B cell-based therapy produces antibodies that inhibit glioblastoma growth

Glioblastoma (GBM) is a highly aggressive and malignant brain tumor with limited therapeutic options and a poor prognosis. Despite current treatments, the invasive nature of GBM often leads to recurrence. A promising alternative strategy is to harness the potential of the immune system against tumor cells. Our previous data showed that the B vax (B-cell-based vaccine) can induce therapeutic responses in preclinical models of GBM. In this study, we aim to characterize the antigenic reactivity of B Vax -derived antibodies and evaluate their therapeutic potential. We performed immunoproteomics and functional assays in murine models and human GBM patient samples. Our investigations revealed that B Vax distributes throughout the GBM tumor microenvironment (TME) and then differentiates into antibody-producing plasmablasts. Proteomic analyses indicate that the antibodies produced by B Vax display unique reactivity, predominantly targeting factors associated with cell motility and the extracellular matrix. Crucially, these antibodies inhibit critical processes such as GBM cell migration and invasion. These findings provide valuable insights into the therapeutic potential of B Vax -derived antibodies for GBM patients, pointing towards a novel direction in GBM immunotherapy.


Introduction
Glioblastoma (GBM) is an aggressive and malignant brain tumor that arises from glial cells (1).
GBM is one of the most common and deadly forms of brain cancer in adults with a median survival time of approximately 15 months after diagnosis.The current standard of care for glioblastoma includes surgery, radiation therapy, and chemotherapy, but the overall prognosis remains poor (2).A major obstacle in treating GBM is its remarkable ability to invade and migrate into surrounding healthy brain tissues, making complete gross total surgical resection impossible, leading to inevitable tumor recurrence (3).Therefore, there is a pressing need to explore alternative therapeutic avenues to inhibit GBM progression and improve patient outcomes.Harnessing the immune system to modulate tumor progression and remote sites of invasion is a compelling strategy for GBM (4).While most immunotherapy efforts have historically focused on T cells, the role of B cells, especially in the context of GBM, remains less explored.
Recent studies suggest that B cells and their secreted antibodies can influence tumor growth, metastasis, and response to treatment (5)(6)(7)(8)(9).The presence of tertiary lymphoid structures (TLSs) in solid cancers, containing B cells undergoing somatic hypermutation, confers a favorable prognosis (10)(11)(12).As such, our laboratory is developing a novel cancer immunotherapeutic approach using activated B cells as a cell-based vaccine (BVax) against GBM (13,14).BVax is a B cell-based vaccine comprising 4-1BBL + B cells activated through CD40 agonism and IFNG stimulation.Advantages to B cell therapy relative to other types of immunotherapies include its antigen-presenting capability (15)(16)(17), shared cognate antigen-specificity with T cells (18), ability to generate tumor-reactive antibodies (Abs) (12), and circulatory mobility enabling tumor and secondary lymphoid organ infiltration (19,20).Moreover, the relative ease as well as timely ex vivo activation and expansion from patient-derived circulating B cells reduces the cost of generating a personalized cell-based therapeutic.As such, B-cell-based vaccines represent a promising, yet under-investigated, immunotherapeutic treatment approach (21)(22)(23)(24) warranting further investigation in GBM.
In this study, we aimed to determine the humoral response induced by BVax, assess the tumor-reactive nature of BVax-derived antibodies (Ab)s, and evaluate their therapeutic potential in preclinical models.This analysis reveals the role of BVax in the immune-tumor interplay and its therapeutic potential for patients with GBM through a blend of molecular and proteomic analyses.

