HIV-specific CD8+ lymphocytes in semen are not associated with reduced HIV shedding.
ABSTRACT Sexual contact with HIV-infected semen is a major driving force behind the global HIV pandemic. Little is known regarding the immune correlates of virus shedding in this compartment, although HIV-1-specific CD8+ T cells are present in semen. We collected blood and semen from 27 chronically HIV-infected, therapy-naive men without common sexually transmitted infections or urethral inflammation and measured HIV-1 RNA viral load and cytokine/chemokine levels in both compartments. HIV-1 RNA levels were 10-fold higher in blood than semen, but discordantly high semen shedding was associated with higher semen levels of the proinflammatory cytokines IL-6, IL-8, IL-12, and IFN-gamma. Virus-specific CD8+ T cell epitopes were mapped in blood by IFN-gamma ELISPOT, using an overlapping HIV-1 clade B peptide matrix, and blood and semen CD8+ T cell responses were then assayed ex vivo using intracellular IFN-gamma staining. HIV-specific CD8+ responses were detected in 70% of semen samples, and their frequency was similar to or higher than blood. There was no correlation between the presence of virus-specific CD8+ T cells in semen and levels of HIV-1 RNA shedding. Among participants with detectable CD8+ IFN-gamma semen responses, their relative frequency was not associated with reduced HIV-1 RNA shedding, and their absolute number was correlated with higher levels of HIV-1 RNA semen shedding (r = 0.6; p = 0.03) and of several proinflammatory cytokines. Neither the presence nor the frequency of semen HIV-specific CD8+ T cell IFN-gamma responses in semen correlated with reduced levels of HIV RNA in semen.
Article: Sexual transmission of HIV.New England Journal of Medicine 05/1997; 336(15):1072-8. · 51.66 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: We examined the influence of viral load in relation to other risk factors for the heterosexual transmission of human immunodeficiency virus type 1 (HIV-1). In a community-based study of 15,127 persons in a rural district of Uganda, we identified 415 couples in which one partner was HIV-1-positive and one was initially HIV-1-negative and followed them prospectively for up to 30 months. The incidence of HIV-1 infection per 100 person-years among the initially seronegative partners was examined in relation to behavioral and biologic variables. The male partner was HIV-1-positive in 228 couples, and the female partner was HIV-1-positive in 187 couples. Ninety of the 415 initially HIV-1-negative partners seroconverted (incidence, 11.8 per 100 person-years). The rate of male-to-female transmission was not significantly different from the rate of female-to-male transmission (12.0 per 100 person-years vs. 11.6 per 100 person-years). The incidence of seroconversion was highest among the partners who were 15 to 19 years of age (15.3 per 100 person-years). The incidence was 16.7 per 100 person-years among 137 uncircumcised male partners, whereas there were no seroconversions among the 50 circumcised male partners (P<0.001). The mean serum HIV-1 RNA level was significantly higher among HIV-1-positive subjects whose partners seroconverted than among those whose partners did not seroconvert (90,254 copies per milliliter vs. 38,029 copies per milliliter, P=0.01). There were no instances of transmission among the 51 subjects with serum HIV-1 RNA levels of less than 1500 copies per milliliter; there was a significant dose-response relation of increased transmission with increasing viral load. In multivariate analyses of log-transformed HIV-1 RNA levels, each log increment in the viral load was associated with a rate ratio of 2.45 for seroconversion (95 percent confidence interval, 1.85 to 3.26). The viral load is the chief predictor of the risk of heterosexual transmission of HIV-1, and transmission is rare among persons with levels of less than 1500 copies of HIV-1 RNA per milliliter.New England Journal of Medicine 04/2000; 342(13):921-9. · 51.66 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Both qualitative and quantitative virologic measurements were compared between blood and genital compartments for 128 men infected with human immunodeficiency virus type 1 (HIV-1) to address several controversial issues concerning HIV-1 shedding in semen and to obtain further information about the distribution of virus between these two compartments. Evidence for viral compartmentalization was suggested by earlier studies that noted the poor correlation between blood and seminal virus load, phenotype, and genotype. Further support for this viral compartmentalization was based on the following observations between semen and blood: lack of association between culturability of virus in semen and viral RNA level in blood, discordant distribution of viral phenotypes, discordant viral RNA levels, a weak correlation between viral RNA level in semen and CD4 cell count in blood, differences in the biologic variability of viral RNA levels, and differences in the virus load response to antiretroviral therapy.The Journal of Infectious Diseases 03/1998; 177(2):320-30. · 5.85 Impact Factor
HIV-Specific CD8?Lymphocytes in Semen Are Not Associated
with Reduced HIV Shedding1,2
Prameet M. Sheth,* Ali Danesh,†Kamnoosh Shahabi,* Anuradha Rebbapragada,*
Colin Kovacs,‡Rowena Dimayuga,‡Roberta Halpenny,‡Kelly S. MacDonald,*†§
Tony Mazzulli,*§David Kelvin,*†?Mario Ostrowski,*†¶and Rupert Kaul3*¶
Sexual contact with HIV-infected semen is a major driving force behind the global HIV pandemic. Little is known regarding the
immune correlates of virus shedding in this compartment, although HIV-1-specific CD8?T cells are present in semen. We
collected blood and semen from 27 chronically HIV-infected, therapy-naive men without common sexually transmitted infections
or urethral inflammation and measured HIV-1 RNA viral load and cytokine/chemokine levels in both compartments. HIV-1 RNA
levels were 10-fold higher in blood than semen, but discordantly high semen shedding was associated with higher semen levels of
the proinflammatory cytokines IL-6, IL-8, IL-12, and IFN-?. Virus-specific CD8?T cell epitopes were mapped in blood by IFN-?
ELISPOT, using an overlapping HIV-1 clade B peptide matrix, and blood and semen CD8?T cell responses were then assayed
ex vivo using intracellular IFN-? staining. HIV-specific CD8?responses were detected in 70% of semen samples, and their
frequency was similar to or higher than blood. There was no correlation between the presence of virus-specific CD8?T cells in
semen and levels of HIV-1 RNA shedding. Among participants with detectable CD8?IFN-? semen responses, their relative
frequency was not associated with reduced HIV-1 RNA shedding, and their absolute number was correlated with higher levels of
HIV-1 RNA semen shedding (r ? 0.6; p ? 0.03) and of several proinflammatory cytokines. Neither the presence nor the frequency
of semen HIV-specific CD8?T cell IFN-? responses in semen correlated with reduced levels of HIV RNA in semen. The Journal
of Immunology, 2005, 175: 4789–4796.
scale of the global pandemic and the fact that most transmission is
sexual, HIV-1 mucosal transmission is surprisingly inefficient,
with a transmission risk of 0.1–1% per sexual contact (1). When
sexual transmission of HIV-1 does occur, it takes place in two
broad steps. First, the virus must be shed in the genital fluids (se-
men or cervicovaginal secretions) of the infected partner. Second,
the virus must cross the mucosal epithelium of the uninfected part-
ner and establish persistent infection. Previous investigations have
found a strong association between levels of virus in blood and the
probability of transmitting HIV-1 to a monogamous sex partner
(2), and this likely relates to the fact that the amount of virus in the
he HIV-1 pandemic has claimed over 20 million lives,
and 42 million people are currently infected (?www.
who.int/hiv/pub/epidemiology/epi2003/en/?). Given the
genital tract tends to correlate with levels in the blood plasma
(3–6). Despite the correlation between levels of virus in the blood
and genital tract, some individuals may shed disproportionately
high or low levels of HIV-1 in the semen in comparison to blood,
a phenomenon that may have profound public health implications
(4, 7). The immune control of HIV-1 shedding in semen, particu-
larly in such “discordant” shedders, is poorly understood.
