Serum elevation of B lymphocyte stimulator does not increase regulatory B cells in glioblastoma patients undergoing immunotherapy

Abstract

Regulatory B cells that secrete IL-10 (IL-10⁺ B_regs) represent a suppressive subset of the B cell compartment with prominent anti-inflammatory capacity, capable of suppressing cellular and humoral responses to cancer and vaccines. B lymphocyte stimulator (BLyS) is a key regulatory molecule in IL-10⁺ B_reg biology with tightly controlled serum levels. However, BLyS levels can be drastically altered upon chemotherapeutic intervention. We have previously shown that serum BLyS levels are elevated, and directly associated, with increased antigen-specific antibody titers in patients with glioblastoma (GBM) undergoing lymphodepletive temozolomide chemotherapy and vaccination. In this study, we examined corresponding IL-10⁺ B_reg responses within this patient population and demonstrate that the IL-10⁺ B_reg compartment remains constant before and after administration of the vaccine, despite elevated BLyS levels in circulation. IL-10⁺ B_reg frequencies were not associated with serum BLyS levels, and ex vivo stimulation with a physiologically relevant concentration of BLyS did not increase IL-10⁺ B_reg frequency. However, BLyS stimulation did increase the frequency of the overall B cell compartment and promoted B cell proliferation upon B cell receptor engagement. Therefore, using BLyS as an adjuvant with therapeutic peptide vaccination could promote humoral immunity with no increase in immunosuppressive IL-10⁺ B_regs. These results have implications for modulating humoral responses in human peptide vaccine trials in patients with GBM.

Introduction

Regulatory B cells (IL-10⁺ B_regs) play an immunosuppressive role in mouse models for autoimmunity and cancer [1–5]. IL-10⁺ B_regs downregulate effector T cell responses through the secretion of immunosuppressive cytokines such as interleukin-10 (IL-10) and through cell-to-cell contact [1, 4, 6] and have been shown to promote cancer progression in murine models by upregulating immunosuppressive cell subsets such as regulatory T cells (T_regs) [7]. However, the regulation of these cells, especially in human systems, is not well understood [8, 9]. B lymphocyte stimulator (BLyS) is a B cell homeostatic cytokine that promotes differentiation and survival of B cells and has been linked to the promotion of IL-10⁺ B_regs in preclinical and clinical studies [10–13]. Examination of BLyS and IL-10⁺ B_regs in the context of vaccination may provide important information on the induction of potentially immunosuppressive and detrimental IL-10⁺ B_regs responses.

Our phase II trial of rindopepimut^®, an epidermal growth factor receptor variant III (EGFRvIII) vaccine consisting of a 14-amino acid peptide containing the EGFRvIII mutation (PEPvIII, H-Leu-Glu–Glu-Lys–Lys-Gln-Asn-Tyr-Val-Val-Thr-Asp-His-Cys-OH) chemically conjugated to keyhole limpet hemocyanin (KLH) for the treatment of patients with glioblastoma (GBM) has shown a survival benefit in a multicenter, single-arm phase II clinical trial (ACT-III) [14], likely through an antibody-dependent mechanism. In newly diagnosed patients with GBM, this vaccine was administered with the standard-of-care (SOC) lymphodepleting chemotherapeutic alkylating agent temozolomide (TMZ) and these patients developed supraphysiological EGFRvIII-specific antibody titers [15–18]. Further analysis revealed elevated serum BLyS levels (likely due to the lymphodepletive effects of TMZ), and these BLyS levels directly correlated with vaccine-induced immunoglobulin levels [19]. This clinical observation is supported by preclinical studies demonstrating a role for BLyS in the promotion of humoral responses [20, 21].

In this study, we examine the association of human IL-10⁺ B_regs with serum BLyS levels in the context of SOC TMZ treatment and PEPvIII-KLH vaccination in patients with GBM [22–24]. Since BLyS levels are elevated in the context of TMZ-induced lymphopenia, we hypothesized that BLyS stimulation may enhance IL-10⁺ B_regs frequencies in these patients and impede the ability to achieve maximal anti-tumor immune responses [19, 25, 26]. Our studies demonstrated that after serial PEPvIII-KLH vaccination in the context lymphopenia, the frequency of IL-10⁺ B_regs in GBM patients did not significantly change. In addition, there was no association between the size of the IL-10⁺ B_reg compartment and BLyS serum levels. Further in vitro studies revealed that the frequency of IL-10⁺ B_regs did not increase upon stimulation with physiologically comparable BLyS levels. However, BLyS enhanced the proliferation and increased the frequency of the global B cell compartment. Taken together, these data suggest that using BLyS as an adjuvant with therapeutic peptide vaccination may not trigger IL-10⁺ B_reg expansion and instead promotes the proliferation of activated B cells and humoral immunity. These results implicate BLyS as a novel adjuvant for enhancing humoral responses in human peptide vaccine trials.

