Oncotarget, Advance Publications 2015
www.impactjournals.com/oncotarget/
Ex vivo generation of myeloid-derived suppressor cells that
model the tumor immunosuppressive environment in colorectal
cancer
Inès Dufait1,2, Julia Katharina Schwarze1, Therese Liechtenstein3,4, Wim Leonard1,
Heng Jiang1, David Escors3,4, Mark De Ridder1,* and Karine Breckpot2,*
1
UZ Brussel, Department of Radiotherapy, Vrije Universiteit Brussel, Brussels, Belgium
2
Laboratory of Molecular and Cellular Therapy, Vrije Universiteit Brussel, Brussels, Belgium
3
Navarrabiomed-Fundaçion Miguel Servet, Immunomodulation group, Pamplona, Spain
4
Division of Infection and Immunity, University College London, London, UK
*
These senior authors contributed equally to this work
Correspondence to: Karine Breckpot, email: kbreckpo@vub.ac.be
Keywords: MDSC, CRC, arginase-1, inducible nitric oxide synthase, GM-CSF
Received: February 18, 2015
Accepted: March 11, 2015
Published: March 29, 2015
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and source are credited.
ABSTRACT
Myeloid-derived suppressor cells (MDSC) are a heterogeneous population of
cells that accumulate in tumor-bearing subjects and which strongly inhibit anticancer immune responses. To study the biology of MDSC in colorectal cancer (CRC),
we cultured bone marrow cells in conditioned medium from CT26 cells, which are
genetically modiied to secrete high levels of granulocyte-macrophage colonystimulating factor. This resulted in the generation of high numbers of CD11b+ Ly6G+
granulocytic and CD11b+ Ly6C+ monocytic MDSC, which closely resemble those
found within the tumor but not the spleen of CT26 tumor-bearing mice. Such MDSC
potently inhibited T-cell responses in vitro, a process that could be reversed upon
blocking of arginase-1 or inducible nitric oxide synthase (iNOS). We conirmed that
inhibition of arginase-1 or iNOS in vivo resulted in the stimulation of cytotoxic T-cell
responses. A delay in tumor growth was observed upon functional repression of
both enzymes. These data conirm the role of MDSC as inhibitors of T-cell-mediated
immune responses in CRC. Moreover, MDSC differentiated in vitro from bone marrow
cells using conditioned medium of GM-CSF-secreting CT26 cells, represent a valuable
platform to study/identify drugs that counteract MDSC activities.
Experimental CRC models such as those based on
murine CT26 cells are often used to evaluate the growing
list of anti-MDSC agents. This model is a valuable
substitute for human CRC as, similarly to CRC in human
patients, it is iniltrated with CTLs that are rendered
inactive due to immunosuppression exerted by MDSC [6].
In mice, MDSC represent a heterogeneous
population comprised of immature myeloid cells. These
are characterized by the expression of CD11b and Gr-1,
and lack markers speciic for monocytes, macrophages and
dendritic cells. MDSC can be subdivided in two subsets,
namely monocytic and granulocytic MDSC on the basis of
Ly6C-Ly6G expression proiles. While monocytic MDSC
express low (or absent) Ly6G levels, granulocytic MDSC
INTRODUCTION
Colorectal cancer (CRC) is characterized by the
iniltration with various immune cell types [1]. The
iniltration of CRC with CD8+ cytotoxic T lymphocytes
(CTLs) has been correlated to a favorable prognosis [2, 3].
However, these CTLs are largely dysfunctional, as a result
of their interaction with myeloid-derived suppressor cells
(MDSC). Therefore, it was suggested that the absence
of MDSC iniltration might serve as a better prognostic
biomarker [4, 5]. Moreover, it was suggested that
pharmacologic blockade of MDSC represents an attractive
strategy to treat CRC.
www.impactjournals.com/oncotarget
1
Oncotarget
express high levels of Ly6G [7]. Various tumor-derived
factors have been described to induce MDSC, these
include but are not limited to granulocyte macrophagecolony stimulating factor (GM-CSF), macrophage-colony
stimulating factor (M-CSF), prostaglandin E2 (PGE2),
vascular endothelial growth factor (VEGF), stem cell
factor (SCF), interleukin-6 (IL-6), IL-10 and IL-1β [7,
8]. Importantly, MDSC use a plethora of mechanisms
to suppress antitumor immunity. One of these is the
depletion of L-arginine mediated by arginase-1 (arg-1)
and inducible nitric oxide synthase (iNOS), expressed in
MDSC. L-arginine depletion was shown to limit T-cell
proliferation and T-cell receptor signaling, and it is still
considered the major mechanism through which MDSC
mediate T-cell dysfunction [9, 10].
The ample evidence on the role of MDSC in cancers
such as CRC has instigated research into the use of
existing drugs as well as the development of novel drugs to
deplete MDSC, block or revert their immunosuppressive
activity. For example, chemotherapy drugs which have
been shown to deplete MDSC include 5-luororacil (5FU) [11], gemcitabine [12, 13] and docetaxel [14]. In these
drug discovery studies, MDSC derived from the spleen
of tumor-bearing animals are most commonly used solely
because they can be obtained in large numbers. However,
splenic MDSC are phenotypically and functionally
different from MDSC derived from within the tumor [15,
16]. Consequently, to ensure reliability and potency of
novel MDSC-targeting drugs, they should be evaluated
on tumor-derived rather than splenic MDSC. However,
studying tumor-derived MDSC poses the technical
challenge of obtaining suficient number of cells at high
purity from a limited number of tumor-bearing animals
[17]. To circumvent this conundrum, researchers have
evaluated various in vitro culture systems to obtain MDSC
that closely resemble those found within the tumor. First
of all, immortalized MDSC cell lines such as MSC-1 and
MSC-2, were constructed using retroviral transduction
but lack the distinct marker of MDSC, namely Gr-1 [18].
However, other ex vivo procedures starting from bone
marrow cells were characterized by a low differentiation
eficiency (up to 40%), resulting in only a limited amount
of MDSC-like cells [19-27]. We recently developed an ex
vivo system to eficiently differentiate bone marrow cells
into MDSC [27, 28]. Herein conditioned medium from
tumor cells that were transduced with lentiviral vectors
encoding GM-CSF is used to differentiate bone marrow
cells. A proof-of-concept on the value of this strategy to
obtain large amounts of MDSC that resemble those found
within B16 melanomas was delivered [28].
In the current study, we demonstrate that the ex
vivo culture procedure is readily applicable to CRC and
could be used as a predictive model as such facilitating the
search for novel anti-MDSC drugs. Here we thoroughly
characterize these ex vivo differentiated CRC-speciic
MDSC, demonstrate that their functions could be
www.impactjournals.com/oncotarget
counteracted by arg-1 and iNOS inhibitors and that these
treatments possess therapeutic activities in vivo.
RESULTS
High levels of GM-CSF are required to eficiently
differentiate bone marrow cells to MDSC
CRC expands MDSC in vivo, which seem to
contribute to tumor staging and poor prognosis [4, 2931]. Usually, a large tumor burden is required to divert
physiological myeloid differentiation towards MDSC
expansion, possibly due to local and systemic GMCSF accumulation. As we aimed to develop an in vitro
culture system to differentiate bone marrow cells to
MDSC resembling those found within CRC tumors, we
irst evaluated using ELISA whether the CRC cell line
CT26 produced high levels of GM-CSF. CT26 tumor
cells produced barely any GM-CSF (Fig. 1A). Therefore,
we decided in analogy to our previous study on in vitrogenerated melanoma MDSC [28], to transduce CT26
tumor cells with lentiviral vectors encoding GM-CSF.