BVax differentiates into plasmablasts and harbors potentially tumor-reactive B-cell receptors.
To determine the humoral responses generated by BVax, we first analyzed the potential of BVax to migrate to the tumor and differentiate into Ab-producing cells (plasmablasts).Using the CD45.1 versus CD45.2 congenic mouse model (Figure 1A), we found that upon intravenous injection, CD45.1 + BVax preferentially migrated to the glioma-bearing brains 72 hours after injection (Figure 1B).Gene set enrichment analysis (GSEA) of BVax showed upregulation of the Gene Ontology (GO) gene set involved in leukocyte migration (Figure 1C and Supplemental Table 1).The potential of BVax to differentiate into Ab-producing cells after migrating into the glioma was confirmed in vivo using CT2A-bearing mice treated with BVax.Approximately 10% of the BVax displayed a CD38 + CD20 -CD19 + plasmablast phenotype (Figure 1D).Analysis of BVax and BNaive heavy chain receptor repertoire (bulk IgH sequence) reveals comparable BCR repertoire diversity between BVax and BNaive cells (Supplementary Figure 1A).However, BVax might present differential reactivity compared to BNaive cells (Figure 1E, Supplementary Table 2 and Supplementary Table 3).Among BVax BCRs, 92 were shared with glioma-infiltrating B cells (tumor-infiltrating (TI) B cells, Figure 1F and Supplementary Table 4).Additionally, around 2% of BVax BCRs overlapped with glioma-infiltrating B cells while being absent in BNaive BCRs (Supplementary Figure 1B).This suggests that BVax may harbor tumor-reactive BCRs.
Characterization of murine BVax-derived immunoglobulins reactivity.To examine the BVax antigenic reactivity, BVax-and BNaive-derived immunoglobulins (Igs) were produced in vivo by adoptive transfer of BVax or BNaive into CT2A-bearing B-cell deficient mice (Figure 2A).After 2 weeks, blood from the experimental groups (BVax and BNaive) was collected, and the Igs were isolated using a Protein A/G Spin Column.Comparable amounts of Igs were obtained in both groups (Figure 2B).
To analyze the reactivity of these Igs, immunoproteomics was performed using immunoprecipitation-mass spectrometry (IP-MS).Intracranial glioma homogenates were produced and fractionated into the membrane and cytosolic fractions (Figure 2C).Proteins enriched in the membrane fraction were then immunoprecipitated with the BVax or BNaive-derived Igs.Commercially available mouse IgG (cIgG) was used as a control.The MS analysis revealed that BVax Igs have a unique reactivity compared to BNaive-derived Igs or control Igs (Figure 2D and Supplementary Table 5).More specifically, the BVax produced Igs with preferential reactivity to factors involved in cell motility, extracellular matrix (ECM), and membrane organization, such as fibrinogen, fibronectin, and myosin, as shown by the GO pathway analysis (Supplementary Figure 2).To determine whether the BVax antibodies specifically target ECM and membrane proteins when isolated from a more controlled environment, we conducted additional immunoprecipitation experiments using the conditioned medium and cell-membrane fractions from CT2A tumor cells cultured in vitro.Consistent with the initial findings, BVax antibodies specifically targeted membrane (such as collagen receptor, integrin-linked kinase (ILK) complex, tight junction proteins, Rho GTPases, vinculin and nischarin) and ECM proteins (such as gelsolin, collagens, fibrinogen, EGFcontaining fibulin-like extracellular matrix protein 2, SPARC-like protein 1, matrix metalloproteinases, caldesmon and asporin) involved in cell adhesion and motility (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39) (Supplementary Figure 3 and Supplementary Table 6), confirming their preferential reactivity.

Characterization of patient BVax-derived Igs reactivity.
To examine the biological relevance of the murine model, we evaluated the BVax antigenic reactivity from GBM patients.BVax from newly diagnosed GBM patients were differentiated into plasmablast in-vitro (Figure 3A).Flow cytometry analysis confirmed that 10 days after activation, approximately 30% of cells generated from either BNaive or BVax showed a CD19 + CD20 -CD38 + plasmablast phenotype (Figure 3, B and C), suggesting that both B-cell types have similar in vitro polyclonal potential to differentiate into plasmablast.Supernatants were then collected every 3 days and the presence of secreted Igs was confirmed by western blot (Figure 3D) and ELISA (Supplementary Figure 4).The production of IgG (both IgH and IgL chains) was detected in both BVax and BNaive conditions, with clear bands appearing from day 6 onward, supporting the efficacy of the ex vivo generation method in producing GBM patient-derived antibodies.
In parallel to the murine glioma analysis, Igs were isolated using a Protein A/G Spin Column from supernatants of the B-cell cultures harvested every 3 days.Due to the limited amount of Igs, the IP-MS was performed from autologous bulk glioma protein homogenates without fractionation (Figure 4A).The patient's serum Igs were included to evaluate the peripheral baseline reactivity level to the autologous tumor.The IP-MS data (Figure 4B and Supplementary Table 7) showed heterogeneity across patients (NU02545, NU0569, and NU02594).However, several regulators of the ECM (such as fibrinogen, versican core protein, and collagens) and cell motility (such as myosin and actin) were found across all patients' BVax-Igs.Furthermore, GO pathway analysis confirmed that BVax shows a preferential reactivity towards biological processes involved in cell migration and motility (Supplementary Figure 5).To corroborate these findings, were conducted additional immunoprecipitation experiments using conditioned medium and cellmembrane fractions from GBM43 tumor cells cultured in vitro.These experiments confirmed that BVax antibodies specifically targeted ECM (such as gelsolin, collagens, matrix metalloproteinases, thrombospondins and laminin) and membrane (such as integrin-linked kinase (ILK) complex, epidermal growth factor receptor (EGFR), flotillin-2 (FLOT2), and Rho GTPases) proteins enriched in these fractions (Supplementary Figure 6 and Supplementary Table 8), confirming their preferential reactivity.Spatial proteomics and histopathological analysis of tumors from these patients illustrate the spatial distribution of BVax-derived Igs-recognized antigens within the ECM of GBM (Figure 5).Gelsolin (GSN), fibronectin (FN1), versican (VCAN), fibrinogen (FGB), myosin type 1 C (MYO1C), and collagen type IV alpha (COL4A) correspond to perivascular clotting and hemorrhagic areas.The diverse staining patterns across patients NU02545 (Figure 5A and Supplementary Figure 7), NU02594 (Figure 5B, Supplementary Figure 8 and Supplementary Video 1), and NU02569 (Figure 5C, Supplementary Figure 9 and Supplementary Video 2) indicate patient-specific variations in ECM composition and immune response.Such heterogeneity could explain the inconsistency of BVax-derived Igs reactivity in different GBM patients.Additionally, B cells were found in regions near BVax-derived Igsrecognized antigens (Supplementary Figure 10), and peritumoral brain displayed a low expression of these extracellular matrix proteins in comparison to the GBM microenvironment itself (Supplementary Figure 11), providing a therapeutic potential for BVax-derived Igs.

