Search tips
Search criteria 


Logo of neuroncolAboutAuthor GuidelinesEditorial BoardNeuro-Oncology
Neuro Oncol. 2012 May; 14(5): 584–595.
Published online 2012 March 9. doi:  10.1093/neuonc/nos014
PMCID: PMC3337302

Soluble factors secreted by glioblastoma cell lines facilitate recruitment, survival, and expansion of regulatory T cells: implications for immunotherapy


In patients with glioma, the tumor microenvironment can significantly impact pro-inflammatory immune cell functions. However, the mechanisms by which this occurs are poorly defined. Because immunosuppressive regulatory T cells (Treg) are over represented in the tumor microenvironment compared with peripheral blood, we hypothesized that the tumor may have an effect on Treg survival, migration, expansion, and/or induction of a regulatory phenotype from non-Treg conventional CD4+ T cells. We defined the impact of soluble factors produced by tumor cells on Treg from healthy patients in vitro to determine mechanisms by which gliomas influence T cell populations. We found that tumor-derived soluble factors allowed for preferential proliferation and increased chemotaxis of Treg, compared with conventional T cells, indicating that these mechanisms may contribute to the increased Treg in the tumor microenvironment. Conventional T cells also exhibited a significantly increased expression of pro-apoptotic transcripts in the presence of tumor-derived factors, indicating that survival of Treg in the tumor site is driven by exposure to soluble factors produced by the tumor. Together, these data suggest that tumor burden may induce increased Treg infiltration, proliferation, and survival, negating productive anti-tumor immune responses in patients treated with immunotherapies. Collectively, our data indicate that several mechanisms of Treg recruitment and retention in the tumor microenvironment exist and may need to be addressed to improve the specificity of immunotherapies seeking to eliminate Treg in patients with glioma.

Keywords: glioma, immunotherapy, regulatory T cell, tumor microenvironment

Traditional treatment of patients with glioblastoma multiforme (GBM) using chemotherapy and radiation can have a devastating effect on a patient's quality of life, making immunotherapy an appealing alternative. The major strength of targeting tumor cells using immunotherapy is the potential to specifically target tumor cells without collateral damage to healthy neural cells. Several studies using mouse models have indicated that immune cell responses to tumor cells are critical to the early elimination of tumor cells.1 Studies using mice deficient in T cells and natural killer (NK) cells or the proinflammatory cytokines that they produce have demonstrated the importance of an intact immune system in tumor surveillance in vivo.24 However, establishment of ex vivo and peripheral immune responses does not guarantee improved clinical outcome.57 Therefore, understanding the mechanisms for poor clinical efficacy of early generation immunotherapies has led researchers to evaluate the impact of the tumor microenvironment on immune cell functions that may inhibit immune responses to the tumor in situ.810

Among cells known to impair productive anti-tumor immunity in many tumor microenvironments are CD4 + CD25 + FoxP3 + Treg.1113 Treg are an endogenous subset of T cells, which play an important role in maintaining peripheral self-tolerance and preventing possible immunopathologies by mitigating potential and ongoing immune reactions. Treg constitute 5%–10% of the peripheral CD4+ T cell population in healthy patients and directly counteract the proliferation and cytokine production by conventional T cells.14 In patients with high-grade glioma, reports demonstrate that up to 60% of the tumor-infiltrating lymphocyte (TIL) population is composed of Treg.15,16 These studies suggest that immunotherapies relying on the effector functions of conventional T cells may be improved by eliminating this population in the tumor microenvironment. Indeed, depletion of Treg has been shown to increase tumor immunity17 and improve survival,18 indicating a role for Treg in preventing immunity to tumors.

Although it has been demonstrated that Treg cell frequencies are vastly overrepresented in the tumor microenvironment, compared with those in circulation,15 the mechanism of recruitment and expansion and/or survival of these cells is controversial. Previous reports have shown the potential for preferential attraction of Treg by chemokines secreted by various tumor types and that tumor-derived chemokines can increase Treg infiltration;12,19 however, it is not clear whether these chemokines are necessary or sufficient to attract a Treg population to the tumor. Glioma-derived TGFβ has also been shown to prevent conventional effector T cell migration to the tumor site, which is associated with a poor prognosis in patients with GBM,20 suggesting a mechanism for Treg overrepresentation in the tumor microenvironment and failure of local anti-tumor responses. More recently, it has been demonstrated that myelomonocytic cells present in the tumor microenvironment of patients with glioma may also promote Treg induction.21 Together, these studies indicate that a tumor may rely on various and potentially redundant mechanisms to recruit and retain Treg within the tumor microenvironment, preventing productive effector T cell functions.

These studies collectively suggest that Treg play a pivotal role in the failure of immune therapies; therefore, we asked whether Treg preferentially migrate, proliferate, or survive after exposure to soluble factors produced by glioma cells. Furthermore, we examined the possibility that a conventional T cell can be converted to a Treg after stimulation in the presence of these secreted factors, as has been recently demonstrated.22,23 Our data suggest that local conversion of conventional T cells to Treg does not significantly contribute to the observed increased frequency but that Treg cells in the tumor microenvironment may preferentially migrate, proliferate, and survive in response to factors that are produced by tumor cells. Collectively, our data support the notion that multiple mechanisms contribute to Treg accumulation in the tumor microenvironment in addition to preferential chemoattraction. Therefore, reduction of Treg in the tumor may be more effective if immunotherapies target additional mechanisms, such as preferential survival and proliferation of Treg, potenitally improving the clinical outcome of GBM treated with immunotherapies.

