Background
Glioblastoma multiforme (GBM) is a highly aggressive, heterogeneous brain tumor with poor prognosis [
1]. Standard of care includes surgical resection followed by radiation and chemotherapy, however median survival is about 15 months with a two-year survival of 30 % and a 5-year survival of <5 % in adults [
2]. Despite breakthroughs in our understanding of the disease, therapeutic options available for GBM have remained largely unchanged over the past three decades. This has led to only marginal increases in overall patient survival and new therapeutic approaches to enhance brain tumor treatment are warranted.
One novel therapeutic approach for GBM involves targeting a phenotypic trait shared by virtually all cancer cells, deregulated metabolism. It has been postulated that metabolic alteration such as that seen with the therapeutic ketogenic diet (KD) may be an effective anti-cancer strategy [
3]. The KD is a high fat, low-carbohydrate/adequate protein nutritional therapy used in the treatment of refractory epilepsy [
4]. We and others have shown that the KD enhances survival in mouse models of malignant glioma [
5‐
8]. We also demonstrated that the KD greatly enhanced survival when administered in combination with radiation [
6]. Mechanistically, the KD alters a variety of processes that influence the tumor microenvironment including hypoxia, inflammation, angiogenesis and vascular permeability [
5,
9]. However, the effect of a KD on the GBM tumor-reactive immune response has yet to be examined.
We have recently shown that an unrestricted KD decreases expression of the hypoxia marker carbonic anhydrase IX (CAIX) and the key mediator of the hypoxic response hypoxia-inducible factor alpha (HIF-1α) in a mouse model of malignant glioma [
9]. Wei et al. demonstrated that hypoxia leads to inhibition of T cell proliferation and effector responses, with induction of CD4 + FoxP3+ T regulatory cells in GBM [
10]. This study also demonstrated that this immunosuppressive effect could be reversed by inhibiting HIF-1α. As tumor hypoxia is linked to the less favorable Th2 immune response [
11], it is possible that by altering the hypoxic response the KD may promote a Th1 type tumor-reactive immune response. Additionally, we previously demonstrated that the KD reduces activation of the pro-inflammatory transcription factor, nuclear factor kappa B (NF-κB) and reduces expression of cyclooxygenase-2 (COX-2) [
5,
9], both of which have been implicated in hypoxia-driven immunosuppression [
12]. Taken together these studies led us to hypothesize that the KD may alter the tumor microenvironment to alleviate immune suppression and enhance anti-tumor immunity.
In this paper we investigated the role that an unrestricted KD plays in alleviating tumor immune suppression in a mouse model of malignant glioma. We studied the direct effects of this metabolic therapy on total infiltration and function of tumor-reactive T cells and natural killer (NK) cells, as well as the indirect benefits of this metabolic therapy on alleviation of immune suppression in the tumor microenvironment.
Methods
Antibodies and cell lines
Fluorochrome-conjugated anti-mouse monoclonal antibodies (Abs) specific for CD8α, CD274, CD279, CTLA-4, CD86, tumor necrosis factor (TNF), interferon gamma (IFNγ), interleukin-2 (IL-2), CD4, FoxP3, NKp46, CD3, and interleukin-10 (IL-10) were purchased from eBiosciences (San Diego, CA) and diluted 1:200 prior to use. Anti-CD8 depletion antibodies were purified from the mouse 2.43 hybridoma cell line purchased from ATCC (Manassas, VA). Bioluminescent GL261-Luc 2 cells were derived and grown as previously described [
6].
Mice and tumor implantation
GL261-Luc2 cells were harvested by trypsinization, washed and resuspended at a concentration of 1–2x10
7 cells/ml in DMEM without FCS and implanted into ten week old B6 (Cg)-
Tyr
c-23
/J (albino C57BL/6) mice (The Jackson Laboratory, Bar Harbor, ME) at an average weight of 19–20 g as previously described [
5,
6,
13]. Briefly, animals were anesthetized by an intraperitoneal injection of ketamine (10 mg/kg) and xylazine (80 mg/kg), placed in a stereotactic apparatus and an incision was made over the cranial midline. A burrhole was made 0.1 mm posterior to the bregma and 2.3 mm to the right of the midline. A needle was inserted to a depth of 3 mm and withdrawn 0.4 mm to a depth of 2.6 mm. Two μl of GL261-luc2 cells (10
7 cells/ml) were infused over the course of 3 min. The burrhole was closed with bonewax and the incision was sutured.
Treatment and animal monitoring
Following implantation surgery, animals were fed standard rodent chow for 3 days. Animals were then randomized to remain on standard diet (SD) or changed to a KD (KetoCal®; Nutricia North America, Gaithersburg, MD). The KD was obtained directly from the manufacturer and is a nutritionally complete diet providing a 4:1 ratio of fats to carbohydrates plus protein (72 % fat, 15 % protein, and 3 % carbohydrate). The KD was prepared by mixing KetoCal® with water (2:1) and fed to the animals each day (
ad libitum). Bioluminescence was analyzed to quantify tumor burden as described [
6]. Serum β-hydroxybutyrate (βHB) and glucose levels were measured using a Precision Xtra® blood monitoring system (Abbott Laboratories, Abbott Park, IL). Animals were weighed every 3–5 days and euthanized upon occurrence of visible symptoms of impending death such as hunched posture, reduced mobility and weight loss [
5,
14]. Measurements of animal body weight, blood βHB, and glucose can be found in (Additional file
1: Figure S1).
