Article

CD8+ Lymphocytes Are Required for Maintaining

Viral Suppression in SIV-Infected Macaques Treated
with Short-Term Antiretroviral Therapy
Highlights Authors
+
d CD8 lymphocyte depletion during ART increases SIV plasma Emily K. Cartwright, Lori Spicer,
viral load (72- to 350-fold) S. Abigail Smith, ..., Jeffrey D. Lifson,
Cynthia A. Derdeyn, Guido Silvestri
d Reconstitution of CD8+ T cells is associated with re-
establishment of viral control Correspondence
d Pre-depletion levels of SIV DNA+ CD4+ T cells correlate with gsilves@emory.edu
viremia after depletion
In Brief
Using the non-human primate model of
simian immunodeficiency virus (SIV)
infection of rhesus macaques, Cartwright
et al. show that CD8+ T cells are required
for maintaining suppression of viremia
during short-term (8–32 week)
antiretroviral therapy (ART). This has
important implications for future
therapeutic interventions for the
treatment of HIV in ART-treated
individuals.

Accession Numbers
KX552042–KX552194

Cartwright et al., 2016, Immunity 45, 656–668

September 20, 2016 ª 2016 Elsevier Inc.
http://dx.doi.org/10.1016/j.immuni.2016.08.018
 Immunity

Article

CD8+ Lymphocytes Are Required for Maintaining

Viral Suppression in SIV-Infected Macaques
Treated with Short-Term Antiretroviral Therapy
Emily K. Cartwright,1 Lori Spicer,1 S. Abigail Smith,1 David Lee,1 Randy Fast,2 Sara Paganini,1 Benton O. Lawson,1
Melon Nega,1 Kirk Easley,3 Joern E. Schmitz,4,6 Steven E. Bosinger,1 Mirko Paiardini,1 Ann Chahroudi,1,5
Thomas H. Vanderford,1 Jacob D. Estes,2 Jeffrey D. Lifson,2 Cynthia A. Derdeyn,1 and Guido Silvestri1,7,*
1Emory Vaccine Center and Yerkes National Primate Research Center, Emory University, Atlanta, GA 30329, USA
2AIDS and Cancer Virus Program, Leidos Biomedical Research Inc., Frederick National Laboratory, Frederick, MD 21702, USA
3Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Atlanta, GA 30329, USA
4Division of Viral Pathogenesis, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, USA
5Department of Pediatrics, Emory University School of Medicine, Atlanta, GA 30322, USA
6Present address: Fraunhofer Institute for Molecular Biology and Applied Ecology IME, Aachen 52074, Germany
7Lead Contact

*Correspondence: gsilves@emory.edu
http://dx.doi.org/10.1016/j.immuni.2016.08.018

SUMMARY SIV infection. First, there is a temporal association between the

Infection with HIV persists despite suppressive anti-
retroviral therapy (ART), and treatment interruption
results in rapid viral rebound. Antibody-mediated
CD8+ lymphocyte depletion in simian immunodefi-
ciency virus (SIV)-infected rhesus macaques (RMs)
shows that these cells contribute to viral control
in untreated animals. However, the contribution of
CD8+ lymphocytes to maintaining viral suppression
under ART remains unknown. Here, we have shown
that in SIV-infected RMs treated with short-term
(i.e., 8–32 week) ART, depletion of CD8+ lymphocytes
resulted in increased plasma viremia in all animals
and that repopulation of CD8+ T cells was associ-
ated with prompt reestablishment of virus control.
Although the number of SIV-DNA-positive cells re-
mained unchanged after CD8 depletion and reconsti-
tution, the frequency of SIV-infected CD4+ T cells
before depletion positively correlated with both the
peak and area under the curve of viremia after deple-
tion. These results suggest a role for CD8+ T cells
in controlling viral production during ART, thus
providing a rationale for exploring immunothera-
peutic approaches in ART-treated HIV-infected
individuals.
INTRODUCTION
CD8+ T cells are important for immune defense against viruses,
intracellular pathogens, and cancers. CD8+ T cell responses to
lentiviruses such as HIV and simian immunodeficiency virus
(SIV) can be detected as early as day 7 after infection (Reimann
et al., 1994; Yasutomi et al., 1993). Several lines of evidence indi-
cate that CD8+ lymphocytes inhibit viral replication during HIV or
expansion of antigen-specific CD8+ lymphocytes and the post-
peak decline of plasma viremia (Borrow et al., 1994; Koup
et al., 1994). Second, there is a clear association between certain
major histocompatibility complex (MHC) class I alleles (i.e., HLA-
B*57, Mamu-B*08, and Mamu-B*17) and disease progression
during both HIV infection of humans and SIV infection of rhesus
macaques (RMs) (Altfeld et al., 2003; Costello et al., 1999; Evans
et al., 1999, 2000; Haynes et al., 1996; Kaslow et al., 1996; Klein
et al., 1998). Third, during both acute and chronic HIV or
SIV infection, immunologic pressure mediated by virus-specific
CD8+ T cells is manifested by the emergence of viral escape mu-
tations (Borrow et al., 1997; Chen et al., 2000; McMichael and
Phillips, 1997). Fourth, HIV-1-infected individuals with the ‘‘elite
controller’’ phenotype exhibit CD8+ lymphocyte responses char-
acterized by polyfunctionality, greater proliferative capacity, and
stronger in vitro killing potential than those observed in normal
progressors (Betts et al., 2006; Migueles et al., 2008; Peris-Per-
tusa et al., 2010). However, this CD8+ T cell response is unable to
clear or even control infection in the overwhelming majority of
HIV-infected individuals.
In the experimental animal model of SIV infection of RMs, the
most direct evidence for the role of CD8+ lymphocytes in viral
control has come from studies in which these cells were tran-
siently depleted in vivo by CD8-specific monoclonal antibodies.
This procedure results in (1) abrogation of the post-peak decline
of viremia when performed during acute SIV infection (Matano
et al., 1998; Schmitz et al., 1999a) and (2) increased viral replica-
tion when performed during chronic SIV infection (Chowdhury
et al., 2015; Jin et al., 1999; Metzner et al., 2000). The fact that
viral loads rapidly return to pre-depletion levels upon reconstitu-
tion of the CD8+ lymphocyte pool further confirms the antiviral
role of these cells. However, the contribution of CD8+ lympho-
cytes to controlling viral replication and/or production during
continuous, highly active ART is unknown. Assessing the poten-
tial antiviral role of CD8+ lymphocytes under ART is important
because it would provide key rationale for exploring immuno-
therapeutic approaches, such as therapeutic vaccines and

656 Immunity 45, 656–668, September 20, 2016 ª 2016 Elsevier Inc.
A

B C

Weeks after depletion

G
D

Weeks after depletion

Weeks after depletion
E H

Weeks after depletion Weeks after depletion

Weeks after depletion

(legend on next page)

Immunity 45, 656–668, September 20, 2016 657
immune checkpoint inhibitors as clinical interventions, to
reduce the viral reservoir in HIV-infected individuals. In this
study, we directly assessed the function of CD8+ lymphocytes
in a cohort of 13 SIV-infected, ART-treated RMs and found
that depletion of CD8+ lymphocytes resulted in increased viral
levels in both plasma and lymphoid tissues in all treated animals.
RESULTS
Administration of Anti-CD8 Antibody MT-807R1 to SIV-
Infected ART-Treated RMs
16 Indian-origin RMs infected intravenously (i.v.) with SIVmac239
started an ART regimen consisting of Tenofovir, Emtricitabine,
Raltegravir, and Darunavir at week 8 after infection and were
treated for 8–32 weeks prior to further intervention (Figure 1A;
see also Table S1). Three RMs were euthanized prior to study
completion as a result of rapid disease progression or side ef-
fects of ART, leaving 13 animals to complete the study. ART sup-
pressed viremia to <60 copies/mL in 12/13 RMs that completed
the protocol (see Experimental Procedures). Seven RMs showed
at least four consecutive time points with viremia below 60
copies/mL (mean = 6.9 weeks, range = 4–10 weeks; i.e., ‘‘persis-
tent suppressors’’), whereas five showed a mix of undetectable
and detectable levels (i.e., ‘‘intermittent suppressors’’). The last
animal never achieved undetectable viremia, even though ART
decreased viremia by more than five logs. As expected, ART
partially restored CD4+ T cells in peripheral-blood mononuclear
cells (PBMCs), lymph nodes (LNs), and rectal biopsies (RBs)
(data not shown). Despite variability in the duration of ART and
kinetics of viremia suppression, at the time of CD8 depletion,
viral load had declined by >99.97% in comparison to pre-ART
levels in all animals (Table S2). Once viral loads were consistently
undetectable (persistent suppressors) or undetectable on at
least three non-consecutive evaluations (intermittent suppres-
sors), we administered one dose of the anti-CD8-depleting anti-
body, MT-807R1, at 50 mg/kg i.v. Animals were followed for
8 weeks after antibody administration and continued on ART
(Figure 1A). Using this method, we were able to deplete >95%
of CD8+ T cells in peripheral blood by day 1 after depletion (Fig-
ure 1B). This depletion was rapid and sustained (mean = 92%
[SD = 9] and mean = 94% [SD = 5] at weeks 1 and 3, respectively)
until approximately 5 weeks after depletion, at which time a var-
iable degree of CD8+ lymphocyte reconstitution in the periphery
was observed (Figures 1B and 1C). Because MT-807R1 was
directed at the CD8a chain, CD3 CD8a+ natural killer (NK) cells
were also depleted during this procedure (Figures S1A–S1C). Of
note, with an average of 70% depletion (SD = 26) at week 1
and 85% depletion (SD = 11) at week 3 after depletion, CD8
depletion in LNs was less complete than in peripheral blood (Fig-
ures S1D and S1E). With an average of 62% depletion (SD = 27)
Figure 1. Treatment of SIV-Infected ART-Treated RMs with the Depleting Anti-CD8 Antibody MT-807R1
(A) Study design.
(B) CD3+CD8+ T cells as a percentage of CD3+ lymphocytes in PBMCs.
(C) CD8+ T cells as a percentage of pre-depletion frequencies (‘‘baseline’’) in PBMCs. Data shown are for 13 SIV-infected ART-treated RMs. Bars are drawn at the
mean, and error bars represent the SEM.
(D–H) Proliferation (Ki67+) of total CD4+ T cells in (D) PBMCs, (E) LNs, and (F) RBs. Proliferation (Ki67+) of CD4+ T cell subsets in (G) PBMCs and (H) LNs. The gap in
the x axis represents 8–32 weeks of continuous ART. Black arrows and dotted vertical lines indicate MT-807R1 administration. Data shown are for 13 SIV-infected
ART-treated RMs. Error bars represent the SEM. NX, necropsy.
658 Immunity 45, 656–668, September 20, 2016
at week 1 after MT-807R1 administration and a more rapid
reconstitution starting by week 3 (mean = 56% depletion [SD =
30]), CD8 depletion was least efficient in RBs (Figures S1F and
S1G). Although MT-807R1 was less effective at depleting CD8+
lymphocytes in tissues than in blood, previous studies have
shown that CD8+ lymphocytes are functionally impaired after
binding of the anti-CD8 antibody even if they are not physically
depleted (Schmitz et al., 1999b). CD8 depletion is followed by
homeostatic proliferation of CD4+ T cells (Fukazawa et al.,
2015; Okoye et al., 2009), and we observed increased CD4+
T cell proliferation starting at week 2 after depletion in PBMCs
(Figure 1D), week 3 in LNs (Figure 1E), and week 1 in RBs
(Figure 1F). This increased CD4+ T cell proliferation involved pre-
dominantly CD4+ effector memory T (Tem) cells, and smaller and
delayed increases were observed in CD4+ stem cell memory T
(Tscm) and central memory T (Tcm) cells (Figures 1G and 1H).
We also examined the frequency of CCR5+, HLA-DR+, and
PD-1+CD4+ T cells both prior to and during CD8 depletion. We
found that CD8 depletion was associated with a moderate
increase in the fraction of CD4+HLA-DR+ and CD4+CCR5+
T cells in LNs (Figure S2). Overall, these results indicate that
administration of the MT-807R1 antibody is effective at depleting
CD8+ T cells in blood and lymphoid tissues of ART-treated
SIV-infected RMs.
CD8 Depletion Is Followed by an Increase in Plasma Viral
Load and SIV RNA in LNs in 100% of ART-Treated RMs
As shown in Figure 2, all 13 SIV-infected RMs showed a meas-
ureable increase in plasma viremia after CD8 depletion; several
animals rebounded as early as day 1 after depletion, and only
one animal remained below 60 copies/mL until week 3 after
depletion, at which point virus was detectable (RGb13). The
seven persistent suppressors all showed at least one time point
with detectable SIV RNA after depletion, and six out of seven
showed at least three time points with viremia above 60
copies/mL (Figure 2A). Similarly, the five intermittent suppres-
sors had detectable viremia at all examined time points after
depletion (Figure 2B). The one animal that never achieved unde-
tectable viremia despite 32 weeks of ART and a decrease in viral
load of more than five logs (i.e., >99.99% decline from baseline)
showed a marked increase in viremia of up to 105 copies/mL
after CD8 depletion (Figure 2C). Of note, CD8+ T cell reconstitu-
tion was consistently associated with a decrease in plasma viral
loads to levels similar to those observed prior to CD8 depletion
(Figures 2A–2C). Interestingly, in three out of four cases where
CD8+ T cells did not reconstitute in peripheral blood during the
8 week follow-up period after depletion, plasma viral loads re-
mained elevated above the limit of detection. To further analyze
the impact of CD8 depletion and reconstitution on SIV viremia,
we next divided the experimental timeline into three periods
 A B

