912 • JID 2006:194 (1 October) • BRIEF REPORT
B R I E F R E P O R T
Repeated Intravaginal Inoculation
with Cell-Associated Simian
Immunodeficiency Virus Results
in Persistent Infection
of Nonhuman Primates
Masahiko Kaizu,2Andrea M. Weiler,2Kim L. Weisgrau,1
Kathleen A. Vielhuber,2Gemma May,1Shari M. Piaskowski,2
Jessica Furlott,2Nicholas J. Maness,1Thomas C. Friedrich,1
John T. Loffredo,2Amy Usborne,3and Eva G. Rakasz1,2
1Immunogenetics and Virology Unit,
University of Wisconsin–Madison, Madison
2AIDS Vaccine Research Laboratory,
3Clinical Pathology Unit, Wisconsin National Primate Research Center,
(See the major article by Subbarao et al., on pages 904–11, and the
editorial commentary by Grant and Wainberg, on pages 874–6.)
The goal of the present study was to develop a nonhuman
primate model of intravaginal human immunodeficiency vi-
rus (HIV) transmission with cell-associated virus. Repro-
ductively mature, cycling cynomolgus macaques with or
without chemically induced, transient ulcers of the lower
female reproductive tract repeatedly received challenge with
a variable amount of in vitro simian immunodeficiencyvirus
mac239–infected peripheral blood mononuclear cells. Per-
sistent viremia was established with surprisingly few in-
fectious lymphocytes containing physiologically relevant
quantities of cell-associated virus. This model will be indis-
pensable for the testing of vaccines and topical agents that
are aimed toward the prevention of heterosexual transmis-
sion of HIV.
Young women are the most vulnerable segment of the popu-
lation affected by the AIDS pandemic . Medical conditions
that compromise the tight epithelial barrier of vaginal and cer-
vical mucosa have been implicated in increased rates of HIV
Received 13 February 2006; accepted 17 April 2006; electronically published 29 August
Financial support: Wisconsin National Primate Research Center (base grant P51 RR00167).
Potential conflicts of interest: none reported.
Reprints or correspondence: Dr. Eva G. Rakasz, AIDS Vaccine Laboratory, WNPRC, 555
Science Dr., Madison, WI 53711 (firstname.lastname@example.org).
The Journal of Infectious Diseases2006;194:912–6
? 2006 by the Infectious Diseases Society of America. All rights reserved.
The semen of HIV-infected men can contain both cell-free
and cell-associated virus, the latter of which is associated with
CD4+T cells and macrophages . In urogenital inflamma-
tions, even in asymptomatic persons, the concentration of lym-
phocytes can be up tocells/mL of ejaculate, and the con-
centration of macrophages can be up to
of ejaculate .
Models of in vivo transmission show that as few as 2 infected
peripheral blood mononuclearcells(PBMCs)cantransmitsim-
ian immunodeficiency virus (SIV) via the intravenous route
. In vitro assays involving human cervical cell lines, as well
as tissue explants, have provided evidence that transmission of
HIV with cell-associated virus is possible if there is a breach
in the mucosal integrity [6, 7]. SCID mice reconstituted with
human PBMCs also support the concept of male-to-female
heterosexual transmission of HIV via cell-associated virus .
However, to date, nonhuman primate models have delivered
contradictory results. In one study in which 4 chimpanzees
were exposed to cell-associated HIV-1LAI(subtype IIIB), one
animal developed persistent viremia, and another animal de-
veloped transient viremia . In another study that involved
rhesus macaques and cell-associatedSIVmac251,alltheanimals
remained virus free . Because the menstrual cycleandfemale
reproductive tract are similar in nonhuman primates and hu-
mans, any result derived from nonhuman primate models
should be relevant to HIV transmission in humans.