This resulted in secretion of high levels of GM-CSF
(Fig. 1A). To examine whether the secreted GM-CSF
was biologically active, GM-CSF-dependent FDCP-1
cells were labeled with CFSE and consequently cultured
in the presence or absence of recombinant murine GMCSF, as well as in conditioned medium (CM) of CT26GM-CSF and CM of non-modiied CT26 tumor cells. In
this assay the proliferation of FDCP-1 cells incubated in
recombinant GM-CSF was comparable to that of FDCP-1
cells incubated with CM of CT26-GM-CSF (Fig. 2B-2C).
This CM was subsequently used to culture bone marrow
cells, demonstrating that after 6 days of culture, cell yields
were consistently higher in the high GM-CSF condition
(Fig. 1D). Moreover, the majority of these cells expressed
CD11b. This was not the case in cultures with CM of
non-modiied CT26 tumor cells (Fig. 1E). To identify
the concentration of GM-CSF necessary to generate
CD11b+ cells, we used CM of non-modiied CT26
tumor cells supplemented with different concentrations
of recombinant GM-CSF to culture bone marrow cells.
High percentages of CD11b+ cells were generated in the
presence of recombinant GM-CSF, without signiicant
differences when using relative high (320 ng/ml) or low
GM-CSF concentrations (20 ng/ml) (Fig. 1E). However,
a signiicant difference was observed in the yield of
CD11b+ cells between the conditions where recombinant
GM-CSF or CM of transduced CT26 tumor cells was
used (Fig. 1F). Since the yield and purity of CD11b+ cells
was highest after differentiation in CM from CT26-GMCSF cells (Fig. 2A), we continued with these culture
conditions. MDSC are known to be a very heterogeneous
population of cells but can be generally divided into a
2
Oncotarget
monocytic (Ly6C+) and a granulocytic (Ly6G+) subset. We
examined the appearance of these subsets in the generated
CD11b+ population (Fig. 2B). The ratio of the different
subsets in the in vitro system coincides with the in vivo
situation. Next, we examined their suppressive capacity
as it is widely accepted that functionality and more
speciically suppression of T-cell responses, is the single
most important marker to identify MDSC. We showed
that sorted CD11b+ Ly6C+ as well as CD11b+ Ly6G+ cells
(Fig. 2C) had a high T-cell suppressive capacity (Fig. 2D-
2E). Consequently, the CD11b+ cells obtained through the
culture of bone marrow cells in CM of CT26-GM-CSF
tumor cells could be considered as MDSC.
Figure 1: Ex vivo myelopoiesis can differentiate bone marrow cells into myeloid cells in the presence of GM-CSF. (A)
Graph representing murine GM-CSF content as measured by ELISA present in the CM of wildtype (no) and transduced (GM-CSF) CT26
tumor cells. (B) Representative histogram showing proliferation, as measured by dilution of CFSE, of the GM-CSF dependent FDCP-1
cells incubated for 72 hours in DMEM with (+) or without (-) recombinant GM-CSF (20 ng/ml) or incubated in CM of non-modiied (no)
and transduced CT26 tumor cells (GM-CSF). (C) Summarizing graph showing the mean luorescence intensity (MFI) of CFSE positive
FDCP-1 cells, a lower MFI representing strong proliferation of the FDCP-1 cells. (D) Fold increase in bone marrow cells incubated for 6
days in CM. (E) Expression of CD11b by bone marrow cells after a 6-day incubation period in CM. (F) Cell yield after 6 days incubation
of 10 x 106 bone marrow cells in CM. Mean of at least 3 experiments +/- SEM is shown in all graphs. Number of asterisks in the igures
indicates the level of statistical signiicance as follows: *, p < 0.05; **, p < 0.01; ***, p < 0.001.
www.impactjournals.com/oncotarget
3
Oncotarget
Figure 2: Differentiated bone marrow cells possess strong suppressive capacities and can be subdivided into both
MDSC subsets. (A) Expression of CD11b by bone marrow cells after a 6-day incubation period in CM as measured by low cytometry.
(B) Summarizing graph of ratio of MDSC subsets (C) Flow cytometry contour plots of in vitro MDSC before and after MACS sort.
Underneath the contour plots of the sorted MDSC, representative pictures showing the morphology of these subsets are depicted. Pictures
were taken with a light microscope at 64 times magniication. (D) Representative experiment showing suppression of CD8+ T cells by
sorted in vitro MDSC (1:4 ratio MDSC to T cell). (E) The graph on the left represents the proliferation inhibition of CD3/CD28-activated
CD8+ T cells (ratio MDSC to T cell as indicated in the graph). The graph on the right shows changes in IFN-γ secretion measured during
the same experiment. Control represents T cells incubated without MDSC. Mean of at least 3 experiments +/- SEM is shown in all graphs.
Number of asterisks in the igures indicates the level of statistical signiicance as follows: ***, p < 0.001.
www.impactjournals.com/oncotarget
4
Oncotarget
In vitro-generated MDSC closely resemble MDSC
found within both wild-type and GM-CSFproducing CT26 tumors
vivo differentiated MDSC, the MDSC-T-cell suppressive
activity is the same as found in mice bearing non-modiied
CT26 tumor cells. Moreover, these results show that our
in vitro-generated MDSC phenotypically and functionally
resemble the MDSC found within the tumor.
To examine whether the in vitro-generated MDSC
resembled those found within CT26 tumors, we then
compared their function and phenotype with MDSC
isolated from the spleen and tumor of CT26-bearing
mice. We also isolated MDSC from mice bearing CT26GM-CSF tumor cells to examine the effect of GM-CSF
overexpression in vivo. We observed that mice bearing
CT26-GM-CSF tumors showed splenomegaly (Fig. 3A3B) and moreover, that CT26-GM-CSF tumors hardly
progressed after day 12, although the growth of unmodiied
CT26 or CT26-GM-CSF tumor cells initially followed a
similar pattern (Fig. 3C). This decline in tumor growth was
not correlated with the presence of tumor-speciic T-cell
responses as evaluated by ELISPOT (data not shown). In
addition, evaluation of the T-cell suppressive activity of
bulk (granulocytic and monocytic) MDSC showed that in
vitro-generated MDSC, as well as MDSC derived from the
spleen or tumor were highly capable of suppressing T-cell
proliferation in a 1:1 MDSC to T cell ratio. No differences
in functionality were observed between MDSC derived
from CT26- and CT26-GM-CSF bearing mice (Fig. 3D).
The CT26-GM-CSF tumors showed high iniltration of
CD45+ cells, which correlated with a signiicant increase
in CD11b+ but not CD11c+ or F4/80+ cells (Fig. 3E-3F).
Similar to previous reports [15, 16], we observed
that MDSC found within the spleen (irrespective of the
level of GM-CSF expression by CT26 cells) showed
signiicant differences to MDSC found within the tumor.
More speciically, a lower expression of CD80 and PD-L1
of splenic-MDSC was observed when compared to tumoriniltrating MDSC (Fig. 3G). This data again conirmed
the observation that MDSC accumulating in the spleen
are distinct and different from tumor MDSC. Moreover,
MDSC found within the spleen of mice bearing CT26GM-CSF cells showed lower expression of MHC II
when compared to MDSC obtained from tumors as well
as the spleen of mice bearing non-modiied CT26 cells.