Time after SIV infection (weeks)

D E

Figure 2. CD8-Depletion Results in Increased Plasma Viremia in ART-Treated SIV-Infected RMs

(A–C) Viral load (red) and CD8+ T cell counts (black) of SIV-infected RMs that were (A) persistently suppressed, (B) intermittently suppressed, and (C) never fully
suppressed on ART. Black arrows and dotted vertical lines indicate administration of anti-CD8 antibody MT-807R1.
(D) Geometric mean viremia for each statistical period for 13 RMs (described fully in Table S3). Bars show the geometric mean with 95% CI; *p < 0.05, ***p < 0.001.
(E) Geometric mean viremia for each statistical period for seven persistently suppressed RMs. Bars show the geometric mean with 95% CI; *p < 0.05. Statistical
analysis was conducted with a linear, mixed-effects model.

defined as follows: period 1 included the last 6 weeks of ART during CD8+ lymphocyte reconstitution when circulating CD8+
before CD8 depletion, period 2 included the post-CD8-depletion T cells were >20% of the baseline (blue, green, and purple lines
time points in which the level of circulating CD8+ T cells in Figure 2, respectively; see also Table S3). Using a mixed
was <10% of the baseline, and period 3 included the time points linear-effects model, we compared viral loads in these three

Immunity 45, 656–668, September 20, 2016 659

A Pre-ART B On ART, pre-depletion
F indicates an SIV-RNA-positive, productively infected
C On ART, post-depletion
TZ
F F
F contribute to maintenance of viral suppres-
F
D B cell follicle T cell zone
2.5
SIV vRNA+ cells / mm2
2.0
1.5
1.0
0.5
0.0
st
st
e
e
Pr
Pr
Po
Po
periods and found that the geometric mean of period 2 (275,
95% confidence interval [CI]: 116–653) was significantly higher
than that of either period 1 (53, 95% CI: 33–83, p < 0.05) or period
3 (90, 95% CI: 45–179, p < 0.001) (Figure 2D). Because of
the many non-detectable time points, we also conducted a bi-
nary analysis (detectable/non-detectable) that did not assume
normality and found that period 2 had significantly more detect-
able time points than either period 1 or 3 (data not shown). When
we applied the mixed linear-effects model to only the seven ART-
treated SIV-infected RMs with persistent suppression of viremia
before depletion (Figure 2A), the geometric mean of period 2
(135, 95% CI: 96–193) was again significantly higher than that
of either period 1 (34, 95% CI: 29–41, p < 0.05) or period 3 (44,
95% CI: 29 65, p < 0.05) (Figure 2E).
To determine whether the viremia trends observed in our
cohort of SIV-infected RMs were also present in lymphoid tis-
sues, we next conducted a sequential analysis of the levels of
SIV RNA production in LNs by using the RNAscope technology
at three time points, i.e., before initiation of ART, during ART
but before CD8 depletion, and during ART after CD8 depletion.
As shown in Figures 3A–3C, for representative animals, we
observed a dramatic decrease in the number of LN SIV-RNA-
positive cells after ART and a rebound of SIV RNA production
after CD8 depletion. A quantitative analysis of the number
of SIV-RNA-positive cells per square millimeter in the B cell
and T cell areas of LNs before and after CD8 depletion is
shown in Figure 3D. Taken together, these data indicate that,
660 Immunity 45, 656–668, September 20, 2016
Figure 3. CD8-Depletion Results in Increased
SIV RNA in the LNs of ART-Treated SIV-
Infected RMs
(A–C) Representative ISH of SIV RNA in the LNs of
(A) chronically infected, (B) ART-treated, pre-CD8-
F depletion, and (C) ART-treated, post-CD8-depletion
TZ RMs. Data presented in (B) and (C) are from the LNs
of two and three different animals, respectively. Red
TZ
cell. Abbreviations are as follows: F, follicle; and TZ,
T cell zone. Images show 203 magnification.
(D) Number of SIV-RNA-positive cells per square
millimeter in the B cell area (left) and T cell area (right)
of the LN before and after CD8 depletion.
F
TZ
in SIV-infected RMs, CD8+ lymphocytes
TZ sion during ART.
Ultrasensitive Viral-Load Assay
Reveals a 72- to 350-fold Increase in
Viral Production after CD8 Depletion
To better assess the level of residual viremia
in the group of persistently suppressed SIV-
infected ART-treated RMs at the time of CD8
depletion, we repeated the measurement of
viral load by using an ultrasensitive viral-
load assay with a detection limit of 3
copies/mL (Del Prete et al., 2014). We found
that viremia was <100 copies/mL in all of
these ART-treated RMs, and five out of the
seven animals showed levels of residual
viremia at %25 copies/mL (Table S2), which is in the range of
viremia levels measured in long-term ART-treated HIV-infected
individuals (Chun et al., 2011; Maldarelli et al., 2007). To more
rigorously evaluate the impact of CD8 depletion on residual
viremia during ART, we next used the same ultrasensitive assay
to measure viral load at all post-CD8-depletion time points in the
five RMs with the lowest residual viremia. As shown in Figure 4A,
these data largely recapitulate the results shown in Figure 2
and indicate that the level of viremia increased from <3–25
copies/mL before depletion (geometric mean = 8.3, 95%
CI: 2–31) to 700–4,900 copies/mL after depletion (geometric
mean = 1,977, 95% CI: 822–4,754, p < 0.05) for a 72- to 350-
fold increase (Figure 4B). Of note, the ultrasensitive assay de-
tected no correlation between pre- and post-CD8-depletion viral
loads (data not shown). Overall, these data confirm that CD8+
lymphocytes contribute to the suppression of viremia in SIV-in-
fected RMs treated with short-term (i.e., 8–32 week) ART.
Pre-depletion SIV-Specific CD8+ T Cells, but Not CD8+
NK Cells, Positively Correlate with the Magnitude of
Plasma Viral Rebound after Depletion
The depleting MT-807R1 is directed at the a chain of the CD8
molecule and depletes both CD8a+ NK cells and CD8+ T cells
in RMs. As expected, the frequency and absolute number of
SIV-specific CD8+ T cells decreased substantially after initiation
of ART (data not shown). However, we were able to detect low
frequencies of Gag-CM9-specific CD8+ T cells (as measured
 A Ultra-sensitive plasma viral load

ART RGb13 ART RLb13 ART RVy10

4 1000 4 1000 4 1000

800 800 800

3 3 3

Absolute number CD8+ T cells

SIV RNA copies per ml of plasma (log10)

600 600 600

2 2 2
400 400 400

1 1 1
200 200 200

0 0 0 0 0 0
-1

-1
-1

7
1

8
2

7
5

7
1

8
1

AB

AB
AB

ART RKq11 ART ROw8

4 2000 4 2000

3 1500 3 1500

2 1000 2 1000

1 500 1 500

0 0 0 0
-1
2

7
1

1
5

8
AB
-1

AB

Time after CD8 depletion (weeks)

B
Ultra sensitive viral load
4
SIV RNA copies per ml
plasma (log10)

0
n

n
io

tio
et

le
pl

ep
de

-d
e-

st
Pr

Po

Figure 4. Ultrasensitive Viral-Load Assay Confirms the Increase in Viremia after CD8 Depletion
(A) Viral load (orange) and CD8+ T cell counts (black) of SIV-infected RMs that were persistently suppressed. Black arrows and dotted vertical lines indicate
administration of anti-CD8 antibody MT-807R1.
(B) Pre-depletion and peak post-depletion viremia (log10) for five RMs. The dotted line indicates the detection limit of the ultrasensitive viral-load assay
(3 copies/mL plasma). The dashed line indicates the detection limit of the standard viral-load assay (60 copies/mL plasma). Paired t test, *p < 0.05.

by tetramer staining) in the PBMCs of Mamu-A*01 RMs at the depletion (Figures S3A and S3B). However, we found no correla-
time of CD8 depletion. First, we determined that the absolute tion between the numbers of circulating CD8+ NK cells before
number of pre-CD8-depletion circulating GAG-CM9-specific depletion and the viral load after depletion at either week 1
CD8+ T cells (but not their percentage of the total CD8+ T cells) or 3 (Figures S3C and S3D). Second, we conducted the same
directly correlated with viral load at both 1 and 3 weeks after analysis described in Figures 2D and 2E by dividing the study

Immunity 45, 656–668, September 20, 2016 661

 A PBMC B Lymph node Figure 5. Pre-depletion Levels of SIV Infec-
tion in CD4+ T Cells Predicts Post-CD8-

SIV DNA copies per 106cells

SIV DNA copies per 106cells

NS NS
105 NS 105 NS depletion Changes in Viremia
(A and B) Fraction of SIV-infected sorted CD4+
104 104 T cells in (A) PBMCs and (B) LNs. Abbreviations are
as follows: WK 3, week 3 after depletion; NX,
necropsy; and NS, not significant (Kruskal-Wallis
103 103
test).
(C and D) Correlation of cell-associated SIV DNA in
102 102 peripheral CD4+ T cells before depletion and the
peak viral rebound (C) and area under the curve (D)
101 101 after depletion. Data are shown for all 13 animals
Pre-dep WK 3 NX Pre-dep WK 3 NX with the Spearman rank correlation. Red circles
represent persistent suppressors, black circles
C D
Peak viral load Area under the curve represent intermittent suppressors, and white
105 r= 0.6044 105 r= 0.6209
circles indicate RMs that were never suppressed.
per 106 CD4+ T cells
SIVgag DNA copies