Methods and Materials

Clinical Specimens

Tumor tissue and blood samples were collected from patients with GBM during surgical resection. Blood samples were also collected from patients free of malignant glioma tumor burden during procedures, such as meningioma resection, as negative controls. Meningioma-resected patients were used as negative control subjects for 2 reasons. The first is that blood samples were collected from both patient subsets at the time of surgery, and accordingly, all patients will have been subjected to a standard algorithm of preoperative management including, but not limited to, antibiotic administration, a bolus of decadron, mannitol, and other medicatons, including anti-convulsant drugs. In addition, to obtain TILs used as comparison for Treg tumor infiltration, the patients must undergo intracranial surgery, limiting the availability of healthy control subjects. All patients gave informed consent for sample acquisition under the University of California, San Francisco (UCSF) Internal Review Board–approved Brain Tumor Research Center protocol CHR# 10-01271.

Cell Lines

Glioma cell lines U87, SF767, and U251 were obtained through the UCSF Brain Tumor Research Center. Cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM H-21) with 10% fetal bovine serum and 1% penicillin-streptomycin.

Tumor Conditioning

Tumor conditioned medium (TCM) was prepared by culturing cell lines in complete T cell medium (RPMI-1640 25 mM Hepes, 2.0 g/L NaHCO3 supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, 1 mM sodium pyruvate, and 10 mM nonessential amino acids) when confluent for 48 h. The conditioned medium was diluted 1:1 with fresh complete T cell medium to replenish nutrients and was used in subsequent assays.

TIL and Circulating Lymphocyte Isolation

Peripheral blood lymphocytes (PBL) were isolated from whole blood using a Ficoll-Paque Plus density gradient solution (GE HealthCare). TILs were isolated from tissue samples by mincing tumor samples and treating them with 1mg/mL Collagenase D (Sigma) for 30 min. The resulting slurry was subjected to a Percoll density gradient (Sigma), as previously described.24 Cells were resuspended in 70% isotonic Percoll (Amersham Biosciences) and centrifuged for 25 min at room temperature at 553 ×gon at a 30%:37%:70% Percoll gradient with densities of ρ = 1.065, ρ = 1.072, and ρ = 1.088, respectively. Treg were isolated from whole PBL using a negative CD4 T cell selection, followed by a CD25 positive selection (Stem Cell Technologies), according to the manufacturer's instructions. Conventional T cells were harvested using the negative fraction remaining from the CD25-positive T cell selection.

Flow Cytometry

T cells were surface stained with anti-human CD4 PerCP (BD Pharmingen) and CD25 APC (eBoscience) on ice for 25 min in PBS/2% BSA. Cells were fixed in 2% paraformaldehyde and permeabilized using CytoPerm (BD Pharmingen). Cells were then intracellularly stained in a saponin-containing buffer with FoxP3 FITC (eBioscience) and TGFβ PE (R&D Systems). Samples were analyzed on FACSCalibur Flow Cytometer using CellQuest Pro (BD Biosciences) and analyzed using FlowJo (Treestar) software.

T Cell Proliferation

Treg were stained using the CellTrace CFSE Cell Proliferation kit (Invitrogen) according to the manufacturer's instructions. In brief, selected Treg were stained at 37°C for 10 min with 10 μM CFSE, quenched with FBS, and washed 3 times with serum-containing media. Conventional CD4+ T cells were stained with PKH26 (Sigma-Aldrich), according to the manufacturer's protocol. In brief, conventional T cells were stained for 5 min at room temperature, quenched with FBS, and washed 3 times with serum-containing media. PKH-labeled T cells were plated alone or with CFSE labeled Treg in a 1:1 ratio for a total cell concentration of 106 cells/mL with 5 μg/mL purified anti-CD3 and 2 μg/mL purified anti-CD28 (eBioscience) in the presence of tumor-conditioned media. The cultures were incubated for 72 h at 37°C and 5% CO2. The cells were harvested and analyzed using flow cytometry, as described above.

T Cell Suppression Assay

Conventional T cells were stained with 10μM CFSE and stimulated with 5 μg/mL purified anti-CD3 and 2 μg/mL purified anti-CD28 in the presence of the indicated number of Treg cells. After 72-h incubation, cells were analyzed using flow cytometry for Treg markers CD3, CD4, CD25, and FoxP3.

T Cell Conversion Assay

Conventional T cells were cultured at 1 × 106 cells/mL in complete T cell medium or U87 TCM and stimulated with 5 μg/mL purified anti-CD3 and 2 μg/mL purified anti-CD28 (eBioscience). Cells were harvested on days 3, 5, 7, and 10; 10 U/mL IL-2 (107 U/mg) was added to cells every 48 h. Brefeldin A (BFA, 10 μg/mL; Sigma-Aldrich), Phorbol 12-myristate 13-acetate (PMA; 50 ng/mL), and ionomycin (500 ng/mL) were added to the coculture for the last 5 h of each time point. Cells were stained as described above and analyzed using flow cytometry.