In contrast, lower expression of Sca-1 was observed in
the granulocytic MDSC subset obtained from the tumor
in comparison to the spleen, while opposite results were
obtained for the monocytic MDSC subset. Importantly,
the expression of MHC II, PD-L1 CD80 and Sca-1 was
not signiicantly different on MDSC isolated from the
tumor of non-modiied CT26 cells when compared to
CT26-GM-CSF tumors (Fig. 3G). The phenotype of in
vitro-generated MDSC showed that they were closely
related to the MDSC found within tumors but not spleen,
as they expressed high levels of MHC II, PD-L1 and
CD80 and low levels of Sca-1 (Fig. 3G). These results
indicate that although high GM-CSF secretion impacts
on the percentage and to a lesser extent phenotype of in
www.impactjournals.com/oncotarget
The in vitro-generated MDSC are a reliable model
to predict the outcome of MDSC-modulating
drugs
A key aim of this study was to prove that the in
vitro-generated MDSC can be used as a platform to predict
the outcome of anti-MDSC drugs. In the literature, it is
described that MDSC-mediated L-arginine depletion by
the expression of arg-1 and iNOS, plays a major role in
their T-cell suppressive capacity. Therefore, we chose
inhibitors of arg-1 and iNOS as ideal candidates to test
the predictive value of our in vitro MDSC-platform.
Interestingly, arg-1 expression was high in tumor and in
vitro-generated MDSC (both monocytic and granulocytic
subsets) but not in splenic MDSC (Fig. 4A).
Similar iNOS expression was observed in the in
vitro-generated and tumor CD11b+Ly6G+ granulocytic
MDSC (Fig. 4A). iNOS expression was lower in in vitrogenerated and tumor CD11b+Ly6C+ monocytic MDSC
compared to their granulocytic counterparts, but still
expressed to a higher extent then in splenic MDSC (Fig.
4A). These results conirm that the in vitro-generated
MDSC closely resemble those found within CT26 tumors.
To evaluate the extent to which arg-1 and iNOS contribute
to the suppressive capacity of the in vitro MDSC, we
performed an in vitro T-cell suppression assay with
sorted CD11b+ Ly6G+ and CD11b+ Ly6C+ MDSC in the
presence or absence of Nor-NOHA, an arg-1 inhibitor,
and AG, an iNOS inhibitor. We showed that both the
T-cell proliferation and IFN-γ production by the T cells
was enhanced in the presence of these inhibitors (Fig 4B).
These results conirmed the previously published role of
arg-1 and iNOS in the T-cell suppressive activity of MDSC
[7, 32] and suggest that the in vitro-generated CRCspeciic MDSC are similar to MDSC obtained from CRC
tumors. To evaluate the predictive value of CRC-speciic
MDSC, this was further conirmed in vivo, in which CT26bearing mice were treated with Nor-NOHA or the speciic
iNOS inhibitor, 1400W, as the only treatment. We showed
that treatment with either one of both inhibitors resulted
in a CT26-speciic cytotoxic immune response (Fig. 4C4D). These data suggested that inhibition of arg-1 or iNOS
enabled CD8+ cytotoxic T cells to escape the MDSCmediated immune suppression, although overall survival
of treated groups was not signiicantly prolonged (Fig.
4E). Moreover, a delay in tumor growth was observed
as long as both inhibitors were administered (Fig. 4F).
Comparison of the CD8+ T-cell iniltration of tumors
of non-treated mice or mice treated with Nor-NOHA
5
Oncotarget
Figure 3: In vitro-generated MDSC closely resemble tumor-MDSC but not splenic-MDSC. (A) Photograph of spleen
of naive mice (control), CT26- (no) and CT26-GM-CSF (GM-CSF)-bearing mice at a tumor size of approximately 5 by 6 mm. (B)
Summarizing graph showing spleen weight. (C) Tumor growth curve of CT26 versus CT26-GM-CSF-bearing mice. Day 1 represents
the day of tumor injection. (D) Graph showing the proliferation inhibition of CD3/CD28-activated CD8+ T cells with bulk MDSC (1:1
MDSC to T cell ratio). (E) Summarizing graph showing CD45 iniltration in the tumor of CT26- versus CT26-GM-CSF-bearing mice.
(F) Summarizing graph showing CD11b, CD11c and F4/80 content in the tumor of CT26- versus CT26-GM-CSF-bearing mice. (G)
Summarizing graphs of different surface markers (CD80, MHC II, PD-L1 and Sca-1) present on MDSC, derived from spleen, tumor or in
vitro-generated. Expression showed in the granulocytic (Ly6G) and monocytic (Ly6C) subset separately. Mean of at least 3 experiments
+/- SEM is shown in all graphs. Number of asterisks in the igures indicates the level of statistical signiicance as follows: *, p < 0.05; ***,
p < 0.001.
www.impactjournals.com/oncotarget
6
Oncotarget
Figure 4: iNOS and arginase-1 can be used as targets to enhance cytotoxic T-cell responses. (A) Summarizing graph
representing arg-1 expression by MDSC at the left and iNOS expression by MDSC at the right. (B) The graph on the left shows the
proliferation inhibition of CD3/CD28-activated CD8+ T cells with sorted in vitro MDSC (1:1 MDSC to T cell ratio). Control represents
Cultures incubated without inhibitors. Cultures were supplemented with Nor-NOHA (300 μM) or AG (1 mM). The graph on the right shows
changes in IFN-γ secretion measured during the same experiment. Mean of at least 3 experiments +/- SEM is shown in all graphs. (C) FACS
graph showing the cytotoxic T-cell response against gp70 peptides of mice treated with Nor-NOHA, 1400W or PBS as a negative control.
Number of mice per group = 3. (D) Summarizing graph of panel C. (E) Overall survival of treated mice. Mice were sacriiced when tumor
diameter reached 15 mm. (F) Tumor growth curve of treated mice. Day 1 represents the irst day of treatment, when tumor diameter reached
approximately 6 mm. (G) Graph summarizing the percentage of CD8+ T cells present in the tumor of treated mice. Values are normalized
to CD45 and CD3. Experiments were performed twice and included 6 mice per group. Mean +/- SEM is shown in all graphs. Number of
asterisks in the igures indicates the level of statistical signiicance as follows: *, p < 0.05; **, p < 0.01; ***, p < 0.001.
www.impactjournals.com/oncotarget
7
Oncotarget
drugs. However, its use is not restricted and they could
be applied in a very broad manner. For example, Van der
Jeught et al. used this ex vivo differentiation protocol to
examine the modulation of the tumor microenvironment
by using mRNA encoding soluble proteins [44].
In this study, we have shown that GM-CSF
overexpression in vivo does initially not lead to changes
in tumor volume, but does, amongst others, cause
splenomegaly. Similar abnormalities were previously
described in GM-CSF transgenic mice [45-47]. The
increased number of myeloid cells in tumor and spleen
is characterized by a CD11b+ MDSC population and can
be seen as an immune-inhibitory iniltrate [48]. These
indings are consistent with the study performed by
Bronte et al. who showed that a population of suppressive
CD11b+/Gr-1+ cells increased when tumor cells were
modiied to produce GM-CSF [33]. However, there is
no clear consensus about the effects of chronic GM-CSF
expression on tumor growth, as studies have shown either
an anti-proliferative effect [49, 50], a tumor-promoting
effect [51, 52] or no signiicant effect on tumor growth
rate [53, 54]. We showed no signiicant effects during the
irst 12 days of tumor growth. Long-lasting follow-up of
tumor growth was impossible in our study, as the mice
had to be sacriiced due to their splenomegaly, the latest at
day 17. Nonetheless, we observed that the size of CT26GM-CSF tumors remained stable from day 12 onwards,
whereas CT26 tumors continued to grow. To our surprise
the lack of continued growth of CT26-GM-CSF tumors
was not correlated to a tumor-speciic T-cell response,
suggesting that other mechanisms are responsible for the
tumor control. Although, chronic GM-CSF expression was
shown in some studies to lead to malignant progression
of the tumor due to enhanced angiogenesis, invasiveness
and migration [55-57], other studies showed that GMCSF production can also lead to improved survival in
CRC [58]. GM-CSF has been studied for a while as a
vaccine adjuvant in cancer immunotherapy due to its
immunostimulatory properties [47, 59]. However, the
results in clinical trials were disappointing in terms of
immune responses and clinical outcome [60, 61]. No link
between GM-CSF-induced MDSC expansion and failure
of GM-CSF in clinical trials has been demonstrated, but
caution in this ield is required as GM-CSF is much more
then solely an immunostimulating cytokine.