Materials and methods

Patient samples

Cryopreserved patient samples were collected by blood draw or leukapheresis from the ACT-II trial, as described in [15]. Briefly, adults with newly diagnosed GBM who had gross total resection of their EGFRvIII-positive tumor and a Karnofsky performance status (KPS) score of ≥80 with no radiographic evidence of progression after radiation therapy were eligible for vaccination. The trial design and informed consent were approved by the FDA (under BB-IND-9944) and the local institutional review boards [15–17]. After tumor resection and conformal external beam radiotherapy (XRT) with concurrent TMZ at a targeted dose of 75 mg/m², informed consent was obtained. The initial 3 vaccinations of a 14-mer peptide conjugated to KLH were given every two weeks starting within 6 weeks of completing radiation [15, 16]. Subsequent vaccines were given until clinical or radiographic evidence of tumor progression or death. Patients were assigned to receive TMZ at a targeted dose of 100 mg/m² for the first 21 days of a 28-day cycle (n = 8).

Patient samples selected for analyses were taken at vaccine 1, prior to any vaccination or TMZ treatment at 100 mg/m² (Pre), and at a patient variable post-vaccination time point that was selected to occur after patient peak serum BLyS levels and near peak anti-EGFRvIII antibody titers (post) [19]. Table 1 includes the serum BLyS levels of GBM patients at the time of IL-10⁺ B_reg analysis. These time points were selected based on our hypothesis that serum BLyS elevation will directly translate into an enhancement of IL-10⁺ B_reg frequency, and based on our available clinical samples. Normal donor samples were collected at Duke University Medical Center.

Table 1.

BLyS serum levels of GBM patients at the time of IL-10⁺ B_reg analysis

All patients	BLyS serum levels (pg/ml)	BLyS (fold increase)
ACTII(B6)	1357	1.346
ACTII(B7)	1786	1.772
ACTII(B10)	2397	2.378
ACTII(B11)	1372	1.361
ACTII(B12)	1688	1.675
ACTII(B14)	1439	1.428
ACTII(B15)	1386	1.375
ACTII(B17)	1337	1.326

Ex vivo evaluation of IL-10 competency

In accordance with the protocols adopted from Iwata et al. [8], and with the studies from Khoder et al. [9], peripheral blood mononuclear cells (PBMCs) were utilized for these studies because of the limited clinical sample availability. PBMCs were rested overnight (o/n) in RPMI Medium 1640 with GlutaMAX™-I (Gibco) plus 10 % fetal bovine serum (FBS, GemCell™). Cells were then harvested and plated at 4.0 × 10⁶ cells/ml in 48-well plate for 43 h in media supplemented with 1 µg/ml CD40L (R&D Systems) and 10 µg/ml lipopolysaccharide (LPS, Sigma-Aldrich catalog no. L4391). They were subsequently stimulated for 5 h with 50 ng/ml phorbol myristate acetate (PMA, Sigma-Aldrich), 1 µg/ml ionomycin (Sigma-Aldrich), 10 µg/ml LPS, and 1ul/ml of Brefeldin A (BFA, BD Biosciences GolgiPlug™—Protein Transport Inhibitor).

Cells were washed and stained with monoclonal fluorophore-conjugated antibodies for the following cell surface markers: extracellular peridinin chlorophyll protein complex (PerCP)-anti-CD19 and fluorescein isothiocyanate (FITC)-anti-CD14 as well as intracellular Allophycocyanin (APC)-anti-IL-10. Intracellular staining was performed using the Cytofix/Cytoperm™ Kit (BD Biosciences) as described by manufacturer.

Ex vivo BLyS-sensitivity assay

In order to examine cell sensitivity to BLyS, cells were stimulated in a similar protocol as used for the determination of IL-10 competency. BLyS (2000 ng/ml; B cell-activating factor (BAFF), R&D Systems) was added to the initial stimulation media as described above, and 4.0 × 10⁶ cells/ml rested PBMCs were stimulated for 67 h in a 48-well plate before being transferred for 5 h to media containing 50 ng/ml PMA (Sigma-Aldrich), 1 µg/ml ionomycin (Sigma-Aldrich), 10 µg/ml LPS (Sigma-Aldrich), and 1 ul/ml of BFA (BD Biosciences). The intracellular staining was performed identically as described above.