Chemoattraction Assay

CFSE-labeled Treg and unlabeled conventional T cells were plated at 105 cells/well in the top chamber in 5-um pore polycarbonate 24-well transwell plates (Corning Incorporated). U87 and SF767 TCM, CCL22 (10 ng/mL or 50 ng/mL), or CCL5 at 30 ng/mL (R&D Systems) were seeded on the bottom chamber of the plates. Concentrations of the negative control, CCL5, were selected based on manufacturer-provided ED50 concentrations. CCL22 concentrations were also selected on the basis of manufacturer-provided ED50 and were chosen to be 1× (10 ng/mL) and 5× (50 ng/mL) concentrations of those used in previously published work.19 Twenty-four h later, cells were harvested, stained with anti-human CD45 APC (eBioscience), and analyzed using flow cytometry. Migration was expressed as percentage of total recovered cells found to migrate through pores and normalized to baseline migration.

Quantitative Polymerase Chain Reaction (PCR)

Whole cell mRNA was extracted using the RNeasy Mini Kit (Qiagen). Total RNA was then extracted and converted to cDNA by random hexamer priming and SuperScript III reverse transcriptase (100–250 ng RNA per reaction; Invitrogen). PCR amplification was performed with SYBR green master mix (5–10 ng cDNA per reaction; Applied Biosystems) via an iQ5 Real-Time PCR thermal cycler (BioRad). Reactions were performed in duplicate; Ct values were normalized to expression levels of hypoxanthine phosphoribosyltransferase (HPRT) and displayed as relative expression units (Table 1).

Table 1.
Primer sets used for quantitative polymerase chain reaction

Statistical Analysis

Data for all the figures were collected from at least 2 independent experiments performed in triplicate. Representative raw experimental data are shown to clarify gating strategy and are consolidated in Excel format. Determinations of statistical significance were made on the basis of integrated experimental data. Statistical significance was determined using a 2-tailed Student t test, with P < .05. Error bars represent ± standard deviation of the mean of repeated experiments. Quantitative PCR was analyzed using the Pfaffl method, and statistical significance was determined using a nonparametric Wilcoxon 2-group test as described elsewhere.25


Because of the well-established role of Treg in suppressing productive immune responses, we hypothesized that the accumulation of Treg in tumor tissue from patients with GBM may be achieved through multiple mechanisms, including preferential Treg chemoattraction.12,16 We first confirmed the findings of previous studies that demonstrate an overrepresentation of Treg in the tumor microenvironment, compared with Treg in circulation.15 Similar to previous reports, we found an increased percentage of Treg in the tumor tissue, compared with that in circulation. Patients with GBM had a mean of 38.5% Tregs, as defined by flow cytometric analysis of CD3, CD4, CD25, and FoxP3 expression at the tumor site (n = 20) (Fig. 1A). This population was significantly more than11.9% in the peripheral blood of the same patients (Fig. 1B). Percentages in the peripheral blood were still higher than those observed in healthy individuals or patients with benign meningioma undergoing intracranial surgery (5%–10%; data not shown), suggesting that GBM tumor burden has both a local and systemic effect on CD4 T cell populations.

Fig. 1.
Patients with GBM have an increased percentage of Treg in circulation relative to healthy controls, and Treg are overrepresented within the tumors. (A) Lymphocytes were isolated from blood and tissue collected during glioblastoma multiforme (GBM) surgical ...

Tumors secrete various soluble factors, including chemoattractant molecules that have the potential to attract Tregs from the periphery to the tumor site. We performed a transwell assay to determine whether CFSE-labeled Treg preferentially migrated in response to tumor cell–derived soluble factors, compared with unlabeled conventional T cells (Fig. 2A). Gliomas have been shown to secrete CCL22,16 and we found that this chemokine preferentially attracted Treg. We also tested a panel of other tumor-derived chemokines, including CCL5, CCL2, CCL10, CXCL10, CCL22, CCL17, CXCL13, and CCL20, and found that, as previously published,19 only CCL22 had a statistically significant effect on Treg migration. SF767 Tumor TCM, containing a variety of chemokines and cytokines, had 2.5-fold preference for Treg migration over that of conventional T cells, compared with a 1.5-fold preference for the highest dose of CCL22 alone (Fig. 2B). Furthermore, antibody blocking of the CCL22 receptor CCR4 did not reduce Treg migration to the level observed for conventional T cells cultured with TCM (data not shown). Together, these data suggest that soluble factors produced by glioma cells, including CCL22, can induce preferential migration of Treg.

Fig. 2.
Tumor-derived soluble factors preferentially attract Treg. (A) 1 × 105 CFSE labeled Treg were cocultured with 1 × 105 unlabeled conventional T cells in 0.5 µm transwell plates. Cells from the bottom chambers were analyzed using ...

Because all human T cells will increase expression of the canonical Treg markers CD25 and FoxP3 after stimulation, we first labeled Treg with CFSE and conventional T cells with PKH26—2 fluorescent dyes that are distinguishable by FACS. Once labeled, we subjected the mixed culture to CD3/CD28 stimulation and coculture with TCM. We found that the 2 cell populations were easily detectable, and the dyes diluted as expected after T cell stimulation (Fig. 3A). To demonstrate that the Treg isolated from healthy individuals were functional, we cocultured stimulated conventional T cells and added increasing numbers of Treg. The addition of as few as 2 × 104 Treg significantly reduced conventional T cell proliferation (Fig. 3B). For subsequent coculture experiments, we therefore modified the ratio of Treg to conventional T cells to 1:1.8. In doing so, we sought to minimize Treg-mediated suppression in coculture and to establish a coculture with physiologically relevant ratios based on tumor-infiltrating T cell data from patients with GBM (Fig. 1). We also analyzed the proliferation of conventional T cells in the absence of any Treg after stimulation and found that conventional T cells cultured in TCM in the absence of Treg proliferate after stimulation (mean = 33.5%) (Fig. 3C) , suggesting that soluble tumor-derived factors inhibit conventional T cell proliferation independently. In contrast, when cultured with TCM, Treg proliferation is significantly greater than that observed for conventional T cells (mean = 45.6%). In the tumor microenvironment, where both Treg and conventional T cells are present in ratios as great as 5 Treg to 1 conventional T cell, it is likely that there is an even greater reduction in the ability of conventional T cells to proliferate in response to stimulation (mean = 25.4%).