Systemic HIV-specific CD8?T cells responses are clearly im-
portant in host HIV-1 immune control, as evidenced by their tem-
poral association with viral control in humans and animal models,
and by the strong immune selection pressure that they place on the
virus (8–10). However, the association between the frequency of
systemic HIV-specific CD8?T cells and levels of plasma virus is
less clear, with different groups reporting an inverse correlation
(11, 12), a positive correlation (13), or no correlation at all (7, 14).
HIV-specific CD8?responses are present in the genital tract of
both HIV-infected and exposed uninfected subjects (15–20), but
their role in controlling levels of viral shedding at mucosal sur-
faces, or in mediating protection against infection, has not been
elucidated. Nonetheless, in animal models it is mucosal, not sys-
temic, CD8?lymphocytes that are most critical in mediating pro-
tection against mucosal viral challenge (21, 22). Furthermore, after
vaginal SIV infection the earliest virus-specific CD8?T cells de-
tected are in the vaginal mucosa (23), and HIV-1 is known to
replicate preferentially within mucosal lymphoid tissue, particu-
larly that of the gut (24, 25). These observations suggest that mu-
cosal CD8?T cell responses may be important in controlling mu-
cosally acquired HIV-1 infection and therefore perhaps in
controlling levels of viral shedding.
Although CD8?T cell responses place a strong immune pres-
sure on HIV-1, excessive systemic immune activation is associated
with higher levels of plasma viremia and with more rapid disease
progression. Negative prognostic markers of immune activation
*Department of Medicine and†Department of Immunology, University of Toronto,
Toronto, Ontario, Canada;‡Canadian Immunodeficiency Research Collaborative, To-
ronto, Ontario, Canada;§Department of Medical Microbiology, Mount Sinai Hospital,
Toronto, Ontario, Canada;¶Canadian Network for Vaccines and Immunotherapeu-
tics; and?Division of Experimental Therapeutics, Toronto General Research Institute,
University Health Network, Toronto, Ontario, Canada
Received for publication January 25, 2005. Accepted for publication July 14, 2005.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported in part by Ontario HIV Treatment Network (OHTN) Grant
454396 and the Canadian Network for Vaccines and Immunotherapeutics. R.K. is
supported by the Canadian Research Chair Program and a Canadian Infectious Dis-
ease Society Bayer Young Investigator Award; K.S.M. is supported by a Career
Scientist Award from the OHTN; and P.M.S. and A.R. are recipients of the OHTN
studentship and postdoctoral awards, respectively.
2These data were presented in part at the 12th Conference on Retroviruses and Op-
portunistic Infections, Abstract 460, Boston, MA, February 22–25, 2005.
3Address correspondence and reprint requests to Dr. Rupert Kaul, Clinical Sci-
ence Division, University of Toronto, 1 Kings College Circle, Medical Science
Building, Room 6356, Toronto, Ontario, M5S 1A8 Canada. E-mail address:
The Journal of Immunology
Copyright © 2005 by The American Association of Immunologists, Inc.0022-1767/05/$02.00
include elevated expression of CD38 and increased levels of in-
flammatory cytokines such as IL-6 and TNF-? (26–28). In addi-
tion, high levels of HIV-1 RNA in the gut mucosa of HIV-1-
infected men taking antiretroviral therapy are associated with
gastrointestinal mucosal cytokine activation (29). However, the
impact of the systemic or genital tract cytokine milieu on levels of
HIV-1 in semen is not known.
To study the role of host immune factors in HIV-1 semen shed-
ding, we have assayed ex vivo CD8?T cell responses in semen for
the first time and have examined the association between CD8?
responses, the local and systemic cytokine milieu, and levels of
HIV-1 RNA in the semen of HIV-1-infected, therapy-naive men.
Materials and Methods
Antiretroviral therapy-naive gay men with chronic HIV-1 infection were
recruited through the Canadian Immunodeficiency Research Collaborative.
Clinicians were asked to enroll participants not expected to require anti-
retroviral therapy within the next 2 years. All subjects provided informed,
written consent. The study protocol was approved by Research Ethics
Boards at the Mount Sinai Hospital, the University Health Network, To-
ronto, and at the University of Toronto.
At the recruitment visit (visit 1), 24 ml of venous blood were collected into
acid citrate dextran for CD8?epitope mapping. At two subsequent visits,
paired blood and semen specimens were collected within an hour of each
other. Semen samples were collected by masturbation into a dry sterile
container at visit 2 and into 10 ml of sterile RPMI 1640 containing anti-
biotic and antimycotic at visit 3. Samples were processed within 2 h of
collection. A first-void urine sample was screened for the presence of leu-
kocytes using a standard urine dipstick (Bayer Diagnostics) and for infec-
tion by either Neisseria gonorrhoeae or Chlamydia trachomatis using the
Amplicor CT/NG assay (Roche Diagnostic Systems). Any participant with
urethral leukocytes, gonorrhea, or chlamydia was excluded from analysis.
Seminal plasma was isolated by centrifugation at 850 ? g for 10 min, and
the semen cell pellet was then resuspended in 10 ml of sterile R10 medium.
PBMCs and seminal fluid mononuclear cells were then isolated by Ficoll-
Hypaque density centrifugation. Blood and seminal plasma were immedi-
ately frozen at ?86°C.
HIV-1 shedding in semen
Blood and semen RNA viral load (VL)4was measured using the Versant
HIV-1 RNA 3.0 assay (bDNA; Bayer Diagnostics) because RT-PCR-based
assays have been found to be inhibited by semen constituents in other
studies (30). VLs were either measured directly in blood and seminal
plasma or in the supernatant of semen samples collected into RPMI 1640.
In the latter case, semen VL was appropriately corrected for the dilution
Epitope mapping using the IFN-? ELISPOT
HIV-1-specific CD8?T cell responses were mapped in blood using an
IFN-? ELISPOT assay, as described previously (31, 32). PBMCs were
incubated at 1 ? 105/well with a matrix of 756 15-mer peptides, overlap-
ping by 11 amino acids, spanning the entire clade B HIV-1 genome (ob-
tained through the AIDS Research and Reference Reagent Program, Divi-
sion of AIDS, National Institute of Allergy and Infectious Diseases,
National Institutes of Health). Each peptide appeared uniquely in two sep-
arate matrix pools at a final working concentration of 2 ?M. All responses
detected using the matrix pools were confirmed using individual 15-mer
peptides. Response frequencies were calculated using an automated ELIS-
POT counter (Cellular Technology), and a positive response was defined as
an HIV peptide-specific response 1) at least 2-fold higher than background
(PBMC ? 2 ?m of DMSO) and 2) ?100 spot-forming units (SFU)/million
cells. All responses were confirmed to be CD8?mediated using IFN-?
intracellular cytokine staining.