BVax-Igs inhibit GBM invasion and migration.
To investigate whether patient BVax-derived Igsrecognized antigens are accessible to the immune system, intra-operative high molecular weight microdialysis, which collects proteins secreted into the interstitial fluid space (40), was performed in the tumor and brain adjacent to the tumor of GBM patients (GBM WT 3, and GBM WT 4) and one patient with grade 4 IDH-mutant astrocytoma (Astro 4-mut 3), followed by mass spectrometry analysis (Figure 6A).High molecular weight catheters (100 kDa) were applied to maximize the volume and diversity of the microdialysis (40).During the resection, catheters were placed in radiographically enhancing tumor (X) and non-enhancing tumor (Y), and relatively normal brain adjacent to tumor (Z).Mass spectrometry of the microdialysates revealed that most BVax-derived Igs-recognized antigens were more abundantly secreted in the enhancing tumors than in normal brains, although one patient showed more BVax-derived Igs-recognized antigens secreted in the non-enhancing tumors (Figure 6B and Supplementary Table 9).Some BVax-derived Igsrecognized antigens, such as myosin (MYH) (Figure 4B), were not identified in the microdialysates (Figure 6B).One possible explanation could be their size surpassing the molecular weight cut-off of the catheters (>100 kDa).Overall, these findings demonstrate the presence of diverse antigenic profiles across heterogeneous zones within each tumor and the potential accessibility of these antigens to the immune system.The presence of these BVax-derived Igs-recognized antigens in the extracellular fluid reiterates the potential of the TME as a reservoir of significant therapeutic targets.
The ECM and cell motility are key biological processes controlling tumor cell migration, invasion, and the epithelial-mesenchymal transition, which are hallmarks of GBM malignant behavior (41)(42)(43)(44).Thus, we hypothesize that BVax-Igs could inhibit ECM and cell motility processes by recognizing these components or regulators.To test this hypothesis, a patient-derived xenograft (PDX) cell line (GBM43) was used, and BVax-Igs from 3 GBM patients (NU03592, NU03614, and NU03636) to test this hypothesis.Igs from autologous plasma samples were used as control.An ex vivo PDX functional assay including invasion and migration assessment was performed using a commercially available matrix that recapitulates mammalian ECM (Matrigel) (Figure 7A).BVax-Igs did not affect the PDX cell viability (assessed by ATP activity, Figure 7B) compared to paired serum Igs (serum Igs).However, BVax-Igs significantly inhibited the PDX from migrating (Figure 7C-E mice (Supplementary Figure 12).Histological analysis of muMT mice treated with BVax cells from CT2A tumor-bearing donors demonstrated a decrease in satellite formation away from the tumor core, supporting the disruption of cell invasion (Figure 8, A and C).Consistent with our previously reported results (13), only mice with orthotopically implanted CT2A treated with BVax-Igs have a significant increase in median survival (Figure 8D).To further investigate the important role of BVax-Igs, we conducted experiments treating CT2A tumor-bearing mice with BVax generated from wildtype (WT) or mice where B cells are deficient in Prdm1 (Cd19 Cre x Prdm1 Flox mice, kindly provided by Dr. Nicole Baumgarth from Johns Hopkins Medicine).Prdm1 encodes Blimp1, a key factor for the development of immunoglobulin-secreting plasma cells (45,46).Notably, Prdm1-deficient BVax-treated mice demonstrated a compromised reduction in survival (Figure 8E), underscoring the essential role of BVax-induced immunoglobulins in mediating the survival benefits observed with BVax treatment.
Based on these results, we concluded that BVax has the potential to produce antibodies reactive to the tumor ECM and components of the cell motility and could interfere with the ability of the tumor to migrate and invade the non-tumor tissue, and ultimately impact the overall tumor growth.