Fig. 3.
Soluble factors produced by glioblastoma cells promote Treg proliferation while suppressing conventional T cell proliferation. (A) Isolated conventional T cells, labeled with PKH26, were cultured alone or with CFSE labeled Treg, in either complete T cell ...

Many recent studies have focused on T cell plasticity and polarizing conditions, which can conditionally induce Treg phenotypes in conventional T cells. If this mechanism is exploited by GBM, immunotherapies that are predicated on initiating T cell responses and recruitment to the tumor site may instead promote tumor evasion through converted Treg-mediated immune suppression. We therefore examined the effects of malignant gliomas on T cell plasticity to determine whether conventional T cells may be converted to a suppressive Treg after recruitment to the tumor microenvironment. We cultured a population of conventional T cells in the presence of tumor-conditioned medium and observed Treg-associated markers FoxP3 and TGFβ over the incubation (Fig. 4A). Initially, there was a significant increase in the percentage of cells expressing both FoxP3 and TGFβ (Fig. 4A and B). The percentages of conventional T cells expressing TGFβ and FoxP3 after stimulation in the presence of TCM were 3.3 and 2.4 times greater, respectively, than those of conventional T cells stimulated without TCM (Fig. 4C). By day 10, however, these percentages were equivalent, suggesting that conversion to a suppressive Treg phenotype is transient and may be apparent only during peak stimulation (Fig. 4C).

Fig. 4.
Conventional T cells transiently increase expression of Treg associated proteins FoxP3 and TGF-β in response to soluble tumor factors and are functionally suppressive. (A) Freshly isolated conventional CD4 + T cells were cultured in complete T ...

Conventional T cells that were transiently induced to increase expression of FoxP3 and TGFβ after TCM culture (converted Treg) were phenotypically similar to Treg (4B and data not shown), but it was unclear whether they were functionally suppressive. To address this, we cultured with autologous CFSE-labeled conventional T cells with converted Treg and found that, similar to thymically derived Treg, converted Treg suppressed conventional T cell proliferation (Fig. 4D). Our data suggest that soluble factors produced by gliomas can upregulate Treg-associated markers in conventional T cells after stimulation and that these cells are functionally suppressive, even if only transiently. Collectively, our data support the notion that patients with a minimal tumor burden would be preferred candidates for T cell–based immunotherapies to prevent even temporary increases in functional Treg during immune responses.

In cases of some immune cells in the tumor microenvironment, such as myelomonocytic cells, tumor cells may actively promote the survival of immune cells that support tumor immune escape and prevent anti-tumor immune responses.26,27 We therefore sought to determine whether tumor-derived soluble factors influenced the survival of Treg, compared with conventional T cells, as determined by changes in the expression of pro- and anti apoptotic genes. We found that expression of anti-apoptotic genes Bcl2, Bcl-XL, IAP-1, and IAP-2 did not vary significantly after stimulation or in the presence of tumor TCM (data not shown). After stimulation in the absence of TCM, both Treg and conventional T cells decreased transcription of pro-apoptotic genes Bak, Bax, and Bim. The most striking reduction was seen in Bax transcription in both Treg (8.2-fold) and conventional T cells (10.6-fold) (Fig. 5A). After culture with TCM, Treg significantly reduced pro-apoptotic genes Bax, Bak, and Bim, even in the absence of stimulation (Fig. 5A). In contrast, conventional T cells increased expression of Bax, Bak, and Bim, suggesting a possible mechanism of preferential Treg survival in the tumor microenvironment. For example, Treg reduced Bax expression from 31.4 to 12.2 (61.2% reduction) relative to HPRT without stimulation, compared with conventional T cells, which increased Bax expression after TCM culture from 26.9 to 43.1, relative to HPRT. Indeed, tumor TCM coculture increased transcript levels for conventional T cells of several pro-apoptotic genes, with a mean 1.98-fold increase after stimulation in the presence of TCM, whereas Treg expression of these genes significantly decreased, compared with unstimulated Treg (Fig. 5B).

Fig. 5.
Conventional CD4 + T cells increase pro-apoptotic mRNA expression in the presence of tumors. Magnetically purified Treg and conventional T cells were cultured in T cell medium (TCM). Following isolation of RNA, whole cell reverse transcription and Quantitative ...