Ex vivo stimulation and intracellular IFN-? staining
A pool of all responding epitopes mapped in blood was used to examine
HIV-specific IFN-? responses in blood and semen (19). Briefly, 1 ? 106
blood mononuclear cells were incubated for 1 h at 37°C in 5% CO2with
medium alone, staphylococcus enterotoxin B (3 ?g/ml), or the HIV epitope
pool (each peptide at 10 ?g/ml); semen mononuclear cells were split in two
vials and incubated with medium alone or the HIV epitope pool. Brefeldin
A (BD Biosciences Immunocytometry Systems) was then added (1 ?g/ml),
and samples were incubated for 5 h, permeabilized, and stained with com-
binations of CD3-FITC, CD3-PerCP, CD8-PE, CD8-PerCP, CD4-PE,
CD69-FITC, CD103-PerCP, CD3-APC, and IFN-?-APC (BD Biosciences
Immunocytometry Systems). Samples were acquired using a FACSCalibur
flow cytometer (BD Systems), and data analysis was performed using
FlowJo analytical software (Tree Star). A positive response was defined as
an HIV-specific response 1) at least 2-fold higher than the unstimulated
control and 2) a CD8?HIV-1-specific response frequency of ?0.05%.
Cytokine bead array (CBA)
Cytokine levels in blood and semen plasma were measured using CBA
(BD Biosciences Immunocytometry Systems), according to the manufac-
turer’s instructions (33). Cytokines/chemokines assayed were IL-2, IL-4,
IL-5, IL-10, TNF-?, IFN-?, IL-1?, IL-8, IL-6, IL-12p70, RANTES, mono-
kine induced by IFN-? (MIG/CXCL9), macrophage chemotactic protein
(MCP1/CCL2), and IFN-inducible protein-10 (IP-10/CXCL10). Blood and
semen plasma samples were incubated for 3 h at room temperature with a
mixture of Ab-coupled beads (50 ?l/sample) and PE-conjugated secondary
Abs (50 ?l/sample) against each cytokine. Samples were then washed with
2% paraformaldehyde, resuspended in 150 ?l of PBS wash buffer, and
analyzed using flow cytometry.
Measurement of innate immune factors
Levels of secretory leukocyte protease inhibitor (SLPI) and lactoferrin
were measured in seminal plasma by ELISA (Quantikine Human SLPI kit,
R&D Systems; Enzyme Immunoassay for Human Lactoferrin, Oxford Bio-
medical Research), according to the manufacturers’ instructions, after di-
lution at 1/64,000 and 1/3,000 in diluent buffer. ELISA plates were read in
a standard 96-well Thermomax reader (Molecular Devices) at 450 nm us-
ing 570 nm as a reference.
SPSS 11 for the Macintosh OS X (SPSS) was used for statistical analysis.
Comparisons between the blood and semen of study subjects were per-
formed using the paired samples t test. The Mantel-Haenszel ?2test with
calculation of likelihood ratios and confidence intervals was used to com-
pare dichotomous variables between study groups, and comparison of
means between groups was performed by one-way ANOVA. Linear asso-
ciation of continuous variables was assessed using the Pearson correlation
coefficient. Independent associations of HIV-1 RNA shedding were exam-
ined in a multivariable linear regression model.
Twenty-seven chronically HIV-1-infected, therapy-naive gay men
consented to take part in the study. Their median CD4?T cell
count was 550/mm3(range, 120–1260/mm3). All participants had
been HIV-1 infected for at least 6 mo. No participant had clinical
urethritis or genital ulcer disease or laboratory evidence of infec-
tion by Treponema pallidum, C. trachomatis, or N. gonorrhoeae.
A first-void urine dipstick for leukocytes (a screen for urethral
inflammation) was negative for all participants.
HIV-1 RNA VL in blood and semen
The median blood plasma VL was 20,302 copies/ml plasma
(range, ?50–401,448 copies). HIV RNA was detected in the se-
men of 21 of 27 participants (78%), and there was no evidence of
PCR inhibition in any assay. Levels of HIV-1 RNA in semen were
highly variable, with a median value of 758 copies/ml semen
plasma (range, ?50–210,350 copies). Only one participant had an
undetectable VL in blood, and this individual also had no virus
detected in semen. Six participants (22%) had an undetectable se-
men VL, and their blood VL was quite variable, with a median of
4Abbreviations used in this paper: VL, viral load; SFU, spot-forming unit; CBA,
cytokine bead array; SLPI, secretory leukocyte protease inhibitor.
4790HIV-SPECIFIC CD8?LYMPHOCYTES IN SEMEN AND HIV SHEDDING
9,893 copies/ml plasma (range, ?50–76,014 copies). As reported
in a preliminary analysis (7), semen VL was positively correlated
with HIV-1 levels in blood plasma (Pearson correlation coefficient,
r ? 0.4; p ? 0.02), although semen levels of virus were ?10-fold
lower than in blood (3.1 vs 4.3 log10copies/ml; p ? 0.001, paired
samples t test). There was an inverse correlation between blood
CD4?T cell counts and the RNA VL in blood (r ? ?0.5; p ?
0.02) but not semen (r ? ?0.3; p ? 0.2).
Although HIV-1 RNA levels in blood and semen were corre-
lated overall, some participants clearly demonstrated dispropor-
tionate shedding of HIV-1 RNA in semen. To study the role of
semen immune factors on discordant HIV-1 shedding, participants
were therefore divided into three groups (7): group 1 had an un-
detectable semen VL (nonshedders; n ? 6); group 2 had a semen
VL ? 60% that of blood plasma (concordant shedders; n ? 17);
and group 3 had a semen viral load ? 60% that in blood plasma
(discordant shedders; n ? 4).
Systemic HIV-specific CD8?IFN-? responses and semen HIV
Peripheral blood HIV-specific CD8?responses were screened us-
ing the IFN-? ELISPOT, and HIV-1-specific responses were de-
tected in 27 of 27 participants. Most epitopes fell within HIV-1
Gag (33%), Pol (26%), Env (17%), or Nef (6.7%). Individuals
responded to a mean of 8.6 epitopes (range, 3–20 epitopes), and
the mean total magnitude (sum) of HIV-1-specific CD8?re-
sponses was 6,833 SFU/million cells (range, 770–25,955 SFU).
All IFN-? responses were confirmed to be CD8?T cell mediated
using intracellular cytokine staining and flow cytometric analysis.
As noted in a preliminary report (7), there was no association
between the frequency of systemic (blood) HIV-1-specific CD8?
IFN-? responses and the RNA VL in either blood (r ? 0.3; p ?