Understanding all the mechanisms employed by HIV during
heterosexual transmission of the virus is critical for designing
effective preventive strategies. In the present study, we dem-
onstrate that intravaginal transmission of SIV is possible with
cell-associated virus in nonhuman primates and that systemic
viremia is established with an unexpectedly small number of
Animals, materials, and methods.
molgus macaques (Macaca fascicularis) were housed at theWis-
consin National Primate Research Center at the University of
Wisconsin–Madison, according to guidelines from theNational
Institutes of Health (NIH) regarding the care and use of lab-
oratory animals . All procedures were performed on an-
esthetized animals according to an experimental protocol ap-
proved by the Research Animal Resource Committee of the
University of Wisconsin–Madison.
The lower female reproductive tract was inspected for signs
of preexisting inflammation, by use of a rigid fiberoptic scope.
Such signs included heavy vaginal discharge, cervicalerythema,
Mature, female cyno-
BRIEF REPORT • JID 2006:194 (1 October) • 913
the cell-associated virus inoculum. A, Proviral DNA copy content of in vitro–infected donor cynomolgus macaque peripheral blood mononuclear cells
at days 7–9 after initiation of the culture. Results are the mean values (?SD) of 3 independent experiments. CD3 antigen expression (B), relative
frequency of Gag p27-positive cells on the second day of viral challenge in macaques at day 8 after in vitro infection (C), and relative frequency of
p27-positive cells among uninfected control cells in the same experiment (D), as assessed by fluorescence-activated cell sorter analysis. Data were
acquired using the FACSCalibur flow cytometer (Becton Dickinson). Between 30,000 and 200,000 events were collected and were displayed in the
lymphocyte gate, as determined by the forward and side scatter parameters. Representative plots are for 1 of 11 independent stainings. APC,
allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin.
Proviral DNA copy content, cell composition, and relative frequency of simian immunodeficiency virus mac239 Gag p27-positive cells of
or disruption of the mucosal epithelium. After inspection, the
vaginal lumen was filled with 1 mL of surgical lubricant that
contained 3% benzalkonium chloride (Sigma). The animals
were kept in prone position for 15 min. One application of the
detergent resulted in visible ulceration 24 h later, which was
detected by fiberoptic scope. The location of the ulcers was
limited to the entry of the vagina and to the cervix. A total of
200 mL of inoculum that contained cell-associated virus was
deposited on the ulcer at the vaginal entry by use of a sterile
plastic 1-mL pipette. To avoid possible trauma to the mucosal
epithelium, control animals were challenged with viruswithout
visualization of the reproductive tract by fiberoptic scope. Viral
inocula were deposited in the vaginal entry. All animals were
kept in a supine position for 15 min before anesthesia was
Samples of mucosal secretion were collected and were stored
at ?80?C by use of premoistened Weck-Cel sponges (Med-
of centrifugation performed at 4?C. The sponges were washed
with 200 mL of PBS and underwent centrifugation again. Cy-
tokine levels were determined using the BD Cytometric Bead
Array Non-Human Primate Th1/Th2 Cytokine kit (BD), ac-
cording to the manufacturer’s instructions, by use of the FACS-
calibur flow cytometer and BD CBA software.
PBMCs obtained from an uninfected donor animal were
stimulated overnight invitrowith5mg/mLphytohemagglutinin
(Sigma) and then were incubated with SIVmac239 at an MOI
of for 4 h. The cells were washed twice, seeded in a
24-well plate at a concentration of
for 7–9 days in the presence of 50 U of interleukin (IL)–2 (the
source of which was the NIH AIDS Research and Reference
Reagent Program). Before inoculation, dead cells wereremoved
cells/well, and incubated
914 • JID 2006:194 (1 October) • BRIEF REPORT
mac239 cell-associated virus inocula and the presence of female reproductive tract inflammation in recipient
macaques at the time of viral challenges.
Parameters regarding the infectious cell content of the individual simian immunodeficiency virus
per inoculum, no.
IL-6 level in cervicovaginal
Viremia Day 1Day 2Day 3Day 0 Day 1Day 2 Day 3
aInflammation was detected by measuring the IL-6 level in cervicovaginal secretions. An IL-6 level of 1150 pg/mL in the cervi-
covaginal secretion in the absence of menses denoted the presence of female reproductive tract inflammation (baseline data not
bMacaque Cy0084 was inoculated only 2 times, because visible signs of menstruation were recorded on day 3.