Well-studied amino acid-consuming enzymes
in MDSC biology are arg-1 and iNOS, both present on
the two MDSC subtypes and important in conferring
immunosuppressive capacities to these cells [7, 36]. As
these molecules are important in the basis MDSC biology,
we used them to further examine our in vitro-generated
CRC-speciic MDSC. Consistent with in vivo MDSC
[10, 21, 62], inhibiting arg-1 and iNOS directly affects
their T-cell immunosuppressive activities. Furthermore,
we inhibited arg-1 or iNOS intratumorally to examine
whether we could observe similar effects in comparison
or iNOS, showed that the number of CD8+ T cells was
enhanced in both treatment conditions (Fig. 3G). These
data show that inhibition of arg-1 and iNOS resulted in
higher numbers of functional CD8+ cytotoxic T cells, as
predicted in the in vitro T-cell suppression assay using in
vitro-generated CRC-speciic MDSC.
DISCUSSION
GM-CSF is one of the most important factors
produced by tumor cells leading to MDSC expansion. In
literature, it is evident that GM-CSF plays an important
role in the accumulation of MDSC [7, 20, 33]. That is
why current ex vivo MDSC differentiation protocols
primarily rely on culturing bone marrow hematopoietic
progenitors with recombinant GM-CSF. But clearly, other
still unknown factors contribute to MDSC differentiation
and expansion, as eficiency rarely surpasses 40%, even
with the addition of various other cytokines, such as IL-4,
IL-13, PGE2, [21, 24, 26].
In this study, we demonstrated the feasibility of
generating in vitro MDSC in a CRC model using the
system described by Liechtenstein et al. in a melanoma
model [28]. They reasoned that endogenous GM-CSF
could have better differentiation eficiency as myelopoiesis
within a tumor microenvironment was simulated.
Obviously, this system does not mimic the complexity
of the in vivo situation, but it may be a good practical
approximation. Indeed, differentiation eficiency up to
90% was achieved in our CRC model, while maintaining
high proliferation capacity. Another advantages of this
protocol compared to previously described methods are the
high MDSC yields, which can not be obtained by merely
supplementing CM of CT26 cells with recombinant GMCSF. Importantly, the high yield of pure CD11b+ cells
obtained in this culture system circumvents the need to
grow tumors and sacriice a large number of mice to obtain
suficient tumor MDSC. Other advantages of the method
presented in this manuscript, is its reproducibility and the
ease at which this technique can be performed.
The one true accepted marker of MDSC is their
suppressive capacity, as these cells otherwise display a
great heterogeneity [34-37]. Our in vitro-generated CRCspeciic MDSC are very potent immunosuppressive cells,
demonstrated by their ability to suppress T cells, even at a
1:8 MDSC to T cell ratio. In addition, our data conirmed
that the in vitro-generated CRC-speciic MDSC are more
similar to tumor MDSC than splenic MDSC. Therefore,
this model would be more relevant in drug discovery
studies than splenic MDSC. Spleen-derived MDSC are
still widely used in MDSC research [38-43], despite
the proof that these cells are very different indeed, both
phenotypically as functionally [15].
The primary aim of the ex vivo MDSC generation
protocol is to use these cells as a predictive tool for highthroughput screening in the search for new anti-MDSC
www.impactjournals.com/oncotarget
8
Oncotarget
colon cancer cell line CT26 and mouse lymphoblast cell
line FDCP-1 were obtained from the American Type
Culture Collection (ATCC, Molsheim Cedex, France) and
cultured according to the recommendations of ATCC.
to the in vitro system. To this end, we used the inhibitors
Nor-NOHA and 1400W, respectively. AG was replaced
with 1400W to serve as the iNOS inhibitor during the in
vivo study, since AG is a general NOS inhibitor, namely an
inhibitor of iNOS but also endothelial NOS and neuronal
NOS. The latter are not present in the in vitro culture, while
1400W speciically inhibits iNOS and is better suited for
an in vivo setting. During treatment, a signiicant delay in
tumor growth was observed and cytotoxic T-cell responses
increased. In contrast, another study using a different arg1 inhibitor, namely N(G)-nitro-L -arginine methyl ester
(L-NAME), reported no increase in endogenous antitumor
immunity [63]. Despite a signiicant increase in cytotoxic
T cells, differences observed in CD8+ T cells were not as
pronounced as anticipated. The lack hereof can probably
be attributed to the experimental design. Treatment was
ceased at day 10 and subsequent tumor growth might have
allowed the ”reconstitution” of the tumor environment.
This experimental setup was chosen, as initially we
were more interested in tumor development and overall
survival. Another possibility is that inhibition of MDSC
immunosuppressive activities may not alter the iniltration
of immune cells (which would depend on cell traficking),
but rather their antitumor properties. We strongly believe
that the immune signature of the tumor during treatment
differs, as characterized by the increase in cytotoxic T
lymphocytes during treatment. No inhibition in tumor
growth was observed when similar experiments, in which
arg-1 was inhibited, were performed in mice lacking
functional T and B cells [64].This suggests that the growth
delay of the tumor caused by the inhibition of arg-1 is
at least partially dependent on the immune system. We
are also aware that only one mechanism, either arg-1 or
iNOS, is studied in this setting. Still, we can modulate the
tumor environment leading to enhanced antitumor T cell
responses. We believe that these T-cell responses could be
further potentiated through combination therapy, ideally
a mix of inhibitory (for example anti-MDSC drugs) and
immunostimulatory molecules (for example vaccination).
The predictive value of in vitro MDSC still has to
be examined further, but preliminary results already give
a good indication about future possibilities of this ex vivo
differentiation system.
Production and characterization of lentiviral
vectors encoding GM-CSF
The packaging plasmid pCMVΔR8.9 and VSV.G
encoding plasmid pMD.G were a gift from Dr. D. Trono
(University of Geneva). The plasmid encoding GMCSF and the puromycin resistance gene was previously
described [28]. The production and characterization of
lentiviral vectors was performed as described before [65].
Transduction of CT26 cells with lentiviral vectors
encoding GM-CSF
Tumor cells, namely CT26, which over express
GM-CSF, were generated by transduction with lentiviral
vectors encoding for both mouse GM-CSF and the
puromycin resistance gene. To that end 2 x 105 CT26
cells were plated in 2 ml culture medium in a 6-well.
One day later, the culture medium was replaced with
2 ml of the lentiviral transduction cocktail containing
15 infectious lentiviral particles per cell and 10 µg/ml
protamine sulphate (Leo Pharma, Lier, Belgium). Three
days later, transduced cells were selected using 3 μg/ml
puromycin (Sigma-Aldrich, Diegem, Belgium). To collect
conditioned medium (CM), cells were plated at 10 x 106
cells in 25 ml culture medium in a T175 cm2 and kept in
culture in the absence of puromycin for 3 days. To verify
the production of GM-CSF, CM was used to culture
FDCP-1 cells. To quantify the amount of GM-CSF, an
ELISA (eBioScience, Vienna, Austria) was performed
according to manufacturer’s instructions.