per 106 CD4+ T cells

*p= 0.032
SIVgag DNA copies

*p= 0.0268
104 104

103 103
Tcm, Tem, and transitional memory T
(Ttm) cells. As shown in Figures 5A and
102 102
5B, CD8 depletion did not induce any
101 101 significant change in the number of SIV-
101 102 103 104 105 106 101 102 103 104 105 106 DNA-positive CD4+ T cells in either blood
SIV RNA copies AUC (copies) or LNs. Similarly, we did not find any signif-
per ml plasma icant change in the amount of SIV DNA in
sorted CD4+ Tscm, Tcm, Ttm, or Tem cells
+
into three periods on the basis of the kinetics of CD8 NK cell in either blood or LNs after CD8 depletion (Figures S5A and S5B).
numbers: period 1 included the last 6 weeks of ART before Although the analysis of SIV DNA changes before and after CD8
CD8 depletion, period 2 included the post-CD8-depletion time depletion did not reveal a consistent pattern, we observed
points in which the level of circulating CD8+ NK cells was several RMs in which the cell-associated SIV DNA had increased
<20% of the baseline, and period 3 included the time points in either blood- or LN-derived CD4+ T cells, thus suggesting that
during CD8+ lymphocyte reconstitution when circulating CD8+ CD8 depletion induced expansion of the viral reservoir under ART
NK cells were >20% of the baseline (blue, green, and purple, in a subset of animals. Interestingly, we observed a significant
respectively, in Figure S4). As shown in Figure S4A for one repre- direct correlation between the amount of cell-associated SIV
sentative RM with persistent suppression of viremia, CD8+ NK DNA in CD4+ T cells before CD8 depletion and both the peak
cells returned to pre-depletion numbers around week 3 after and the area under the curve of plasma viremia after CD8 deple-
depletion, whereas CD8+ T cells were still significantly depleted. tion (Figures 5C and 5D), suggesting that the size of the reservoir
Using the same mixed linear-effects model of Figure 2, we found under ART is a key determinant of the resultant amount of viral
that whereas the geometric mean of the viral load in period 2 production when CD8+ lymphocytes are removed from the
(199, 95% CI: 136–321) was significantly higher than that in system. Finally, we sought to determine whether the amount of
period 1 (53, 95% CI: 44–70, p < 0.05), the geometric mean of viral production after CD8 depletion (i.e., peak and/or area under
the viral load in period 3 (221, 95% CI: 122–382) was unchanged the curve) correlated with the observed changes in the CD4+
in comparison to that in period 2 (Figure S4B). We found similar T cell proliferation and activation, as measured by Ki67, CCR5,
results when we analyzed only the persistently suppressed RMs HLA-DR, and PD-1 expression, in various cell subsets and tis-
(Figure 2A), whose viral loads remained elevated despite recon- sues. As shown in Table S4, we found no significant correlations
stitution of the CD8+ NK cell pool (Figure S4C). Finally, no corre- between the change in the fraction of proliferating (i.e., Ki-67+)
lation was observed between either the length of ART or the time cells in any of the examined CD4+ T cell subsets after CD8 deple-
to suppression and the change in viremia after CD8 depletion tion and the amount of viremia at the same time point. We also
(data not shown). These data, together with those shown in Fig- found no correlation between the fraction of proliferating CD4+
ures 2D and 2E, indicate that the depletion and repopulation of T cells and the amounts of cell-associated SIV DNA after CD8
CD8+ T cells, but not the repopulation of CD8+ NK cells, is depletion (data not shown). This lack of correlation suggests
temporally associated with the control of viremia under ART. that the homeostatic CD4+ T cell proliferation that is associated
with CD8 depletion is unlikely to be a key factor responsible for
Pre-depletion Amounts of Cell-Associated SIV DNA the observed increase in SIV viremia. However, we found a direct
Positively Correlate with Plasma SIV RNA after correlation between changes in CD4+HLA-DR+ and CD4+PD-1+
Depletion T cells (but not CD4+Ki-67+ or CD4+CCR5+ T cells) and viremia,
To determine whether CD8 depletion had an impact on the overall although it was significant only at the earliest time points after
viral reservoir in ART-treated SIV-infected RMs, we sequentially depletion (i.e., days 1 and 2) (Table S5), suggesting that the initial
measured the amount of total, cell-associated SIV DNA by RT- activation of CD4+ T cells that occurs after CD8 depletion might
PCR in bulk CD4+ T cells, as well as in subsets of CD4+ Tscm, contribute to the observed increase in SIV viremia.