Having demonstrated that preferential migration, proliferation, and survival may contribute to the suppressive tumor microenvironment, we sought to determine whether the effects of tumor-mediated increases in Treg are restricted to the tumor microenvironment or whether the tumor may also exert a systemic effect. Increased survival, trafficking, and proliferation of Treg may in part explain the increase in Treg percentages in the circulation of patients harboring a GBM tumor. Therefore, we sought to determine whether tumor burden is associated with circulating Treg percentages in patients with GBM. To test this, we compared MRIs used to determine tumor burden and analyzed percentages of Treg by flow cytometry in patients within 48 h of the image acquisition. We analyzed a small cohort of patients prior to tumor resection, a mean of 34.8 days after the procedure, and again prior to surgery for tumor recurrence. We observed a significant decrease in circulating Treg populations after a gross total tumor resection (more than 0%) from 16.1% to 5.97%, which rebounded to 14.5% by the time of recurrence (Fig. 6A). We then examined the trend in a subset of patients opting for repeat surgery at the time of recurrence. Of the 7 patients analyzed, we found that, although not always statistically significant, the percentage of circulating regulatory T cells decreases after gross total resection and increases at the time of surgery for tumor recurrence (Fig. 6B). Together, these data suggest an association between tumor burden in patients with glioma and percentages of circulating Treg. Current experiments seek to evaluate a larger and prospectively analyzed cohort of patients to determine whether the observed trend is reproducible. These data suggest that circulating Treg may be useful as an indicator of tumor burden in patients with GBM.

Fig. 6.
Treg populations in circulation correlate with tumor burden. Magnetic resonance imaging showing T1 contrast with FLARE indicate the tumor burden at the time of PBL analysis (Left). Percentage of Treg relative to total CD4 + T cells in PBL were measured ...


The accumulation of Treg in the tumor microenvironment in patients with GBM has been a focus of those seeking to improve immune therapy because of the influence of tumor-associated Treg on local and systemic immunity.28 However, the source of Treg in the tumor microenvironment remains controversial. Similar to previous reports, our data suggest that preferential migration of Treg in response to tumor-derived soluble factors contributes to the observed accumulation at the tumor site. Our data indicate that increased proliferation and survival also contribute to Treg accumulation, particularly in the context of T cell activation.

Although increased Treg chemoattraction in response to CCL22 has been demonstrated,19 our data suggest that other soluble factors secreted by glioma cells may also contribute to Treg accumulation in the tumor microenvironment, because blocking of CCL22 receptor did not completely abrogate the observation of preferential Treg migration. Future studies are needed to identify additional chemokines responsible for Treg recruitment to and retention in the tumor microenvironment.

Recent work has also examined proliferation rates in response to tumor-derived microvesicles and has shown Treg proliferation in response to microvesicles in head and neck squamous cell carinoma cells.29,30 Our data support the observation that tumor-derived secreted factors increase proliferation of Treg, compared with conventional T cells. In the present study, we extend the observation to show that this increased proliferation occurs in response to glioma-derived soluble factors and is specific to cells defined as Treg prior to stimulation. Identification of human Treg is difficult after stimulation, because stimulated CD4 cells increase expression of all canonical Treg markers (Fig. 5). Future experiments will examine whether glioma-derived microvesicles originally described by Bastida et al. in 198431 have a similar effect on Treg expansion. More recently, Coffelt et al. described a preferential expansion in Treg population in the tumor environment using an angiogenic mammary tumor model,32 suggesting that this phenomenon may extend to many types of tumors, independent of location.

We also sought to evaluate the potential for stable conversion of conventional T cells to Treg in patients with glioma, which is a highly controversial subject because of demonstration in some animal models.22,33,34 Although we found a transient induction of the suppressive phenotype after activation, our data did not indicate stable conversion in vitro. One possible explanation is that Treg conversion in vitro may be much less stable because of decreased methylation on the FoxP3 regulatory site.35 Cell-cell interactions, such as stabilizing interactions from CD103+ dendritic cells and retinoic acid found in vivo, have been shown to enhance FoxP3 induction.36 Recent data from Wainwright et al. demonstrate that the Treg that predominate the glioma microenvironment are thymically derived rather than stably converted,37 a finding that is supported by our data of only temporary Treg conversion after stimulation.

Our data also indicate that, in contrast to conventional T cells, Treg suppress the transcription of pro-apoptotic genes. We previously demonstrated that conventional T cells undergo apoptosis in response to programmed cell death ligand-1 (PD-L1) expression by tumor cells through the interaction with the receptor programmed cell death-1 (PD-1).38 Recently, Terawaki et al. demonstrated that IFN-α mediates the expression of the PD-1 gene by T cells.39 Taken with the increase in pro-apoptotic proteins shown here, these data suggest that soluble factors in the tumor microenvironment during therapy may therefore preferentially reduce the number of conventional T cells in the microenvironment relative to Treg by increasing conventional T cell apoptosis.

To date, significant improvements in prognosis have not been established with traditional or experimental therapies for patients with malignant glioma. Our data suggest that immune therapies may be improved if the effects of tumors on the local immune environment are improved. In support of this hypothesis, Pellagrata et al. recently showed that improved antigen presentation in the tumor microenvironment and pro-inflammatory cytokine production improve survival in mouse models of glioma, specifically through the reduction of Treg in the microenvironment.40

The data presented in this manuscript are not without limitations, and future experiments are necessary to determine whether our findings are consistent with mechanisms of Treg recruitment and retention in the tumor microenvironment in patients. Although the recent use of well-characterized glioma cell lines serves as a reproducible system for the study of tumor initiation,41 tumor cell survival,42 angiogenesis,43 invasion,44 and influence on immune cells both in vitro45 and in vivo,46 it should be noted that these cell lines have been cultured for several years and may not be representative of tumor cell behavior in vivo. For example, the system may be improved by the use of autologous tumor samples grown in a manner that better sustains the heterogeneity found in tumor masses.47 This is of particular interest because it relates to the variability of tumor cell properties, such as oncogenic pathway activation,48 epigenetic abnormalities,49 immune evasion50,51 and resistance to therapies.52 The use of autologous human serum, rather than serum of bovine origin, may provide further insight into mechanisms of our findings. Although no xenogeneic effects of bovine serum have been specifically demonstrated in this context, soluble factors produced in each patient that cross the blood-brain barrier and influence circulating T cells would remain throughout in vitro culture and may therefore influence Treg function. Finally, we seek to validate our data by prospectively analyzing a larger cohort of patients to determine whether or not Treg percentage may indeed serve as a peripheral biomarker of glioblastoma tumor burden.