Discussion
Herein, we show that BVax elicits anti-tumor reactivity, as evidenced by selective migration to glioma-bearing brains, differentiation into plasmablasts, and secretion of specific Igs.BVax-derived Igs bind to factors predominantly involved in cell motility and extracellular matrix, essential for GBM invasion and motility (42,(47)(48)(49).These antibodies are also functionally active against the key processes of cancer progression revealing a strategy for developing novel immunotherapeutic strategies against GBM.
The BVax-derived Igs recognition of specific ECM components, such as gelsolin, fibronectin, fibrinogen, versican, and collagens, is particularly intriguing.Traditionally seen as a physical scaffold for cells, the ECM is now increasingly recognized for its role in modulating tumor behavior, progression, and response to therapy (43,44).The ECM and the hypoxic microenvironment orchestrate the mesenchymal transition, a biological process associated with the aggressive pathological properties of GBM and therapeutic resistance (49)(50)(51).Recently, the finding of collagen 1 A1 (COL1A1)-abundant oncostreams (52) and cancer-associated fibroblasts (53) in GBM and their protumor effects reinforce the concept that components of the ECM might not be a mere bystander but could actively participate in the progression of GBM.The preferential reactivity observed in murine models and human GBM samples highlighting its translational potential.
To date, relatively few antigens recognized by tumor-infiltrating B cell (TIB)-derived antibodies have been identified in other cancer models, partially limited by the low TIB number from fresh tumors (54)(55)(56)(57)(58)(59)(60).Here, by expanding and differentiating BVax into antibody-secreting plasmablast, we were able to identify potential antigens recognized by BVax-derived antibodies in mouse and human GBM models.This method can also be used in other cancer models, expanding the repertoire of TIB target antigens.
Most TIB target antigens identified in patient tissue samples span across the nuclear, cytoplasmic, and extracellular compartments (24).Our data demonstrated that cytoplasmic and extracellular proteins could both be recognized by BVax antibodies.While it is expected that B-cellderived Igs could recognize extracellular proteins, the detection of intracellular cytoplasmic proteins was surprising.However, the analysis of the GBM secretome obtained from microdialysis of enhancing tumor, non-enhancing tumor, and normal brain showed that cytoplasmic proteins can be detected in the extracellular fluid, hinting that cytoplasmic proteins are also accessible to the immune system.However, the mechanism behind this is still not clear.One could hypothesize that soluble antigens from the interstitial fluid could travel to regional lymph nodes, such as the deep cervical lymph nodes (61) and potentially activate B cells.
Although our data indicate that BVax-Igs contribute significantly to improved survival, it is essential to consider the potential contributions of other immune mechanisms of BVax, such as CD8 + T cell activation, which also plays an important role in the BVax-conferred therapeutic effect (13,62).Further investigations are needed to delineate the interplay between BVax-Igs and other immune mechanisms of BVax, and how these mechanisms collectively enhance patient outcomes.
The heterogeneity observed in the BVax antigenic reactivity across different GBM patients reminds us of the complex and diverse nature of GBM tumors.Although certain regulators like fibrinogen, myosin, and collagens emerged consistently, future research should aim at identifying and characterizing other potential tumor antigens bound by BVax-Igs using a larger cohort of patient samples.In conclusion, our research highlights the therapeutic potential of BVax-derived Igs in GBM therapy.Through their unique tumor-reactive nature, BVax and the antibodies they produce offer a novel and promising strategy against this formidable malignancy.Future studies should focus on how to unleash the full potential of BVax in an immune-suppressive GBM TME.

Methods
Sex as a Biological Variable.In this study, both male and female mice were used in the animal model experiments to investigate the effects of B cell-based therapy on glioblastoma growth.However, sex was not considered as a biological variable in any of the analyses.For studies involving human samples, sex was also not considered as a biological variable due to limitations in sample availability and the specific focus of the study.Intracranial Tumor Implantation.Each mouse was implanted with 1x10 5 glioma cells in a total volume of 2.5µl of PBS.Mice were anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg) via intraperitoneal injection.After shaving the surgical site and disinfection with povidoneiodine and 70% ethanol, an incision was made at the midline for access to the skull.A 1 mmdiameter burr hole was drilled 2mm posterior to coronal suture and 2 mm lateral to the sagittal suture.Injections were performed using a Hamilton syringe fitted with a 26-gauge blunt needle at a depth of 3.5mm.The injection site was then sutured closed.
GBM Patient-Derived BVax/BNaive immunoglobins (Igs) Generation.BVax/BNaive cells were cultured using the ImmunoCult™ Human B Cell Expansion Kit (100-0645, STEMCELL).Cells were seeded at a density of 5x10 5 cells/mL in B cell expansion medium and incubated at 37°C in a CO₂ incubator.Every 3 days, supernatants were harvested and quantified for IgG using the IgG (Total) Human Uncoated ELISA Kit (88-50550-22, Invitrogen).The supernatants were then stored at -80°C, and cells were replenished with fresh expansion medium.After several rounds of supernatant collection, all supernatants were pooled, and the Igs were purified using a NAb Protein A/G Spin Column (89962, Thermo Fisher).
Murine BVax/BNaive Igs Generation.The murine Igs generation is depicted in Figure 2A.Briefly, BVax/BNaive cells were adoptively transferred into B cell knockout mice (muMT mice).After 2 weeks, blood from BVax/BNaive recipient mice was collected, and the Igs were isolated using a Protein A/G Spin Column.
Murine GBM Tumor Lysate for IP-MS Using Cellular Fractionation.CT2A cells were implanted intracranially into C57BL/6 mice.After 21 days, the brains were collected, and the tumors were dissected and flash frozen.Per manufacturer guidelines, cellular fractionation was performed with the Thermo Fisher Mem-PER Plus Kit (Catalog # 89842Y).Protein quantification was performed using a Bradford Assay (Biorad Catalog # 500-0006) per manufacturer guidelines.