0.1) or semen (r ? 0.2; p ? 0.4). In addition, there was no asso-
ciation between the breadth or specificity of blood CD8?T cells
responses and VL at either site. The weak inverse association pre-
viously described between the magnitude and breadth of HIV-1
Tat-specific responses and semen HIV-1 shedding was no longer
apparent with the increased sample size (p ? 0.8 and p ? 0.9,
Semen proinflammatory cytokines and HIV-1 RNA shedding
Cytokine/chemokine levels were assayed directly in semen and
blood plasma using the CBA. Semen levels of several proinflam-
matory cytokines were associated with discordant shedding of
HIV-1. In particular, disproportionate semen HIV-1 RNA shed-
ding was associated with increased levels of IL-6 (p ? 0.008; Fig.
1a), IFN-? (p ? 0.04), IL-12 (p ? 0.008), and IL-8 (p ? 0.02;
Fig. 1b). No association was seen between HIV-1 RNA shedding
in semen and the other cytokines or chemokines assayed in semen,
and there was no correlation between levels of cytokines/chemo-
kines blood plasma and the HIV-1 RNA VL in blood or semen.
Ex vivo measurement of HIV-specific CD8?T lymphocytes
Based on the association of local inflammatory cytokines, includ-
ing IFN-?, with HIV RNA shedding, we went on to examine the
association of HIV-specific, IFN-?-producing CD8?T cells in se-
men with HIV shedding. At a follow-up clinic visit, semen sam-
ples were collected from 20 participants by masturbation into 10
ml of sterile RPMI 1640 to preserve lymphocyte viability (with
correction for the resulting dilution factor when assaying semen
RNA VL and cytokine levels).
Semen mononuclear cells were split into two equal volumes,
one used as a negative control and the second sample to evaluate
semen HIV-specific responses assayed by IFN-? intracellular
staining after short-term ex vivo stimulation with a pool of HIV-1
epitopes previously mapped in blood. Cells were permeabilized
and stained, and the median number of gated CD3?T lymphocytes
acquired per preparation was 2,862 cells (range, 476–70,030 T
cells). Expression of CD103 is a marker for mucosal lymphocytes
in both the male and female genital tracts and the gastrointestinal
tract (19, 34, 35). CD103?expression was measured for 12 sub-
jects and was more common on semen than blood CD3?T cells
(4.4 vs 0.1%; p ? 0.003, paired samples t test; Fig. 2). CD8?T
cells constituted a higher proportion of T cells in the semen than in
blood (69.9 vs 52.9%; p ? 0.001, paired samples t test), as has
been described in the semen of HIV-uninfected men (36, 37). Sem-
inal CD8?T cells were also more activated than those in blood,
with much higher levels of spontaneous IFN-? release (1.3% of
shedding. Participants (n ? 27) were divided into three shedding groups:
nonshedders (group 1; n ? 6) had an undetectable semen VL; concordant
shedders (group 2; n ? 17) had a semen VL ? 60% that in blood plasma;
and discordant shedders (group 3; n ? 4) had a semen viral load ? 60%
that in blood plasma. There was a stepwise association between discordant
HIV-1 RNA shedding and semen levels of several proinflammatory cyto-
kines, including IL-6 (p ? 0.008; a) and IL-8 (p ? 0.02; b).
Semen proinflammatory cytokines and discordant HIV
4791The Journal of Immunology
semen vs 0.05% of blood CD8?T cells; p ? 0.005, paired samples
t test; Fig. 3).
HIV-1-specific CD8?IFN-? T cell responses were detected in
the peripheral blood of all participants (20 of 20), and a response
to the same pool of HIV-1 epitopes was detected in most semen
samples (14 of 20; 70%; Fig. 3 for representative example and
summary data). After correction for background IFN-? release,
HIV-1-specific CD8?responses were present in semen at a sig-
nificantly higher frequency than blood (2.6 vs 0.7% of CD8?T
cells; p ? 0.04). There were no detectable HIV-1-specific CD8?
responses in the semen of 6 of 20 participants (30%) despite a
strong CD8?response in blood (Fig. 3a, lower panel: representa-
tive example). In four of six cases, this may have been due to a low
numbers of gated T cell events (?1000 gated cells), but in two of
six cases, there was no semen response seen despite relatively high
numbers of gated CD3?T cells (mean, 5.5 ? 103gated T cells).
Lack of response was unlikely to be due to anergy of semen T cells
because staphylococcus enterotoxin B stimulation induced a robust
IFN-? response in the semen of both a representative HIV re-
sponder (4% of CD3/CD8?T cells) and a nonresponder (11.5% of
CD3/CD8?T cells). The frequency of HIV-1-specific IFN-?
CD8?T cells in blood was similar in participants with and without
CD8?responses detected in semen (0.7 vs 1.3% of CD8?T cells,
respectively; p ? 0.3). No HIV-specific responses were seen in the
blood or semen of five HIV-uninfected controls after stimulation
with a pool of the eight HIV epitopes most commonly recognized
by infected participants.
Semen HIV-1-specific CD8?T cell responses and HIV-1 RNA
Semen levels of HIV-1 RNA at the time of the semen CD8?assay,
a mean of 177 days after the baseline assessment (range, 73–313
days), were highly correlated with HIV-1 RNA levels measured at
enrollment (r ? 0.7; p ? 0.001; data not shown), confirming the
robust nature of our semen VL assay. There was no association be-
tween the presence/absence of detectable HIV-1-specific CD8?T cell
IFN-? responses in semen and the level of HIV-1 RNA: the mean
semen VL was 4.0 log10RNA copies/ml in nonresponders (n ? 6)
and 3.7 log10copies/ml in responders (n ? 14; p ? 0.6). Among
participants with a detectable semen CD8?response (n ? 14), no
association was seen between the relative frequency of semen HIV-
specific CD8?T cells and the semen VL when the frequency of
HIV-1-specific CD8?T cells was expressed as a percentage of all
semen CD8?T cells (r ? ?0.2; p ? 0.5; Fig. 4a).
Because the total number of T cells in semen varied widely be-
tween study participants, the absolute number of HIV-1-specific
CD8?T cells in a given semen sample was also calculated for each
participant, and this absolute number was then transformed into a
logarithmic scale. When analyzed in this fashion, there was a positive
correlation between the semen CD8?T cell response and levels of
HIV-1 RNA so that a higher absolute number of HIV-1-specific, IFN-
?-producing CD8?T cells in semen was associated with higher se-
men HIV-1 RNA shedding (r ? 0.6; p ? 0.03; Fig. 4b). When the
total number of CD3?T cells in each semen sample was calculated
and then transformed into a logarithmic scale, there was also a trend
to a positive correlation with the semen HIV-1 RNA load (r ? 0.5;
p ? 0.07), and the absolute number of HIV-1-specific CD8?T cells
was closely related to the total number of CD3?T cells in a given
sample (r ? 0.8; p ? 0.001).
Participants were again divided into three groups based on the
concordance of HIV-1 RNA levels in blood and semen, as de-
scribed earlier: nonshedders (n ? 5), concordant shedders (n ? 8),
and discordant shedders (n ? 7). There was no significant associ-
ation between discordant HIV-1 semen shedding and the fre-
quency of semen CD8?T cell responses (percentage of semen
CD8?T cells 2.5, 2.0, and 1.8, respectively; p ? 0.8) or the ab-
solute number of virus-specific CD8?T cells (log10absolute num-
ber of responding cells 1.6, 1.8, and 2.0, respectively; p ? 0.6).