Benzalk., benzalkonium chloride; IL, interleukin; ND, not done; +, present; ?, absent.
by ficoll-Histopaque (Sigma) density-gradient centrifugation.
To remove cell-free virus, the cells were washed 2 times. A
fraction of the inoculum was processed for intracellular
SIVmac239 Gag p27 analysis, as described elsewhere .
Serial 2-fold dilutions (in duplicate) of the SIV-infected
PBMCs were cocultured with CEM?174 host cells in10%FBS-
RPMI 1640 medium. The cocultures were maintained for 4
weeks and were monitored weekly for the production of p27
Gag antigen by use of the SIV p27 Core Antigen ELISA kit
(Beckman Coulter). The end point results were expressed as
the TCID per the number of PBMCs in the inocula.
Genomic DNA was isolated using the DNeasy Tissue kit
(Qiagen) according to the manufacturer’s recommendations.
Proviral DNA was detected by real-time polymerase chain re-
action amplification of the viral gag gene, by use of the primers
GTCTCTGCACTATGTGTTTTG-3?and probe 5?-6FAM-CTT-
LightCycler TaqMan Master kit (Roche Diagnostics) was used
for amplification on the LightCycler 2.0, with primers at a final
concentration of 400 nM and with probes at a concentration
of 100 nM. Cycling conditions were as follows: 95?C for 10 s;
followed by 45 cycles at 95?C for 10 s, 60?C for 30 s, and 72?C
for 1 s; and a final cycle at 40?C for 30 s. Viral DNA was
quantified by comparison to a standard curve of log dilutions
of DNA standards ranging from 15 to
were run in triplicate.
To determine the mean number of cells that contained pro-
viral DNA, we used a parallel amplificationof?-globin.Because
genomic DNA from a cell contains 2 copies of ?-globin, we
could determine the number of cellular equivalents in each
reaction by amplifying this gene. Reaction conditions were
identical to those detailedabove. Weusedprimers5?-TGGCAA-
3?and probe 5?-6FAM-TGCAGGCTGCCTGGCAGAAGC-
Plasma and cellular elements of the EDTA anticoagulated
blood were separated by centrifugation performed for 10 min
at 3000 rpm. The plasma was stored at ?80?C, until it un-
derwent further processing for viral load quantification. The
plasma virus concentration was determined as described else-
Challenges of animals with cell-free virus are usu-
ally performed using frozen aliquots of defined virus stock.
However, cryopreservation of SIVmac239-infected PBMCs de-
creases the number of CD4+T cells by ∼30% (our unpublished
observation). Freezing may also exert unknown effects on the
functional integrity of adhesion receptors and cell migratory
functions. Therefore, we performed viral challenges using non-
cryopreserved, SIV-infected PBMCs 7–9 days after initiation of
the tissue culture. Because the animals were inoculatedonmul-
tiple occasions, we characterized each viral inoculum.
The proviral DNA content of such cultures was between
andcopies/million cells (figure 1A). The inocula mainly
consisted of T cells (190% CD3+T cells), and 35%–45% of
the cells expressed CD4 coreceptor (figure 1B and 1C). A total
of 4%–10% of the cells in the lymphocyte gate were positive
for the SIVmac239 Gag p27 antigen. The majority of these cells
BRIEF REPORT • JID 2006:194 (1 October) • 915
did not express detectable levels of CD3 and CD4 T cells (figure
1C and 1D), likely because of nef-mediated down-regulation
of these proteins. The number of infectious cells ranged from
as few as 7 cells to as many as 8192 cells, depending on the
actual number of PBMCs in the inoculum (table 1).