Immunoprecipitation (IP).
Cell/tissue lysates were incubated and tumbled overnight at 4°C with either BVax-derived Igs or IgG control.BVax-derived Igs were generated in vivo for the murine samples and ex vivo for patient-derived samples as described in Figure 2A and 3A.Pulldown experiments were then performed with magnetic Dynabeads Protein G (ThermoFisher Scientific).
For GBM patient derived samples, we used 1-2mg tumor lysate with 20-60µg Igs.For murine intracranial GBM tumor samples, we used 1.5mg tumor lysate with 30µg Igs.For GBM43 cell derived samples, we used 2-4mg membrane or conditioned medium proteins with 40-120µg Igs.
For CT2A cell derived samples, we used 8mg membrane or conditioned medium proteins with 160µg Igs.Bead/Protein complexes were isolated, then washed 3 times with lysis buffer and 2 times with Triton-X buffer (HEPES 40mM, NaCl 120mM, EDTA 1mM, NaPP 10mM, NaF 50mM, Triton-X 0.5% and beta-glycerophosphate 10mM).The beads were then boiled for 10 minutes at 95°C in 2X Laemmli sample buffer, and 20% of each sample was resolved by SDS-PAGE, and 80% was submitted to the proteomics core for mass spectroscopy analysis.
Proteomic Immunoprecipitation Analysis using LC-MS/MS.LC-MS/MS was performed as previously described (63).For visualization of protein targets identified through IP followed by mass spectrometry, R programming language was used leveraging specialized packages such as 'ggplot2' and 'pheatmap'.Our datasets encompassed protein measurements derived from three tumor samples from GBM patients, in vitro cultures and mouse samples, which included protein targets extracted from control Ig, BNaive or Serum Ig, and BVax Ig groups.For heatmap visualization, the Ward's method (64) was adopted to cluster samples, ensuring a more coherent and intuitive presentation.
Flow Cytometry.Flow cytometry was performed as described previously (13).The following antihuman Abs were used (all from BioLegend): 4-1BBL PerCP-Cy5.Western Blot.BVax was generated ex vivo from GBM-patient peripheral blood samples and subsequently activated as described previously (13).Supernatants were collected every few days during the activation protocol and prepared for western blot in non-reduced and reduced fractions.
RNA isolation from BVax, BNaive, and TI B cells was performed using Trizol (Invitrogen).BCR sequencing and bioinformatic analysis were performed by Adaptive Biotechnologies using the ImmunoSEQ platform.For optimal concentration and the best signal/noise ratio, all antibodies were tested at three different dilutions, starting with the manufacturer-recommended dilution (MRD), then MRD/2 and MRD/4.Secondary Alexa fluorophore 555 (Thermo Fisher Scientific A32727) and Alexa fluorophore 647 (Thermo Fisher Scientific A32733) were used at 1/200 and 1/400 dilutions, respectively.The optimizations and full runs of the multiplex panel were executed using the seqIF™ methodology integrated into the Lunaphore COMET™ platform (characterization 2 and 3 protocols, and seqIF™ protocols, respectively (65)).The staining can be performed at a maximum of 4 tissue slides simultaneously, where automated cycles of 2 antibodies staining at a time, followed by imaging and elution are fully automated and no sample manipulation is required.
All reagents were diluted in Multistaining Buffer (BU06, Lunaphore Technologies).The elution step lasted 2min for each cycle and was performed with Elution Buffer (BU07-L, Lunaphore Technologies) at 37°C.Quenching lasted for 30sec and was performed with Quenching Buffer (BU08-L, Lunaphore Technologies).Staining incubation times lasted 4min for all primary antibodies and 2min for secondary antibodies.Imaging was performed in Imaging Buffer (BU09, Lunaphore Technologies) with an integrated epifluorescent microscope at 20x magnification.
Image registration was performed immediately after concluding the staining and imaging procedures by COMET™ Control Software.Each seqIF™ protocol resulted in a multi-layer OME-TIFF file where the imaging outputs from each cycle were stitched and aligned.COMET™ OME-TIFF files contain a DAPI image, intrinsic tissue autofluorescence in TRITC and Cy5 channels, and a single fluorescent layer per marker.Markers were subsequently pseudocolored for visualization of markers in the Viewer from Lunaphore.
