Fc-Optimized Anti-CD25 Depletes Tumor-Infiltrating Regulatory T Cells and Synergizes with PD-1 Blockade to Eradicate Established Tumors

Abstract

CD25 is expressed at high levels on regulatory T (Treg) cells and was initially proposed as a target for cancer immunotherapy. However, anti-CD25 antibodies have displayed limited activity against established tumors. We demonstrated that CD25 expression is largely restricted to tumor-infiltrating Treg cells in mice and humans. While existing anti-CD25 antibodies were observed to deplete Treg cells in the periphery, upregulation of the inhibitory Fc gamma receptor (FcγR) IIb at the tumor site prevented intra-tumoral Treg cell depletion, which may underlie the lack of anti-tumor activity previously observed in pre-clinical models. Use of an anti-CD25 antibody with enhanced binding to activating FcγRs led to effective depletion of tumor-infiltrating Treg cells, increased effector to Treg cell ratios, and improved control of established tumors. Combination with anti-programmed cell death protein-1 antibodies promoted complete tumor rejection, demonstrating the relevance of CD25 as a therapeutic target and promising substrate for future combination approaches in immune-oncology.

Full text

Report
Fc-Optimized Anti-CD25 Depletes Tumor-Infiltrating
Regulatory T Cells and Synergizes with PD-1
Blockade to Eradicate Established Tumors
Graphical Abstract
Highlights
dCD25 expression is largely restricted to Treg cells in mice and
humans
dFcgRIIb inhibits anti-CD25-mediated depletion of intra-
tumoral Treg cells
dFc-optimized anti-CD25 efficiently depletes intra-tumoral
Treg cells
dAnti-CD25 synergizes with PD-1 blockade to reject
established tumors
Authors
Frederick Arce Vargas,
Andrew J.S. Furness,
Isabelle Solomon, ..., Teresa Marafioti,
Karl S. Peggs, Sergio A. Quezada
Correspondence
k.peggs@ucl.ac.uk (K.S.P.),
s.quezada@ucl.ac.uk (S.A.Q.)
In Brief
Anti-CD25 antibodies have displayed
only modest therapeutic activity against
established tumors. Arce Vargas et al.
demonstrate that existing anti-CD25
antibodies fail to deplete intra-tumoral
Treg cells due to upregulation of FcgRIIb
within tumors. Fc-optimized anti-CD25
mediates effective depletion of tumor-
infiltrating Treg cells and synergizes with
PD-1 blockade to promote tumor
eradication.
Arce Vargas et al., 2017, Immunity 46, 577–586
April 18, 2017 ª2017 Elsevier Inc.
http://dx.doi.org/10.1016/j.immuni.2017.03.013

Immunity
Report
Fc-Optimized Anti-CD25 Depletes Tumor-Infiltrating
Regulatory T Cells and Synergizes with PD-1
Blockade to Eradicate Established Tumors
Frederick Arce Vargas,
1,2,11
Andrew J.S. Furness,
1,2,3,11
Isabelle Solomon,
1,2
Kroopa Joshi,
1,2,3
Leila Mekkaoui,
4
Marta H. Lesko,
1,2
Enrique Miranda Rota,
4
Rony Dahan,
5
Andrew Georgiou,
1,2
Anna Sledzinska,
1,2
Assma Ben Aissa,
1,2
Dafne Franz,
1,2
Mariana Werner Sunderland,
1,2
Yien Ning Sophia Wong,
1,2
Jake Y. Henry,
1,2
Tim O’Brien,
6
David Nicol,
3
Ben Challacombe,
6
Stephen A. Beers,
7
Melanoma TRACERx Consortium, Renal TRACERx Consortium,
Lung TRACERx Consortium, Samra Turajlic,
3,8
Martin Gore,
3
James Larkin,
3
Charles Swanton,
8,9
Kerry A. Chester,
4
Martin Pule,
2
Jeffrey V. Ravetch,
5
Teresa Marafioti,
10
Karl S. Peggs,
1,2,
*and Sergio A. Quezada
1,2,12,
*
1
Cancer Immunology Unit, University College London Cancer Institute, London WC1E 6DD, UK
2
Research Department of Haematology, UCL Cancer Institute, London WC1E 6DD, UK
3
The Royal Marsden NHS Foundation Trust, London SW3 6JJ, UK
4
Research Department of Oncology, UCL Cancer Institute, London WC1E 6DD, UK
5
Leonard Wagner Laboratory of Molecular Genetics and Immunology, The Rockefeller University, New York, NY 10065, USA
6
Guy’s and St. Thomas’ NHS Foundation Trust, London SE1 9RT, UK
7
Antibody and Vaccine Group, Cancer Sciences Unit, University of Southampton, Faculty of Medicine, Southampton SO17 1BJ, UK
8
The Francis Crick Institute, London NW1 1AT, UK
9
Translational Cancer Therapeutics Laboratory, UCL Cancer Institute, London WC1E 6DD, UK
10
Department of Cellular Pathology, University College London Hospital, London NW1 2BU, UK
11
These authors contributed equally
12
Lead Contact
*Correspondence: k.peggs@ucl.ac.uk (K.S.P.), s.quezada@ucl.ac.uk (S.A.Q.)
http://dx.doi.org/10.1016/j.immuni.2017.03.013
SUMMARY
CD25 is expressed at high levels on regulatory
T (Treg) cells and was initially proposed as a target
for cancer immunotherapy. However, anti-CD25 anti-
bodies have displayed limited activity against estab-
lished tumors. We demonstrated that CD25 expres-
sion is largely restricted to tumor-infiltrating Treg
cells in mice and humans. While existing anti-CD25
antibodies were observed to deplete Treg cells in
the periphery, upregulation of the inhibitory Fc
gamma receptor (FcgR) IIb at the tumor site pre-
vented intra-tumoral Treg cell depletion, which may
underlie the lack of anti-tumor activity previously
observed in pre-clinical models. Use of an anti-
CD25 antibody with enhanced binding to activating
FcgRs led to effective depletion of tumor-infiltrating
Treg cells, increased effector to Treg cell ratios,
and improved control of established tumors. Com-
bination with anti-programmed cell death protein-1
antibodies promoted complete tumor rejection,
demonstrating the relevance of CD25 as a therapeu-
tic target and promising substrate for future combi-
nation approaches in immune-oncology.
INTRODUCTION
Regulatory T (Treg) cells are generally regarded as one of the
major obstacles to the successful clinical application of tumor
immunotherapy. It has been consistently demonstrated that
Treg cells contribute to the early establishment and progression
of tumors in murine models and that their absence results in
delay of tumor progression (Elpek et al., 2007; Golgher et al.,
2002; Jones et al., 2002; Onizuka et al., 1999; Shimizu et al.,
1999). In humans, high tumor infiltration by Treg cells and,
more importantly, a low ratio of effector T (Teff) cells to Treg cells,
is associated with poor outcomes in multiple solid cancers
(Shang et al., 2015). Conversely, a high Teff/Treg cell ratio is
associated with favorable responses to immunotherapy in both
humans and mice (Hodi et al., 2008; Quezada et al., 2006). To
date, most studies support the notion that targeting Treg cells,
either by depletion or functional modulation, may offer significant
therapeutic benefit, particularly in combination with other im-
mune modulatory interventions such as vaccines and check-
point blockade (Bos et al., 2013; Goding et al., 2013; Quezada
et al., 2008; Sutmuller et al., 2001).
Defining appropriate targets for selective interference with
Treg cells is therefore a critical step in the development of effec-
tive therapies. In this regard, CD25, also known as the inter-
leukin-2 high-affinity receptor alpha chain (IL-2Ra), was the first
surface marker used to identify and isolate Treg cells (Sakaguchi
et al., 1995) prior to the discovery of their master regulator, tran-
scription factor forkhead box P3 (FoxP3). It is also the most exten-
sively studied target for mediating Treg cell depletion. Whereas
CD25 is constitutively expressed on Treg cells and absent on
naive Teff cells, transient upregulation has been described
upon activation of Teff cells, although these observations derive
largely from in vitro studies (Boyman and Sprent, 2012).
A number of pre-clinical studies in mice have used the anti-
CD25 antibody clone PC-61 (rat IgG1, l), which partially depletes
Treg cells in the blood and peripheral lymphoid organs (Setiady
Immunity 46, 577–586, April 18, 2017 ª2017 Elsevier Inc. 577