Our data indicate that glioma cells rely on a variety of methods mediated by soluble, tumor-derived factors to recruit, expand, and promote the survival of Treg in the microenvironment and in circulation. A better understanding and prevention of these mechanisms may improve the efficacy of future immune therapies for patients with GBM. Therefore, we propose that a better understanding of Treg recruitment and their crosstalk with tumor cells and the soluble factors that they produce will improve our ability to design more effective immune therapies for patients with GBM.


This work was supported by the Brain Tumor Special Program of Research Excellence Grant: Project 5 and the Accelerated Brain Cancer Cure (CA097257-06).


We thank Valerie A. Kivett and Anne Fedoroff for clinical trial management; UCSF neuro-oncologists Michael D. Prados and, Nicholas Butowski; UCSF Brain Tumor Research Center Investigators Jennifer Clarke and Susan Chang; clinical support staff in the UCSF Brain Tumor Research Center; and C. David James and Russell O. Pieper for their critical review of the data presented in this manuscript.

Conflict of interest statement: None declared.


1. Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion. Science. 2011;331:1565–1570. doi:10.1126/science.1203486. [PubMed]
2. Dunn GP, Old LJ, Schreiber RD. The immunobiology of cancer immunosurveillance and immunoediting. Immunity. 2004;21:137–148. doi:10.1016/j.immuni.2004.07.017. [PubMed]
3. Stutman O. Tumor development after 3-methylcholanthrene in immunologically deficient athymic-nude mice. Science. 1974;183:534–536. doi:10.1126/science.183.4124.534. [PubMed]
4. Dighe AS, Richards E, Old LJ, Schreiber RD. Enhanced in vivo growth and resistance to rejection of tumor cells expressing dominant negative IFN gamma receptors. Immunity. 1994;1:447–456. doi:10.1016/1074-7613(94)90087-6. [PubMed]
5. Klebanoff CA, Acquavella N, Yu Z, Restifo NP. Therapeutic cancer vaccines: are we there yet. Immunol Rev. 2011;239:27–44. doi:10.1111/j.1600-065X.2010.00979.x. [PMC free article] [PubMed]
6. Rosenberg SA, Sherry RM, Morton KE, et al. Tumor progression can occur despite the induction of very high levels of self/tumor antigen-specific CD8+ T cells in patients with melanoma. J Immunol. 2005;175:6169–6176. doi: [PubMed]
7. Rosenberg SA, Yang JC, Restifo NP. Cancer immunotherapy: moving beyond current vaccines. Nat Med. 2004;10:909–915. doi:10.1038/nm1100. [PMC free article] [PubMed]
8. Malyguine AM, Strobl SL, Shurin MR. Immunological monitoring of the tumor immunoenvironment for clinical trials. Cancer Immunol Immunother. 2012;61:239–247. [PubMed]
9. Liao D, Liu Z, Wrasidlo WJ, et al. Targeted therapeutic remodeling of the tumor microenvironment improves an HER-2 DNA vaccine and prevents recurrence in a murine breast cancer model. Cancer Res. 2011;71:5688–5696. doi:10.1158/0008-5472.CAN-11-1264. [PubMed]
10. Tran Thang NN, Derouazi M, Philippin G, et al. Immune infiltration of spontaneous mouse astrocytomas is dominated by immunosuppressive cells from early stages of tumor development. Cancer Res. 2010;70:4829–4839. doi:10.1158/0008-5472.CAN-09-3074. [PubMed]
11. Kryczek I, Wu K, Zhao E, et al. IL-17+ regulatory T cells in the microenvironments of chronic inflammation and cancer. J Immunol. 2011;186:4388–4395. doi:10.4049/jimmunol.1003251. [PubMed]
12. Wei J, Wu A, Kong LY, et al. Hypoxia potentiates glioma-mediated immunosuppression. PLoS One. 2011;6:e16195. doi:10.1371/journal.pone.0016195. [PMC free article] [PubMed]
13. Curran MA, Montalvo W, Yagita H, Allison JP. PD-1 and CTLA-4 combination blockade expands infiltrating T cells and reduces regulatory T and myeloid cells within B16 melanoma tumors. Proc Natl Acad Sci USA. 2010;107:4275–4280. doi:10.1073/pnas.0915174107. [PubMed]
14. Sakaguchi S. Naturally arising Foxp3-expressing CD25+CD4+ regulatory T cells in immunological tolerance to self and non-self. Nat Immunol. 2005;6:345–352. doi:10.1038/ni1178. [PubMed]
15. Heimberger AB, Abou-Ghazal M, Reina-Ortiz C, et al. Incidence and prognostic impact of FoxP3+ regulatory T cells in human gliomas. Clin Cancer Res. 2008;14:5166–172. doi:10.1158/1078-0432.CCR-08-0320. [PubMed]
16. Jacobs JF, Idema AJ, Bol KF, et al. Prognostic significance and mechanism of Treg infiltration in human brain tumors. J Neuroimmunol. 2010;225:195–199. doi:10.1016/j.jneuroim.2010.05.020. [PubMed]