In vitro-differentiation
suppressor cells
myeloid-derived
Bone marrow cells were extracted from the femur
and tibia of Balb/c mice, after which 10 x 106 bone
marrow cells were cultured in 75% CM and 25% Iscove’s
Modiied Dulbecco’s medium (IMDM, Sigma-Aldrich)
supplemented with 10% fetal clone I (FCI, GE Health Care
Life Sciences, Hyclone Laboratories, Utah, USA), 100 U/
ml penicillin, 100 μg/ml streptomycin (Sigma-Aldrich)
and 2 mM L-glutamine (Sigma-Aldrich) for 6 days. Cell
viability and cell numbers were evaluated by trypan blue
staining (Sigma-Aldrich). To evaluate the morphology of
the cells, 5 x 105 sorted MDSC were ixed on glass slides
using the cytospin technique and were centrifuged at a
speed of 1000 rpm for 5 minutes. Cytospin slides, ilter
cards, sample chambers, and metal clips were all obtained
MATERIALS AND METHODS
Mice and cell lines
Female, 6 to 8 weeks old Balb/c mice (Charles
River Laboratories, L’Arbresle Cedex, France) were
treated according to the European guidelines for animal
experimentation. Experiments were reviewed by the
Ethical Committee for use of laboratory animals of the
Vrije Universiteit Brussel (Jette, Belgium). The mouse
www.impactjournals.com/oncotarget
of
9
Oncotarget
cell suspension was incubated at 37°C, 5% CO2 for 10
minutes, washed in serum free Optimem (Invitrogen, Life
Technologies), centrifuged for 7 minutes at 1500 rpm and
suspended in 5 ml Optimem. Cells were plated at 1 x 105
cells in 100 μl in a 96-well. Subsequently, the cells were
either left unstimulated or were stimulated with a 1/800
dilution of anti-CD3/anti-CD28 coated beads (Invitrogen).
Enriched Ly6G+ or Ly6C+ MDSC were obtained using the
Myeloid-Derived Suppressor Cell Isolation Kit (Miltenyi
Biotec). Sorted MDSC were added to the stimulated
T cells at the indicated MDSC to T cell ratios. When
indicated, speciic inhibitors for arg-1 (Nω-hydroxy-norArginine (Nor-NOHA), 300 μM) (Enzo Life Sciences,
Antwerpen, Belgium) or iNOS (aminoguanidine (AG), 1
mM) (Sigma-Aldrich) were added. Dilution of CFSE was
evaluated 3 days later by low cytometry as a measure of
T-cell proliferation. To that end, T cells were additionally
stained with anti-CD3-PercP-Cy5.5 (Biolegend). Data
were collected using the FACSCanto Flow Cytometer
(Becton Dickinson) and were analyzed with FlowJo 7.6
(Treestar Inc.). During the analysis, cells were gated
according to their forward and side scatter distribution
and to CD3 expression. Alternatively, supernatants were
collected and screened for IFN-γ content using a standard
ELISA (Thermo scientiic) according to manufacturer’s
instructions.
from Thermo scientiic (Massachusetts, USA). Cytospins
were air dried for 2 hours and afterwards stained with
hematoxylin and eosin.
Isolation of in vivo differentiated myeloid-derived
suppressor cells
In order to grow tumors, Balb/c mice received a
subcutaneous injection of 1 x 105 CT26 tumor cells. When
the tumor diameter exceeded 15 mm, mice were sacriiced
and single cell suspensions from the tumor and spleen
were obtained as previously described [66]. To enrich
Ly6G+ or Ly6C+ MDSC, we sorted the MDSC using the
Myeloid-Derived Suppressor Cell Isolation Kit according
to the manufacturer’s instructions (Miltenyi Biotec,
Bergisch-Gladbach, Germany).
Cell staining and low cytometry
Staining of cell surface markers was performed
as described [67]. The following antibodies were used:
anti-CD11b-eF450, anti-MHC II-PE, anti-F4/80-APC-H7
(eBioScience), anti-CD11b-FITC, anti-Ly6G-AF647,
anti-Ly6C-Pe-Cy7, anti-CD80-BV421, anti-PD-L1-PE,
anti-CD3-PercP-Cy5.5, anti-CD11c-AF647 (Biolegend,
London, United Kingdom), anti-Ly6G-PE-CF594, antiCD8-FITC (Becton Dickinson, Erembodegem, Belgium)
and anti-CD45-VioBlue (130-102-775) (Mitenyi Biotec).
For intracellular staining, cells were treated with inside
FIX and incubated for 20 minutes at room temperature.
Cells were incubated with PERM (Miltenyi Biotec)
and the anti-arginase-1-PE (R&D systems, Abingdon,
United Kingdom) or anti-iNOS-PercP-Cy5.5 (Santa
Cruz Biotechnology, Heidelberg, Germany) antibody
for 20 minutes at room temperature. Subsequently, cells
were washed. Cells stained with isotype matched control
antibodies served as a control. Cells were acquired using
the LSR Fortessa (Becton Dickinson) and analysis was
performed using FlowJo 7.6 (Treestar Inc Oregon, United
States of America).
Therapy
Balb/c mice received a subcutaneous injection of
1 x 105 CT26 tumor cells. When the tumors reached a
diameter of 6 mm, mice were treated for 10 consecutive
days with an intratumoral injection of 50 µl Nor-NOHA
(80 mg/kg) (Enzo Life Sciences), 1400W hydrochloride
(20 mg/kg) (Sigma- Aldrich) or PBS. The tumor volume
was measured on a daily basis using a caliper. Mice were
sacriiced by neck dislocation when the tumor diameter
exceeded 15 mm.
In vivo cytotoxicity assay
An in vivo cytotoxicity assay was performed after
5 consecutive treatments with Nor-NOHA, 1400W or
PBS (see above) to evaluate the stimulation of cytotoxic
antitumor immune responses. The assay was performed
as described by Van Lint et al. [68] using gp70 peptide
(Thermo Electron GmbH, Ulm, Germany) pulsed cells as
targets.
In vitro T-cell suppression assay
To evaluate the suppressive activity of MDSC, we
performed an in vitro T-cell suppression assay. To that
end, CD8+ T lymphocytes were isolated from the spleen
of Balb/c mice using the CD8α+ T cell Isolation Kit II
(Miltenyi Biotec). These CD8α+ T lymphocytes were labeled
with carboxyluorescein diacetate succinimidyl ester
(CFSE, Life Technologies, Gent, Belgium). First, cells
were washed and suspended in 5 ml phosphate buffered
saline (PBS, Sigma-Aldrich) containing 0.1% bovine
serum albumin (BSA, Life Technologies). Five ml of 0.5
μM CFSE were added to the cell suspension. The single
www.impactjournals.com/oncotarget
ELISPOT
Lymph nodes (LN) and spleens were isolated and
single cell suspensions were prepared as described before
by Goyvaerts et al. [66]. Enzyme-linked immunospot
10
Oncotarget
(ELISPOT) plates (Millipore, Brussels, Belgium) were
coated with 100 μl puriied anti-IFN-γ antibodies and
incubated overnight at 4°C. Wells were then blocked with
100 μl blocking buffer (RPMI supplemented with 10%
FCI). A total of 1 x 105 MACS sorted CD8+ splenocytes
(Miltenyi Biotec.) or 2 x 105 unsorted LN cells were plated
per well (in duplicate). Cells were either left unstimulated
and served as a negative control or cells were treated with
gp70 peptide. Concanavalin A (Sigma- Aldrich) stimulated
T cells served as a positive control. ELISPOT plates were
incubated at 37°C, 5% CO for 24h. Next, the ELISPOT
plates were developed according to the manufacturer’s
instructions (Diaclone, Besaņon, France). Spots were
counted using an ELISPOT counter (Autoimmun
Diagnostika GmbH, Straβberg, Germany) and software
(Autoimmun Diagnostika ELISPOT Reader 5.0).