HIV-1-specific CD8?T cell responses in semen and local
Because our baseline analysis had found that discordant HIV-1
semen shedding was associated with higher local levels of the
proinflammatory cytokines, we examined the relationship among
semen CD8?T cell responses, proinflammatory cytokines, and
semen HIV-1 RNA shedding. Levels of several proinflammatory
cytokines in semen were again positively correlated with higher
HIV-1 shedding, including IFN-? (r ? 0.5; p ? 0.02), TNF-? (r ?
0.4; p ? 0.08), IL-6 (r ? 0.4; p ? 0.05), and IL-8 (r ? 0.5; p ?
0.04; Fig. 5a). In addition, there was a positive correlation between
the log10absolute number of HIV-1-specific semen CD8?T cell
IFN-? responses and local levels of the proinflammatory cytokines
in blood and semen. Expression of the mucosal marker
CD103 (?E?7) was measured on CD3/CD8?T cells
isolated from the blood and semen of 12 participants.
CD103 expression was more common on semen than
blood T cells (4.4 vs 0.1%; p ? 0.003, paired samples t
Expression of CD103 on CD8?T cells
4792HIV-SPECIFIC CD8?LYMPHOCYTES IN SEMEN AND HIV SHEDDING
TNF-? (r ? 0.5; p ? 0.02), IL-6 (r ? 0.6; p ? 0.006), and IL-8
(r ? 0.5; p ? 0.02; Fig. 5b). Therefore, the absolute number of
HIV-1-specific, IFN-?-producing CD8?T cells in semen was as-
sociated with both increased semen HIV-1 RNA shedding and in-
creased levels of proinflammatory cytokines in semen.
Soluble innate factors in semen and HIV-1 RNA levels
Both SLPI and lactoferrin were detected in the semen of all par-
ticipants. The mean level of SLPI was 29.0 ?g/ml (median, 27.3
?g/ml; range, 1.1–72.5 ?g/ml) and of lactoferrin was 168.4 ng/ml
(median, 169.2 ng/ml; range, 5.1–566.4 ng/ml). There was no as-
sociation between SLPI levels and semen HIV-1 RNA detection
(33.1 ?g/ml in HIV-1 shedders vs 22.3 ?g/ml in nonshedders; p ?
0.3), semen HIV-1 RNA levels (r ? 0.24; p ? 0.3), or any of the
proinflammatory cytokines (data not shown). Although lactoferrin
levels did not vary based on the detection of HIV-1 RNA (195.1
ng/ml in HIV-1 shedders vs 115.3 in nonshedders; p ? 0.3), there
was a positive correlation between lactoferrin and levels of HIV-1
RNA (r ? 0.54; p ? 0.01), and lactoferrin also tended to be higher
in discordant shedders (115.3 ng/ml in nonshedders, 145.0 ng/ml
in concordant shedders, 259.6 ng/ml in discordant shedders; p ?
0.07). No association was found between levels of lactoferrin and
Although contact during sex with semen containing HIV-1 is a
major driving force in the global pandemic, little is known about
the immune correlates of HIV-1 shedding in semen. There is no
question that systemic levels of virus are an important predictor of
semen virus load (4, 5, 38–41), but our study confirms the obser-
vation that discordance between levels of virus in the blood and
semen is relatively common (4). Therefore, understanding the im-
mune basis of HIV-1 semen shedding, and particularly of discor-
dant shedding, is important for the development of both immuno-
therapeutics and rational public health policy. Virus-specific
CD8?T cell responses are critical in the control of HIV-1 by an
infected person (9, 10), but there is controversy as to whether there
is a direct correlation between the frequency of systemic virus-
specific CD8?T cells and the blood VL (12, 13, 42, 43), and the
optimal assay technique to measure these responses is also unclear
cific, IFN-?-producing CD8?T lym-
phocytes in semen. Ex vivo HIV-spe-
cific CD8?IFN-? responses were
detected in the blood of 20 of 20 and
the semen of 14 of 20 HIV-infected,
therapy naive men after short-term
stimulation with a pool of HIV epitope
peptides mapped in blood. High IFN-?
background and strong HIV-specific
responses are demonstrated in a repre-
sentative semen responder OM125 (a,
top panels). No semen responses were
demonstrated in 6 of 20 participants
despite strong blood responses, as
demonstrated in OM 129 (a, bottom
panels). b, Spontaneous and HIV-1-
specific IFN-? production in the blood
and semen of all participants.
Ex vivo HIV-1 spe-
4793 The Journal of Immunology
(44). HIV-1-specific CD8?T cells are present in the mucosa of the
female cervix (15, 19), in semen (20), and in the gastrointestinal
mucosa (45, 46). However, previous assay techniques have not
allowed precise quantitation of responses in semen and have not
attempted to correlate mucosal response levels with shedding of
For the first time, we have been able to assay functional CD8?
responses directly ex vivo in the semen of HIV-1-infected men and
to examine their association with semen VL. HIV-1-specific, IFN-
?-producing CD8?T cells were present in the semen of most
participants, and their frequency in semen exceeded that measured
in blood. However, no association was seen between the detection
of semen CD8?IFN-? responses and HIV-1 RNA shedding in
semen. Indeed, among those with a detectable CD8?semen re-
sponse, higher absolute numbers of virus-specific CD8?T cells in
semen correlated with increased levels of HIV-1 RNA. The fact
that CD8?responses were also associated with higher semen
proinflammatory cytokines, including IL-6, IFN-?, and IL-8, sug-
gests a possible mechanism for this observation. Although limited
semen lymphocyte numbers necessitated screening semen re-
sponses using epitopes that had been mapped in blood, this should
not affect our results because the ontogeny and Ag specificity of
mucosal HIV-specific CD8?T cells closely mirror that of blood
(16, 17, 20, 45).
These observations do not prove that HIV-1-specific CD8?T
cells are themselves inducing inflammation and increased viral
shedding in semen. Our study shows a clear association between
semen inflammation, HIV-1 RNA shedding, and CD8?T cell re-
sponses, but we cannot demonstrate causation. Increased levels of
inflammatory cytokines in the blood have been associated with
higher VL in blood plasma and in the gastrointestinal mucosa with
higher HIV-1 levels in the mucosal tissues of the gut (29). There-
fore, it is possible, and perhaps more intuitive, that inflammatory
cytokines in semen might induce viral shedding and that the in-
creased virus-specific CD8?T cell numbers are secondary to in-
flammation and/or the resultant higher levels of virus. The absolute
number of HIV-1-specific CD8?T cells in semen tended to mirror
the total number of CD3?T cells so that any inflammatory process
and levels of HIV-1 RNA shedding. In participants with an HIV-1-specific
CD8?T cell response in semen, there was no correlation between the
frequency of these responses and the semen HIV-1 RNA load when CD8?
responses were expressed as a percentage of all CD8?T cells (a; r ? ?0.2;
p ? 0.5). When CD8?T cell responses were expressed as log10absolute
number of semen HIV-specific CD8?T cells, there was a positive corre-
lation between CD8?responses and the level of semen HIV-1 RNA (b; r ?