Previous attempts to infect macaques with 1 intravaginal
inoculation of 10,000 infectious cells yielded negative results
. Therefore, our viral challenge experiments included 3 an-
imal cohorts (table 1). The first cohort was exposed to 500,000
in vitro–infected PBMCs on 3 consecutive days. The number
of infectious cells in these inocula was 384–8192 cells. The
second cohort was inoculated with 5000 PBMCs on 3 consec-
utive days (7–82 infectious cells/inoculum). The third cohort
was challenged only once with an inoculum of 5000 PBMCs
(4–8 infectious cells/inoculum). Each cohort involved 1 or 2
One of the control animals (macaque Cy0083) had heavy vag-
inal discharge, which is one type of evidence of preexisting
inflammation of the female reproductive tract. Two of theother
control animals (macaques Cy0094 and Cy0084) developedmi-
nor inflammation by the second or third day of receiving in-
oculations; this inflammation was only detected on the basis
of slightly increased IL-6 levels in cervicovaginal secretions (ta-
ble 1). Because macaque Cy0084 started menstruating on the
third day of receiving inoculations, her elevated IL-6 level in
cervicovaginal secretions can be attributed to this physiological
All the macaques with ulcers became persistently infected
after receiving repeated inoculations. The infection was char-
acterized with plasma viral load kinetics typical of mucosal
challenge with SIVmac239: peak plasma viral loads of
copies/mL 2 weeks after viral challenge and set pointplasma
viral loads of– copies/mL.
Surprisingly, 2 of the control animals that were inoculated
2 times (macaque Cy0084) or 3 times (macaque Cy0094) also
became infected with similar viral load kinetics. None of the
animals that were exposed to inoculum only once showed signs
of persistent viremia.
The purpose of the present studies was to in-
troduce a nonhuman primate model of intravaginal transmis-
sion of HIV with cell-associated virus. Previous reports have
implicated cell-associated virus in the sexual transmission of
HIV [6–8], but experiments involving nonhuman primate
model systems have not provided definitive results [5, 9]. An
intact genital mucosa may be a formidable obstacle to HIV,
because several in vitro experiments have demonstrated the
importance of the integrity of the mucosal epithelium. Indeed,
previous nonhuman primate models have focused on main-
taining the integrity of the mucosal epithelium during viral
inoculation, which may explain their difficulty in achieving
systemic infection with both cell-free and, in particular, cell-
associated virus inocula. We designed a model that involved
the presence of ulcers of the female reproductive tract. Un-
expectedly, some of the animals that had no clinical evidence
of inflammation became infected. These results suggest that
although major inflammation may increase the probability of
infection, it is not an absolute requisite for this route of trans-
mission. The presence of minor epithelial trauma in thecontrol
animals cannot be entirely excluded. However, the precedent
merits consideration, because it may resemble certain con-
ditions in humans when the probability of HIV transmission
might be elevated. For example, in thefemalereproductivetract
in healthy women, minor trauma of the mucosa that was in-
stigated or exacerbated by a heterosexual encounter is well doc-
umented in several publications [12, 13].
Our data, similar to findings from a recent rhesus macaque
model using intravaginal challenges with cell-free SIVmac251,
show that repeated exposures to low-dose inocula result in
systemic viremia within a short time . In both of these low-
dose models, a high transmission rate was achieved in the ab-
sence of hormonal thinning of the vaginal epithelium, dem-
onstrating that the thick epithelial barriercanbeovercomewith
frequent exposures to virus .
Semen from HIV-infected men with urethritis can contain
10–104copies of proviral DNA/mL of ejaculate [3, 4]. One
infected cell can harbor a mean number of 3–4 copies/cell ,
suggesting that 1 mL of ejaculate could comprise 3–3000 HIV-
infected cells. Our inocula of 5000 PBMCs contained up to
copies of proviral DNA. Because ∼4%–10% of the
lymphocytes were found to be positive for the SIVmac239 Gag
p27 antigen, at least 200–500 cells harbored virus in theseprep-
arations. The cell-associated viral burden of our inocula was
therefore similar to that of the semen of HIV-infected men.
There are several mechanisms that could promote mucosal
cell-associated HIV/SIVtransmission. Infectedcellscanmigrate
through epithelial abrasion, or they may adhere to the mucosal
epithelium, serving as a continued source of budding virions.