662 Immunity 45, 656–668, September 20, 2016

Figure 6. Highlighter Plots of SGA-Derived Env Amino Acid Sequences
Highlighter plots were generated to illustrate variation in amino acid composition within Env over the course of the experimental protocol in three animals: RLb13
(left), ROw8 (center), and RKq11 (right). Sequences were isolated at peak viral load, pre-ART initiation, and post-CD8 depletion (range = weeks 1–6). Colored ticks
indicate deviation from the master sequence.
Longitudinal Viral Sequencing by Single-Genome
Analysis
To gain insight into the dynamics of viral replication and/or
reactivation after CD8 depletion, we performed a longitudinal
study of plasma virus by single-genome analysis (SGA) of
env of SIVmac239 at three time points: (1) day 10 after SIV infec-
tion, i.e., peak viremia; (2) day 56 after SIV infection, i.e., imme-
diately prior to ART initiation; and (3) after CD8 depletion. We
conducted this analysis on three representative RMs (RKq11,
ROw8, and RLb13) with persistent SIV suppression under ART
(Table S2). As shown in Figure 6, circulating viruses sequenced
at day 10 after infection were relatively homogeneous within
each RM (99.9%–100% amino acid identity). In contrast,
sequences isolated on day 56 after infection showed several
fixed mutations that most likely represent escape from either
T-cell-mediated or antibody responses. Interestingly, the viral
sequences derived from plasma after CD8 depletion were
more similar to those derived at peak viremia than to those
derived immediately prior to ART (i.e., peak versus pre-ART
amino acid identity = 99.5%–99.7% and peak versus post-
CD8-depletion amino acid identity = 99.8–99.9). Importantly,
the previously fixed escape mutations observed at the pre-ART
time point represent only a small fraction of the circulating vi-
ruses after CD8 depletion. In addition, we conducted a MiSeq
deep-sequence analysis of the circulating cell-associated SIV
DNA immediately prior to ART initiation and immediately before
CD8 depletion. As shown in Figure S6, this analysis revealed a
consistent increase in inoculum-like sequences (i.e., similar to
those derived at peak viremia) under ART. This latter observation
is consistent with the finding that viruses similar to the initial in-
fecting variants reemerge in plasma during the early stages of
ART (data not shown) and suggests the presence of a pool of
cells that are infected with inoculum-like virus and outlive cells in-
fected with virus harboring escape mutations. Of note, when we
conducted the same longitudinal sequence analysis of plasma
virus by SGA in the SIV-infected RM that never achieved
undetectable viremia (ROn13), we found throughout the three
examined time points a complex pattern of viral quasi-species
diversity consistent with continuous viral replication (data not
shown). Overall, these data suggest that the plasma viruses
that are present after CD8 depletion are produced from long-
lived cells that were infected prior to ART initiation and before
the generation of immune escape mutants.
DISCUSSION
This study demonstrates that CD8+ lymphocytes can contribute
to maintaining suppression of plasma viremia in SIV-infected
RMs that are treated with short-term (i.e., 8–32 week) ART. To
date, the in vivo suppressive effect of CD8+ lymphocytes has
Immunity 45, 656–668, September 20, 2016 663
primarily been characterized in studies involving untreated ani-
mals (Chowdhury et al., 2015; Jin et al., 1999; Matano et al.,
1998; Schmitz et al., 1999a); only one report has described the
treatment of three SIV-infected RMs with non-suppressive teno-
fovir monotherapy (Van Rompay et al., 2004). Indeed, to our
knowledge, the current work represents the first experiment in
which SIV-infected RMs have undergone depletion of CD8+ lym-
phocytes in the setting of highly active ART (i.e., >99.97% sup-
pression of viremia). Although the majority of the RMs included
in this study showed some residual viral production under
ART, these levels of viremia were often within the range observed
in long-term ART-treated HIV-infected individuals (Maldarelli
et al., 2007). Importantly, the magnitude of the viral increase after
CD8 depletion (i.e., 72- to 350-fold) and the fact that control of
viremia was promptly reestablished upon CD8+ T cell reconstitu-
tion strongly support the hypothesis of an important (and yet pre-
viously unrecognized) role for CD8+ lymphocytes in cooperating
with ART to maintain viral suppression. As such, the current data
provide an evidence base for exploring immune-based interven-
tions, such as therapeutic vaccines and immune checkpoint
inhibitors, in ART-treated HIV-infected individuals. It should be
noted, however, that interventions aimed at improving the
antiviral CD8+ lymphocyte-mediated response under ART will
be expected to be effective only against infected cells with suf-
ficient expression of viral antigen, suggesting that combining
viral induction and immune-based approaches could enhance
the effectiveness of such interventions.
In ultrasensitive viral-load assays, residual viremia below the
clinical limit of detection is present in many ART-treated HIV-in-
fected individuals (Chun et al., 2011; Dornadula et al., 1999; Mal-
darelli et al., 2007). Whether and to what extent this residual
viremia is due to continuous, low-level de novo cycles of replica-
tion as opposed to either viral reactivation in latently infected
cells or the presence of persistently virus-expressing cells re-
mains an active area of research. Recent evidence suggests
that clonal expansion of virally infected cells during ART is
another important contributor to HIV persistence (Maldarelli
et al., 2014; von Stockenstrom et al., 2015; Wagner et al.,
2014). Similarly, at this time we do not know to what extent the
increase in viremia after CD8 depletion in SIV-infected, ART-
treated RMs reflects de novo viral replication, reactivation of
latent virus, or increased viral production from persistently,
non-latently infected cells. Regardless of the mechanisms
responsible for the very low viremia of ART-treated HIV-infected
individuals and SIV-infected RMs, the current study suggests a
direct in vivo role for CD8+ lymphocytes in maintaining very low
to undetectable viral loads under ART. However, the current
study does not define whether the antiviral role of CD8+ lympho-
cytes under ART involves classical cytotoxic T lymphocyte activ-
ity, non-cytolytic mechanisms (i.e., chemokine production and
inhibition of viral transcription; Cocchi et al., 1995; Walker
et al., 1986), or both.
In this study, CD8 depletion was followed by various degrees
of CD4+ T cell activation and proliferation, which could at least in
part contribute to the observed increase in SIV viremia after CD8
depletion through reactivation of viral production from latently
infected, previously resting CD4+ T cells. Interestingly, we
observed a direct correlation between the changes in HLA-DR
and PD-1 expression (but not Ki-67 or CCR5 expression) on
664 Immunity 45, 656–668, September 20, 2016
CD4+ T cells and the level of viremia early after CD8 depletion.
In particular, animal RGb13, which showed the greatest sup-
pression of viremia with ART (fewer than three copies for three
consecutive measurements), experienced a delayed rebound
of viremia that was coincident with the peak of CD4+ T cell acti-
vation after CD8 depletion. In this context, however, we wish to
emphasize that the relationship among CD8 depletion, increased
SIV viremia, and increased CD4+ T cell activation is very complex
and that the exact contribution of this latter factor will need to be
assessed in future in vivo interventional studies in which CD4+
T cell activation is selectively inhibited in SIV-infected ART-
treated RMs at the time of CD8 depletion.
In this cohort of ART-treated SIV-infected RMs, the increased
viremia that followed CD8 depletion was not consistently associ-
ated with specific changes (i.e., increases or decreases) in the
levels of cell-associated SIV DNA. However, a subset of CD8-
depleted SIV-infected RMs showed an increase in the level of
SIV DNA, consistent with de novo rounds of viral replication
and/or homeostatic proliferation of SIV-infected cells. In this
experimental setting, the change in the fraction of SIV-DNA-pos-
itive cells in blood and LNs after CD8 depletion is the result of
factors such as (1) de novo infection of cells, (2) the lifespan of
cells that have reactivated viral production, and (3) the prolifera-
tion levels and lifespan of cells that harbor integrated SIV DNA
but do not express the virus. An additional complication of these
studies is that the level of cell-associated SIV DNA is measured
as the number of copies per fixed amount of cells (typically one
million), whereas it is virtually impossible to measure the actual
total number of SIV-DNA-positive cells in the body. Clarifying
the relative contribution of these mechanisms to the changes
in the size of viral reservoirs after CD8 depletion is a challenging
scientific problem that is the object of intensive studies in our
laboratories.
In an attempt to elucidate the impact of CD8 depletion on the
viral dynamics in our cohort of ART-treated SIV-infected RMs,
we performed, in a subset of animals, a longitudinal sequence
analysis of plasma SIV RNA and cell-associated SIV DNA by
SGA and deep sequencing, respectively. Our SGA analysis re-
vealed a clear and consistent pattern of viral sequence dynamics
in three of the SIV-infected RMs with the best viral suppression
under ART. This analysis showed that (1) certain amino acid
substitutions become ‘‘fixed’’ in the circulating population of vi-
ruses in the interval between peak viremia and ART initiation and
(2) the virus that emerges after CD8 depletion shows sequences
that largely overlap with the peak viremia quasi-species and not
the pre-ART quasi-species. In addition, the analysis of SIV DNA
revealed a concomitant increase in the fraction of ‘‘inoculum-
like’’ sequences during ART treatment. The most parsimonious
explanation for these findings is that the de novo viral production
after CD8 depletion was mainly derived from a subset of long-
lived cells that, at the time of ART initiation, were either latently
infected or produced only a small minority of circulating virions.
Potential mechanisms responsible for the increased viral pro-
duction by these putatively long-lived SIV-infected cells after
CD8 depletion include the absence of CD8+ lymphocyte-medi-
ated cytolytic activity, lack of non-cytolytic viral suppression
(Klatt et al., 2010; Wong et al., 2010), and increased CD4+
T cell activation. Of note, the striking similarity between the inoc-
ulum (i.e., peak viremia) sequences and the post-CD8-depletion
sequences does not support the hypothesis that the observed
increase in viremia after CD8 depletion is simply the conse-
quence of an ‘‘amplification’’ of cellular and/or anatomic pockets
of residual viral replication under ART.
Although the effect of CD8 depletion on the levels of plasma
viremia in ART-treated SIV-infected RMs is clear, there are
important caveats to this study. The first is that the used anti-
body depletes both CD8+ T cells and CD8a-expressing NK
cells. However, our analysis of the correlation between viremia
after CD8 depletion and either the numbers of CD8+ T cells and
CD8+ NK cells before depletion or their levels during depletion
and repopulation strongly suggests that CD8+ T cells, rather
than NK cells, are the main contributors to the control of
viremia under ART. In addition, we observed reactivation
of viral production in lymphoid tissues in which the levels of
CD8+ NK cells were extremely low even before CD8 depletion.
Further studies using a CD8-b-specific antibody that depletes
only CD8ab+ T cells will formally establish the specific contribu-
tion of CD8+ T cells to the control of viremia in ART-treated
SIV-infected RMs. The second caveat is that the period of
ART suppression was relatively short even in the SIV-infected
RMs defined as persistent suppressors, and therefore the level
of viral suppression might not be as complete as in long-term
ART-treated HIV-infected humans. However, we wish to point
out that the level of viral suppression was >99.97% in all
treated RMs (average 99.99%), and five animals showed
viremia levels similar to those observed in long-term ART-
treated HIV-infected humans (Maldarelli et al., 2007). Further
studies in which SIV-infected RMs are treated for longer pe-
riods of time (i.e., >1 year) will determine whether and to
what extent the viral suppression that occurs in the setting of
long-term ART is affected by CD8 depletion. We wish to point
out that even if future studies reveal that CD8+ cells are no
longer needed for maintaining suppression of viremia when
ART is administered for years, the current results would still
provide rationale for clinical interventions in which CD8-
enhancing therapies (immune checkpoint inhibitors, therapeutic
vaccines, etc.) are performed during the first 6 to 8 months
of ART. The third caveat of the current study is that the impact
of CD8 depletion on the reservoir was measured by a relatively
insensitive assay, i.e., total cell-associated SIV DNA, which
does not provide a functional analysis of the reservoir of repli-
cation-competent virus. Although interesting, analyses such as
the quantitative viral-outgrowth assay require a large number of
cells and are not feasible in animals undergoing an already
heavy schedule of blood collections. For these reasons, the
current work can be considered a pilot, hypothesis-generating
study that should prompt additional in-depth investigation of
this intriguing and previously unrecognized role of CD8 deple-
tion in inducing increased viral production in ART-treated
SIV-infected RMs.
Our current set of findings might seem to contrast with the
observation that immunosuppressive treatment is not consis-
tently associated with increased viremia in ART-treated HIV-in-
fected individuals. However, these clinical instances are different
from CD8 depletion because they can involve depletion of all
T cells, as in the case of aggressive chemotherapy regimens, or
markedly reduce CD4+ T cell activation, as in the case of several
commonly used immunomodulators. For instance, cyclosporine
and Tacrolimus (FK506) are calcineurin inhibitors that inhibit
NFAT-mediated CD4+ T cell activation and have a well-charac-
terized direct antiviral effect via inhibition of cyclophilin (Bartz
et al., 1995; Braaten et al., 1996; Emmel et al., 1989; Franke
and Luban, 1996; Franke et al., 1994; Schaller et al., 2011; Sokol-
skaja et al., 2010; Streblow et al., 1998). Similarly, mycophenolate
induces apoptosis of activated CD4+ T lymphocytes and inhibits
HIV replication (Chapuis et al., 2000; Margolis et al., 2002),
whereas rapamycin (sirolimus) decreases LTR-driven transcrip-
tion of HIV (Roy et al., 2002) and reduces expression of the
main HIV co-receptor CCR5 (Gilliam et al., 2007). These caveats
notwithstanding, the current data have important implications for
understanding how the host antiviral cellular immune response
works in concert with antiretroviral drugs in suppressing SIV repli-
cation and for defining the rationale for therapeutic interventions
aimed at boosting the virus-specific CD8+ lymphocyte response
in ART-treated HIV-infected individuals.
EXPERIMENTAL PROCEDURES
Animals
16 Indian-origin RMs were enrolled in this study. They were all infected i.v. with
3,000 TCID50 of SIVmac239. Three animals were euthanized because of severe
weight loss prior to CD8 depletion and were thus excluded from analysis. The
remaining 13 animals consisted of six females and seven males and ranged in
age from 4 to 13 years at the time of infection. Nine were Mamu-A*01 positive,
and all 16 were Mamu-B*08 and Mamu-B*17 negative. All animals were
housed at the Yerkes National Primate Research Center of Emory University
and maintained in accordance with US NIH guidelines. Anesthesia was used
for all blood and tissue collections. All studies were approved by the Emory
University Institutional Animal Care and Usage Committee.
Antiretroviral Therapy
All RMs were put on ART at 8 weeks after SIV infection. The drug regimen
started as 20 mg/kg PMPA (Tenofovir) and 30 mg/kg FTC (Emtricitabine)
administered once a day by subcutaneous injection along with 100 mg Ralte-
gravir and 400 mg Darunavir orally twice daily. During the first 8 weeks, all RMs
received this treatment regimen. Three animals remained on this regimen for
the duration of the study. In the remaining ten animals, we increased the
dose of Darunavir and Raltegravir as described in Table S1 until suppression
was achieved. Animals were given oral antiretroviral treatment via orogastric
tube on days when anesthesia was also administered as a result of fasting
and nausea. Suppression of plasma viremia to below the detection limit of
our assay (60 copies/mL of plasma) was achieved in 12 out of 13 animals,
and persistent suppression was observed in 7 out of 13 animals. Of note,
the original study design planned for CD8 depletion after two consecutive
undetectable cases of viremia were measured 2 weeks apart. However,
some animals who had not reached this definition showed side effects of
ART. As such, we decided to perform the depleting treatment in all enrolled
animals to generate as much data as possible regarding CD8 depletion in
ART-treated RMs.
Depletion of CD8+ Lymphocytes
We administered one dose of MT-807R1 (from the NIH NHP Reagent
Resource Program) i.v. at 50 mg/kg (<0.25 EU/mL of endotoxin by a gel-clot
limulous amoebocyte lysate test). The extent of CD8 depletion was determined
(1) in peripheral blood by flow cytometric staining and complete blood counts
as both the number of cells/mm3 of blood and the percentage of CD3+ T cells
and (2) in tissues by flow cytometry and measurement of the level of CD8+
T cells as a fraction of their pre-depletion frequency.
Sample Collection and Tissue Processing
Blood was collected in EDTA tubes. Plasma was obtained by centrifugation.
PBMCs were obtained by density-gradient centrifugation using 90% Lympho-
cyte Separation Media from Lonza. LN biopsies were taken at day 56 before
Immunity 45, 656–668, September 20, 2016 665
ART initiation, 1 week before CD8 depletion, weeks 1 and 3 after CD8 Statistical Analysis
depletion, and at necropsy. LNs were cut in half, and one half was put in 4% The data presented in Figures 2D and 2E and Figures S4B and S4C
paraformaldehyde and then embedded in paraffin. LNs were ground over a were analyzed by a mixed-effects linear model that allows estimation of
70 mm cell strainer to produce LN-derived cells. RBs were digested in mean viral load by period. Comparisons between frequency of infection before
0.75 mg/mL collagenase (Sigma-Aldrich) and 0.15 mL DNase in 10% fetal and after depletion were carried out with either a Kruskal-Wallis test (Figures
calf serum, 1% penicillin-streptomycin, and 1% L-glutamine at 37 C with 5A and 5B) or a Wilcoxon matched-pairs signed-rank test (Figures S5A and
gentle shaking. After 2 hr, tissues were mechanically separated with a plastic S5B). Correlations were determined with the non-Gaussian Spearman correla-
cannula and run over a 70 mm filter. tion. Significance was attributed at p < 0.05. Analysis was done with GraphPad
Prism 6.0 and R 3.1.3.
Immunophenotype by Flow Cytometry
Multiparametric flow cytometry was performed on PBMCs, LN mononuclear ACCESSION NUMBERS
cells, and cells isolated from RBs with fluorescently labeled monoclonal anti-
bodies cross-reactive in RMs. The following antibodies were used: APC-Cy7 The accession numbers for the sequence data reported in this paper are
mouse anti-human CD3 (SP34-2), PE-CF594 mouse anti-human CD4 (L200), GenBank: KX552042–KX552194.
BV711 mouse anti-human CD8 (RPA-T8), FITC mouse anti-human CD197
(CCR7) (150503), PE-Cy7 mouse anti-human CD45RA (L48), PE-Cy5 mouse
SUPPLEMENTAL INFORMATION
anti-human CD95 (DX2), Alexa Fluor 700 mouse anti-Ki-67 (B56), BV650
mouse anti-human CD14 (M5E2), BV605 mouse anti-human CD56
Supplemental Information includes Supplemental Experimental Procedures,
(NCAM16.2), PE mouse anti-human CD62L (SK11), PerCP-Cy5.5 mouse
six figures, and five tables and can be found with this article online at http://
anti-human HLA-DR (G46-6), and APC mouse anti-human CCR5 (3A9) from
dx.doi.org/10.1016/j.immuni.2016.08.018.
BD Biosciences; CD28 (human) (CD28.2) from Beckman Coulter; Brilliant
Violet 421 anti-human CD16 (3G8), Brilliant Violet 650 anti-human CD4
AUTHOR CONTRIBUTIONS
(OKT4), and Brilliant Violet 421 anti-human CD279 (PD-1) (EH12.2H7) from
BioLegend; and mouse anti-human CD8/APC (DK25) from DAKO. All speci-
E.K.C. and G.S. designed and performed the experiments. L.S., S.A.S.,
mens were acquired on an LSR II (BD Biosciences), and the acquired data
S.P., and C.D. performed single-genome sequencing of SIV RNA. D.L.,
were analyzed with FlowJo software (Tree Star).
S.E.B., and T.H.V. performed deep sequencing of cell-associated SIV DNA.
R.F. and J.D.L. performed ultrasensitive viral-load analyses. B.O.L., M.N.,
Cell Sorting
and T.H.V. analyzed viral load and cell-associated DNA. K.E. performed statis-
After isolation, cells were resuspended in PBS containing 2 mM EDTA and
tical analyses. J.D.E. performed RNAscope analysis. J.E.S., M.P., A.C., C.D.,
spun for the removal of contaminating platelets. Prior to sorting, CD4+
and S.E.B. discussed the data. E.K.C., A.C., and G.S. wrote the manuscript.
T cells were enriched with the use of magnetic beads and column purification
(Miltenyi Biotec). Enriched cells were then stained with previously determined
ACKNOWLEDGMENTS
volumes of APC-Cy7 mouse anti-human CD3 (SP34-2), Brilliant Violet 650
anti-human CD4 (OKT4), Brilliant Violet 605 anti-human CD8a (RPA-T8),
We thank J. Levy, D.G. Carnathan, M. Mavigner, C.S. McGary, and G. Mylva-
LIVE/DEAD aqua, APC-H7 mouse anti-human CD45RA (5H9), PE-Cy7 rat
ganam for helpful discussions; B. Cervasi, K.P. Gill, and J.P. Mackel for tech-
anti-human CD197 (CCR7) (3D12), PE-Cy5 mouse anti-human CD95 (DX2),
nical support; S. Ehnert for study organization and scheduling; Yerkes National
CD28 (human) (CD28.2), and PE mouse anti-human CD62L (SK11). Popula-
Primate Research Center (YNPRC) veterinary staff, especially S. Jean, for car-
tions for sorting were defined as follows: Tscm, CD45RA+CCR7+CD95+
ing for the animals and daily ART administration; Merck (I.MTA.14.232) Gilead
CD62L+; Tcm, CD45RA CD95+CD28+CCR7+CD62L+; Ttm, CD95+CCR7+
Sciences (I.MTA.14.230) for generously providing Raltegravir, PMPA, and
CD62L ; and Tem, CD95+CCR7 CD62L . Sorting was performed on a
FTC; and Janssen R&D Ireland (I.MTA.14.231) for providing Darunavir. We
FACSAria LSR II (BD Biosciences) equipped with FACS Diva software.
thank the Emory CFAR Virology Core for viral loads and the Emory CFAR
Immunology Core for use of the MiSeq platform. This work was funded in
Plasma Viral Load and Cell-Associated SIV gag DNA
part by NIH grants R01-AI90797 (to G.S.), RR000165/OD011132 (to the
Plasma viral load was quantified as described previously (Taaffe et al., 2010).
YNPRC), and P30-AI-50409, R01 AI-58706, and U19 AI-109633 (to C.D.), the
Quantification of SIVmac gag DNA was performed as previously described
Emory CFAR, and federal funds from the NIH National Cancer Institute (con-
(Chahroudi et al., 2014). Additional details can be found in the Supplemental
tract HHSN261200800001E with J.L.).
Experimental Procedures.

