incident group in 1994 and was first described
in this cohort.
27
Additionally, these data support and confirm
the predominance of recombinant subtype CRF02_AG in West
Africa, as documented in both the HIV-1 incident and preva-
lent groups of our study.
The continent of Africa bears the worlds highest HIV infec-
tion burden (UNAIDS, 2004), as well as the highest diversity
of HIV-1 subtypes.
28
While HIV-1 subtype B continues to re-
main the most characterized strain, other subtypes have perhaps
greater importance, for example, subtype C infection is re-
sponsible for over 50% of all infections worldwide.
28
Previous
studies have suggested higher transmission risk from subtype
CRF01_AE than subtype B.
29
This population has been followed now for over two decades,
and new studies continue to evolve from this cohort and our
long-term collaborative research program. Our analysis has re-
vealed striking changes in the composition of the Senegalese
HIV epidemic over the study period thus far. Tracking the
global spread and relative success of different HIV strains and
subtypes will remain important focuses of future research, treat-
ment, and prevention endeavors. These prevalence and subtype
MOLECULAR EPIDEMIOLOGY OF HIV IN SENEGAL 1195
data significantly add to the knowledge of the molecular epi-
demiology of HIV in Senegal, and advances the goal of un-
derstanding the importance of multiple subtypes in the natural
history of HIV in West Africa.
REFERENCES
1. Kanki PJ, Hamel DJ, Sankale JL, Hsieh C, Thior I, Barin F, et al.:
Human immunodeficiency virus type 1 subtypes differ in disease
progression. J Infect Dis 1999;179:68–73.
2. Vasan A, Renjifo B, Hertzmark E, Chaplin B, Msamanga G, Es-
sex M, et al.: Different rates of disease progression of HIV type 1
infection in Tanzania based on infecting subtype. Clin Infect Dis
2006;42:843–852.
3. Baeten JM, Chohan B, Lavreys L, Chohan V, McClelland RS, Cer-
tain L, et al.: HIV-1 Subtype D infection is associated with faster
disease progression than subtype A in spite of similar plasma
HIV-1 loads. J Infect Dis 2007;195:1177–1180.
4. Barin F, Mboup S, Denis F, Kanki P, Allan JS, Lee TH, and Es-
sex M: Serological evidence for virus related to simian T-lym-
photropic retrovirus III in residents of West Africa. Lancet
1985;ii:1387–1390.
5. Kanki P, M’Boup S, Marlink R, Travers K, Hsieh CC, Gueye A,
et al.: Prevalence and risk determinants of human immunodefi-
ciency virus type 2 (HIV-2) and human immunodeficiency virus
type 1 (HIV-1) in west African female prostitutes. Am J Epidemiol
1992;136:895–907.
6. Meda N, Ndoye I, M’Boup S, Wade A, Ndiaye S, Niang C, et al.:
Low and stable HIV infection rates in Senegal: Natural course of
the epidemic or evidence for success of prevention? AIDS 1999;13:
1397–1405.
7. Toure-Kane C, Montavon C, Faye MA, Gueye PM, Sow PS, Ndoye
I, et al.: Identification of all HIV type 1 group M subtypes in Sene-
gal, a country with low and stable seroprevalence. AIDS Res Hum
Retroviruses 2000;16:603–609.
8. Carr JK, Torimiro JN, Wolfe ND, Eitel MN, Kim B, Sanders-Buell
E, et al.: The AG recombinant IbNG and novel strains of group M
HIV-1 are common in Cameroon. Virology 2001;286:168–181.
9. Esteves A, Parreira R, Piedade J, Venenno T, and Canas-Ferreira
WF: Genetic characterization of HIV type 1 and type 2 from Bis-
sau, Guinea-Bissau (West Africa). Virus Res 2000;68:51–61.
10. Ishikawa K, Janssens W, Brandful J, Heyndrickx L, Takebe Y, Am-
pofo W, et al.: Genetic and phylogenetic analysis of HIV type I
env subtypes in Ghana, West Africa. AIDS Res Hum Retroviruses
1996;12:1575–1578.
11. McCutchan FE, Carr JK, Bajani M, Sanders-Buell E, Harry TO,
Stoeckli TC, et al.: Subtype G and multiple forms of A/G inter-
subtype recombinant human immunodeficiency virus type 1 in
Nigeria. Virology 1999;254:226–234.
12. Vidal N, Koyalta D, Richard V, Lechiche C, Ndinaromtan T, Dji-
masngar A, et al.: High genetic diversity of HIV-1 strains in Chad,
West Central Africa. J Acquir Immune Defic Syndr 2003;33:
239–246.