CONFLICTS OF INTEREST
We declare that all authors participated in the
proposed work and that this manuscript is not under
consideration for review elsewhere. The contents of this
manuscript will not be copyright, submitted or published
elsewhere while acceptance by Oncotarget is under
consideration. David Escors and Therese Liechtenstein
are inventors of the MDSC production method (European
patent 14166221.3-1405). The other authors declare no
potential conlicts of interest.
REFERENCES
1.
Pagès F, Berger A, Camus M, Sanchez-Cabo F, Costes
Anne, Molidor R, Mlecnik B, Kirilovsky A, Nilsson
M, Damotte D, Meatchi T, Bruneval P, Cugnenc P et al.
Effector memory T cells, early metastasis, and survival in
colorectal cancer. N Engl J Med. 2005; 353: 2654–66.
A t-test and a one-way ANOVA followed by a
Bonferroni’s multiple comparison test was performed.
Sample sizes and number of times experiments were
repeated are indicated in the igure legends. Number of
asterisks in the igures indicates the level of statistical
signiicance as follows: *, p < 0.05; **, p < 0.01; ***, p <
0.001. The results are shown in column graphs as the mean
± standard error of the mean (SEM). Survival of mice in
the therapy experiment was visualized in a Kaplan-Meier
plot. Differences in survival were analyzed by the log-rank
test. All data were analyzed using GraphPad Prism 5.
2.
Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik
B, Lagorce-Pagès C, Tosolini M, Camus M, Berger A,
Wind P, Zinzindohoué F, Bruneval P, Cugnenc P et al.
Type, density, and location of immune cells within human
colorectal tumors predict clinical outcome. Science. 2006;
313: 1960–64.
3.
Tosolini M, Kirilovsky A, Mlecnik B, Fredriksen T,
Mauger S, Bindea G, Berger A, Bruneval P, Fridman W,
Pagès F and Galon J. Clinical impact of different classes
of iniltrating T cytotoxic and helper cells (Th1, Th2, Treg,
Th17) in Patients with colorectal Cancer. Cancer Res. 2011;
71: 1263–71.
ACKNOWLEDGEMENTS
4.
Zhang B, Wang Z, Wu L, Zhang M, Li W, Ding J, Zhu
J, Wei H and Zhao K. Circulating and tumor-iniltrating
myeloid-derived suppressor cells in patients with colorectal
carcinoma. PLoS ONE. 2013; 8: e57114.
5.
Sermeus A, Leonard W, Engels B and De Ridder M.
Advances in radiotherapy and targeted therapies for rectal
cancer. World J Gastroenterol. 2014; 20: 1–5.
6.
Lechner MG, Karimi SS, Barry-Holson K, Angell TE,
Murphy KA, Church CH, Ohlfest JR, Hu P and Epstein AL.
Immunogenicity of murine solid tumor models as a deining
feature of in vivo behavior and response to immunotherapy.
J Immunother. 2013; 36: 477–89.
7.
Gabrilovich DI and Nagaraj S. Myeloid-derived suppressor
cells as regulators of the immune system. Nat Rev Immunol.
2009; 9: 162–74.
8.
Lechner MG, Liebertz DJ and Epstein AL. Characterization
of cytokine-induced myeloid-derived suppressor cells
from normal human peripheral blood mononuclear cells. J
Immunol. 2010; 185: 2273–84.
9.
Talmadge JE. Pathways mediating the expansion and
immunosuppressive activity of myeloid-derived suppressor
cells and their relevance to cancer therapy. Clin Cancer Res.
2007; 13: 5243–48.
Statistical analyses
We thank Ka Lun Law and Angelo Willems for their
technical assistance.
This work was supported by grants from the
Interuniversity Attraction Poles Program-Belgian StateBelgian Science Policy, the National Cancer Plan of the
Federal Ministry of Health, the “Stichting tegen Kanker”,
the “Vlaamse Liga tegen Kanker”, an Integrated Project
and EU FP7-funded Network of Excellence, an IWTTBM program, the “FWO-Vlaanderen”, the Scientiic
Fund Willy Gepts of the University Hospital Brussels.
ID is funded by the IWT. TL is funded by a University
College London Overseas PhD Scholarship. WL is funded
by Scientiic Fund Willy Gepts of the University Hospital
Brussels. DE has been funded by an Arthritis Research UK
Career Development Fellowship (18433) and currently by
a Miguel Servet Fellowship from the Instituto de Salud
Carlos III, Spain. MDR is funded by the VLK.
David Escors and Therese Liechtenstein are
inventors of the MDSC production method (European
patent 14166221.3-1405).
www.impactjournals.com/oncotarget
11
Oncotarget
10. Gabrilovich DI, Ostrand-Rosenberg S and Bronte V.
Coordinated regulation of myeloid cells by tumours. Nat
Rev Immunol. 2012; 12: 253–68.
21. Youn J-I, Nagaraj S, Collazo M and Gabrilovich DI.
Subsets of myeloid-derived suppressor cells in tumorbearing mice. J Immunol. 2008; 181: 5791–802.
11. Vincent J, Mignot G, Chalmin F, Ladoire S, Bruchard M,
Chevriaux A, Martin F, Apetoh L, Rébé C and Ghiringhelli
F. 5-Fluorouracil selectively kills tumor-associated
myeloid-derived suppressor cells resulting in enhanced T
cell-dependent antitumor immunity. Cancer Res. 2010; 70:
3052–61.
22. Marigo I, Bosio E, Solito S, Mesa C, Fernandez A, Dolcetti
L, Ugel S, Sonda N, Bicciato S, Falisi E, Calabrese F, Basso
G, Zanovello P et al. Tumor-induced tolerance and immune
suppression depend on the C/EBPbeta transcription Factor.
Immunity. 2010; 32: 790–802.
23. Zhou Z, French DL, Ma G, Eisenstein S, Cheng Y, Divino
CM, Keller G, Chen S and Pan P. Development and
function of myeloid-derived suppressor cells generated
from mouse embryonic and hematopoietic stem cells. Stem
Cells. 2010; 28: 620-32.
12. Suzuki E, Kapoor V, Jassar AS, Kaiser LR and Albelda
SM. Gemcitabine selectively eliminates splenic Gr-1+/
CD11b+ myeloid suppressor cells in tumor-bearing animals
and enhances antitumor immune activity. Clin Cancer Res.
2005; 11: 6713–21.
24. Obermajer N and Kalinski P. Generation of myeloidderived suppressor cells using prostaglandin E2. Transplant
Res. 2012; 1: 15.
13. Le HK, Graham L, Cha E, Morales JK and Bear HD.
Gemcitabine directly inhibits myeloid derived suppressor
cells in BALB/c mice bearing 4T1 mammary carcinoma
and augments expansion of T cells from tumor-bearing
mice. Int Immunopharmacol. 2009; 9: 900–09.
25. Xiang X, Poliakov A, Liu C, Liu Y, Deng Z, Wang J, Cheng
Z, Shah SV, Wang G, Zhang L, Grizzle WE, Mobley J and
Zhang H. Induction of myeloid-derived suppressor cells by
tumor exosomes. Int J Cancer. 2008; 124: 2621–33.
14. Kodumudi KN, Woan K, Gilvary DL, Sahakian E, Wei S
and Djeu JY. A novel chemoimmunomodulating property
of docetaxel: suppression of myeloid-derived suppressor
cells in tumor bearers. Clin Cancer Res. 2010; 16: 4583–94.