Intra-operative microdialysis.Intra-operative microdialysis was performed at Mayo Clinic under an approved IRB protocol 19-004694 during three standard-of-care glioma resections based on previously published methods (40), including one for a recurrent grade 4 IDH-mutant astrocytoma (Astro 4-mut 3), one primary GBM (GBM WT 3), and one recurrent GBM (GBM WT 4).Briefly, three high molecular weight microdialysis catheters (100 kDa; M Dialysis 71 High-cut off brain microdialysis catheters) were inserted into radiographically diverse regions (enhancing, non-enhancing, and normal brain) based on stereotactic neuronavigation.Microdialysis was performed at a flow rate of 2 ul/min using the 107 microdialysis pump and perfusion fluid with 3% Dextran 500 kDa to improve analyte recovery.Catheters were flushed prior to insertion to minimize dead space, and microvials were then changed every 20 minutes until the sampling area needed to be resected.
Microdialysates were split into two aliquots and then placed on dry ice.The third aliquot after catheter insertion was sent for mass spectrometry proteomic analysis at the Mayo Clinic Proteomics Core via LC-MS for label-free relative quantitation by iBAQ (intensity Based Abundance Quantity).
Migration assay.This assay used a silicone insert with a defined cell-free gap (80206, IBIDI), as described previously (66).Briefly, cells were seeded at 3.5 × 10 4 per well of culture insert chamber (80206, IBIDI).Cells were left undisturbed for approximately 12h in a 37°C CO2 incubator.After successful attachment, cells were then washed with Dulbecco's phosphate-buffered saline (DPBS) and the silicone insert was carefully lifted using sterile forceps.Culture medium was switched to cDMEM containing 100µg/ml BVax or serum Abs.At different timepoints, imaging was obtained by bright field via Leica Microscope until visible gap closed.Distance or area between cells was measured via pixels on Fiji ImageJ.Migration index was calculated as follows: Migration index (%) = (Wound area at 0h-Wound area at 24h)/ Wound area at 0h × 100.
Briefly, after rehydration of the chamber, 5 × 10 4 cells were suspended in DMEM and 100µg/ml BVax or serum Abs and seeded into the invasion chamber.The bottom chambers were filled with 750µl cDMEM, which contains DMEM, 10% FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin.After 24h incubation, invaded cells were fixed overnight in 4% formaldehyde, then washed and stained with DAPI (P36931, Thermo Fisher) as described previously (67).The membrane was imaged, and invading cells were counted using Fiji ImageJ.
Cell viability assay.CellTiter-Glo® 2.0 Cell Viability Assay kit (G9242, Promega) was used to assess viability of PDX cells after treatment with BVax-derived Abs or serum Abs (100µg/ml), as described by the manufacturer.Because muMT mice do not have endogenous Igs, and by using anti-mouse Ig will detect BVax or BNaive-derived antibodies.Nearby sections were stained for H&E to check the organization of the tumors.Immunofluorescence images were taken from peritumoral region, intratumoral region, and relatively normal brain.For quantification of immunoglobulin intensity, 10-15 images were taken around the peritumoral region of each mouse.Mean Fluorescence Intensity (MFI) of antimouse IgG and IgM (red) in each image were quantified as previously using ImageJ (68,69).For evaluation of the invasive feature of CT2A tumor cells after treatment, satellites away from the CT2A tumor core were quantified based on H&E images from each mouse (70).In GL261 glioma model, BVax or BNaive cells from GL261 tumor-bearing C57BL/6 mice were adoptively transferred into GL261 tumor-bearing muMT (B cell knockout) mice 7 days after tumor injection.Each mouse received 2×10 6 cells every 3 days.7 days after the 2 nd treatment, brain tissues from recipient mice were harvested, stained and quantified as in CT2A glioma model.