17. Lee H, Kwon Y, Lee JH, et al. Methyl gallate exhibits potent antitumor activities by inhibiting tumor infiltration of CD4+CD25+ regulatory T cells. J Immunol. 2010;185:6698–6705. doi:10.4049/jimmunol.1001373. [PubMed]
18. Pabst C, Zustin J, Jacobsen F, et al. Expression and prognostic relevance of MAGE-C1/CT7 and MAGE-C2/CT10 in osteolytic lesions of patients with multiple myeloma. Exp Mol Pathol. 2010;89:175–181. doi:10.1016/j.yexmp.2010.06.011. [PubMed]
19. Jordan JT, Sun W, Hussain SF, DeAngulo G, Prabhu SS, Heimberger AB. Preferential migration of regulatory T cells mediated by glioma-secreted chemokines can be blocked with chemotherapy. Cancer Immunol Immunother. 2008;57:123–131. doi:10.1007/s00262-007-0336-x. [PubMed]
20. Lohr J, Ratliff T, Huppertz A, et al. Effector T-cell infiltration positively impacts survival of glioblastoma patients and is impaired by tumor-derived TGF-beta. Clin Cancer Res. 2011;17:4296–4308. doi:10.1158/1078-0432.CCR-10-2557. [PubMed]
21. Xu L, Xiao H, Xu M, Zhou C, Yi L, Liang H. Glioma-derived T cell immunoglobulin- and mucin domain-containing molecule-4 (TIM4) contributes to tumor tolerance. J Biol Chem. 2011;286:36694–36699. doi:10.1074/jbc.M111.292540. [PMC free article] [PubMed]
22. Oliveira VG, Caridade M, Paiva RS, Demengeot J, Graca L. Sub-optimal CD4+ T-cell activation triggers autonomous TGF-beta-dependent conversion to Foxp3+ regulatory T cells. Eur J Immunol. 2011;41:1249–1255. doi:10.1002/eji.201040896. [PubMed]
23. Ye J, Su X, Hsueh EC, et al. Human tumor-infiltrating Th17 cells have the capacity to differentiate into IFN-gamma+ and FOXP3+ T cells with potent suppressive function. Eur J Immunol. 2011;41:936–951. doi:10.1002/eji.201040682. [PubMed]
24. Ford AL, Goodsall AL, Hickey WF, Sedgwick JD. Normal adult ramified microglia separated from other central nervous system macrophages by flow cytometric sorting. Phenotypic differences defined and direct ex vivo antigen presentation to myelin basic protein-reactive CD4+ T cells compared. J Immunol. 1995;154:4309–4321. [PubMed]
25. Yuan JS, Reed A, Chen F, Stewart CN., Jr Statistical analysis of real-time PCR data. BMC Bioinformatics. 2006;7:85. doi:10.1186/1471-2105-7-85. [PMC free article] [PubMed]
26. Almand B, Clark JI, Nikitina E, et al. Increased production of immature myeloid cells in cancer patients: a mechanism of immunosuppression in cancer. J Immunol. 2001;166:678–689. [PubMed]
27. Meyer C, Sevko A, Ramacher M, et al. Chronic inflammation promotes myeloid-derived suppressor cell activation blocking antitumor immunity in transgenic mouse melanoma model. Proc Natl Acad Sci USA. 2011;108:17111–17116. doi:10.1073/pnas.1108121108. [PubMed]
28. Kennedy BC, Maier LM, D'Amico R, et al. Dynamics of central and peripheral immunomodulation in a murine glioma model. BMC Immunol. 2009;10:11. doi:10.1186/1471-2172-10-11. [PMC free article] [PubMed]
29. Szajnik M, Czystowska M, Szczepanski MJ, Mandapathil M, Whiteside TL. Tumor-derived microvesicles induce, expand and up-regulate biological activities of human regulatory T cells (Treg) PLoS One. 2010;5:e11469. doi:10.1371/journal.pone.0011469. [PMC free article] [PubMed]
30. Wieckowski EU, Visus C, Szajnik M, Szczepanski MJ, Storkus WJ, Whiteside TL. Tumor-derived microvesicles promote regulatory T cell expansion and induce apoptosis in tumor-reactive activated CD8+ T lymphocytes. J Immunol. 2009;183:3720–3730. doi:10.4049/jimmunol.0900970. [PubMed]
31. Bastida E, Ordinas A, Escolar G, Jamieson GA. Tissue factor in microvesicles shed from U87MG human glioblastoma cells induces coagulation, platelet aggregation, and thrombogenesis. Blood. 1984;64:177–184. [PubMed]
32. Coffelt SB, Chen YY, Muthana M, et al. Angiopoietin 2 stimulates TIE2-expressing monocytes to suppress T cell activation and to promote regulatory T cell expansion. J Immunol. 2011;186:4183–4190. doi:10.4049/jimmunol.1002802. [PubMed]
33. Lu Y, Suzuki J, Guillioli M, Umland O, Chen Z. Induction of self-antigen-specific Foxp3+ regulatory T cells in the periphery by lymphodepletion treatment with anti-mouse thymocyte globulin in mice. Immunology. 2011;134:50–59. doi:10.1111/j.1365-2567.2011.03466.x. [PubMed]