et al., 2010), inhibits tumor growth, and improves survival when
administered before or soon after tumor challenge (Golgher
et al., 2002; Jones et al., 2002; Onizuka et al., 1999; Quezada
et al., 2008; Shimizu et al., 1999). However, the use of anti-
CD25 as a therapeutic intervention against established tumors
fails to delay tumor growth or prolong survival (Golgher et al.,
2002; Jones et al., 2002; Onizuka et al., 1999; Shimizu et al.,
1999). This has been attributed to several factors, including
poor T cell infiltration of the tumor (Quezada et al., 2008) and po-
tential depletion of activated effector CD8
+
and CD4
+
T cells that
upregulate CD25 (Onizuka et al., 1999). Early-phase clinical
studies exploring the use of vaccines in combination with dacli-
zumab (a humanized IgG1 anti-human CD25 antibody) (Jacobs
et al., 2010; Rech et al., 2012) or denileukin difitox (a recombinant
fusion protein combining human IL-2 and a fragment of diptheria
toxin) (Dannull et al., 2005; Luke et al., 2016) demonstrate a var-
iable impact on the number of circulating Treg cells and vaccine-
induced immunity. However, the limited indirect data assessing
intra-tumoral FoxP3 transcript levels provide no clear evidence
that Treg cells in the tumor microenvironment are effectively
reduced and anti-tumor activity has appeared disappointing
across all studies, with no demonstrable survival benefit.
The modest therapeutic activity in pre-clinical and clinical set-
tings and concern regarding potential depletion of activated Teff
cells has contributed to limited enthusiasm for the further evalu-
ation of anti-CD25 antibodies in combination with novel immu-
notherapies. However, recent data demonstrate the contribution
of intra-tumoral Treg cell depletion to the activity of immune
modulatory antibody-based therapies and the relevance of the
antibody isotype in this setting (Bulliard et al., 2014; Coe et al.,
2010; Selby et al., 2013; Simpson et al., 2013). We therefore
re-evaluated CD25 as target for Treg cell depletion and tumor
immunotherapy in vivo. We demonstrated that the lack of ther-
apeutic activity of the widely used anti-CD25 antibody (PC-61)
against established mouse tumors results from a failure to effec-
tively deplete intra-tumoral Treg cells. Optimizing FcgR binding
and antibody-dependent cell-mediated cytotoxicity (ADCC)
resulted in superior intra-tumoral Treg cell depletion and potent
synergy when combined with programmed cell death protein-1
(PD-1) blockade. We demonstrated high levels of CD25
expression on Treg but not Teff cells in human tumors,
highlighting this receptor as a clinical target and anti-CD25 as
a promising therapeutic strategy in combination with novel
immunotherapies.
RESULTS
CD25 Is Highly Expressed on Murine Tumor-Infiltrating
Treg Cells
We sought to evaluate the relative expression of CD25 on indi-
vidual T lymphocyte subsets within tumors (TILs) and draining
lymph nodes (LNs) of mice 10 days after tumor challenge.
CD25 expression appeared consistent across multiple models
of transplantable tumor cell lines of variable immunogenicity
including MCA205 sarcoma, MC38 colon adenocarcinoma,
B16 melanoma, and CT26 colorectal carcinoma, with a higher
percentage of CD25-expressing CD4
+
FoxP3
+
Treg cells
relative to CD4
+
FoxP3

and CD8
+
Teff cells (Figure 1A). In
contrast to in vitro studies, minimal expression of CD25 on the
Teff cell compartment was observed in vivo and the per-
centage of CD25-expressing Teff cells (CD8
+
= 3.08%–8.35%,
CD4
+
FoxP3