26. Highill SL, Rodriguez PC, Zhou Q, Goetz CA, Koehn BH,
Veenstra R, taylor PA, Panoskaltsis-Mortari A, Serody JS,
Munn DV, Tolar J, Ochoa AC and Blazer BR. Bone marrow
myeloid-derived suppressor cells (MDSCs) inhibit graftversus-host disease (GVHD) via an arginase-1-dependent
mechanism that is up-regulated by interleukin-13. Blood.
2010; 116: 5738–47.
15. Maenhout SK, Van Lint S, Emeagi PU, Thielemans K
and Aerts JL. Enhanced suppressive capacity of tumoriniltrating myeloid-derived suppressor cells compared with
their peripheral counterparts. Int J Cancer. 2014; 134: 107790.
27. Escors D, Liechtenstein T, Perez-Janices N, Schwarze
J, Dufait I, Goyvaerts C, Lanna A, Arce F, BlancoLuquin I, Kochan G, Guerrero-Setas D and Breckpot K.
Assessing T-cell responses in anticancer immunotherapy:
Dendritic cells or myeloid-derived suppressor cells?
Oncoimmunology. 2013; 2: e26148.
16. Corzo CA, Condamine T, Lu L, Cotter MJ, Youn J, Cheng
P, Cho H, Celis E, Quiceno DG, Padhya T, McCaffery TV,
McCaffrey JC and Gabrilovich D. HIF-1 regulates function
and differentiation of myeloid-derived suppressor cells in
the tumor microenvironment. J Exp Med. 2010; 207: 2439–
53.
28. Liechtenstein T, Perez-Janices N, Gato M Caliendo F,
Kochan G, Blanco-Luquin A, Van Der Jeught K, Arce F,
Guerrero-Setas D, Fernandez-Irigoyen J, Santamari
E, Breckpot K and Escors D. A highly eficient tumoriniltrating MDSC differentiation system for discovery of
anti-neoplastic targets, which circumvents the need for
tumor establishment in mice. Oncotarget. 2014; 5: 7843–57.
17. Thaci B, Ahmed AU, Ulasov IV, Wainwright DA, Nigam
P, Aufinger B, Tobias AL, Han Y, Zhang L, Moon KS and
Lesniak MS. Depletion of myeloid-derived suppressor cells
during interleukin-12 immunogene therapy does not confer
a survival advantage in experimental malignant glioma.
Cancer Gene Ther. 2014; 21: 38–44.
18. Apolloni E, Bronte V, Mazzoni A, Seraini P, Cabrelle
A, Segal DM, Young A and Zanovello P. Immortalized
myeloid suppressor cells trigger apoptosis in antigenactivated T lymphocytes. J Immunol. 2000; 165: 6723–30.
29. Solito S, Falisi E, Diaz-Montero CM, Doni A, Pinton L,
Rosata A, Francescato S, Basso G, Zanovello P, Onicescu
G, Garrett-Mayer E, Montero AJ, Bronte V et al. A
human promyelocytic-like population is responsible for
the immune suppression mediated by myeloid-derived
suppressor cells. Blood. 2011; 118: 2254–65.
19. Lutz M, Kukutsch N, Menges M, Rössner S and Schuler G.
Culture of bone marrow cells in GM-CSF plus high doses
of lipopolysaccharide generates exclusively immature
dendritic cells which induce alloantigen- speciic CD4 T
cell anergy in vitro. Eur J Immunol. 2000; 30: 1048-52.
30. Sun HL, Zhou X, Xue YF, Wang K, Shen Y, Mao J, Guo H
and Miao Z. Increased frequency and clinical signiicance
of myeloid-derived suppressor cells in human colorectal
carcinoma. World J Gastroenterol. 2012; 18: 3303–09.
20. Morales JK, Kmieciak M, Knutson KL, Bear HD and
Manjili MH. GM-CSF is one of the main breast tumorderived soluble factors involved in the differentiation of
CD11b-Gr1- bone marrow progenitor cells into myeloidderived suppressor cells. Breast Cancer Res Treat. 2009;
123: 39–49.
www.impactjournals.com/oncotarget
31. Choi J, Suh B, Ahn YO, Kim TM, Lee JO, Lee SH and Heo
DS. CD15+/CD16low human granulocytes from terminal
cancer patients: granulocytic myeloid-derived suppressor
cells that have suppressive function. Tumor Biol. 2011; 33:
121–29.
12
Oncotarget
32. Kusmartsev S, Nagaraj S and Gabrilovich DI. Tumorassociated CD8+ T cell tolerance induced by bone marrowderived immature myeloid cells. J Immunol. 2005; 175:
4583–92.
Immunol. 2010; 22: 238–44.
44. Van der Jeught K, Tjok Joe P, Bialkowski L, Heirman C,
Daszkiewicz L, Liechtenstein T, Escors D, Thielemans K
and Breckpot K. Intratumoral administration of mRNA
encoding a fusokine consisting of IFN-β and the ectodomain
of the TGF-β receptor II potentiates antitumor immunity.
Oncotarget. 2014; 5: 10100-13.
33. Bronte V, Chappell DB, Apolloni E, Cabrelle A, Wang
M, Hwu P and Restifo NP. Unopposed production of
granulocyte-macrophage colony-stimulating factor by
tumors inhibits CD8+ T cell responses by dysregulating
antigen-presenting cell maturation. J Immunol. 1999; 162:
5728–37.
45.
34. Gabrilovich DI, Bronte V, Chen SH, Colombo MP, Ochoa
A, Ostrand-Rosenberg S and Schreiber H. The terminology
issue for myeloid-derived suppressor cells. Cancer Res.
2007; 67: 425.
46. Lang RA, Metcalf D, Cuthbertson RA, Lyons I, Stanley E,
Kelso A, Kannourakis G, Williamson DJ, Klinthworth GK
and Gonda TJ. Transgenic mice expressing a hemopoietic
growth factor gene (GM-CSF) develop accumulations of
macrophages, blindness, and a fatal syndrome of tissue
damage. Cell. 1987; 51: 675-86.
35. Ostrand-Rosenberg S and Sinha P. Myeloid-derived
suppressor cells: linking inlammation and cancer. J
Immunol. 2009; 182: 4499–506.
36. Seraini P, Borrello I and Bronte V. Myeloid suppressor
cells in cancer: Recruitment, phenotype, properties, and
mechanisms of immune suppression. Seminars in Cancer
Biology. 2006; 16: 53–65.
47. Dranoff G, Jaffee E, Lazenby A, Golumbek P, Levitsky H,
Brose K, Jackson V, Hamada H, Pardoll D and Mulligan
RC. Vaccination with irradiated tumor cells engineered to
secrete murine granulocyte-macrophage colony-stimulating
factor stimulates potent, speciic, and long-lasting antitumor immunity. Proc Natl Acad Sci USA. 1993; 90: 353943.
37. Ribechini E, Greifenberg V, Sandwick S and Lutz MB.
Subsets, expansion and activation of myeloid-derived
suppressor cells. Med Microbiol Immunol. 2010; 199:
273–81.
48. Lesokhin AM, Hohl TM, Kitano S, Cortez C, HirschhornCymerman D, Avogadri F, Rizzuto GA, Lazarus JJ,
Pamer EG, Houghton AN, Merghoub T and Wolchok
JD. Monocytic CCR2+ myeloid-derived suppressor cells
promote immune escape by limiting activated CD8 T-cell
iniltration into the tumor microenvironment. Cancer Res.
2012; 72: 876-86.
38. Zheng Y, Xu M, Li X, Jia J, Fan K and Lai G. Cimetidine
suppresses lung tumor growth in mice through proapoptosis
of myeloid-derived suppressor cells. Mol Immunol. 2013;
54: 74–83.
39. Mao Y, Sarhan D, Steven A, Seliger B, kiessling R and
Lundqvist A. Inhibition of tumor-derived prostaglandin-E2
blocks the induction of myeloid-derived suppressor cells
and recovers natural killer cell activity. Clin Cancer Res.