Received: March 24, 2016

In Situ Hybridization
For in situ hybridization (ISH), we used RNAscope (Advanced Cell Diagnos-
tics), a next-generation, ultra-sensitive RNA ISH technology, as described
previously (Smedley et al., 2014).
Ultrasensitive Viral-Load Assay
Ultrasensitive measurements of plasma viral RNA were performed essentially
as described previously (Del Prete et al., 2014).
SGA of Circulating SIV
SGA of plasma-derived SIV was performed essentially as described previously
(Burton et al., 2015; Smith et al., 2015). Additional details can be found in the
Supplemental Experimental Procedures.
Deep Sequencing of Cell-Associated SIV DNA
Deep sequencing of two amplicons covering nucleotides 6,679–7,103 and
8,748–9,196 of the wild-type SIVmac239 genome (NCBI: M33262.1) and corre-
sponding to Env amino acids 35–158 and 724–845 was performed essentially
as described previously (Vanderford et al., 2011). Additional details can be
found in the Supplemental Experimental Procedures.
Revised: July 1, 2016
Accepted: July 7, 2016
Published: September 20, 2016
REFERENCES
Altfeld, M., Addo, M.M., Rosenberg, E.S., Hecht, F.M., Lee, P.K., Vogel, M.,
Yu, X.G., Draenert, R., Johnston, M.N., Strick, D., et al. (2003). Influence of
HLA-B57 on clinical presentation and viral control during acute HIV-1 infection.
AIDS 17, 2581–2591.
Bartz, S.R., Hohenwalter, E., Hu, M.K., Rich, D.H., and Malkovsky, M. (1995).
Inhibition of human immunodeficiency virus replication by nonimmunosup-
pressive analogs of cyclosporin A. Proc. Natl. Acad. Sci. USA 92, 5381–5385.
Betts, M.R., Nason, M.C., West, S.M., De Rosa, S.C., Migueles, S.A.,
Abraham, J., Lederman, M.M., Benito, J.M., Goepfert, P.A., Connors, M.,
et al. (2006). HIV nonprogressors preferentially maintain highly functional
HIV-specific CD8+ T cells. Blood 107, 4781–4789.
Borrow, P., Lewicki, H., Hahn, B.H., Shaw, G.M., and Oldstone, M.B. (1994).
Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control