0.6; p ? 0.03).
HIV-1-specific, IFN-?-producing CD8?T cells in semen
creased HIV-1 RNA shedding and with absolute numbers of virus-specific
CD8?T cells. Increased semen levels of several inflammatory cytokines
correlated with HIV-specific CD8?T cell responses in this compartment
(IL-8 is shown in a; r ? 0.5; p ? 0.05) and also with log10semen VL (b;
r ? 0.5; p ? 0.04).
Proinflammatory cytokines in semen are associated with in-
4794 HIV-SPECIFIC CD8?LYMPHOCYTES IN SEMEN AND HIV SHEDDING
recruiting CD3?T cells might increase both HIV-1 shedding and
local numbers of CD3?T cells, including HIV-1-specific CD8?T
cells. If both HIV-1 semen shedding and CD8?T cell responses
were secondary to increased levels of proinflammatory cytokines,
then it will be important to identify the cause of local semen in-
flammation. The semen cytokine milieu was not simply a reflection
of systemic cytokine levels because there was no significant cor-
relation of proinflammatory cytokine levels in semen and blood.
Men were screened for gonorrhea and chlamydia, as well as for
urethral inflammation due to other factors, so classical sexually
transmitted infections should not have been responsible.
Several innate mucosal factors, including SLPI and lactoferrin,
have anti-HIV activity and are found at high levels at many mu-
cosal surfaces. Lactoferrin can block HIV-1 replication in T cells,
block dendritic cell uptake of HIV-1 by binding to DC-SIGN, and
may reduce breast milk transmission of HIV-1 (47, 48). SLPI has
direct anti-HIV activity and is induced in the oral mucosa by
HIV-1, and levels of SLPI in the female genital tract and breast
milk have been correlated with reduced mother-child transmission
of HIV-1 (49–51). We did not find that either factor was associated
with reduced HIV-1 semen RNA: indeed, lactoferrin was associ-
ated with increased HIV shedding. This may relate to different
levels or effects of these innate factors at different mucosal sites or
could perhaps reflect induction of lactoferrin as an antiviral de-
fense in response to higher local levels of HIV-1.
Our CBA analyses involved the measurement of levels of 14
different cytokines/chemokines in semen, and so without correc-
tion for multiple comparisons, one might expect at least one sig-
nificant association with HIV shedding, using a two-sided ? error
of 0.05. However, the hypothesis of an association between HIV-1
RNA levels and semen inflammation was generated at the first
study visit and was then confirmed at a second separate visit, mak-
ing it extremely unlikely that this represents a chance finding. In
addition, an association arising through chance would be unlikely
to result in the repeated clustering of associations with only the
proinflammatory cytokines and not with the other cytokines/che-
Although this cross-sectional study cannot prove that HIV-1-
specific CD8?responses in semen increase semen HIV-1 RNA
load, they were clearly not associated with reduced virus shedding
in chronic HIV infection. Whether the induction or boosting of
semen CD8?responses by therapeutic vaccines or other immuno-
therapeutics would affect semen shedding or transmission cannot
be addressed by our study. Although the association of semen
CD8?responses with higher HIV shedding might imply that this
will not be a useful therapeutic strategy, our use of IFN-? produc-
tion as a means of identifying HIV-1-specific CD8?T cells might
bias to finding an association between inflammatory cytokines and
CD8?responses. Future studies might consider alternate means to
measure semen responses, including MHC class I peptide tetram-
ers, Ag-specific proliferation, expression of the lytic marker
CD107, or the production of alternate cytokines/chemokines, par-
ticularly IL-2 or MIP-1?. The ability of blood CD8?T cells to
proliferate in response to cognate epitope, rather than IFN-? pro-
duction, has been linked to enhanced viral control in HIV-1-in-
fected long-term nonprogressors (52). Similar studies may also be
useful in elucidating the role of semen CD8?T cells in control of
HIV-1 shedding and transmission. Nonetheless, our results do im-
ply that when therapeutic HIV vaccines are developed that boost
HIV-1-specific CD8?T cell immunity, the impact of such vac-
cines on semen CD8?T cell responses and levels of HIV shedding
should be carefully monitored.
We offer our thanks to the study participants for their support and partic-
ipation. Overlapping HIV-1 clade B peptide pools were obtained through
the AIDS Research and Reference Reagent Program, Division of AIDS,
National Institute of Allergy and Infectious Diseases, National Institutes of
The authors have no financial conflict of interest.
1. Royce, R. A., A. Sena, W. Cates, Jr., and M. S. Cohen. 1997. Sexual transmission
of HIV. N. Engl. J. Med. 336: 1072–1078.
2. Quinn, T. C., M. J. Wawer, N. Sewankambo, D. Serwadda, C. Li,
F. Wabwire-Mangen, M. O. Meehan, T. Lutalo, and R. H. Gray. 2000. Viral load
and heterosexual transmission of human immunodeficiency virus type 1: Rakai
Project Study Group. N. Engl. J. Med. 342: 921–929.
3. Coombs, R. W., C. E. Speck, J. P. Hughes, W. Lee, R. Sampoleo, S. O. Ross,
J. Dragavon, G. Peterson, T. M. Hooton, A. C. Collier, et al. 1998. Association
between culturable human immunodeficiency virus type 1 (HIV-1) in semen and
HIV-1 RNA levels in semen and blood: evidence for compartmentalization of
HIV-1 between semen and blood. J. Infect. Dis. 177: 320–330.
4. Tachet, A., E. Dulioust, D. Salmon, M. De Almeida, S. Rivalland,
L. Finkielsztejn, I. Heard, P. Jouannet, D. Sicard, and C. Rouzioux. 1999. De-
tection and quantification of HIV-1 in semen: identification of a subpopulation of
men at high potential risk of viral sexual transmission. AIDS 13: 823–831.
5. Bourlet, T., C. Cazorla, P. Berthelot, F. Grattard, F. Cognasse, A. Fresard,
C. Defontaine, F. R. Lucht, C. Genin, and B. Pozzetto. 2001. Compartmental-
ization of HIV-1 according to antiretroviral therapy: viral loads are correlated in
blood and semen but poorly in blood and saliva. AIDS 15: 284–285.
6. Coombs, R. W., P. S. Reichelderfer, and A. L. Landay. 2003. Recent observations
on HIV type-1 infection in the genital tract of men and women. AIDS 17:
7. Sheth, P., K. Shahabi, A. Rebbapragada, C. Kovacs, R. Dimayuga,
S. Chakkalackal, K. MacDonald, T. Mazzulli, and R. Kaul. 2004. HIV viral
shedding in semen: lack of correlation with systemic virus-specific CD8 re-
sponses. AIDS 18: 2202–2205.
8. Horton, H., T. Vogel, D. O’Connor, L. Picker, and D. I. Watkins. 2002. Analysis
of the immune response and viral evolution during the acute phase of SIV in-
fection. Vaccine 20: 1927–1932.