B cell proliferation assay

To assess the B cell proliferative response to BLyS stimulation, we stained normal donor PBMCs with carboxyfluorescein succinimidyl ester (CFSE, Life Technologies), a fluorescent stain used to track cell proliferation, and stimulated with goat anti-immunoglobulin (Ig)M (Southern Biotech) and BLyS (R&D; 2000 ng/ml) in a 48-well plate (BD Falcon) for 72 h. The cells were then stained for CD19 (PerCP) and analyzed for proliferation by flow cytometry.

Statistical analysis

Based on previous studies demonstrating the impact of BLyS on increasing the IL-10⁺ B_reg frequency among B cells [10–13], our a priori hypothesis was that due to the elevated serum BLyS levels experienced in our GBM patient population [19], this effect will directly translate into an enhancement of IL-10⁺ B_reg frequency within the B cells compartment. Since this hypothesis is unidirectional, a paired, one-tailed Student’s t test was used to compare IL-10⁺ B_regs frequency in pre- and post-vaccination samples. A Pearson linear regression was used to assess associations and statistical significance when examining associations with various other trial parameters. An unpaired, one-tailed Student’s t test was used when comparing BLyS-induced IL-10⁺ B_regs frequencies in GBM patients to normal donor samples. Statistical significance was defined at p < 0.05. All statistical analysis was completed in GraphPad Prism 5.00.

Results

The frequency of IL-10⁺ B_regs is unchanged following TMZ and peptide vaccine treatment

To understand the effect of tumor-specific peptide vaccination and TMZ-induced lymphopenia, on IL-10-secreting IL-10⁺ B_regs and based on the hypothesis that serum BLyS elevation will directly translate into an enhancement of IL-10⁺ B_reg frequency, we examined PBMCs from cryopreserved samples from PEPvIII-KLH-vaccinated GBM patients undergoing lymphodepletive TMZ treatment, utilizing a functional IL-10⁺ B_reg assay adopted from Iwata et al. [8] and Khoder et al. [9] depicted in Fig. 1a, b. IL-10⁺ B_reg frequency was analyzed at vaccine 1, prior to any vaccination or TMZ treatment at 100 mg/m² (Pre), and was also analyzed post-vaccination at a time point chosen on a patient by patient basis that was selected to occur after peak serum BLyS levels and near peak anti-EGFRvIII antibody titers (post) [19]. Table 1 includes the serum BLyS levels of GBM patients at the time of IL-10⁺ B_reg analysis. In accordance with previous studies published from this patient population [15], analysis of the B cell population demonstrated an expected reduction in the frequency of B cells (Fig. 1c; *p = 0.0466). This reduction is largely attributed to the lymphodepletive effect of TMZ chemotherapy [15]. Of interest, while the B cell frequency was reduced, there was no significant difference in IL-10⁺ B_reg frequency between these time points (Fig. 1d).

Fig. 1.

IL-10⁺ B_regs frequency remains constant in glioblastoma patients undergoing TMZ and peptide vaccine. a PBMCs were rested overnight (o/n) in media (RPMI Medium 1640 with GlutaMAX™-I plus 10 % FBS). Cells were harvested and stimulated in a 48-well plate at 4.0 × 10⁶ cells/ml for 43 h in media supplemented with 1 µg/ml CD40L and 10 µg/ml LPS. They were subsequently stimulated for 5 h in media with 50 ng/ml PMA, 1 µg/ml ionomycin, 10 µg/ml LPS, and 1× Brefeldin A (BFA) before being stained for CD19, CD14, IL-10, and, in some experiments, the EGFRvIII tetramer. b A representative flow cytometry sample from a patient sample is shown, compared to the appropriate negative control. c In the same experiments, the frequency of total B cells was found to decrease over the course of vaccination (*p = 0.0466). d No significant difference was observed in the frequency of IL-10⁺ B_regs in patient PBMC samples (N = 8) from the ACT-II trial (p = 0.20165). Paired one-tailed t test analyses were used to determine statistical significance (*p < 0.05)

The frequency of IL-10⁺ B_regs does not associate with anti-EGFRvIII antibody titers or patient survival

Published data indicate a potential role for IL-10⁺ B_reg in regulating humoral and cellular responses to vaccines as well as a regulatory role in immunosuppression and cancer progression. We therefore decided to analyze the relationship of IL-10⁺ B_reg frequency with peak anti-EGFRvIII antibody titers and patient survival outcomes. We found that post-vaccination IL-10⁺ B_reg frequency is not related to peak anti-EGFRvIII antibody titers (Fig. 2a; r ² = 0.03185; p = 0.6724). Also, there does not appear to be any association with overall patient survival (Fig. 2b; r ² = 0.06793; p = 0.5330) progression-free survival, or time to radiographic progression of disease (Fig. 2c; r ² = 0.09114; p = 0.4674) measured in months (mo).