= 14.11%–26.87%) was significantly lower than
on Treg cells (83.66%–90.23%) (p < 0.001) (Figure 1B). CD25
expression was also observed on Treg cells present in LNs
and blood (data not shown). However, the level of expression,
based on mean fluorescence intensity (MFI), was significantly
lower than that observed on tumor-infiltrating Treg cells (Fig-
ure 1C). Based on these data, CD25 appeared an attractive
target for preferential depletion of Treg cells.
Anti-CD25-Mediated Depletion of Treg Cells Is Limited
to Lymph Nodes and Blood
Based on evidence demonstrating the contribution of intra-tu-
moral Treg cell depletion to the activity of immune modulatory
antibodies (Bulliard et al., 2014; Coe et al., 2010; Selby et al.,
2013; Simpson et al., 2013), we sought to compare the impact
of anti-CD25 (clone PC-61 rat IgG1, aCD25-r1) on the frequency
of Teff and Treg cells in the blood, LNs, and TILs of mice with es-
tablished tumors. We focused our analyses on the MCA205
model because of its higher immunogenicity in order to deter-
mine any potential negative impact of aCD25 on activated Teff
cells within tumors.
As previously described (Onizuka et al., 1999; Setiady et al.,
2010), administration of 200 mgofaCD25-r1 on days 5 and 7 after
tumor challenge resulted in a reduced frequency of CD25
+
cells
in all analyzed sites (Figures 1D and 1E) and a reduction in the
frequency of CD4
+
FoxP3
+
Treg cells in blood and LN (Figure 1F).
However, aCD25-r1 failed to deplete tumor-infiltrating Treg cells,
which demonstrated a CD4
+
FoxP3
+
CD25

phenotype after
therapy. Their frequency remained comparable to that of un-
treated mice (Figure 1F), potentially explaining the lack of effi-
cacy observed against established tumors in previous studies
despite an apparent reduction in CD25
+
T cells within the tumor
(Golgher et al., 2002; Jones et al., 2002; Onizuka et al., 1999;
Quezada et al., 2008; Shimizu et al., 1999).
We next investigated whether an antibody with optimized
ADCC activity could efficiently deplete intra-tumoral Treg cells
without significant impact on Teff cells. We replaced the
constant regions of the original aCD25 obtained from clone
PC-61 with murine IgG2a and kconstant regions (aCD25-
m2a), the classical mouse isotype associated with ADCC,
and compared its activity to that of aCD25-r1 in vivo. While
both antibody variants resulted in reduced expression of
CD25 on T cells and a reduction in the number of Treg cells
in blood and LNs, only aCD25-m2a resulted in depletion of
tumor-infiltrating Treg cells to levels comparable to those
observed with anti-cytotoxic T lymphocyte associated pro-
tein-4 (aCTLA-4, clone 9H10), which is known to preferentially
deplete Treg cells in the tumor but not the periphery (Figures
1D–1F; Selby et al., 2013; Simpson et al., 2013). In keeping
with these observations, both aCD25 isotypes resulted in an
increased Teff/Treg cell ratio in circulating lymphocytes and
LN, but only aCD25-m2a increased the intra-tumoral ratio in
a similar manner to aCTLA-4 (Figure 1G). Despite a reduction
in the number of circulating and LN-resident Treg cells, no
macroscopic, microscopic, or biochemical evidence of toxicity
was observed in the skin, lungs, or liver after multiple doses of
aCD25-m2a (Figures S1A–S1C).
578 Immunity 46, 577–586, April 18, 2017

High Expression of FcgRIIb Inhibits aCD25-r1-Mediated
Treg Cell Depletion in the Tumor
Anti-CD25-r1 has been described to deplete circulating Treg
cells by FcgRIII-mediated ADCC (Setiady et al., 2010). However,
its intra-tumoral activity has not been investigated. To determine
this, we characterized the expression of Fc-gamma receptors
(FcgRs) on different leukocyte subpopulations in the blood,
spleen, LN, and tumor of mice bearing MCA205 tumors
A
B
C
D
E
F
G
Figure 1. Anti-CD25-r1-Mediated Depletion of CD25
+
Regulatory T Cells Is Restricted to Blood and Lymph Nodes
(A–C) Mouse LNs and TILs were analyzed by flow cytomery 10 days after MCA205 (n = 10), MC38 (n = 5), B16 (n = 3), or CT26 (n = 3) tumor implantation.
(A) CD25 expression on T cell subsets in representative mice. Dotted lines indicate the gate.
(B and C) Percentage (B) and MFI (C) of CD25 in each T cell subset. Error bars show standard error of the mean (SEM). p values obtained by two-way analysis of
variance (ANOVA).
(D–G) Tumor-bearing mice were injected with 200 mgofaCD25-r1, aCD25-m2a, or aCTLA-4 on days 5 and 7 after MCA205 tumor implantation. Blood, LNs, and
TILs were harvested and processed on day 9 for flow cytometry analysis.
(D) Representative plots showing expression of CD25 (detected with antibody clone 7D4) and FoxP3 in CD3
+
CD4
+
T cells. Numbers show percentage of cells in
each quadrant.
(E) MFI of CD25 in CD4
+
FoxP3
+
Treg cells.
(F) Percentage of FoxP3
+
Treg cells of total CD3
+
CD4
+
T cells.
(G) CD8
+
/Treg cell ratios (n = 10). Experiment was repeated three times.
Immunity 46, 577–586, April 18, 2017 579