9. Leslie, A. J., K. J. Pfafferott, P. Chetty, R. Draenert, M. M. Addo, M. Feeney,
Y. Tang, E. C. Holmes, T. Allen, J. G. Prado, et al. 2004. HIV evolution: CTL
escape mutation and reversion after transmission. Nat. Med. 10: 282–289.
10. McMichael, A. J., and S. L. Rowland-Jones. 2001. Cellular immune responses to
HIV. Nature 410: 980–987.
11. Musey, L., J. Hughes, T. Schacker, T. Shea, L. Corey, and M. J. McElrath. 1997.
Cytotoxic-T cell responses, viral load, and disease progression in early human
immunodeficiency virus type 1 infection. N. Engl. J. Med. 337: 1267–1274.
12. Ogg, G. S., X. Jin, S. Bonhoeffer, P. R. Dunbar, M. A. Nowak, S. Monard,
J. P. Segal, Y. Cao, S. L. Rowland-Jones, V. Cerundolo, et al. 1998. Quantitation
of HIV-1-specific cytotoxic T lymphocytes and plasma load of viral RNA. Sci-
ence 279: 2103–2106.
13. Betts, M. R., D. R. Ambrozak, D. C. Douek, S. Bonhoeffer, J. M. Brenchley,
J. P. Casazza, R. A. Koup, and L. J. Picker. 2001. Analysis of total human
immunodeficiency virus (HIV)-specific CD4?and CD8?T cell responses: rela-
tionship to viral load in untreated HIV infection. J. Virol. 75: 11983–11991.
14. Addo, M. M., X. G. Yu, A. Rathod, D. Cohen, R. L. Eldridge, D. Strick,
M. N. Johnston, C. Corcoran, A. G. Wurcel, C. A. Fitzpatrick, et al. 2003. Com-
prehensive epitope analysis of human immunodeficiency virus type 1 (HIV-1)-
specific T cell responses directed against the entire expressed HIV-1 genome
demonstrate broadly directed responses, but no correlation to viral load. J. Virol.
15. Shacklett, B. L., S. Cu-Uvin, T. J. Beadle, C. A. Pace, N. M. Fast, S. M. Donahue,
A. M. Caliendo, T. P. Flanigan, C. C. Carpenter, and D. F. Nixon. 2000. Quan-
tification of HIV-1-specific T cell responses at the mucosal cervicovaginal sur-
face. AIDS 14: 1911–1915.
16. Musey, L., Y. Ding, J. Cao, J. Lee, C. Galloway, A. Yuen, K. R. Jerome, and
M. J. McElrath. 2003. Ontogeny and specificities of mucosal and blood human
immunodeficiency virus type 1-specific CD8?cytotoxic T lymphocytes. J. Virol.
17. Musey, L., Y. Hu, L. Eckert, M. Christensen, T. Karchmer, and M. J. McElrath.
1997. HIV-1 induces cytotoxic T lymphocytes in the cervix of infected women.
J. Exp. Med. 185: 293–303.
18. Kaul, R., F. A. Plummer, J. Kimani, T. Dong, P. Kiama, T. Rostron, E. Njagi,
K. S. MacDonald, J. J. Bwayo, A. J. McMichael, and S. L. Rowland-Jones. 2000.
HIV-1 specific mucosal CD8?lymphocyte responses in the cervix of HIV-1
resistant prostitutes in Nairobi. J. Immunol. 164: 1602–1611.
19. Kaul, R., P. Thottingal, J. Kimani, P. Kiama, C. W. Waigwa, J. J. Bwayo,
F. A. Plummer, and S. L. Rowland-Jones. 2003. Quantitative ex vivo analysis of
functional virus-specific CD8 T lymphocytes in the blood and genital tract of
HIV-infected women. AIDS 17: 1139–1144.
20. Quayle, A. J., W. M. Coston, A. K. Trocha, S. A. Kalams, K. H. Mayer, and
D. J. Anderson. 1998. Detection of HIV-1-specific CTLs in the semen of HIV-
infected individuals. J. Immunol. 161: 4406–4410.
4795The Journal of Immunology
21. Murphey-Corb, M., L. A. Wilson, A. M. Trichel, D. E. Roberts, K. Xu,
S. Ohkawa, B. Woodson, R. Bohm, and J. Blanchard. 1999. Selective induction
of protective MHC class I-restricted CTL in the intestinal lamina propria of
rhesus monkeys by transient SIV infection of the colonic mucosa. J. Immunol.
22. Belyakov, I. M., J. D. Ahlers, B. Y. Brandwein, P. Earl, B. L. Kelsall, B. Moss,
W. Strober, and J. A. Berzofsky. 1998. The importance of local mucosal HIV-
specific CD8?cytotoxic T lymphocytes for resistance to mucosal viral transmis-
sion in mice and enhancement of resistance by local administration of IL-12.
J. Clin. Invest. 102: 2072–2081.
23. Ma, Z. M., K. Abel, Y. C. Wang, and C. Miller. 2004. A period of transient
viremia and occult infection precedes persistent viremia and anti-viral immune
responses during multiple low-dose intravaginal SIV infections. In AidsVaccine
04, Lausanne, Switzerland, p. 37.
24. Mehandru, S., M. A. Poles, K. Tenner-Racz, A. Horowitz, A. Hurley, C. Hogan,
D. Boden, P. Racz, and M. Markowitz. 2004. Primary HIV-1 infection is asso-
ciated with preferential depletion of CD4?T lymphocytes from effector sites in
the gastrointestinal tract. J. Exp. Med. 200: 761–770.
25. Brenchley, J. M., T. W. Schacker, L. E. Ruff, D. A. Price, J. H. Taylor,
G. J. Beilman, P. L. Nguyen, A. Khoruts, M. Larson, A. T. Haase, and
D. C. Douek. 2004. CD4?T cell depletion during all stages of HIV disease
occurs predominantly in the gastrointestinal tract. J. Exp. Med. 200: 749–759.
26. Clerici, M., C. Balotta, A. Salvaggio, C. Riva, D. Trabattoni, L. Papagno,
A. Berlusconi, S. Rusconi, M. L. Villa, M. Moroni, and M. Galli. 1996. Human
immunodeficiency virus (HIV) phenotype and interleukin-2/interleukin-10 ratio
are associated markers of protection and progression in HIV infection. Blood 88:
27. Rizzardini, G., D. Trabattoni, M. Saresella, S. Piconi, M. Lukwiya, S. Declich,
M. Fabiani, P. Ferrante, and M. Clerici. 1998. Immune activation in HIV-infected
African individuals: Italian-Ugandan AIDS cooperation program. AIDS 12:
28. Srikanth, P., R. C. Castillo, G. Sridharan, T. J. John, A. Zachariah, D. Mathai, and
D. H. Schwartz. 2000. Increase in plasma IL-10 levels and rapid loss of CD4?T
cells among HIV-infected individuals in south India. Int. J. STD AIDS. 11:
29. McGowan, I., J. Elliott, M. Fuerst, P. Taing, J. Boscardin, M. Poles, and
P. Anton. 2004. Increased HIV-1 mucosal replication is associated with gener-
alized mucosal cytokine activation. J. Acquir. Immune Defic. Syndr. 37:
30. Dunne, A. L., F. Mitchell, K. M. Allen, H. W. Baker, S. Garland, G. N. Clarke,
A. Mijch, and S. M. Crowe. 2003. Analysis of HIV-1 viral load in seminal plasma
samples. J. Clin. Virol. 26: 239–245.