2014; 20: 4096-106.
49. Ruff MR, Farrar WL and Pert CB. Interferon gamma and
granulocyte/macrophage colony-stimulating factor inhibit
growth and induce antigens characteristic of myeloid
differentiation in small-cell lung cancer cell lines. Proc Natl
Acad Sci USA. 1986; 83: 6613-17.
40. Mikysková R, Indrová M, Vlková V, Bieblová J, Simová
J, Paracková Z, Pajtasz-Piasecka E, Rossowskai J and
Reinis M. DNA demethylating agent 5-azacytidine inhibits
myeloid-derived suppressor cells induced by tumor growth
and cyclophosphamide treatment. J Leukoc Biol. 2014; 95:
743- 53.
50. Yamashita Y, Nara N and Aoki N. Antiproliferative and
differentiative effect of granulocyte-macrophage colonystimulating factor on a variant human small cell lung cancer
cell line. Cancer Res. 1989; 49: 5334-8.
41. Alizadeh D, Trad M, Hanke NT, Larmonier CB,
Janikashvill N, Bonnotte B, Katsanis E and Larmonier N.
Doxorubicin eliminates myeloid-derived suppressor cells
and enhances the eficacy of adoptive T-cell transfer in
breast cancer. Cancer Res. 2014; 74: 104–18.
51. Berdel WE, Danhauser-Riedl S, Steinhauser G and
Winton EF. Various human hematopoietic growth factors
(interleukin-3, GM-CSF, G-CSF) stimulate clonal growth
of nonhematopoietic tumor cells. Blood. 1989; 73: 80-83.
52. Dedhar S, Gaboury L, Galloway P and Eaves C. Human
granulocyte-macrophage
colony-stimulating
factor
is a growth factor active on a variety of cell types of
nonhemopoietic origin. Proc Natl Acad Sci USA. 1988; 85:
9253-57.
42. Chakraborty P, Das S, Banerjee K, Sinha A, Roy S,
Chatterjee M and Choudhuri SK. A copper chelate
selectively triggers apoptosis in myeloid-derived
suppressor cells in a drug-resistant tumor model and
enhances antitumor immune response. Immunopharmacol
Immunotoxicol. 2014; 36: 165–75.
53. Berdel WE, Zafferani M, Senekowitsch R, Kreuser ED
and Thiel E. Effect of interleukin-3 and granulocytemacrophage colony-stimulating factor on growth of
xenotransplanted human tumour cell lines in nude mice.
Eur J Cancer. 1992; 28-377-80.
43. Peranzoni E, Zilio S, Marigo I, Dolcetti L, Zanovello P,
Mandruzzato S and Bronte V. Myeloid-derived suppressor
cell heterogeneity and subset deinition. Curr Opin
www.impactjournals.com/oncotarget
Burke B, Pridmore A, Harraghy N, Collick A, Brown J
and Mitchell T. Transgenic mice showing inlammationinducible overexpression of granulocyte macrophage
colony-stimulating factor. Clin Diagn Lab Immunol. 2004;
11: 588-98.
13
Oncotarget
54. Foulke RS, Marshall MH, Trotta PP and Von Hoff DD. In
vitro assessment of the effects of granulocyte-macrophage
colony-stimulating factor on primary human tumors and
derived lines. Cancer Res. 1990; 50: 6264-67.
Robays L, Heirman C, Reiser J, Zhang XY, De Baestelier
P, Raes G and Breckpot K. Development of the Nanobody
display technology to target lentiviral vectors to antigenpresenting cells. Gene Therapy. 2012; 19: 1133–40.
55. Obermueller E, Vosseler S, Fusenig NE and Mueller
MM. Cooperative autocrine and paracrine functions of
granulocyte colony-stimulating factor and granulocytemacrophage colony-stimulating factor in the progression of
skin carcinoma cells. Cancer Res. 2001; 64: 7801-12.
66. Emeagi PU, Thielemans K and Breckpot K. The role of
SMAC mimetics in regulation of tumor cell death and
immunity. Oncoimmunology. 2012; 1: 965–67.
67. Breckpot K, Dullaers M, Bonehill A, Meirvenne SV,
Heirman C, Greef CD, Van der bruggen P and Thielemans
K. Lentivirally transduced dendritic cells as a tool for
cancer immunotherapy. J Gene Med. 2003; 5: 654–67.
56. Gutschalk CM, Herold-Mende CC, Fusening NE and
Mueller MM. Granulocyte Colony-Stimulating Factor
and Granulocyte-Macrophage Colony-Stimulating Factor
Promote Malignant Growth of Cells from Head and Neck
Squamous Cell Carcinomas In vivo. Cancer Res. 2006; 66:
8026-36.
68. Van Lint S, Goyvaerts C, Maenhout S, Goethals L, Disy A,
Benteyn D, Pen J, Bonehill A, Heirman C, Breckpot K and
Thielemans K. Preclinical evaluation of triMix and antigen
mRNA-based antitumor therapy. Cancer Res. 2012; 72:
1661–71.
57. Gutschalk CM, Yanamandra AK, Linde N, Meides A,
Depner S and Mueller MM. GM-CSF enhances tumor
invasion by elevated MMP-2, -9, and -26 expression.
Cancer Med. 2012; 2: 117-29.
58. Nebiker C, Han J, Eppenberger-Castori S, Iezzi G,
Amicarella F, Cremonesi E, Huber X, Padovan E, Angrisani
B, Droeser RA, Rosso R, Bolli M, Oertli D et al. GM-CSF
production by tumor cells is associated with improved
survival in colorectal cancer. Clin Cancer Res. 2014; 20:
3094-106.
59. Le DT, Pardoll DM and Jaffee EM. Cellular vaccine
approaches. Cancer Journal. 2010; 16: 304-10.
60. Faries MB, Hsueh EC, Ye X, Hoban M and Morton DL.
Effect of granulocyte/macrophage colony-stimulating factor
on vaccination with an allogeneic whole-cell melanoma
vaccine. Clin Cancer Res. 2009; 15: 7029-35.
61. Slingluff CL, Petroni GR, Olson WC, Smolkin ME, Ross
MI, Haas NB, Grosh WW, Boisvert ME, Kirkwood JM and
Chianese-Bullock KA. Effect of granulocyte/macrophage
colony-stimulating factor on circulating CD8+ and CD4+
T-cell responses to a multipeptide melanoma vaccine:
outcome of a multicenter randomized trial. Clin Cancer Res.
2009; 15: 7036-44.
62. Jayaraman P, Alfarano MG, Svider PF, Parikh F, Lu G,
Xiong H and Sikora AG. iNOS expression in CD4+ T cells
limits T-reg induction by repressing TGFb-1: combined
iNOS inhibition and T-reg depletion unmask endogenous
anti-tumor immunity. Clin Cancer Res. 2014; 20(23).
63. Rigamonti N, Capuano G, Ricupito A, Jachetti E, Grioni
M, Generoso L, Freschi M and Bellone M. Modulators of
arginine metabolism do not impact on peripheral T-cell
tolerance and disease progression in a model of spontaneous
prostate cancer. Clin Cancer Res. 2011; 17: 1012–23.
64.
Rodriguez PC, Ernstoff MS, Hernandez C Atkins M,
Zabaleta J, Sierra R and Ochoa AC. Arginase I-producing
myeloid-derived suppressor cells in renal cell carcinoma
are a subpopulation of activated granulocytes. Cancer Res.
2009; 69: 1553–60.
65. Goyvaerts C, De Groeve K, Dingemans J, Van Lint S,
www.impactjournals.com/oncotarget
14
Oncotarget