(Figures 2A and S2). The percentage of FcgR-expressing cells
appeared higher on tumor-infiltrating myeloid cells (granulocytic
cells, dendritic cells, and monocyte/macrophages) relative to all
other studied organs (Figures 2A and 2B). We then analyzed the
binding affinity of the two Fc variants of aCD25 to FcgRs (Fig-
ure 2C). As previously described (Nimmerjahn and Ravetch,
2005), the mIgG2a isotype binds to all FcgR subtypes with a
high activatory to inhibitory ratio (A/I). In contrast, the rIgG1
isotype binds with a similar affinity to a single activatory FcgR,
FcgRIII, as well as the inhibitory FcgRIIb, resulting in a low A/I
ratio (<1) (Figure 2C).
To determine which specific FcgRs were involved in aCD25-
mediated Treg cell depletion, we quantified the number of tu-
mor-infiltrating Treg cells in mice lacking expression of different
FcgRs (Figures 2D–2G). Analysis of Fcer1g
/
mice, which lack
expression of activating FcgRs (I, III, and IV), demonstrated a
complete absence of Treg cell depletion. Treg cell elimination
by aCD25-r1 in the periphery and by aCD25-m2a in the periph-
ery and tumor therefore results from FcgR-mediated ADCC and
not blocking of IL-2 binding to CD25 (Figure 2D). Depletion by
aCD25-m2a was not dependent on any individual activatory
FcgR, with Treg cell elimination maintained in both Fcgr3
/
and Fcgr4
/
mice (Figures 2E and 2F). In keeping with previous
studies (Setiady et al., 2010), we confirmed that depletion of pe-
ripheral Treg cells by aCD25-r1 depends on FcgRIII (data not
shown), but it fails to deplete in the tumor despite high intra-tu-
moral expression of this receptor (Figure 2E). Intra-tumoral
Treg cell depletion was, however, effectively restored in mice
AB
CGDFE
Figure 2. FcgRIIb Inhibits aCD25-r1-Mediated Treg Cell Depletion in Tumors
(A and B) Expression of FcgRs was measured by flow cytometry in leukocytes from blood, spleen, LNs, and MCA205 tumors (TIL) 10 days after tumor im-
plantation.
(A) Expression of FcgRs on granulocytes (CD11b
+
Ly6G
+
), conventional dendritic cells (cDCs) (CD11c
hi
MHC-II
+
), and monocyte/macrophages (Mono/M4)
(CD11b
+
Ly6G

NK1.1

CD11c
lo/neg
). Dotted lines indicate the gate, numbers show the percentage of positive cells.
(B) Cumulative data of FcgR expression in cell subpopulations (n = 3). Error bars represent SEM; the experiment was repeated three times.
(C) Binding affinity of rat IgG1 and mouse IgG2a isotypes to individual mouse FcgRs as determined by surface plasmon resonance (SPR).
(D–G) Percentage of CD4
+
FoxP3
+
Treg cells of total CD4
+
T cells in TILs of wild-type (WT, n = 5–10), Fcer1g
/
(n = 10), Fcgr3
/
(n = 5), Fcgr4
/
(n = 10), or
Fcgr2b
/
(n = 5) mice treated as in Figures 1D–1G.
580 Immunity 46, 577–586, April 18, 2017

A
DCB
EF
HG
Figure 3. Synergistic Effect of Anti-CD25-m2a and Anti-PD-1 Combination Results in Eradication of Established Tumors
Tumor-bearing mice were treated with 200 mgofaCD25 on day 5 and 100 mgofaPD-1 on days 6, 9, and 12 after tumor implantation.
(A) Growth curves of individual MCA205 tumors, showing the product of three orthogonal tumor diameters. The number of tumor-free survivors is shown in
each graph.
(legend continued on next page)
Immunity 46, 577–586, April 18, 2017 581

lacking expression of the inhibitory receptor FcgRIIb. In this
setting, intra-tumoral Treg cell depletion was comparable be-
tween aCD25-r1 and aCD25-m2a (Figure 2G). Therefore, the
lack of Treg cell depletion by aCD25-r1 in the tumor is explained
by its low A/I binding ratio and high intra-tumoral expression of
FcgRIIb. FcgRIIb has been associated with modulation of
ADCC in tumors (Clynes et al., 2000), and in this case inhibits
ADCC mediated by the single activatory receptor engaged by
the aCD25-r1 isotype.
Anti-CD25-m2a Synergizes with Anti-PD-1 to Eradicate
Established Tumors
To determine whether the enhanced intra-tumoral Treg cell-
depleting activity of aCD25-m2a could improve therapeutic out-
comes, we compared the anti-tumor activity of aCD25-m2a and
-r1 against established tumors. We administered a single dose
of aCD25 5 days after subcutaneous implantation of MCA205
cells, when tumors were established with an average diameter
of 4–5 mm. Consistent with the observed lack of capacity to
deplete intra-tumoral Treg cells (Figure 1F) and previous studies
(Golgher et al., 2002; Jones et al., 2002; Onizuka et al., 1999;
Quezada et al., 2008; Shimizu et al., 1999), aCD25-r1 failed to
control tumor growth. Conversely, growth delay and long-term
survival was observed in a proportion of mice receiving
aCD25-m2a (15.4%) (Figures 3A and 3B).
Based on its role in T cell regulation within the tumor microen-
vironment and the observed clinical activity of agents targeting
the PD-1-PD-L1 axis, we hypothesized that depletion of CD25
+
Treg cells and PD-1 blockade might be synergistic in combina-
tion. In the same model, blocking anti-PD-1 antibody (aPD-1,
clone RMP1-14) at a dose of 100 mg every 3 days was ineffective
in the treatment of established MCA205 tumors when used as
monotherapy or in combination with aCD25-r1 (Figures 3A and
3B). However, a single dose of aCD25-m2a followed by aPD-1
therapy eradicated established tumors in 78.6% of the mice, re-
sulting in long-term survival of more than 100 days (Figures 3A
and 3B). This activity was significantly reduced in the absence
of CD8
+
T cells (Figures S3A and S3B), demonstrating that tumor
elimination depends on the impact of the aPD-1 and aCD25
combination on both CD8
+
and Treg cell compartments, and
that overall effector T cell responses are not negatively impacted
by a depleting aCD25 antibody.
Similar findings were observed in MC38 and CT26 tumor
models, where aCD25-m2a had a partial therapeutic effect
that synergized with aPD-1 therapy (Figures 3C and 3D). Ac-
tivity was also observed against the poorly immunogenic
B16 melanoma tumor model when aCD25-m2a and aPD-1
were combined with a granulocyte-macrophage colony stimu-
lating factor (GM-CSF)-expressing whole tumor cell vaccine
(Gvax). As previously described, in this system, Gvax alone
failed to extend survival of tumor-bearing mice (Quezada
et al., 2006; van Elsas et al., 2001). Combination therapy
with aCD25-m2a and aPD-1 translated into a modest increase
in survival, which was not observed with aCD25-r1 and aPD-1
(Figure S4).
To understand the mechanisms underpinning the observed
synergy, we evaluated the phenotype and function of TILs in
MCA205 tumors at the end of the treatment protocol, 24 hr after
the third dose of aPD-1 (Figures 3E–3H). Monotherapy with
aPD-1 did not impact upon Teff cell proliferation (Figure 3E)
nor the number infiltrating the tumor, where a persisting high fre-
quency of Treg cells was observed (data not shown), resulting in
a low Teff/Treg ratio (Figure 3F) and lack of therapeutic activity.
Conversely, intra-tumoral Treg cell depletion with aCD25-m2a
resulted in a higher proportion of proliferating and interferon-g
(IFN-g)-producing CD4
+
and CD8
+
T cells in the tumor, corre-
sponding to a high Teff/Treg cell ratio and anti-tumor activity
(Figures 3E–3H). This effect was further enhanced in combina-
tion with aPD-1, which yielded even higher proliferation and a
1.6-fold increase in the number of IFN-g-producing CD4
+
and
CD8
+
T cells compared to aCD25-m2a alone. In contrast, the
observed lack of Treg cell depletion with aCD25-r1 resulted in
no change in Teff cell proliferation or IFN-gproduction, when
used as monotherapy or in combination with aPD-1 (Figures
3E–3H). Combination of aCD25 and aPD-1 therefore appeared
highly effective at rejecting established tumors, but only when
intra-tumoral Treg cells were efficiently depleted by aCD25 of
appropriate isotype.
CD25 Expression Profiles in Human Cancers Validate Its
Use as Target for Therapeutic Treg Cell Depletion
To validate the translational value of CD25 as a target for Treg
cell depletion, we analyzed the expression of CD25 on periph-
eral blood mononuclear cells (PBMCs) and TILs in patients with
advanced melanoma, early-stage non-small cell lung carci-
noma (NSCLC), and renal cell carcinoma (RCC) by flow cy-
tometry and multiplex immunohistochemistry (IHC). Despite
heterogeneity in clinical characteristics both within and
between studied cohorts (Tables S1–S3), CD25 expression
remained largely restricted to CD4
+
FoxP3
+
Treg cells (mean
% CD25
+
= 54.8% of Treg, 7.5% of CD4
+
FoxP3