CD8 depletion in vivo
Supernatant from 2.43 hybridoma cells was precipitated in saturated ammonium sulfate to 45 % (v/v) overnight at 4 °C and dialyzed against PBS for 24 h. The concentration of dialyzed antibody was determined by UV spectroscopy, and 0.3 mg of purified antibody was administered via intraperitoneal injection twice before tumor inoculation (day −5 and −3), and continued every three days after inoculation until euthanasia. CD8 T cell depletion was confirmed by flow cytometry analysis of peripheral blood mononuclear cells, as previously described [
15]. Confirmation of CD8 depletion can be found in (Additional file
2: Figure S2).
Tissue preparation
When mice became symptomatic they were anesthetized with 80 mg/kg ketamine, 10 mg/kg xylazine followed by cardiac perfusion with ice-cold RPMI media just prior to euthanization. Tumor tissue and non-tumor contralateral brain were collected in RPMI media and run through a 70 μm filter. Tumor-infiltrating cells were isolated from tumor tissue by centrifugation over a 30/70 % Percoll gradient (Sigma-Aldrich, St. Louis, MO) before antibody staining and analysis of cell populations on an LSRFortessa flow cytometer (BD Biosciences, San Jose, CA). Flow cytometry data were analyzed with FlowJo8.8 (Tree Star Inc., Ashland, OR) and graphs were generated using Prism 5 software (GraphPad Software, La Jolla, CA). Gating strategies and isotype controls can be found in the Additional file
3: Figure S3 and Additional file
4: Figure S4 section.
Intracellular cytokine staining
Lymphocytes were cultured alone or stimulated with GL261-Luc2 cells at a density of 10
6 cells per well (6-well plate). GolgiStop (BD Biosciences) was added at 1 h to inhibit export of cytokines and after a further 5 h of incubation, cells were stained for extracellular proteins. Permeabilization and intracellular staining for cytokines was done according to manufacturer’s instructions using the Cytofix/Cytoperm kit (BD Biosciences). Gating strategies and isotype controls can be found in the Additional file
5: Figure S5 and Additional file
6: Figure S6 section.
Cytotoxicity ELISA
Lymphocytes were isolated from tumor tissue, and cultured alone or with GL261-Luc2 cells at varying effector to target cell ratios. Lactate dehydrogenase (LDH) ELISA was performed using CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI). Absorbance was recorded at 490 nm.
Animals and virus
Six to 8-week-old female C57BL/6 mice were obtained from The Jackson Laboratory. All experiments were conducted under Arizona State University IACUC approval and followed all relevant federal guidelines and institutional policies. The Armstrong and clone 13 strains of Lymphocytic Choriomeningitis Virus (LCMV) were grown as previously described [
16]. Mice were infected with 2 x 10
5 PFU of LCMV (Armstrong) injected intraperitoneally or 2 x 10
6 PFU of LCMV (clone 13) injected intravenously.
Statistical methods
Statistical analyses were performed using GraphPad Prism 5 (GraphPad Software, San Diego, CA). All values are represented as the mean ± SD and significance was determined using both the Student’s t test and the Mann Whitney non-parametric test. P < 0.05 was considered statistically significant. For the Kaplan Meier survival data the log-rank (Mantel-Cox) test was used to assess statistical significance.
Discussion
Activated effector immune responses against glioblastoma multiforme (GBM) may provide benefits in patient survival; however these tumors exert a variety of immunosuppressive pressures on the surrounding microenvironment [
17,
18]. These include increased induction of CD8 + FOXp3+ regulatory T cells (Tregs), elevated immunosuppressive cytokine levels, diminished CD4+ helper T cell populations, tolerized antigen presenting cells and upregulated immune inhibitory checkpoints [
19]. For example, Tregs suppress immune responses by secreting cytokines such as IL-10 and facilitating inactivation of CD8+ cytotoxic T cells by direct cell-to-cell interactions [
20]. A key observation in immunosuppressed GBM patients is a decrease in CD4+ T cells with an increased proportion of Tregs and increased IL-10 levels [
19,
21]. The current study demonstrated that tumors from animals maintained on the KD had a significantly increased CD4+ T cell population and a decreased proportion of Tregs when compared to control animals. Further the Tregs isolated from animals maintained on the KD produced significantly less IL-10 when stimulated with tumor cells. Similar results were demonstrated in a study using a pancreatic cancer model which showed increased CD4+ T cells and decreased Tregs when animals were fed a KD [
22].