Fig. 2.

The frequency of IL-10⁺ B_regs is not associated with anti-EGFRvIII Ig titer or patient survival. Associations between IL-10⁺ B_reg frequencies and (a) patient peak serum anti-EGFRvIII titer (r ² = 0.03185; p = 0.6724), (b) patient overall survival in months (mo) (r ² = 0.06793; p = 0.5330), and (c) patient progression-free survival in months (r ² = 0.09114; p = 0.4674) were evaluated. The Pearson correlation statistics were used to determine statistical significance

BLyS elevation does not affect the frequency of IL-10⁺ B_regs but increases the frequency and proliferation of B cells

Since previous preclinical studies have demonstrated a role for BLyS in promoting IL-10⁺ B_reg frequencies, we decided to evaluate whether BLyS serum levels were associated with IL-10⁺ B_reg frequencies in our GBM patients at time of IL-10⁺ B_reg analysis (Table 1). Therefore, we compared IL-10⁺ B_reg frequencies to corresponding patient BLyS serum levels. There was no significant association (Fig. 3a, R ² = 0.3253; p = 0.1399) between IL-10⁺ B_reg and patient BLyS serum levels. Since our previous findings suggested a strong relationship between serum BLyS and B cell activity [19], we investigated these seemingly incongruous results by stimulating healthy, normal donor PBMC samples with BLyS in addition to ex vivo B cell receptor stimulation before intracellular IL-10 staining. We found no significant difference in IL-10⁺ B_reg frequency when cells were cultured with physiologically relevant levels of BLyS (Fig. 3b; p = 0.4882). However, both the frequency of B cells (Fig. 3c; *p = 0.0283) and the level of proliferation as determined by carboxyfluorescein succinimidyl ester (CFSE) did increase significantly (Fig. 3d; *p = 0.0284). Thus, BLyS does not appear to affect the IL-10⁺ B_reg compartment in ex vivo samples at the time of peak antibody titer or in vitro healthy donor samples, but does increase global B cell frequency and proliferation.

Fig. 3.

BLyS elevation does not impact frequency of IL-10⁺ B_regs, but significantly impacts B cell levels and proliferation. a There was no significant relationship observed between the frequency of IL-10⁺ B_regs and patient serum BLyS levels at the time of IL-10⁺ B_regs analysis (R ² = 0.3253; p = 0.1399). Dotted line represents the average serum levels of BLyS in normal donors. b Ex vivo experiments with healthy donor samples found no significant change in the frequency of IL-10⁺ B_regs in healthy donor PBMC samples in the context of soluble BLyS (2000 ng/ml). c However, ex vivo experiments demonstrated a significant change in the frequency of B cells in healthy donor PBMC samples upon soluble BLyS addition (p = 0.0283). d B cells stimulated with anti-IgM and BLyS proliferated significantly more, as measured by a CFSE proliferation assay (p = 0.0284). Paired one-tailed t test analysis was used to determine statistical significance (*p < 0.05)

Discussion

In the context of tumor-specific vaccination during lymphodepletive TMZ chemotherapy in patients with GBM, we found that IL-10⁺ B_reg frequency was unchanged despite elevated BLyS levels at the time of analysis. While this therapy elicits robust production of anti-EGFRvIII antibodies and results in an increased overall survival when compared to historical controls [14, 15, 24], IL-10⁺ B_reg frequency was not associated with peak anti-EGFRvIII antibody titers, patient survival outcomes, or BLyS levels. These results were unexpected due the mouse and human literature demonstrating a link between BLyS levels and IL-10⁺ B_reg promotion [10–12].

However, recent studies suggest that IL-10⁺ B_reg subsets may not have the same activation-program response to BLyS observed in the greater CD19⁺ B cell compartment [27], which could account for the IL-10⁺ B_reg dynamics observed in our study. Another important point is the utilization of rested PBMCs for the IL-10⁺ B_reg assay. Many studies utilize purified B cells in stimulation assays in order to specifically stimulate the pure B cell population and to determine the IL-10⁺ B_reg fraction [8, 28]. As we utilized PBMCs, there is the possibility of an indirect B cell stimulation by factors secreted by non-B cell subsets which might contribute to IL-10⁺ B_reg differentiation. Caveats of this study are the limited availability of the clinical samples as well as the decreased lymphocyte counts within these samples due to patient treatment with lymphodepletive TMZ chemotherapy. However, a recent study by Khoder et al. [9] demonstrated that IL-10⁺ B_regs can be effectively examined from whole PBMCs. Therefore, we believe that our results represent the normal physiology of the IL-10⁺ B_reg compartment and that this compartment is not increased under elevated levels of serum BLyS.