Direct evidence of
Essential Role of BVax Induced Immunoglobulins.CT2A-bearing B deficient mice were treated with purified BVax or BNaive Igs (12.5 μg/mouse/injection) as described previously (13), and their survival was monitored.Additionally, CT2A-bearing B deficient mice were treated with10 6          and IgM (red) in each image were quantified using ImageJ as described previously (68,69).PBS ) and invading (Figure 7F-G).Additionally, a migration assay using BVax-Igs from patient NU03762 and the corresponding PDX line (NU03762 PDX) was conducted.Consistent with initial findings, BVax-Igs significantly inhibited the migration of the paired PDX line (Figure 7H-I).These results support the potential of BVax-Igs to interfere with key processes involved in GBM progression.To provide direct evidence of BVax induced immunoglobulins in tumors, we conducted additional experiments where BVax cells from healthy or CT2A tumor-bearing C57BL/6 mice were adoptively transferred into CT2A tumor-bearing muMT (B cell knockout) mice.Following treatment, we harvested brain tissues from recipient mice and performed immunofluorescence (IF) staining for anti-mouse IgG and IgM to assess the presence and localization of BVax-derived antibodies.Our results demonstrated that mice receiving BVax cells from CT2A tumor-bearing donors exhibited significant IgG and IgM staining in the peritumoral region compared to those treated with BVax cells derived from healthy donors or BNaive cells (Figure 8, A and B).Similar results were observed in the GL261 model where BVax or BNaive cells from GL261 tumor-bearing C57BL/6 mice were adoptively transferred into GL261 tumor-bearing muMT (B cell knockout) Human Specimens.The Nervous System Tumor Bank collected all human tissue samples at Northwestern University under the Institutional Review Board protocol number STU00202003, involving nine patients with IDH-wildtype glioblastoma.The Mayo Clinic Cancer Center collected all human tissue samples at Mayo Clinic under the Institutional Review Board protocol number 19-004694.All patients who contributed to this study signed a written consent form, and the study was conducted using the ethical standard from the U.S. Common Rule of Ethical Standards.Samples collected from high-grade glioma patients included freshly resected tumors, peripheral blood, frozen tumors, and paraffin-embedded tissue sections.A neuropathologist reviewed all H&E sections to confirm at least 50% presence of tumor based on cellularity.CellLines.CT2A cells were obtained from Millipore (Sigma-Aldrich), GL261 cells were obtained from NCI (National Cancer Institute), and GBM43 patient-derived xenograft (PDX) glioma cell lines were obtained from Dr. David James at Northwestern University.NU03762 PDX cells were obtained from Dr. Craig M. Horbinski at Northwestern University.NU03762 PDX cells were maintained in complete RPMI (RPMI supplemented with 10% fetal bovine serum (FBS; Hyclone), 100 U/ml penicillin (Corning), and 100 mg/ml streptomycin (Corning), 0.1% 2-mercaptoethanol (Sigma Millipore), 2 mM L-glutamine (Invitrogen), 25 mM HEPES (Invitrogen) and 1 mM sodium Pyruvate (Invitrogen)).The rest cell lines were maintained in Dulbecco's modified Eagle's medium (DMEM) (Corning) supplemented with 10% fetal bovine serum (FBS; Hyclone), 100 U/ml penicillin (Corning), and 100 mg/ml streptomycin (Corning) and incubated at 37°C in 5% CO2.Every 2 months cell lines were tested for Mycoplasma contamination using the Universal Mycoplasma Detection Kit (ATCC 30-1012K).Murine Models.Mice used in this study included C57BL/6 and B-cell Knockout (muMT) purchased from The Jackson Laboratory and bred for use in experiments.Cd19 Cre x Prdm1 Flox mice, kindly provided by Dr. Nicole Baumgarth from Johns Hopkins Medicine, Baltimore.Animal experiments were initiated when the mice were six to eight weeks old.The Institutional Animal Care and Use Committee (IACUC) at Northwestern University approved all procedures conducted in this study under the protocol number ISO16696.All animals were housed at the Simpson Querry Center for Comparative Medicine in a dedicated pathogen-free animal facility with 12-hour light and 12-hour dark cycles and ad libitum access to food and water.For in vivo studies, the sample size for each experiment is indicated in the figure legend.The investigators were not blinded for any experiments.This study incorporated sex as a biological variable by including both male and female mice.
BVax generated from WT or Cd19 Cre x Prdm1 Flox mice (B cells deficient in Prdm1 kindly provided by Dr. Nicole Baumgarth from Johns Hopkins Medicine), and their survival was monitored.Statistics.GraphPad Prism version 8 (GraphPad Software, Inc.) and R version 4.2.3 (R Foundation for Statistical Computing) were used for all statistical analysis.The sample size for the experiments was ≥ 3. Results are represented as the mean +/the standard deviation unless otherwise indicated.Comparisons between two groups were conducted using two-tailed Student's t test.Comparisons between more than two groups were conducted using one-way ANOVA with Tukey's or Dunnett's post hoc multiple comparison tests.For animal survival experiments, Kaplan-Meier survival curves were generated, and a log-rank test was applied to compare survival distributions.All reported P values were two-sided and considered statistically significant at *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.Study Approval.All animal experiments conducted in this study were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at Northwestern University under protocol number ISO16696.The study protocols adhered to the IACUC guidelines to ensure the ethical treatment of animals.For studies involving human samples, the research was reviewed and approved by the Institutional Review Board (IRB) at Northwestern University (IRB number STU00202003), and the Mayo Clinic (IRB number 19-004694).Written informed consent was obtained from all human subjects prior to their participation in the study.