666 Immunity 45, 656–668, September 20, 2016

of viremia in primary human immunodeficiency virus type 1 infection. J. Virol.
68, 6103–6110.
Borrow, P., Lewicki, H., Wei, X., Horwitz, M.S., Peffer, N., Meyers, H., Nelson,
J.A., Gairin, J.E., Hahn, B.H., Oldstone, M.B., and Shaw, G.M. (1997). Antiviral
pressure exerted by HIV-1-specific cytotoxic T lymphocytes (CTLs) during
primary infection demonstrated by rapid selection of CTL escape virus. Nat.
Med. 3, 205–211.
Braaten, D., Aberham, C., Franke, E.K., Yin, L., Phares, W., and Luban, J.
(1996). Cyclosporine A-resistant human immunodeficiency virus type 1
mutants demonstrate that Gag encodes the functional target of cyclophilin
A. J. Virol. 70, 5170–5176.
Burton, S.L., Kilgore, K.M., Smith, S.A., Reddy, S., Hunter, E., Robinson, H.L.,
Silvestri, G., Amara, R.R., and Derdeyn, C.A. (2015). Breakthrough of SIV strain
smE660 challenge in SIV strain mac239-vaccinated rhesus macaques despite
potent autologous neutralizing antibody responses. Proc. Natl. Acad. Sci. USA
112, 10780–10785.
Chahroudi, A., Cartwright, E., Lee, S.T., Mavigner, M., Carnathan, D.G.,
Lawson, B., Carnathan, P.M., Hashempoor, T., Murphy, M.K., Meeker, T.,
et al. (2014). Target cell availability, rather than breast milk factors, dictates
mother-to-infant transmission of SIV in sooty mangabeys and rhesus ma-
caques. PLoS Pathog. 10, e1003958.
Chapuis, A.G., Paolo Rizzardi, G., D’Agostino, C., Attinger, A., Knabenhans,
C., Fleury, S., Acha-Orbea, H., and Pantaleo, G. (2000). Effects of mycophe-
nolic acid on human immunodeficiency virus infection in vitro and in vivo.
Nat. Med. 6, 762–768.
Chen, Z.W., Craiu, A., Shen, L., Kuroda, M.J., Iroku, U.C., Watkins, D.I., Voss,
G., and Letvin, N.L. (2000). Simian immunodeficiency virus evades a dominant
epitope-specific cytotoxic T lymphocyte response through a mutation result-
ing in the accelerated dissociation of viral peptide and MHC class I.
J. Immunol. 164, 6474–6479.
Chowdhury, A., Hayes, T.L., Bosinger, S.E., Lawson, B.O., Vanderford, T.,
Schmitz, J.E., Paiardini, M., Betts, M., Chahroudi, A., Estes, J.D., and
Silvestri, G. (2015). Differential impact of in vivo CD8+ T-lymphocyte depletion
in controller versus progressor Simian Immunodeficiency Virus-infected
macaques. J. Virol. 89, 8677–8686.
Chun, T.W., Murray, D., Justement, J.S., Hallahan, C.W., Moir, S., Kovacs, C.,
and Fauci, A.S. (2011). Relationship between residual plasma viremia and the
size of HIV proviral DNA reservoirs in infected individuals receiving effective
antiretroviral therapy. J. Infect. Dis. 204, 135–138.
Cocchi, F., DeVico, A.L., Garzino-Demo, A., Arya, S.K., Gallo, R.C., and Lusso,
P. (1995). Identification of RANTES, MIP-1 alpha, and MIP-1 beta as the major
HIV-suppressive factors produced by CD8+ T cells. Science 270, 1811–1815.
Costello, C., Tang, J., Rivers, C., Karita, E., Meizen-Derr, J., Allen, S., and
Kaslow, R.A. (1999). HLA-B*5703 independently associated with slower
HIV-1 disease progression in Rwandan women. AIDS 13, 1990–1991.
Del Prete, G.Q., Shoemaker, R., Oswald, K., Lara, A., Trubey, C.M., Fast, R.,
Schneider, D.K., Kiser, R., Coalter, V., Wiles, A., et al. (2014). Effect of suber-
oylanilide hydroxamic acid (SAHA) administration on the residual virus pool in
a model of combination antiretroviral therapy-mediated suppression in
SIVmac239-infected indian rhesus macaques. Antimicrob. Agents
Chemother. 58, 6790–6806.
Dornadula, G., Zhang, H., Shetty, S., and Pomerantz, R.J. (1999). HIV-1 virions
produced from replicating peripheral blood lymphocytes are more infectious
than those from nonproliferating macrophages due to higher levels of intravi-
rion reverse transcripts: implications for pathogenesis and transmission.
Virology 253, 10–16.
Emmel, E.A., Verweij, C.L., Durand, D.B., Higgins, K.M., Lacy, E., and
Crabtree, G.R. (1989). Cyclosporin A specifically inhibits function of nuclear
proteins involved in T cell activation. Science 246, 1617–1620.
Evans, D.T., Knapp, L.A., Jing, P., Mitchen, J.L., Dykhuizen, M., Montefiori,
D.C., Pauza, C.D., and Watkins, D.I. (1999). Rapid and slow progressors differ
by a single MHC class I haplotype in a family of MHC-defined rhesus ma-
caques infected with SIV. Immunol. Lett. 66, 53–59.
Evans, D.T., Jing, P., Allen, T.M., O’Connor, D.H., Horton, H., Venham, J.E.,
Piekarczyk, M., Dzuris, J., Dykhuzen, M., Mitchen, J., et al. (2000). Definition
of five new simian immunodeficiency virus cytotoxic T-lymphocyte epitopes
and their restricting major histocompatibility complex class I molecules:
evidence for an influence on disease progression. J. Virol. 74, 7400–7410.
Franke, E.K., and Luban, J. (1996). Inhibition of HIV-1 replication by cyclo-
sporine A or related compounds correlates with the ability to disrupt the
Gag-cyclophilin A interaction. Virology 222, 279–282.
Franke, E.K., Yuan, H.E., and Luban, J. (1994). Specific incorporation of cyclo-
philin A into HIV-1 virions. Nature 372, 359–362.
Fukazawa, Y., Lum, R., Okoye, A.A., Park, H., Matsuda, K., Bae, J.Y., Hagen,
S.I., Shoemaker, R., Deleage, C., Lucero, C., et al. (2015). B cell follicle
sanctuary permits persistent productive simian immunodeficiency virus infec-
tion in elite controllers. Nat. Med. 21, 132–139.
Gilliam, B.L., Heredia, A., Devico, A., Le, N., Bamba, D., Bryant, J.L., Pauza,
C.D., and Redfield, R.R. (2007). Rapamycin reduces CCR5 mRNA levels in
macaques: potential applications in HIV-1 prevention and treatment. AIDS
21, 2108–2110.
Haynes, B.F., Pantaleo, G., and Fauci, A.S. (1996). Toward an understanding
of the correlates of protective immunity to HIV infection. Science 271,
324–328.
Jin, X., Bauer, D.E., Tuttleton, S.E., Lewin, S., Gettie, A., Blanchard, J., Irwin,
C.E., Safrit, J.T., Mittler, J., Weinberger, L., et al. (1999). Dramatic rise in
plasma viremia after CD8(+) T cell depletion in simian immunodeficiency
virus-infected macaques. J. Exp. Med. 189, 991–998.
Kaslow, R.A., Carrington, M., Apple, R., Park, L., Muñoz, A., Saah, A.J.,
Goedert, J.J., Winkler, C., O’Brien, S.J., Rinaldo, C., et al. (1996). Influence
of combinations of human major histocompatibility complex genes on the
course of HIV-1 infection. Nat. Med. 2, 405–411.
Klatt, N.R., Shudo, E., Ortiz, A.M., Engram, J.C., Paiardini, M., Lawson, B.,
Miller, M.D., Else, J., Pandrea, I., Estes, J.D., et al. (2010). CD8+ lymphocytes
control viral replication in SIVmac239-infected rhesus macaques without
decreasing the lifespan of productively infected cells. PLoS Pathog. 6,
e1000747.
Klein, M.R., van der Burg, S.H., Hovenkamp, E., Holwerda, A.M., Drijfhout,
J.W., Melief, C.J., and Miedema, F. (1998). Characterization of HLA-B57-
restricted human immunodeficiency virus type 1 Gag- and RT-specific cyto-
toxic T lymphocyte responses. J. Gen. Virol. 79, 2191–2201.
Koup, R.A., Safrit, J.T., Cao, Y., Andrews, C.A., McLeod, G., Borkowsky, W.,
Farthing, C., and Ho, D.D. (1994). Temporal association of cellular immune
responses with the initial control of viremia in primary human immunodefi-
ciency virus type 1 syndrome. J. Virol. 68, 4650–4655.
Maldarelli, F., Palmer, S., King, M.S., Wiegand, A., Polis, M.A., Mican, J.,
Kovacs, J.A., Davey, R.T., Rock-Kress, D., Dewar, R., et al. (2007). ART
suppresses plasma HIV-1 RNA to a stable set point predicted by pretherapy
viremia. PLoS Pathog. 3, e46.
Maldarelli, F., Wu, X., Su, L., Simonetti, F.R., Shao, W., Hill, S., Spindler, J.,
Ferris, A.L., Mellors, J.W., Kearney, M.F., et al. (2014). HIV latency. Specific
HIV integration sites are linked to clonal expansion and persistence of infected
cells. Science 345, 179–183.
Margolis, D.M., Kewn, S., Coull, J.J., Ylisastigui, L., Turner, D., Wise, H.,
Hossain, M.M., Lanier, E.R., Shaw, L.M., and Back, D. (2002). The addition
of mycophenolate mofetil to antiretroviral therapy including abacavir is associ-
ated with depletion of intracellular deoxyguanosine triphosphate and a
decrease in plasma HIV-1 RNA. J. Acquir. Immune Defic. Syndr. 31, 45–49.
Matano, T., Shibata, R., Siemon, C., Connors, M., Lane, H.C., and Martin, M.A.
(1998). Administration of an anti-CD8 monoclonal antibody interferes with the
clearance of chimeric simian/human immunodeficiency virus during primary
infections of rhesus macaques. J. Virol. 72, 164–169.
McMichael, A.J., and Phillips, R.E. (1997). Escape of human immunodeficiency
virus from immune control. Annu. Rev. Immunol. 15, 271–296.
Metzner, K.J., Jin, X., Lee, F.V., Gettie, A., Bauer, D.E., Di Mascio, M.,
Perelson, A.S., Marx, P.A., Ho, D.D., Kostrikis, L.G., and Connor, R.I. (2000).
Effects of in vivo CD8(+) T cell depletion on virus replication in rhesus
macaques immunized with a live, attenuated simian immunodeficiency virus
vaccine. J. Exp. Med. 191, 1921–1931.
Immunity 45, 656–668, September 20, 2016 667
Migueles, S.A., Osborne, C.M., Royce, C., Compton, A.A., Joshi, R.P., Weeks,
K.A., Rood, J.E., Berkley, A.M., Sacha, J.B., Cogliano-Shutta, N.A., et al.
(2008). Lytic granule loading of CD8+ T cells is required for HIV-infected cell
elimination associated with immune control. Immunity 29, 1009–1021.
Okoye, A., Park, H., Rohankhedkar, M., Coyne-Johnson, L., Lum, R., Walker,
J.M., Planer, S.L., Legasse, A.W., Sylwester, A.W., Piatak, M., Jr., et al. (2009).
Profound CD4+/CCR5+ T cell expansion is induced by CD8+ lymphocyte
depletion but does not account for accelerated SIV pathogenesis. J. Exp.
Med. 206, 1575–1588.
Peris-Pertusa, A., López, M., Rallón, N.I., Restrepo, C., Soriano, V., and
Benito, J.M. (2010). Evolution of the functional profile of HIV-specific CD8+
T cells in patients with different progression of HIV infection over 4 years.
J. Acquir. Immune Defic. Syndr. 55, 29–38.
Reimann, K.A., Tenner-Racz, K., Racz, P., Montefiori, D.C., Yasutomi, Y., Lin,
W., Ransil, B.J., and Letvin, N.L. (1994). Immunopathogenic events in acute
infection of rhesus monkeys with simian immunodeficiency virus of macaques.
J. Virol. 68, 2362–2370.
Roy, J., Paquette, J.S., Fortin, J.F., and Tremblay, M.J. (2002). The immuno-
suppressant rapamycin represses human immunodeficiency virus type 1 repli-
cation. Antimicrob. Agents Chemother. 46, 3447–3455.
Schaller, T., Ocwieja, K.E., Rasaiyaah, J., Price, A.J., Brady, T.L., Roth, S.L.,
Hué, S., Fletcher, A.J., Lee, K., KewalRamani, V.N., et al. (2011). HIV-1
capsid-cyclophilin interactions determine nuclear import pathway, integration
targeting and replication efficiency. PLoS Pathog. 7, e1002439.
Schmitz, J.E., Kuroda, M.J., Santra, S., Sasseville, V.G., Simon, M.A., Lifton,
M.A., Racz, P., Tenner-Racz, K., Dalesandro, M., Scallon, B.J., et al.
(1999a). Control of viremia in simian immunodeficiency virus infection by
CD8+ lymphocytes. Science 283, 857–860.
Schmitz, J.E., Simon, M.A., Kuroda, M.J., Lifton, M.A., Ollert, M.W., Vogel,
C.W., Racz, P., Tenner-Racz, K., Scallon, B.J., Dalesandro, M., et al.
(1999b). A nonhuman primate model for the selective elimination of CD8+ lym-
phocytes using a mouse-human chimeric monoclonal antibody. Am. J. Pathol.
154, 1923–1932.
Smedley, J., Turkbey, B., Bernardo, M.L., Del Prete, G.Q., Estes, J.D., Griffiths,
G.L., Kobayashi, H., Choyke, P.L., Lifson, J.D., and Keele, B.F. (2014).
Tracking the luminal exposure and lymphatic drainage pathways of intravagi-
nal and intrarectal inocula used in nonhuman primate models of HIV transmis-
sion. PLoS ONE 9, e92830.
Smith, S.A., Kilgore, K.M., Kasturi, S.P., Pulendran, B., Hunter, E., Amara,
R.R., and Derdeyn, C.A. (2015). Signatures in Simian Immunodeficiency
Virus SIVsmE660 Envelope gp120 Are Associated with Mucosal
668 Immunity 45, 656–668, September 20, 2016
Transmission but Not Vaccination Breakthrough in Rhesus Macaques.
J. Virol. 90, 1880–1887.
Sokolskaja, E., Olivari, S., Zufferey, M., Strambio-De-Castillia, C., Pizzato, M.,
and Luban, J. (2010). Cyclosporine blocks incorporation of HIV-1 envelope
glycoprotein into virions. J. Virol. 84, 4851–4855.
Streblow, D.N., Kitabwalla, M., Malkovsky, M., and Pauza, C.D. (1998).
Cyclophilin a modulates processing of human immunodeficiency virus
type 1 p55Gag: mechanism for antiviral effects of cyclosporin A. Virology
245, 197–202.
Taaffe, J., Chahroudi, A., Engram, J., Sumpter, B., Meeker, T., Ratcliffe, S.,
Paiardini, M., Else, J., and Silvestri, G. (2010). A five-year longitudinal analysis
of sooty mangabeys naturally infected with simian immunodeficiency virus re-
veals a slow but progressive decline in CD4+ T-cell count whose magnitude is
not predicted by viral load or immune activation. J. Virol. 84, 5476–5484.
Van Rompay, K.K., Singh, R.P., Pahar, B., Sodora, D.L., Wingfield, C., Lawson,
J.R., Marthas, M.L., and Bischofberger, N. (2004). CD8+-cell-mediated
suppression of virulent simian immunodeficiency virus during tenofovir treat-
ment. J. Virol. 78, 5324–5337.
Vanderford, T.H., Bleckwehl, C., Engram, J.C., Dunham, R.M., Klatt, N.R.,
Feinberg, M.B., Garber, D.A., Betts, M.R., and Silvestri, G. (2011). Viral CTL
escape mutants are generated in lymph nodes and subsequently become
fixed in plasma and rectal mucosa during acute SIV infection of macaques.
PLoS Pathog. 7, e1002048.
von Stockenstrom, S., Odevall, L., Lee, E., Sinclair, E., Bacchetti, P., Killian, M.,
Epling, L., Shao, W., Hoh, R., Ho, T., et al. (2015). Longitudinal Genetic
Characterization Reveals That Cell Proliferation Maintains a Persistent HIV
Type 1 DNA Pool During Effective HIV Therapy. J. Infect. Dis. 212, 596–607.
Wagner, T.A., McLaughlin, S., Garg, K., Cheung, C.Y., Larsen, B.B., Styrchak,
S., Huang, H.C., Edlefsen, P.T., Mullins, J.I., and Frenkel, L.M. (2014). HIV
latency. Proliferation of cells with HIV integrated into cancer genes contributes
to persistent infection. Science 345, 570–573.
Walker, C.M., Moody, D.J., Stites, D.P., and Levy, J.A. (1986). CD8+ lympho-
cytes can control HIV infection in vitro by suppressing virus replication.
Science 234, 1563–1566.
Wong, J.K., Strain, M.C., Porrata, R., Reay, E., Sankaran-Walters, S., Ignacio,
C.C., Russell, T., Pillai, S.K., Looney, D.J., and Dandekar, S. (2010). In vivo
CD8+ T-cell suppression of siv viremia is not mediated by CTL clearance of
productively infected cells. PLoS Pathog. 6, e1000748.
Yasutomi, Y., Reimann, K.A., Lord, C.I., Miller, M.D., and Letvin, N.L. (1993).
Simian immunodeficiency virus-specific CD8+ lymphocyte response in
acutely infected rhesus monkeys. J. Virol. 67, 1707–1711.
Immunity, Volume 45

Supplemental Information

CD8+ Lymphocytes Are Required for Maintaining

Viral Suppression in SIV-Infected Macaques
Treated with Short-Term Antiretroviral Therapy
Emily K. Cartwright, Lori Spicer, S. Abigail Smith, David Lee, Randy Fast, Sara
Paganini, Benton O. Lawson, Melon Nega, Kirk Easley, Joern E. Schmitz, Steven E.
Bosinger, Mirko Paiardini, Ann Chahroudi, Thomas H. Vanderford, Jacob D.
Estes, Jeffrey D. Lifson, Cynthia A. Derdeyn, and Guido Silvestri
Supplemental Experimental Procedures

Cell-associated SIVgag DNA

DNA was extracted from sorted peripheral and lymph node CD4+ T cells and CD4+ T cell memory subsets using the
Blood DNA Mini Kit (QIAGEN). Quantification of SIVmac gag DNA was performed as previously described on the
extracted cell-associated DNA by quantitative PCR using the 5′ nuclease (TaqMan) assay with an ABI7500 system
(PerkinElmer Life Sciences) (Chahroudi et al., 2014). The sequence of the forward primer for SIVmac gag was 5′-
GCAGAGGAGGAAATTACCCAGTAC-3′; the reverse primer sequence was 5′-
CAATTTTACCCAGGCATTTAATGTT-3′; and the probe sequence was 5′-6 FAM-
TGTCCACCTGCCATTAAGCCCGA-TAMRA-3′. For cell number quantification, quantitative PCR was
performed simultaneously for monkey albumin gene copy number.