, and 1.9%
of CD8
+
; p < 0.0001) (Figures 4A and 4B). Similar to murine
models, the level of CD25 expression, as assessed by MFI,
was significantly higher on CD4
+
FoxP3
+
Treg cells relative to
CD4
+
FoxP3

and CD8
+
T cells within all studied tumor sub-
types (mean MFI Treg = 190.0, CD4
+
FoxP3
+
= 34.5 and
CD8
+
= 17.9; p < 0.0001) (Figure 4C).
We further performed longitudinal assessment of CD25
expression in the context of immune modulation. Core biopsies
were performed on the same lesion at baseline and after either
four cycles of nivolumab (3 mg/kg Q2W) or two cycles of
pembrolizumab (200 mg Q3W) in patients with advanced kidney
cancer and melanoma, respectively (Table S4). Despite systemic
immune modulation, CD25 expression remained restricted to
(B) Survival of mice shown in (A).
(C and D) Survival of mice with MC38 or CT26 tumors treated as described above (n = 10 per condition).
(E) Percentage of Ki67
+
cells in tumor-infiltrating CD4
+
FoxP3

and CD8
+
T cells.
(F) CD4
+
FoxP3

/CD4
+
FoxP3
+
and CD8
+
/CD4
+
FoxP3
+
cell ratios.
(G and H) Representative histograms (G) and percentage (H) of IFN-g-producing CD4
+
and CD8
+
TILs in MCA205 tumors determined by intracellular staining after
ex vivo re-stimulation with PMA and ionomycin. Graphs show cumulative data of two separate experiments (n = 10).
582 Immunity 46, 577–586, April 18, 2017

FoxP3
+
Treg cells, even in areas of dense CD8
+
T cell infiltrate
evaluated by multiplex immunohistochemistry (Figures 4D and
4E). These findings confirmed the translational value of the
described pre-clinical data, lending further support to the
concept of selective therapeutic targeting of Treg cells via
CD25 in human cancers.
DISCUSSION
We have demonstrated that CD25 is an attractive target
for Treg cell depletion owing to its expression profile on
tumor-infiltrating T cells in both mice and humans. Contrary
to in vitro studies, minimal expression of CD25 on the effector
AB
C
ED
Figure 4. CD25 Is Highly Expressed on Treg Cell Infiltrating Human Tumors
(A) Representative histograms demonstrating CD25 expression on circulating (PBMC) and tumor-infiltrating (TIL) CD8
+
, CD4
+
FoxP3

, and CD4
+
FoxP3
+
T cell
subsets. Dotted lines indicate the gate.
(B and C) Quantification of CD25 expression (percentage [B] and MFI [C]) on individual T cell subsets inhuman melanoma (n = 11) , NSCLC (n = 9), and RCC (n = 8).
Error bars represent SEM; p values obtained by two-way ANOVA.
(D) Longitudinal analysis of CD25 expression in human melanoma and RCC lesions prior to (‘‘Ba seline’’) and during PD-1 blockade (‘‘On therapy’’). CD8 staining is
displayed in red, FoxP3 in blue, and CD25 in brown.
(E) Percentage of CD25 expression on CD8
+
and FoxP3
+
T cells at baseline and during PD-1 blockade. Plotted values derive from analysis of 10 340 high-power
fields per patient at each time point.
Immunity 46, 577–586, April 18, 2017 583