13. Sankale JL, Hamel D, Woolsey A, Traore T, Dia TC, Gueye-Ndi-
aye A, et al.: Molecular evolution of human immunodeficiency
virus type 1 subtype A in Senegal: 1988–1997. J Hum Virol
2000;3:157–164.
14. Gueye-Ndiaye A, Clark R, Samuel K, NDour-Sarr AN, Ouangre
A, Sangare L, et al.: Cost-effective diagnosis of HIV-1 and HIV-2
by recombinant-expressed env peptide (566/966) dot blot analysis.
AIDS 1993;7:481–495.
15. Kanki PJ, Travers KU, Mboup S, Hsieh CC, Marlink RG, Gueyen-
diaye A, et al.: Slower heterosexual spread of HIV-2 than HIV-1.
Lancet 1994;343:943–946.
16. Marlink R, Kanki P, Thior I, Travers K, Eisen G, Siby T, et al.:
Reduced rate of disease development after HIV-2 infection as com-
pared to HIV-1. Science 1994;265:1587–1590.
17. Wolinsky S, Wike CM, and Korber BT: Selective transmission of
human immunodeficiency virus type-1 variants from mothers to in-
fants. Science 1992;255:1134–1137.
18. Thompson JD, Higgins DG, and Gibson TJ: CLUSTAL W: Im-
proving the sensitivity of progressive multiple sequence alignment
through sequence weighting, position-specific gap penalties and
weight matrix choice. Nucleic Acids Res 1994;22:4673–4680.
19. Myers G, Korber B, Hahn B, Eds. Human Retroviruses and AIDS
1995: A Compilation and Analysis of Nucleic Acid and Amino Acid
Sequences. Los Alamos National Laboratory, Theoretical Biology
and Biophysics Group, Los Alamos, NM, 1995.
20. Myers G, Korber B, Wain-Hobson S, Smith RF, Pavlakis GN, Eds.
Human Retroviruses and AIDS 1992. Los Alamos National Labo-
ratory, Theoretical BIology and Biophysics Group, Los Alamos,
NM, 1992.
21. Meloni ST, Kim B, Sankale JL, Hamel DJ, Tovanabutra S, Mboup
S, et al.: Distinct human immunodeficiency virus type 1 subtype
A virus circulating in West Africa: Sub-subtype A3. J Virol
2004;78:12438–12445.
22. Leitner T KB, Daniels M, Calef C, and Folely B: HIV-1 subtype
and circulating recombinant form (CRF) reference sequences,
2005. In: HIV Sequence Compendium 2005;2005:41–48.
23. van der Loeff MF, Awasana AA, Sarge-Njie R, van der Sande M,
Jaye A, Sabally S, et al.: Sixteen years of HIV surveillance in a
West African research clinic reveals divergent epidemic trends of
HIV-1 and HIV-2. Int J Epidemiol 2006;35:1322–1328.
24. Sarr AD, Sankale JL, Hamel DJ, Travers KU, Gueye-Ndiaye A,
Essex M, et al.: Interaction with human immunodeficiency virus
(HIV) type 2 predicts HIV type 1 genotype. Virology 2000;268:
402–410.
25. Anderson RM and May RM: The population biology of the inter-
action between HIV-1 and HIV-2: Coexistence or competitive ex-
clusion? AIDS 1996;10:1663–1673.
26. Travers K, MBoup S, Marlink R, Gueye-NDiaye A, Siby T, Thior
I, et al.: Natural protection against HIV-1 infection provided by
HIV-2. Science 1995;268:1612–1615.
27. McCutchan FE, Sankale JL, M’Boup S, Kim B, Tovanabutra S,
Hamel DJ, et al.: HIV type 1 circulating recombinant form
CRF09_cpx from west Africa combines subtypes A, F, G, and may
share ancestors with CRF02_AG and Z321. AIDS Res Hum Retro-
viruses 2004;20:819–826.
28. McCutchan FE: Global epidemiology of HIV. J Med Virol
2006;78(Suppl 1):S7–S12.
29. Kunanusont C, Foy HM, Kreiss JK, Rerks-Ngarm S, Phanuphak
P, Raktham S, et al.: HIV-1 subtypes and male-to-female trans-
mission in Thailand. Lancet 1995;345:1078–1083.
Address reprint requests to:
Phyllis Kanki
Department of Immunology and Infectious Diseases
Harvard School of Public Health
651 Huntington Ave. FXB-405
Boston, Massachusetts 02115
E-mail: pkanki@hsph.harvard.edu
HAMEL ET AL.
1196