Single genome analysis of circulating SIV

Viral RNA was extracted from plasma samples collected at Day 10, Day 56, Day 79 or Day 93, and various time
points between 1-6 weeks post CD8 depletion, using either the QIAamp RNA isolation kit or the QIAGEN
Ultrasens Virus kit (QIAGEN) following manufacturer’s instructions. Reverse transcription of viral RNA into
cDNA was performed using the Superscript III kit (Invitrogen) with SIVsm/macEnvR1 5′-
TGTAATAAATCCCTTCCAGTCCCCCC-3’. Single genome amplification by nested PCR was performed on each
cDNA using a limiting dilution that yielded one-third or less positive amplicons (Burton et al., 2015; Smith et al.,
2016). First-round PCR was performed in a 15 µl total volume using Phusion Hotstart II high-fidelity DNA
polymerase (Thermo Scientific) with the following primers: SIVsm/macEnvF1 5′-CCTCCCCCTCCAGGACTAGC-
3’, and SIVsm/macEnvR1 5′-TGTAATAAATCCCTTCCAGTCCCCCC-3′. Cycling conditions were as follows:
94°C for 2 min; 35 cycles of 94°C for 15 sec, 55°C for 30 s, and 68°C for 4 min; 68°C for 10 min; and 4°C hold.
Second-round PCR was performed with the same enzyme in a 10 µl volume using 1 µl of the first round PCR
product with SIVmacEnvF2 5′-TATAATAGACATGGAGACACCCTTGAGGGAGC-3′ and SIVsmEnvR2 5′-
ATGAGACATRTCTATTGCCAATTTGTA-3′. Cycling conditions were 94°C for 2 min; 45 cycles of 94°C for 15
s, 55°C for 30 s, and 68°C for 4 min; 68°C for 10 min; and 4°C hold. Positive bands were gel purified via QIAquick
Gel Extraction Kit (QIAGEN). Purified samples were sequenced with the following primers: SIVmac251seqF1 5’-
GGGATATGTTATGAGCAGTCACG-3’; SIVmac251seqF2 5’-ATCCAAGAGTCTTGTGACAAGC-3’;
SIVmac251seqF3 5’-AAGAGAGGGAGACCTCACG-3’; SIVmac251seqF4 5’-AGGCCAGTGTTCTCTTCC-3’;
SIVmac251seqR1 5’-CTTGTTCCAAGCCTGTGC-3’; SIVmac251seqR2 5’-CCTCTGCAATTTGTCCACATG-3’;
SIVmac251seqR3 5’-TCCAAGAAGTCAACCTTTCGC-3’; SIVmac251seqR4 5’-AGCTGGGTTTCTCCATGG-3’.
Sequencher v5 was used to generate nucleotide sequence contigs, and sequences with no ORF or evidence of mixed
peaks were omitted from the analysis. Geneious v6.1.7 was used to translate nucleotide sequences and create
alignments. Highlighter plots were generated with Highlighter for Amino Acids v1.3.4
(www.
hiv.lanl.gov/content/sequence/HIGHLIGHT/highlighter_top.html) (Keele et al., 2008). SGA derived Env amino
acid sequences have been submitted under Genbank accession numbers KX552042–KX552194.

Deep sequencing of cell-associated SIV-DNA

Deep-sequencing of two amplicons covering nucleotides 6679-7103 and 8748-9196 of the wildtype SIVmac239
genome (NCBI accession#: M33262.1) and corresponding to Env amino acids 35-158 and 724-845 were amplified
from peripheral blood cell DNA by nested PCR as previously described (Vanderford et al., 2011). Primers were
designed using each animal’s SGA plasma viral RNA sequences to avoid potential mismatches. For the pre-ART
amplicons, two different primer combinations were compared to rule out PCR or primer-induced bias in genotype
frequencies in our PCR protocol. Bands of the correct size were gel purified and libraries were generated from 10 ng
of amplicon DNA using the Kapa HyperPlus DNA kit. The amplicons were end-repaired, A-tailed and single index
barcodes and sequencing primers were added according to the manufacturer protocol. Libraries were assessed by
Bioanalyzer capillary electrophoresis, quantified, pooled and sequenced on an Illumina MiSeq system as 300 bp
paired-end reactions. Each sample was sequenced to an approximate depth of 1 million reads. Sequence reads were
extracted from the Illumina MiSeq run and samples were identified by barcoded adaptors. After adaptors were
trimmed, paired-end sequence reads were trimmed for quality and length using Trimmomatic (Bolger et al., 2014)
using the recommended parameters in paired-end mode. The inner nested PCR primers were identified near the
beginning of both reads and trimmed. Reads were then aligned to the original SIVmac239 genomic sequence using
BLAT (Kent, 2002). Reads were included in the analysis only if they (i) overlapped by more than 10 nucleotides,
(ii) contained both inner nested primers prior to trimming, (iii) contained no ambiguous base calls, and (iv) uniquely
aligned to the wildtype SIVmac239 genome without insertions or deletions. To account for the low level of
sequencing-induced mutations in the pool of reads, baseline amino acid mutation frequencies for each position in the
reconstructed amplicons were estimated at likely invariant sites. A conservative cutoff amino acid mutation
frequency of 0.01 was chosen under which a particular site was considered invariant within each sample. The
frequency of reads identical to the original SIVmac239 genomic sequence was then calculated allowing for
mutations only at variant sites. Similar results were obtained if all sites were allowed to vary in the analysis. All
sequence analysis was performed Python v3.4, except where noted.

Supplemental References

Bolger, A.M., Lohse, M., and Usadel, B. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data.
Bioinformatics 30, 2114-2120.
Burton, S.L., Kilgore, K.M., Smith, S.A., Reddy, S., Hunter, E., Robinson, H.L., Silvestri, G., Amara, R.R., and
Derdeyn, C.A. (2015). Breakthrough of SIV strain smE660 challenge in SIV strain mac239-vaccinated rhesus
macaques despite potent autologous neutralizing antibody responses. Proceedings of the National Academy of
Sciences of the United States of America 112, 10780-10785.

Chahroudi, A., Cartwright, E., Lee, S.T., Mavigner, M., Carnathan, D.G., Lawson, B., Carnathan, P.M.,
Hashempoor, T., Murphy, M.K., Meeker, T., et al. (2014). Target cell availability, rather than breast milk factors,
dictates mother-to-infant transmission of SIV in sooty mangabeys and rhesus macaques. PLoS pathogens 10,
e1003958.

Keele, B.F., Giorgi, E.E., Salazar-Gonzalez, J.F., Decker, J.M., Pham, K.T., Salazar, M.G., Sun, C., Grayson, T.,
Wang, S., Li, H., et al. (2008). Identification and characterization of transmitted and early founder virus envelopes
in primary HIV-1 infection. Proceedings of the National Academy of Sciences of the United States of America 105,
7552-7557.

Kent, W.J. (2002). BLAT--the BLAST-like alignment tool. Genome Res 12, 656-664.
Smith, S.A., Kilgore, K.M., Kasturi, S.P., Pulendran, B., Hunter, E., Amara, R.R., and Derdeyn, C.A. (2016).
Signatures in Simian Immunodeficiency Virus SIVsmE660 Envelope gp120 Are Associated with Mucosal
Transmission but Not Vaccination Breakthrough in Rhesus Macaques. Journal of virology 90, 1880-1887.

Vanderford, T.H., Bleckwehl, C., Engram, J.C., Dunham, R.M., Klatt, N.R., Feinberg, M.B., Garber, D.A., Betts,
M.R., and Silvestri, G. (2011). Viral CTL escape mutants are generated in lymph nodes and subsequently become
fixed in plasma and rectal mucosa during acute SIV infection of macaques. PLoS pathogens 7, e1002048.
Supplemental Figures and Tables
A Frequency D Lymph node E Lymph node
ART ART 100
50 80

%CD8+ of baseline
80

%CD8+ of CD3+
(of lymphocytes)
%CD8+ NK cells

40 60 60
30
40 40
20
20 20
10
0
0 0

NX
PD

PD
p
de
AB
6

8
-1

3
e-
AB

1
2
3

5
6
-1

D5
7
8

K
Pr

W
Weeks post depletion

B Absolute count
F G
ART Rectal biopsy
2000 Rectal biopsy
ART 100
Absolute number of

1500
100

%CD8+ of baseline
CD8+ NK cells

1000
80
%CD8+ of CD3+

500 80
200 60
60
150
100 40
40
50
0 20
20
AB

1
2
3

5
6
-1

7
8

0 0

NX
PD

PD
p
1

8
AB
D24
8

6
-1
V-

de
C
D1

D5
SI

3
Percent of baseline

e-

K
Weeks post depletion

Pr

W
ART
800
(% of pre-depletion)

600
Absolute count

400
200
100
80
60
40
20
0
AB

1
2
3

5
6
-1

7
8

Time post-depletion (weeks)

Figure S1. Related to Figure 1. MT-807R1 administration also depletes CD8α+ NK cells. Cells
were gated on lymphocytes, then CD3-CD20-. (A) Frequency of CD8α+ NK cells as a proportion
of total lymphocytes and (B) absolute count of CD8α+ NK cells in PBMC pre- and post-MT-
807R1 administration. (C) Absolute count of CD8+ NK cells as a proportion of pre-depletion
CD8α+ NK cell counts. CD8+ T cells in lymph nodes as a percentage of (D) CD3+ lymphocytes
and (E) percentage of pre-depletion frequencies (“baseline”). CD8+ T cells in rectal biopsies as
a percentage of (F) CD3+ lymphocytes and (G) percentage of pre-depletion frequencies
(“baseline”). Gap in x-axis represents 8-32 weeks of continuous ART. Black arrow and dotted
vertical line indicate MT-807R1 administration. NX=necropsy. Data shown is for 13 SIV-infected
ART treated rhesus macaques. (E, G) Bars drawn at the mean, error bars represent standard
error of mean (SEM).
 Activation of CD4+ T cells
PBMC Lymph node Rectal biopsy

A 100 ART
PD-1 B ART
PD-1 C PD-1
ART
80 100
%PD-1+ of CD4+

%PD-1+ of CD4+
80

%PD-1 of +CD4+
60 80
60 60
40
40 40
20
20 20
0 0 0

AB
AB

8
4
8

1
2
3

5
6
7
8

-1
-1
D1 -

AB
14
28

56

8
-1
V-
V

D5
D2

D5
SI

SI

D
D

D
Weeks post depletion Weeks post depletion Weeks post depletion

HLA-DR HLA-DR HLA-DR

ART ART ART
6 4 10
%HLA-DR+ of CD4+

%HLA-DR+ of CD4+
%HLA-DR+ of CD4+

8
3
4
6
2
2 4
1
2
0 0 0
AB
4
8

1
2
3

5
6
7
8
-1
D1 -

AB
6

8
-1
V

AB
4
8

8
-1
V-
D2

D5

D5

D1
D2

D5
SI

SI
Weeks post depletion Weeks post depletion Weeks post depletion

CCR5 ART
CCR5 CCR5
ART 20 ART
15 80
%CCR5+ of CD4+
%CCR5+ of CD4+

% CCR5+ of CD4+
15 60
10
10 40
5
5 20
†
† 0
0 0
AB

AB
4
8

1
2
3

5
6
7
8
-1

8
D1 -

-1

AB
4
8

8
V

-1
V-
D2

D5

D5

D1
D2

D5
SI

SI

Weeks post depletion Weeks post depletion Weeks post depletion

Figure S2. Related to Figure 1. Administration of anti-CD8 antibody MT-807R1 to SIV-infected

ART treated RM results in variable increases in CD4+ T cell activation. Expression of activation
markers PD-1, HLA-DR, and CCR5 on CD4+ T cells in (A) PBMC, (B) Lymph Node and (C)
Rectal Biopsy. Gap in x-axis represents 8-32 weeks of continuous ART. Data shown is for 13
SIV-infected ART treated RM. Black arrow and dotted vertical line indicate MT-807R1
administration.
 Pre-depletion GAG-CM9
A Week 1 B Week 3