31. Kaul, R., and S. L. Rowland-Jones. 1999. Methods of detection of HIV-specific
CTL and their role in protection against HIV infection. In HIV Immunology
Database. B. T. M. Korber, B. F. Haynes, R. Koup, C. Brander, J. P. Moore,
B. D. Walker, and D. I. Watkins, eds. Theoretical Biology and Biophysics Group
T-10, Los Alamos National Laboratory, Los Alamos, NM, p. IV 27–36.
32. Kaul, R., T. Dong, F. A. Plummer, J. Kimani, T. Rostron, P. Kiama, E. Njagi,
E. Irungu, B. Farah, J. Oyugi, et al. 2001. CD8?lymphocytes respond to different
HIV epitopes in seronegative and infected subjects. J. Clin. Invest. 107:
33. Morgan, E., R. Varro, H. Sepulveda, J. A. Ember, J. Apgar, J. Wilson, L. Lowe,
R. Chen, L. Shivraj, A. Agadir, et al. 2004. Cytometric bead array: a multiplexed
assay platform with applications in various areas of biology. Clin. Immunol. 110:
34. Pudney, J., and D. J. Anderson. 1995. Immunobiology of the human penile ure-
thra. Am. J. Pathol. 147: 155–165.
35. Shacklett, B. L., C. A. Cox, J. K. Sandberg, N. H. Stollman, M. A. Jacobson, and
D. F. Nixon. 2003. Trafficking of human immunodeficiency virus type 1-specific
CD8?T cells to gut-associated lymphoid tissue during chronic infection. J. Virol.
36. Wolff, H., and D. J. Anderson. 1988. Immunohistologic characterization and
quantitation of leukocyte subpopulations in human semen. Fertil. Steril. 49:
37. Wolff, H., and D. J. Anderson. 1988. Male genital tract inflammation associated
with increased numbers of potential human immunodeficiency virus host cells in
semen. Andrologia 20: 404–410.
38. Barroso, P. F., M. Schechter, P. Gupta, M. F. Melo, M. Vieira, F. C. Murta,
Y. Souza, and L. H. Harrison. 2000. Effect of antiretroviral therapy on HIV
shedding in semen. Ann. Intern. Med. 133: 280–284.
39. Fiscus, S. A., P. L. Vernazza, B. Gilliam, J. Dyer, J. J. Eron, and M. S. Cohen.
1998. Factors associated with changes in HIV shedding in semen. AIDS Res.
Hum. Retroviruses 14: S27–S31.
40. Vernazza, P. L., B. L. Gilliam, J. Dyer, S. A. Fiscus, J. J. Eron, A. C. Frank, and
M. S. Cohen. 1997. Quantification of HIV in semen: correlation with antiviral
treatment and immune status. AIDS 11: 987–993.
41. Speck, C. E., R. W. Coombs, L. A. Koutsky, J. Zeh, S. O. Ross, T. M. Hooton,
A. C. Collier, L. Corey, A. Cent, J. Dragavon, et al. 1999. Risk factors for HIV-1
shedding in semen. Am. J. Epidemiol. 150: 622–631.
42. Betts, M. R., J. F. Krowka, T. B. Kepler, M. Davidian, C. Christopherson,
S. Kwok, L. Louie, J. Eron, H. Sheppard, and J. A. Frelinger. 1999. Human
immunodeficiency virus type 1-specific cytotoxic T lymphocyte activity is in-
versely correlated with HIV type 1 viral load in HIV type 1-infected long-term
survivors. AIDS Res. Hum. Retroviruses 15: 1219–1228.
43. Edwards, B. H., A. Bansal, S. Sabbaj, J. Bakari, M. J. Mulligan, and
P. A. Goepfert. 2002. Magnitude of functional CD8?T cell responses to the gag
protein of human immunodeficiency virus type 1 correlates inversely with viral
load in plasma. J. Virol. 76: 2298–2305.
44. Lieberman, J. 2004. Tracking the killers: how should we measure CD8 T cells in
HIV infection? AIDS 18: 1489–1493.
45. Ibarrondo, F. J., P. A. Anton, M. Fuerst, H. L. Ng, J. T. Wong, J. Matud, J. Elliott,
R. Shih, M. A. Hausner, C. Price, et al. 2005. Parallel human immunodeficiency
virus type 1-specific CD8?T-lymphocyte responses in blood and mucosa during
chronic infection. J. Virol. 79: 4289–4297.
46. Shacklett, B. L., T. J. Beadle, P. A. Pacheco, J. H. Grendell, P. A. Haslett,
A. S. King, G. S. Ogg, P. M. Basuk, and D. F. Nixon. 2000. Characterization of
HIV-1-specific cytotoxic T lymphocytes expressing the mucosal lymphocyte in-
tegrin CD103 in rectal and duodenal lymphoid tissue of HIV-1-infected subjects.
Virology 270: 317–327.
47. Groot, F., T. B. Geijtenbeek, R. W. Sanders, C. E. Baldwin, M. Sanchez-Hernandez,
R. Floris, Y. van Kooyk, E. C. de Jong, and B. Berkhout. 2005. Lactoferrin prevents
dendritic cell-mediated human immunodeficiency virus type 1 transmission by
blocking the DC-SIGN–gp120 interaction. J. Virol. 79: 3009–3015.
48. Nduati, R., and G. John. 1995. Breast milk transmission of HIV-1. NARESA
49. Jana, N. K., L. R. Gray, and D. C. Shugars. 2005. Human immunodeficiency
virus type 1 stimulates the expression and production of secretory leukocyte
protease inhibitor (SLPI) in oral epithelial cells: a role for SLPI in innate mucosal
immunity. J. Virol. 79: 6432–6440.
50. Farquhar, C., T. C. VanCott, D. A. Mbori-Ngacha, L. Horani, R. K. Bosire,
J. K. Kreiss, B. A. Richardson, and G. C. John-Stewart. 2002. Salivary secretory
leukocyte protease inhibitor is associated with reduced transmission of human
immunodeficiency virus type 1 through breast milk. J. Infect. Dis. 186:
51. Pillay, K., A. Coutsoudis, A. K. Agadzi-Naqvi, L. Kuhn, H. M. Coovadia, and
E. N. Janoff. 2001. Secretory leukocyte protease inhibitor in vaginal fluids and
perinatal human immunodeficiency virus type 1 transmission. J. Infect. Dis. 183:
52. Migueles, S. A., A. C. Laborico, W. L. Shupert, M. S. Sabbaghian, R. Rabin,
C. W. Hallahan, D. V. Baarle, S. Kostense, F. Miedema, M. McLaughlin, et al.
2002. HIV-specific CD8?T cell proliferation is coupled to perforin expression
and is maintained in nonprogressors. Nat. Immunol. 3: 1061–1068.
4796HIV-SPECIFIC CD8?LYMPHOCYTES IN SEMEN AND HIV SHEDDING