compartment was observed in vivo. The efficacy of aCD25 as
an anti-tumor therapy depends on Treg cell depletion in the tu-
mor microenvironment, which can be achieved only by using an
antibody isotype optimized for engagement of activating
FcgRs, capable of inducing ADCC. Our results demonstrated
that the limited efficacy observed in pre-clinical studies using
the aCD25 PC-61 monoclonal antibody with a rat IgG1 isotype
relates to ineffective or suboptimal intra-tumoral Treg cell
depletion, a consequence of its low A/I binding ratio and high
intra-tumoral expression of inhibitory FcgRIIb. This may also
explain the modest results observed in early clinical trials using
the anti-human CD25 antibody daclizumab. However, the
impact of aCD25 antibodies of varying IgG subclass remains
to be evaluated in humans.
Local depletion of tumor-infiltrating Treg cells by aCD25
monotherapy mediated only partial tumor control, suggesting
that further intervention is necessary to increase the intra-
tumoral Teff/Treg cell balance and promote effector T cell
activity. These data mirror those previously demonstrated for
aCTLA-4 antibodies, where targeting solely the Treg cell
compartment was ineffective in eradicating established
tumors, while targeting both Treg and Teff cell compartments
resulted in effective therapeutic synergy (Peggs et al., 2009).
Increased regulation of Teff cell responses by co-inhibitory im-
mune checkpoints in the tumor microenvironment might also
explain the modest responses observed in early-stage clinical
trials evaluating aCD25 antibodies in cancer patients (Jacobs
et al., 2010; Rech et al., 2012). Our data suggest that such re-
sponses could be enhanced through combination with thera-
pies that address this regulation including immune checkpoint
blockade or agonistic antibodies targeting immune co-stimula-
tory receptors.
Treg cell depletion can be achieved by targeting other
molecules highly expressed on Treg cells (Bulliard et al., 2014;
Coe et al., 2010; Selby et al., 2013; Simpson et al., 2013). While
combined blocking and depleting activity of specific immune
modulatory antibodies is effective against certain target mole-
cules, such as CTLA-4, it can also be deleterious owing to
simultaneous high expression on Teff cells. Differential expres-
sion is therefore critical; for example, in addition to its expres-
sion on Treg cells, PD-1 is highly expressed on activated
CD8
+
T cells. Anti-PD-1 antibodies therefore lose anti-tumor
activity when a depleting antibody isotype is employed (Dahan
et al., 2015).
Anti-PD-1 therapy now forms a key part of the treatment
paradigm for multiple solid malignancies, with response rates
varying between 20% and 30% when used as monotherapy
(Topalian et al., 2015). However, the majority of responses
are partial. This could be explained in part by tumor infiltration
with CD25
+
FoxP3
+
Treg cells that are unaffected by non-
depleting aPD-1 antibodies. In this setting another target
molecule specific to Treg cells is required in order to achieve
potential synergy through Treg cell depletion. Combination of
aCTLA-4 and aPD-1 therapy has achieved superior response
rates to either agent alone in patients with advanced melanoma
(Larkin et al., 2015). This may be the result of the cell-intrinsic
immune modulatory activity of aCTLA-4 and aPD-1 antibodies
and concomitant depletion of Treg cells by aCTLA-4, although
this second activity has not been demonstrated in vivo. Combi-
nation therapy results in higher immune-related toxicity, under-
scoring the need for alternative combinations balancing
maximal activity with minimal toxicity. We have demonstrated
that aCD25 therapy synergizes with blocking aPD-1 therapy,
provided Treg cells are depleted locally in the tumor.
Combining aPD-1 with aCD25-depleting antibodies might
improve the therapeutic window compared to the aCTLA-4
combination, as aCD25 lacks the additional cell-intrinsic im-
mune modulatory activity of aCTLA-4. Such hypotheses are
further supported by our model, in which only transient Treg
cell depletion was required for effective synergy, with no evi-
dence of immune-related toxicity. These data support further
evaluation of Fc-optimized aCD25 as a combination partner
in clinical trials.
EXPERIMENTAL PROCEDURES
Antibodies and Antibody Production
The sequence of the variable regions of the heavy and light chains of aCD25
were resolved from the PC-61.5.3 hybridoma by rapid amplification of cDNA
ends (RACE), cloned into the constant regions of murine IgG2a and kchains
and expressed in a stable K562 cell line generated by co-transduction with
murine leukemia virus-derived retroviral vectors encoding both chains. The
antibody was initially purified from supernatants with a protein G HiTrap
MabSelect column (GE Healthcare), dialyzed in phosphate-buffered saline
(PBS), concentrated, and filter-sterilized. For subsequent experiments, anti-
body production was outsourced to Evitria AG. Anti-CD25-r1 (PC-61.5.3),
aCTLA-4 (9H10), aPD-1 (RMP1-14), and aCD8 (2.43) were supplied by
BioXcell. The binding affinity of isotype variants to FcgRs was measured
by SPR in the Ravetch laboratory as described before (Nimmerja hn and Rav-
etch, 2005).
Tumor Experiments
Details of mouse strains, cell lines and flow cytometry antibodies are shown in
Supplemental Experimental Procedures. Mice were injected subcutaneously
with 5 310
5
MCA205, MC38, or CT26 cells or 5 310
4
B16 cells re-suspended
in PBS. Therapeutic antibodies were administered intraperitoneally at the time
points and doses shown in figure legends. Cell suspensions for flow cytometry
were prepared as described previously (Simpson et al., 2013). Tumors were
measured twice weekly and mice were euthanized when any orthogonal tumor
diameter reached 150 mm.
Human Study Oversight
Human data derives from three translational studies approved by local institu-
tional review board and Research Ethics Committee (Melanoma, REC no.
11/LO/0003; NSCLC, REC no.13/LO/1546; RCC, REC no. 11/LO/1996). All
were conducted in accordance with the provisions of the Declaration of Hel-
sinki and with Good Clinical Practice guidelines as defined by the International
Conference on Harmonization. All patients (or their legal representatives) pro-
vided written informed consent before enrollment.
Analysis of Human Tissue
For flow cytometry, cell suspensions were prepared with the same protocol
employed for mouse tissues (Simpson et al., 2013). Leukocytes were enriched
by gradient centrifugation with Ficoll-paque (GE Healthcare). Isolated live cells
were frozen at 80C and stored in liquid nitrogen until analysis.
Histopathology protocols are described in Supplemental Experimental
Procedures.
Data Analysis
Flow cytometry data were analyzed with FlowJo v10.0.8 (Tree Star). Statistical
analyses were done with Prism 6 (GraphPad Software); p values were calcu-
lated using Kruskall-Wallis and Dunn’s post hoc tests, unless otherwise indi-
cated (ns = p > 0.05; *p %0.05; **p %0.01; ***p %0.001; ****p %0.0001).
Kaplan-Meier curves were analyzed with the log-rank test.
584 Immunity 46, 577–586, April 18, 2017