Uninfected, virus-trapping cells can also prolong the presence
of infectious virus in the vaginal cavity by adhering to the
mucosa. Undoubtedly, major inflammation of the reproductive
tract creates an environment that is immunologically different
than that associated with minor mucosal abrasions.
In the present study, we have provided evidence for in vivo
vaginal SIV transmission in a nonhuman primate species with
a clinically relevant amount of cell-associated virus. We have
also found that reproductive tract inflammation manifested by
overt clinical symptoms is not essential for this mode of trans-
mission. Because cell-associated HIV is particularly difficult to
neutralize, our viral challenge model will be critical for the
916 • JID 2006:194 (1 October) • BRIEF REPORT Download full-text
testing of preventive measures against heterosexual transmis-
sion of HIV infection.
We thank the animal care and veterinary staff of the Wisconsin National
Primate Research Center at the University of Wisconsin–Madison. We are
indebted to Dr. David Watkins for guidance and helpful discussions.
1. Quinn TC, Overbaugh J. HIV/AIDSinwomen:anexpandingepidemic.
2. Galvin SR, Cohen MS. The role of sexually transmitted diseases in
HIV transmission. Nat Rev Microbiol 2004;2:33–42.
3. Quayle AJ, Xu C, Mayer KH, Anderson DJ. T lymphocytes and mac-
rophages, but not motile spermatozoa, are a significant source of hu-
man immunodeficiency virus in semen. J Infect Dis 1997;176:960–8.
4. Coombs RW, Reichelderfer PS, Landay AL. Recent observations on
HIV type-1 infection in the genital tract of men and women. AIDS
5. Sodora DL, Gettie A, Miller CJ, Marx PA. Vaginal transmission of SIV:
assessing infectivity and hormonal influences in macaques inoculated
with cell-free and cell-associated viral stocks. AIDS Res Hum Retro-
6. Shattock RJ, Griffin GE, Gorodeski GI. In vitro models of mucosal
HIV transmission. Nat Med 2000;6:607–8.
7. Greenhead P, Hayes P, Watts PS, Laing KG, Griffin GE, Shattock RJ.
Parameters of human immunodeficiency virus infection of humancer-
vical tissue and inhibition by vaginal virucides. J Virol 2000;74:5577–86.
8. Di Fabio S, Giannini G, Lapenta C, et al. Vaginal transmission of HIV-
1 in hu-SCID mice: a new model for the evaluation of vaginal micro-
bicides. AIDS 2001;15:2231–8.
9. Girard M, Mahoney J, Wei Q, et al. Genital infection of female chim-
panzees with human immunodeficiency virus type 1. AIDS Res Hum
10. National Research Council. Guide for the care and use of laboratory
animals. Washington, DC: National Academy Press, 1996:1–117.
11. Loffredo JT, Rakasz EG, Giraldo JP, et al. Tat28–35SL8-specific CD8+
T lymphocytes are more effective than Gag181–189CM9-specific CD8+
T lymphocytes at suppressing simian immunodeficiency virus repli-
cation in a functional in vitro assay. J Virol 2005;79:14986–91.
12. Fraser IS, Lahteenmaki P, Elomaa K, et al. Variations in vaginal epi-
thelial surface appearance determined by colposcopic inspection in
healthy, sexually active women. Hum Reprod 1999;14:1974–8.
13. McClelland RS, Lavreys L, Hassan WM, Mandaliya K, Ndinya-Achola
J, Baeten JM. Vaginal washing and increased risk of HIV-1 acquisition
among African women: a 10-year prospective study. AIDS 2006;20:
14. Ma ZM, Abel K, Rourke T, et al. A period of transient viremia and
occult infection precedes persistent viremia and antiviral immune re-
sponses during multiple low-dose intravaginal simian immunodefi-
ciency virus inoculations. J Virol 2004;78:14048–52.
15. Marx PA, Spira AI, Gettie A, et al. Progesterone implants enhance SIV
vaginal transmission and early virus load. Nat Med 1996;2:1084–9.
16. Jung A, Maier R, Vartanian JP, et al. Multiply infected spleen cells in
HIV patients. Nature 2002;418:144–5.