IL-10⁺ B_regs represent a suppressive compartment of the B cell lineage. While several markers exist to identify different B_reg subsets in humans, the reported phenotype of these cells varies between studies [8, 28, 29]. A B cell population with the phenotype CD19⁺ CD24^high CD38^high has been shown to secrete large amounts of IL-10 upon stimulation through CD40 [28]. Another human B_reg cell subset equivalent to murine B10 cells, which expresses CD24 and the memory-related marker CD27 [8, 30], has also been identified. While different subsets exist for IL-10⁺ B_regs in both humans and mice, the IL-10⁺ B_reg subset evaluated in this study universally displays the phenotypic markers of CD19⁺ CD24^high CD38^high. Therefore, our study focuses on the evaluation of CD19⁺ CD24^high CD38^high IL-10⁺ B_regs, which are largely considered to be the main contributors of IL-10 secretion and regulation. Our study demonstrates that this population might not be as sensitive to the homeostatic cytokine BLyS as previously expected from studies that demonstrate the role of BLyS on IL-10⁺ B_reg homeostasis. It remains to be elucidated whether another distinct regulatory B cell subset characterized by CD1d^high CD5⁺ expression, which was previously determined to be sensitive to BLyS-induced expansion [10–13], will be elevated in these patients with elevated serum BLyS levels.

These data indicate that in the context of TMZ-induced lymphopenia and humoral immunity-targeting vaccines, IL-10⁺ B_regs may not be as responsive to BLyS as nonregulatory B cell subsets. These discrepancies may pertain to differences in BLyS receptor expression or the expression of intracellular adaptor molecules known to regulate receptor-mediated IL-10 induction [13]. Alternatively, in response to BLyS, serial vaccinations with PEPvIII-KLH-inducing anti-EGFRvIII immunoglobulins may differentially affect activation and maturation between regulatory and nonregulatory B cell subsets. Unlike the T_reg compartment [31–34], our current and previous findings [19] indicate that there may be diminished IL-10⁺ B_reg responses to tumor-specific vaccines.

These results have implications for interpreting and more effectively modulating humoral responses in human peptide vaccine trials. By delineating the interplay between BLyS, IL-10⁺ B_regs, and antigen-specific humoral responses, we may ultimately be able to deliver tangible benefits for patients with refractory tumors (i.e., GBM) expressing tumor-specific surface antigens. Importantly, our findings indicate that IL-10⁺ B_regs may not be a regulatory barrier in the use of BLyS as a humoral adjuvant for vaccine-based cancer immunotherapy.

Abbreviations

APC

Allophycocyanin

B_regs

Regulatory B cells

BAFF

B cell-activating factor

BFA

Brefeldin A

BLyS

B lymphocyte stimulator

CFSE

Carboxyfluorescein succinimidyl ester

EGFRvIII

Epidermal growth factor receptor variant III

FBS

Fetal bovine serum

FDA

Food and Drug Administration

FITC

Fluorescein isothiocyanate

GBM

Glioblastoma

Ig

Immunoglobulin

IL-10

Interleukin-10

IL-10⁺ B_regs

Regulatory B cells that secrete IL-10

KLH

Keyhole limpet hemocyanin

KPS

Karnofsky performance status

LPS

Lipopolysaccharide

mo

Months

o/n

Overnight

PBMCs

Peripheral blood mononuclear cells

PEPvIII

Peptide containing the EGFRvIII mutation

PerCP

Peridinin chlorophyll protein complex

PMA

Phorbol myristate acetate

SOC

Standard of care

T_regs

Regulatory T cells

TMZ

Temozolomide

XRT

External beam radiotherapy

Compliance with ethical standards

Conflict of interest

Employment or leadership position: none; consultant or advisory role: John H. Sampson, Celldex Therapeutics©; stock ownership: none; honoraria: none; research funding: John H. Sampson, Celldex Therapeutics©; expert testimony: none; other remuneration: John H. Sampson received funding from license fees paid by Celldex Therapeutics©. All other authors do not have any conflict of interest.