SUPPLEMENTAL INFORMATION
Supplemental Information includes four figures, four tables, Supplemental
Experimental Procedures, and consortia memberships and can be found
with this article online at http://dx.doi.org/10.1016/j.immuni.2017.03.013.
AUTHOR CONTRIBUTIONS
S.A.Q. and K.S.P. conceived the project. F.A.V., A.J.S.F., K.S.P., and S.A.Q.
designed the experiments, analyzed the data, and wrote the manuscript.
F.A.V. and A.J.S.F. performed the experiments. I.S., K.J., L.M., M.H.L., A.G.,
A.S., A.B.A., D.F., M.W.S., Y.N.S.W., and J.Y.H. contributed experimentally.
E.M.R., R.D., S.A.B., K.A.C., M.P., and J.V.R. provided reagents and contrib-
uted scientifically. T.M. performed the histology analyses. T.O., D.N., B.C.,
S.T., M.G., J.L., C.S., and the TRACERx consortia coordinated clinical trials
and provided patient samples.
ACKNOWLEDGMENTS
We thank Josep Linares for his technical expertise. S.A.Q. is a Cancer
Research U.K. (CRUK) Senior Fellow (C36463/A22246) and is funded by a
Cancer Research Institute Investigator Award and a CRUK Biotherapeut ic Pro-
gram Grant (C36463/A20764). K.S.P. receives funding from the NIHR BTRU for
Stem Cells and Immunotherapies (167097), of which he is the Scientific
Director. None of the animal work described was funded by NIHR. This work
was undertaken at UCL Hospitals/UCL with support from the CRUK-UCL
Centre (C416/A18088), CRUK’s Lung Cancer Centre of Excellence (C5759/
A20465), the CRUK and Engineering and Physical Sciences Research Council
at King’s College London and UCL (C1519/A16463), the Cancer Immuno-
therapy Accelerator Award (CITA-CRUK) (C33499/A20265), CRUK’s Lung
TRACERx study (led by C. Swanton) (C11496/A17786), the Sam Keen Founda-
tion/RMH NIHR Biomedical Research Centre, Bloodwise (formerly Leukaemia
and Lymphoma Research) (08022/P4664), the Department of Health, and
CRUK funding schemes for National Institute for Health Research Biomedical
Research Centres and Experimental Cancer Medicine Centres.
Received: August 27, 2016
Revised: January 26, 2017
Accepted: February 9, 2017
Published: April 11, 2017
REFERENCES
Bos, P.D., Plitas, G., Rudra, D., Lee, S.Y., and Rudensky, A.Y. (2013). Transient
regulatory T cell ablation deters oncogene-driven breast cancer and enhances
radiotherapy. J. Exp. Med. 210, 2435–2466.
Boyman, O., and Sprent, J. (2012). The role of interleukin-2 during homeosta-
sis and activation of the immune system. Nat. Rev. Immunol. 12, 180–190.
Bulliard, Y., Jolicoeur, R., Zhang, J., Dranoff, G., Wilson, N.S., and Brogdon,
J.L. (2014). OX40 engagement depletes intratumoral Tregs via activating
FcgRs, leading to antitumor efficacy. Immunol. Cell Biol. 92, 475–480.
Clynes, R.A., Towers, T.L., Presta, L.G., and Ravetch, J.V. (2000). Inhibitory
Fc receptors modulate in vivo cytotoxicity against tumor targets. Nat. Med.
6, 443–446.
Coe, D., Begom, S., Addey, C., White, M., Dyson, J., and Chai, J.G. (2010).
Depletion of regulatory T cells by anti-GITR mAb as a novel mechanism for
cancer immunotherapy. Cancer Immunol. Immunother. 59, 1367–1377.
Dahan, R., Sega, E., Engelhardt, J., Selby, M., Korman, A.J., and Ravetch, J.V.
(2015). FcgRs modulate the anti-tumor activity of antibodies targeting the
PD-1/PD-L1 axis. Cancer Cell 28, 285–295.
Dannull, J., Su, Z., Rizzieri, D., Yang, B.K., Coleman, D., Yancey, D., Zhang, A.,
Dahm, P., Chao, N., Gilboa, E., and Vieweg, J. (2005). Enhancement of vac-
cine-mediated antitumor immunity in cancer patients after depletion of regula-
tory T cells. J. Clin. Invest. 115, 3623–3633.
Elpek, K.G., Lacelle, C., Singh, N.P., Yolcu, E.S., and Shirwan, H. (2007).
CD4+CD25+ T regulatory cells dominate multiple immune evasion mecha-
nisms in early but not late phases of tumor development in a B cell lymphoma
model. J. Immunol. 178, 6840–6848.
Goding, S.R., Wilson, K.A., Xie, Y., Harris, K.M., Baxi, A., Akpinarli, A., Fulton,
A., Tamada, K., Strome, S.E., and Antony, P.A. (2013). Restoring immune
function of tumor-specific CD4+ T cells during recurrence of melanoma.
J. Immunol. 190, 4899–4909.
Golgher, D., Jones, E., Powrie, F., Elliott, T., and Gallimore, A. (2002). Depletion
of CD25+ regulatory cells uncovers immune responses to shared murine tumor
rejection antigens. Eur. J. Immunol. 32, 3267–3275.
Hodi, F.S., Butler, M., Oble, D.A., Seiden, M.V., Haluska, F.G., Kruse, A.,
Macrae, S., Nelson, M., Canning, C., Lowy, I., et al. (2008). Immunologic and
clinical effects of antibody blockade of cytotoxic T lymphocyte-associa ted an-
tigen 4 in previously vaccinated cancer patients. Proc. Natl. Acad. Sci. USA
105, 3005–3010.
Jacobs, J.F.M., Punt, C.J.A., Lesterhuis, W.J., Sutmuller, R.P.M., Brouwer,
H.M.-L.H., Scharenborg, N.M., Klasen, I.S., Hilbrands, L.B., Figdor, C.G., de
Vries, I.J.M., and Adema, G.J. (2010). Dendritic cell vaccination in combination
with anti-CD25 monoclonal antibody treatment: a phase I/II study in metastatic
melanoma patients. Clin. Cancer Res. 16, 5067–5078.
Jones, E., Dahm-Vicker, M., Simon, A.K., Green, A., Powrie, F., Cerundolo, V.,
and Gallimore, A. (2002). Depletion of CD25+ regulatory cells results in sup-
pression of melanoma growth and induction of autoreactivity in mice.
Cancer Immun. 2,1.
Larkin, J., Chiarion-Sileni, V., Gonzalez, R., Grob, J.J., Cowey, C.L., Lao, C.D.,
Schadendorf, D., Dummer, R., Smylie, M., Rutkowski, P., et al. (2015).
Combined nivolumab and ipilimumab or monotherapy in untreated melanoma.
N. Engl. J. Med. 373, 23–34.
Luke, J.J., Zha, Y., Matijevich, K., and Gajewski, T.F. (2016). Single dose deni-
leukin diftitox does not enhance vaccine-induced T cell responses or effec-
tively deplete Tregs in advanced melanoma: immune monitoring and clinical
results of a randomized phase II trial. J. Immunother. Cancer 4,35.
Nimmerjahn, F., and Ravetch, J.V. (2005). Divergent immunoglobulin g sub-
class activity through selective Fc receptor binding. Science 310, 1510–1512.
Onizuka, S., Tawara, I., Shimizu, J., Sakaguchi, S., Fujita, T., and Nakayama, E.
(1999). Tumor rejection by in vivo administration of anti-CD25 (interleukin-2
receptor alpha) monoclonal antibody. Cancer Res. 59, 3128–3133.
Peggs, K.S., Quezada, S.A., Chambers, C.A., Korman, A.J., and Allison, J.P.
(2009). Blockade of CTLA-4 on both effector and regulatory T cell compart-
ments contributes to the antitumor activity of anti-CTLA-4 antibodies.
J. Exp. Med. 206, 1717–1725.
Quezada, S.A., Peggs, K.S., Curran, M.A., and Allison, J.P. (2006). CTLA4
blockade and GM-CSF combination immunotherapy alters the intratumor
balance of effector and regulatory T cells. J. Clin. Invest. 116, 1935–1945.
Quezada, S.A., Peggs, K.S., Simpson, T.R., Shen, Y., Littman, D.R., and
Allison, J.P. (2008). Limited tumor infiltration by activated T effector cells re-
stricts the therapeutic activity of regulatory T cell depletion against established
melanoma. J. Exp. Med. 205, 2125–2138.
Rech, A.J., Mick, R., Martin, S., Recio, A., Aqui, N.A., Powell, D.J., Jr., Colligon,
T.A., Trosko, J.A., Leinbach, L.I., Pletcher, C.H., et al. (2012). CD25 blockade
depletes and selectively reprograms regulatory T cells in concert with immuno-
therapy in cancer patients. Sci. Transl. Med. 4, 134ra62.
Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., and Toda, M. (1995).
Immunologic self-tolerance maintained by activated T cells expressing IL-2
receptor alpha-chains (CD25). Breakdown of a single mechanism of self-toler-
ance causes various autoimmune diseases. J. Immunol. 155, 1151–1164.
Selby, M.J., Engelhardt, J.J., Quigley, M., Henning, K.A., Chen, T., Srinivasan,
M., and Korman, A.J. (2013). Anti-CTLA-4 antibodies of IgG2a isotype
enhance antitumor activity through reduction of intratumoral regulatory
T cells. Cancer Immunol. Res. 1, 32–42.
Setiady, Y.Y., Coccia, J.A., and Park, P.U. (2010). In vivo depletion of
CD4+FOXP3+ Treg cells by the PC61 anti-CD25 monoclonal antibody is medi-
ated by FcgammaRIII+ phagocytes. Eur. J. Immunol. 40, 780–786.
Immunity 46, 577–586, April 18, 2017 585