In addition to increasing Tregs and IL-10 production in the microenvironment, tumors exploit immune inhibitory signaling pathways involving direct cell-to-cell interactions. Key mediators of this system include cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and programmed death-1 (PD-1) which are found on the surface of activated effector T cells and act as checkpoints to regulate immune proliferation and activation. For example, when PD-1 binds its ligand, programmed death ligand 1 (PD-L1), activated CD8+ and CD4+ T cells are suppressed [
23]. Increased PD-L1 expression has been observed on tumor cells and immune cells within the GBM microenvironment [
24‐
27] and leads to direct inactivation of CD8+ T cells [
28,
29]. The current study demonstrates significantly decreased expression of PD-1 and CTLA-4 on tumor-infiltrating CD8+ T cells and decreased expression of their ligands (PD-L1 and CD86, respectively) on dissociated tumor cells from animals maintained on the KD when compared to control animals. Blockade of the CTLA-4 and PD1 immune checkpoints represents a potentially important anti-glioma strategy that has proven effective in preclinical models of glioma [
30‐
34] and has warranted exploration in ongoing clinical trials [
35].
The current study suggests that the KD may shift the immunological landscape from inflammatory, non-protective immune responses to cytotoxic Th1 responses and promotion of immune mediated killing at the tumor site. Shifting the balance toward a Th1 immune response leads to a general change in cytokine milieu at the tumor site which alters antigen presenting cell maturation and amount of overall immune cell activation [
36‐
41]. This may explain results seen in this manuscript including increased NK and CD8+ T cell function, changes in CD4+ T cell recruitment, reduction in immune inhibitory receptor expression, and ligand availability on the tumor cells themselves. It should be noted that increased CD4 to CD8 T cell ratio may be indicative of a Th2 type immune response at the tumor site [
42], which may promote an immune tolerance state; however, greater CD8 T cell activation in the tumors from mice maintained on a KD suggests this is not the case.
It is known that activated T cells undergo metabolic reprogramming in which glycolysis is required to support proliferation and efficient growth [
43‐
46]. Recent evidence also suggests that reduced glucose availability and increased fatty acid oxidation favors T regulatory cells over effector T cells [
47]. However, tumor-infiltrating T cells from mice fed the KD are still able to mount effective responses, undergo appropriate differentiation, and retain function even with the characteristic drop in glucose availability that accompanies the KD. It is currently unclear how the KD alters the metabolic activity of lymphocytes and why this effect appears to be specific to the lymphocytes isolated from the tumor microenvironment. It is possible that T-cells can utilize ketones as a primary energy source in place of glucose in a way similar to that of normal cells in the brain [
48,
49]. Recent work has suggested that tumor cells may outcompete other cells in the microenvironment for glucose and other nutrients, thereby reducing the activation of anti-tumor effector T cells [
50,
51]. By providing ketones as an alternative energy source for lymphocytes it can be postulated that the KD may alleviate immunosuppression mediated by nutrient competition. Further studies are needed to explore this question and determine the precise role of ketones in T cell metabolism.
While the effect of the KD on tumor-infiltrating lymphocytes has only recently been explored, existing preclinical
in vitro and
in vivo data as well as case reports and anecdotal information have generated increased support for clinical testing. Prospective Phase I and II clinical trials have been initiated to assess the safety, efficacy and tolerability of the KD in patients with recurrent GBM (ClinicalTrials.gov; NCT01754350; NCT01535911; NCT01865162; NCT02149459). In addition, we have initiated a phase I/II trial assessing the tolerability and efficacy of the KD up-front, concurrently with radiation and temozolomide in newly diagnosed GBM patients (NCT02046187) based on our preclinical data demonstrating that the KD, when given in combination with radiation, dramatically enhances survival when compared to radiation treatment alone [
6]. The mechanisms underlying this effect are still under investigation; however, as radiation-induced tumor killing is known to expose the immune system to a greater diversity of tumor antigens, increased antigen processing, and increased immunogenic cytotoxicity it is possible that the KD as an adjuvant can work to augment the effect of radiation in part by enhancing immunity against GBM.
Conclusions
In summary, the KD may work as an immune adjuvant in the glioma microenvironment by reducing immune suppression, and promoting Th1 type immune responses against the tumor. These data provide additional support for the use of the KD in combination with the current standard of care and newer therapies for the treatment of brain tumors.
Ethics statement
This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of St. Joseph’s Hospital and Medical Center (protocol number 334 (A3510-01)). All surgery was performed under ketamine/xylazine anesthesia, and every effort was made to minimize suffering.
Consent for publication
Not applicable.
Availability of data and materials
The datasets supporting the conclusions of this article are included within the article and its supplementary files.
Acknowledgements
The authors thank Nutricia North America for providing KetoCal®, the Remi Savioz Glut1 Foundation for providing blood glucose and βHB testing strips, and Dr. Phillip Stafford at Arizona State University for assisting with statistical analysis.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
DML, ECW and ACS conceived of the study and participated in its design and coordination. DML, ECW, JLJ and KSB performed the experiments. DML, ECW, KSB, JLJ, JNB and ACS analyzed the data. JNB and ACS contributed reagents, materials and analysis tools. DML and ECW wrote the manuscript. All authors have read and approved the manuscript.