3 4 r = 0.9643
r = 0.9286

ml of plasma (log10)
SIV RNA copies per
**p= 0.0028
ml of plasma (log10)
SIV RNA copies per

**p= 0.0067
3
2
2

1 1
0 2 4 6 8 10 0 2 4 6 8 10
#GAG-CM9+ CD8 T cells #GAG-CM9+ CD8 T cells
(per µL blood) (per µL blood)
Pre-depletion CD8+ NK cell
C Week 1 D Week 3
4 5
ml of plasma (log10)
SIV RNA copies per

ml of plasma (log10)
SIV RNA copies per

4
3
3
2
2

1 1
0 200 400 600 0 200 400 600
# CD8+NK cells # CD8+NK cells
(per µL blood) (per µL blood)

Figure S3. Related to Figure 2. Number of pre-depletion SIV-specific CD8+ T cells predicts post
depletion viral loads. Panels A-B: Presence of a significant direct correlation of number of GAG-
CM9 specific CD8+ T cells in PBMCs pre-depletion with plasma viral load at (A) week 1 and (B)
week 3 post-depletion. Data shown in (A) and (B) is for 7 SIV-infected ART treated Mamu-A*01
rhesus macques. Panels C-D: Lack of correlation of number of NK cells in PBMCs pre-depletion
with plasma viral load at (C) week 1 and (D) week 3 post-depletion. Data shown in (C) and (D)
is for 13 SIV-infected ART treated rhesus macaques. Spearman rank correlation, two tailed.
 A
ROw8
ART
SIV RNA copies per ml of plasma

Absolute number CD8+ NK cells

109 600

108

107

400
106

105

104
200

103

†
102

101 0

AB
4

24

25
11

13

15
16

18
19
20

22
23
Time post-SIV infection (weeks)

B Viral load
* NS
106
SIV RNA copies per

105
ml of plasma

104

103

102

101
1

3
d

d
rio

rio

rio
Pe

Pe

Pe

C
Persistently suppressed
* NS
104
SIV RNA copies per
ml of plasma

103

102

101
1

3
d

d
rio

rio

rio
Pe

Pe

Pe

Figure S4. Related to Figure 2. NK cell reconstitution does not coincide with viral load decline in
ART-treated SIV-infected RMs. (A) Viral load (red line) and CD8+ NK cell counts (black line) of
a representative SIV-infected RM on ART. Black arrow and dotted line indicates anti-CD8 MT-
807R1 administration. (B) Geometric mean viremia for each statistical period (NK cell) for all 13
animals. Bars show the geometric mean with 95% CI, *p<0.05. (C) Geometric mean viremia for
each NK cell statistical period for 7 persistently suppressed animals. Bars shown at the
geometric mean with 95% CI. Dashed horizontal line represents limit of detection of standard
viral load assay (60 copies/ml). Statistical analysis (B, C) was conducted using a linear, mixed
effects model, *p<0.05.
 A B
PBMC Lymph node
SCM SCM
SIV DNA copies per 106cells

SIV DNA copies per 106cells

106 NS 106 NS

105 105

104 104

103 103

102 102

101 101
Pre-dep Post-dep Pre-dep Post-dep

CM SIV DNA copies per 106cells

CM
SIV DNA copies per 106cells

106 NS 106 NS

105 105

104 104

103 103

102 102

101 101
Pre-dep Post-dep Pre-dep Post-dep
TM TM
SIV DNA copies per 106cells

SIV DNA copies per 106cells

106 NS 106 NS

105 105

104 104

103 103

102 102

101 101
Pre-dep Post-dep Pre-dep Post-dep
EM EM
SIV DNA copies per 106cells
SIV DNA copies per 106cells

106 NS 106 NS

105 105

104 104

103 103

102 102

101 101
Pre-dep Post-dep Pre-dep Post-dep

Figure S5. Related to Figure 5. No significant changes in total cell-associated SIV-DNA in

sorted memory CD4+ T cell subsets of SIV-infected ART-treated RM before and after CD8
depletion. Fraction of SIV-infected sorted TSCM, TCM, TTM and TEM (as defined in Methods) in (A)
PBMC (data shown for all animals with >10,000 sorted cells (TSCM=6, TCM=6, TTM=11, TEM=11)
and (B) lymph node (data shown for all animals with >10,000 sorted cells (TSCM=6, TCM=6,
TTM=11, TEM=11). Dashed horizontal line represents limit of detection of SIV DNA assay (100
copies/106 cells). Statistical analysis was conducted using the Wilcoxon matched-pairs signed
rank test.
Figure S6. Related to Figure 6. ART adminstration results in an increase in the frequency of
virus bearing the SIVmac239 inoculum amino acid sequence. Two amplicons covering Env
amino acids 35-158 and 724-845 were amplified from peripheral blood cell DNA and sequenced
via Illumina MiSeq to verify the presence of amino acid mutations identified in the circulating
plasma virus. The frequency of the wildtype sequence was estimated for each sample prior to
ART administration and just prior to CD8 depletion.
Number of ART Regimen Duration (post-ART
animals initiation)

3 20 mg/kg PMPA s.c., 30 mg/kg FTC s.c., 100 mg From 0 to 16 weeks

Raltegravir and 400 mg Darunavir orally b.i.d.

5 20 mg/kg PMPA s.c., 30 mg/kg FTC s.c., 100 mg From 0 to 8 weeks

Raltegravir and 400 mg Darunavir orally b.i.d.

followed by

20 mg/kg PMPA s.c., 30 mg/kg FTC s.c., 150 mg From 8 to 24 weeks

Raltegravir and 600 mg Darunavir orally b.i.d.

5 20 mg/kg PMPA s.c., 30 mg/kg FTC s.c., 100 mg From 0 to 8 weeks

Raltegravir and 400 mg Darunavir orally b.i.d.

followed by

20 mg/kg PMPA s.c., 30 mg/kg FTC s.c., 150 mg From 8 to 24 weeks

Raltegravir and 600 mg Darunavir orally b.i.d.

followed by

20 mg/kg PMPA s.c., 30 mg/kg FTC s.c., 150 mg From 24 to 40 weeks

Raltegravir and 800 mg Darunavir orally b.i.d.

Table S1. Related to Figure 1. Antiretroviral drug regimen and length of treatment for SIV-
infected ART-treated RM. All 13 animals were treated for at least 16 weeks. After 8 weeks of
continuous treatment, 10 of the animals were given increased doses of Raltegravir and
Darunavir. After another 16 weeks of increased Raltegravir and Darunavir, 5 animals received
a further increased dose of Darunavir.
 Magnitude of SIV suppression during ART

Post-ART (Pre CD8

Pre-ART Depl.) % Suppression
(copies/ml of plasma) (copies/ml of plasma)
RGb13 433 Below 3 100

RLb13 233,000 7 99.992

RVy10 2,940,000 10 99.998

ROw8 124,000 15 99.985

RKq11 599,000 25 99.993

RBv13 1,580,000 65 99.994

RWj14 496,000 90 99.979

RSj14 921,000 80 99.989

RDh10 3,400,000 90 99.997

RLc10 1,940,000 160 99.985

RAz12 2,270,000 300 99.982

RYf14 5,870,000 330 99.991

Table S2. Related to Figure 1 and 2. Magnitude of virus suppression during ART, prior to CD8+
lymphocyte depletion is >99.9% in all animals where at least 1 undetectable time point was
achieved. Animals in red were “persistently suppressed”, animals in black were “intermittently
suppressed.”
 Code Period 1 Period 2 Period 3

RGb13 6 weeks 3 weeks 4 weeks

RLb13 6 weeks 3 weeks 4 weeks

RVy10 6 weeks 3 weeks 1 week

RKq11 6 weeks 7 weeks 1 week

RBv13 6 weeks 3 weeks 4 weeks

RWj14 6 weeks 3 weeks 4 weeks

RSj14 6 weeks 3 weeks 4 weeks

RYf14 6 weeks 6 weeks 2 weeks

RAz12 6 weeks 5 weeks 3 weeks

RLc10 6 weeks 7 weeks n/a

ROw8 6 weeks 7 weeks n/a

RDh10 6 weeks 8 weeks n/a

ROn13 6 weeks 8 weeks n/a

Table S3. Related to Figure 2. Length of the periods 1-3 used for statistical analysis as
described in Figure 2 for each SIV-infected ART-treated RM. Period 1 is pre-CD8 depletion
(blue), Period 2 is post-CD8 depletion, pre-CD8 T cell recovery (green), and Period 3 is during
CD8 T cell recovery (purple). CD8+ T-cell reconstitution above 20% of baseline (i.e., pre-
depletion) levels was not acheived in four animals (RLc10, ROw8, RDh10, and ROn13).
 Subset Time Point Tissue p

Total CD4 WK 1 Post Depletion PBMC NS

CD95+ Memory WK 1 Post Depletion PBMC NS

TSCM WK 1 Post Depletion PBMC NS

TCM WK 1 Post Depletion PBMC NS

TTM WK 1 Post Depletion PBMC NS

TEM WK 1 Post Depletion PBMC NS

Total CD4 WK 1 Post Depletion Lymph node NS

CD95+ Memory WK 1 Post Depletion Lymph Node NS
TSCM WK 1 Post Depletion Lymph Node NS
.
TCM WK 1 Post Depletion Lymph Node NS
TTM WK 1 Post Depletion Lymph Node NS

TEM WK 1 Post Depletion Lymph Node NS

Total CD4 WK 1 Post Depletion Rectal Biopsy NS

CD95+ Memory WK 1 Post Depletion Rectal Biopsy NS

Total CD4 WK 3 Post Depletion PBMC NS

CD95+ Memory WK 3 Post Depletion PBMC NS

TSCM WK 3 Post Depletion PBMC NS

TCM WK 3 Post Depletion PBMC NS

TTM WK 3 Post Depletion PBMC NS

TEM WK 3 Post Depletion PBMC NS

Total CD4 WK 3 Post Depletion Lymph node NS

CD95+ Memory WK 3 Post Depletion Lymph Node NS

TSCM WK 3 Post Depletion Lymph Node NS

TCM WK 3 Post Depletion Lymph Node NS

TTM WK 3 Post Depletion Lymph Node NS

TEM WK 3 Post Depletion Lymph Node NS

Total CD4 WK 3 Post Depletion Rectal Biopsy NS
CD95+ Memory WK 3 Post Depletion Rectal Biopsy NS

Table S4. Related to Figure 5. No correlation between the frequency of proliferating (Ki67+)
CD4+ T-cells and plasma viral load at weeks 1 and 3 post-CD8 depletion. Spearman rank
correlation.
 Time Point (Post Depletion) Tissue r, p
CCR5+CD4+ Day 1 PBMC NS
Day 2 PBMC NS
Week 1 PBMC NS
Week 1 Lymph Node NS
Week 3 Lymph Node NS
Week 1 Rectal Biopsy NS
Week 3 Rectal Biopsy NS
PD-1+CD4+ Day 1 PBMC r=0.6599, p=0.0120
Day 2 PBMC r=0.5144, p=0.0120
Week 1 PBMC NS
Week 1 Lymph Node NS
Week 3 Lymph Node NS
Week 1 Rectal Biopsy NS
Week 3 Rectal Biopsy NS
HLA-DR+CD4+ Day 1 PBMC r=0.6308, p=0.0163
Day 2 PBMC NS
Week 1 PBMC NS
Week 1 Lymph Node NS
Week 3 Lymph Node NS
Week 1 Rectal Biopsy r=0.6099, p=0.0151
Week 3 Rectal Biopsy NS

Table S5. Related to Figure 5. Correlation between PD-1+CD4+ and HLA-DR+CD4+ T cells in
PBMC and RB and plasma viral load at early time points (Day 1, Day 2, and Week 1) post-
depletion. Spearman rank correlation.