Shang, B., Liu, Y., Jiang, S.J., and Liu, Y. (2015). Prognostic value of tumor-
infiltrating FoxP3+ regulatory T cells in cancers: a systematic review and
meta-analysis. Sci. Rep. 5, 15179.
Shimizu, J., Yamazaki, S., and Sakaguchi, S. (1999). Induction of tumor
immunity by removing CD25+CD4+ T cells: a common basis between tumor
immunity and autoimmunity. J. Immunol. 163, 5211–5218.
Simpson, T.R., Li, F., Montalvo-Ortiz, W., Sepulveda, M.A., Bergerhoff, K.,
Arce, F., Roddie, C., Henry, J.Y., Yagita, H., Wolchok, J.D., et al. (2013).
Fc-dependent depletion of tumor-infiltrating regulatory T cells co-defines the
efficacy of anti-CTLA-4 therapy against melanoma. J. Exp. Med. 210,
1695–1710.
Sutmuller, R.P., van Duivenvoorde, L.M., van Elsas, A., Schumacher, T.N.,
Wildenberg, M.E., Allison, J.P., Toes, R.E., Offringa, R., and Melief, C.J.
(2001). Synergism of cytotoxic T lymphocyte-associated antigen 4 blockade
and depletion of CD25(+) regulatory T cells in antitumor therapy reveals alter-
native pathways for suppression of autoreactive cytotoxic T lymphocyte re-
sponses. J. Exp. Med. 194, 823–832.
Topalian, S.L., Drake, C.G., and Pardoll, D.M. (2015). Immune checkpoint
blockade: a common denominator approach to cancer therapy. Cancer Cell
27, 450–461.
van Elsas, A., Sutmuller, R.P., Hurwitz, A.A., Ziskin, J., Villasenor, J., Medema,
J.P., Overwijk, W.W., Restifo, N.P., Melief, C.J., Offringa, R., and Allison, J.P.
(2001). Elucidating the autoimmune and antitumor effector mechanisms of a
treatment based on cytotoxic T lymphocyte antigen-4 blockade in combina-
tion with a B16 melanoma vaccine: comparison of prophylaxis and therapy.
J. Exp. Med. 194, 481–489.
586 Immunity 46, 577–586, April 18, 2017