34. Sharma MD, Hou DY, Baban B, et al. Reprogrammed foxp3(+) regulatory T cells provide essential help to support cross-presentation and CD8(+) T cell priming in naive mice. Immunity. 2010;33:942–954. doi:10.1016/j.immuni.2010.11.022. [PMC free article] [PubMed]
35. Liu B, Tahk S, Yee KM, Fan G, Shuai K. The ligase PIAS1 restricts natural regulatory T cell differentiation by epigenetic repression. Science. 2010;330:521–525. doi:10.1126/science.1193787. [PMC free article] [PubMed]
36. Paidassi H, Acharya M, Zhang A, et al. Preferential Expression of Integrin alphavbeta8 Promotes Generation of Regulatory T Cells by Mouse CD103(+) Dendritic Cells. Gastroenterology. 2011;141:1813–1820. [PMC free article] [PubMed]
37. Wainwright DA, Sengupta S, Han Y, Lesniak MS. Thymus-derived rather than tumor-induced regulatory T cells predominate in brain tumors. Neuro Oncol. 2011;13:1308–1323. [PMC free article] [PubMed]
38. Parsa AT, Waldron JS, Panner A, et al. Loss of tumor suppressor PTEN function increases B7-H1 expression and immunoresistance in glioma. Nat Med. 2007;13:84–88. doi:10.1038/nm1517. [PubMed]
39. Terawaki S, Chikuma S, Shibayama S, et al. IFN-alpha directly promotes programmed cell death-1 transcription and limits the duration of T cell-mediated immunity. J Immunol. 2011;186:2772–2779. doi:10.4049/jimmunol.1003208. [PubMed]
40. Pellegatta S, Poliani PL, Stucchi E, et al. Intra-tumoral dendritic cells increase efficacy of peripheral vaccination by modulation of glioma microenvironment. Neuro Oncol. 2010;12:377–388. doi:10.1093/neuonc/nop024. [PMC free article] [PubMed]
41. Watanabe A, Ogiwara H, Ehata S, et al. Homozygously deleted gene DACH1 regulates tumor-initiating activity of glioma cells. Proc Natl Acad Sci USA. 2011;108:12384–12389. doi:10.1073/pnas.0906930108. [PubMed]
42. Li G, Xu Y, Guan D, Liu Z, Liu DX. HSP70 protein promotes survival of C6 and U87 glioma cells by inhibition of ATF5 degradation. J Biol Chem. 2011;286:20251–20259. doi:10.1074/jbc.M110.211771. [PMC free article] [PubMed]
43. Kavitha CV, Agarwal C, Agarwal R, Deep G. Asiatic acid inhibits pro-angiogenic effects of VEGF and human gliomas in endothelial cell culture models. PLoS One. 2011;6:e22745. doi:10.1371/journal.pone.0022745. [PMC free article] [PubMed]
44. Badiga AV, Chetty C, Kesanakurti D, et al. MMP-2 siRNA inhibits radiation-enhanced invasiveness in glioma cells. PLoS One. 2011;6:e20614. doi:10.1371/journal.pone.0020614. [PMC free article] [PubMed]
45. Curran CS, Evans MD, Bertics PJ. GM-CSF production by glioblastoma cells has a functional role in eosinophil survival, activation, and growth factor production for enhanced tumor cell proliferation. J Immunol. 2011;187:1254–1263. doi:10.4049/jimmunol.1001965. [PMC free article] [PubMed]
46. Zhu X, Fujita M, Snyder LA, Okada H. Systemic delivery of neutralizing antibody targeting CCL2 for glioma therapy. J Neurooncol. 2011;104:83–92. doi:10.1007/s11060-010-0473-5. [PMC free article] [PubMed]
47. Wakimoto H, Mohapatra G, Kanai R, et al. Maintenance of primary tumor phenotype and genotype in glioblastoma stem cells. Neuro Oncol. 2012;14:132–144. [PMC free article] [PubMed]
48. Rodriguez-Escudero I, Oliver MD, Andres-Pons A, Molina M, Cid VJ, Pulido R. A comprehensive functional analysis of PTEN mutations: implications in tumor- and autism-related syndromes. Hum Mol Genet. 2011;20:4132–4142. doi:10.1093/hmg/ddr337. [PubMed]
49. Molina JR, Agarwal NK, Morales FC, et al. PTEN, NHERF1 and PHLPP form a tumor suppressor network that is disabled in glioblastoma [published online ahead of print August 1, 2011] Oncogene. 2011 doi:10.1038/onc.2011.324. [PMC free article] [PubMed]
50. Lemke D, Pfenning PN, Sahm F, et al. Costimulatory protein 4IgB7H3 drives the malignant phenotype of glioblastoma by mediating immune escape and invasiveness. i Clin Cancer Res. 2012;18:105–117. [PubMed]
51. Sippel TR, White J, Nag K, et al. Neutrophil Degranulation and Immunosuppression in Patients with GBM: Restoration of Cellular Immune Function by Targeting Arginase I. Clin Cancer Res. 2011;17:6992–7002. doi:10.1158/1078-0432.CCR-11-1107. [PubMed]
52. Paolini A, Pasi F, Facoetti A, et al. Cell Death Forms and HSP70 Expression in U87 Cells after Ionizing Radiation and/or Chemotherapy. Anticancer Res. 2011;31:3727–3731. [PubMed]

Articles from Neuro-Oncology are provided here courtesy of Society for Neuro-Oncology and Oxford University Press