Figure 2 .
Figure 2. Characterization of murine BVax-derived Ig reactivity.(A) Schema demonstrating how in vivo BVax-derived Igs are produced from mice bearing CT2A gliomas.(B) Amount of BVaxderived Igs generated from mice bearing GBM tumors.n=7 for each group.(C) Schema depicting the protocol for the murine immunoprecipitation-mass spectrometry (IP-MS) experiments used to identify tumor specific antigens recognized by BVax-derived Igs.(D) Heatmap revealing hierarchical clustering of GBM tumor antigens recognized by BVax-derived Igs.Each triplicate corresponds to an independent IP-MS experiment.In Triplicate 1, BVax-derived Igs were pooled from 10 mice, and BNaive-derived Igs were pooled from 11 mice.In Triplicate 2, BVax-derived Igs were pooled from 11 mice, and BNaive-derived Igs were pooled from 10 mice.Triplicate 3 involved BVax-derived Igs pooled from 10 mice and BNaive-derived Igs pooled from 12 mice.Data are the mean ± s.d. and were analyzed by Student's t test.ns represents no significance.

Figure 3 .
Figure 3. Production of GBM patient BVax-derived Igs.(A) Schema of the ex vivo generation of GBM patient BVax-derived antibodies.(B) Dot plots of flow cytometry analysis of CD20 and CD38 expression during different stages in the BVax activation protocol to generate GBM patientderived antibodies ex vivo.(C) Box and whisker plot of the percentage of plasmablasts generated at day 10 of BVax/BNaive activation in GBM patients.n=5 for each group.(D) Western blot confirming presence of antibody in the media during various stages of the GBM patient ex vivo BVax activation protocol.Data are the mean ± s.d. and were analyzed by two-tailed Student's t test.ns represents no significance.

Figure 4 .
Figure 4. Characterization of GBM patient BVax-derived Igs reactivity.(A) Schema depicting the protocol for the human IP-MS experiments used to identify tumor-specific antigens recognized by BVax-derived antibodies.(B) Heatmap revealing hierarchical clustering of GBM tumor antigens recognized by BVax-derived Igs.n=3.Targets related to adhesion, motility, or the extracellular matrix (ECM) are shown in red.

Figure 6 .
Figure 6.BVax-derived Igs-recognized antigens are detected in the extracellular fluid with brain microdialysis.(A) T1-post gadolinium axial and T2 FLAIR coronal MRIs demonstrating the stereotactic target location of each catheter in enhancing and non-enhancing tumor, and normal brain (patient GBM WT 3 shown).(B) Heatmap reveals the relative intensity of high-grade glioma antigens recognized by BVax-derived Igs in the micro dialysate.Samples were collected from 3 distinct patients, different from those in Figure 5.

Figure 7 .
Figure 7. BVax-derived Igs inhibit tumor invasion and migration.(A) Schema depicting the protocol for ex vivo functional assay.(B) Cell viability of GBM43 cells after treated with serum-or BVax-derived Igs from GBM patients (NU03592, NU03614 and NU03636).n=3.(C) Representative images of wound areas (marked by yellow lines) on confluent monolayers of GBM43 cells at 0h and 24h treated with serum-or BVax-derived Igs from a GBM patient (NU03592).(D) Quantification of the wound area of GBM43 cells at different time points treated with serum-or BVax-derived Igs from a GBM patient (NU03592).(E) Quantification of the migration index of GBM43 cells at 24h treated with serum-or BVax-derived Igs from GBM patients (NU03592, NU03614, and NU03636).n=3.(F) Representative images and (G) quantification of invading GBM43 cells (DAPI + ) at 24h treated with serum-or BVax-derived Igs from GBM patients (NU03592, NU03614 and NU03636).n=3.Each white dot represents a single invaded cell, scale bars =250µm.(H) Representative images of wound areas (marked by yellow lines) on confluent monolayers of PDX cells at 0h and

Figure 8 .
Figure 8. BVax from tumor-bearing mice have a superior ability to produce antibodies that localize to the peritumoral region and promote GBM-bearing mice survival.(A) Representative images of H&E and immunofluorescence (IF) staining for anti-mouse IgG and IgM to assess the presence and localization of BVax-derived Igs.BVax or BNaive cells from healthy or CT2A tumor-bearing C57BL/6 mice were adoptively transferred into CT2A tumor-bearing muMT (B cell knockout) mice.Following treatment, brain tissues were harvested from recipient mice and stained for anti-mouse IgG and IgM (red).H&E images show the organization of the tumors and the locations where the IF images were taken: peritumoral region (dotted green line), intratumoral region (orange box), and relatively normal brain (purple box).(B) Quantification of the relative intensity of BVax-Igs in the peritumoral region.10-15 images were taken around the peritumoral region of each mouse (dotted green line).Mean Fluorescence Intensity (MFI) of anti-mouse IgG and IgM (red) in each image were quantified using ImageJ as described previously(68,69).PBS