Blocking VPS34 protein boosts neuroblastoma immunotherapy efficacy

A team at Penn State College of Medicine has identified a method to make high-risk neuroblastoma cells more visible to the immune system. By blocking the PIK3C3/VPS34 protein, researchers raised surface levels of GD2, a key target for immunotherapy drugs. This strategy improved treatment outcomes in cell experiments and mouse models, addressing the challenge of cancer cells evading detection.

The mechanism behind enhanced visibility

High-risk neuroblastoma is an aggressive solid tumor affecting children, originating in the nervous system. Although targeted immunotherapies have improved survival rates, the American Cancer Society notes that cancer can still hide from the immune system, leading to relapse or resistance. The new study, published in the journal Autophagy, demonstrates that inhibiting VPS34 makes tumor cells easier to detect and more vulnerable to destruction by monoclonal antibodies.

These lab-made proteins are engineered to bond specifically to GD2 on the tumor surface, marking the cells for attack by the patient’s own immune cells. Hong-Gang Wang, the corresponding author and Lois High Berstler Professor of Pediatrics at Penn State, stated that VPS34 plays a critical role in whether these tumors survive and respond to therapy. He noted a potential path to boost the effectiveness of existing anti-GD2 drugs, such as dinutuximab and naxitamab, though further research is required.

Surprising results from autophagy inhibition

VPS34 is an enzyme first discovered decades ago in yeast cells, known for sorting and transporting proteins to and from the cell membrane. It later became recognized for its role in autophagy, a process where cells break down and recycle nutrients to survive stress. Jiawen Zhang, a biomedical sciences doctoral student and first author, explained that prior work linked autophagy to both tumor survival and immune response. Tumors growing rapidly often rely on autophagy for nutrients, especially in areas with poor blood supply.

"We had previously found that when we block autophagy, no matter the stage, it inhibited tumor growth, regardless if there was an immune response or not. But it had better efficacy in the models with working immune systems, which suggested it is probably also triggering the immune system to attack the tumors," Zhang said.

In this study, the team observed a strong link between VPS34 inhibition and increased anti-GD2 immunotherapy effectiveness. However, the mechanisms were unexpected. When VPS34 was blocked in lab cell experiments—where cancer cells have ample oxygen and nutrients and do not heavily rely on autophagy—the tumor cells died anyway, independent of any immunotherapy treatment.

Shifting GD2 distribution on cell surfaces

Blocking VPS34 sharply increased the amount of GD2 on the surface of tumor cells, as measured by flow cytometry. This laboratory technique analyzes specific molecules by tagging them with fluorescent markers and passing individual cells through a laser to identify those carrying higher levels of GD2. While the total amount of GD2 within the cell rose, the surface concentration increased disproportionately, shifting the balance toward the outside of the cell.

"The most exciting finding for us in this study was that VPS34 inhibition increased GD2 presentation, which would make the tumor more visible to the antibody," Wang said. Neuroblastoma cells with more surface GD2—a fatty molecule with a small sugar chain—are generally more "visible" to antibodies, providing more targets for them to latch onto.

To isolate the effect, researchers conducted control experiments blocking ATG14, a separate protein that partners with VPS34 for autophagy but leaves VPS34’s other functions untouched. The impact on GD2 was dramatically reduced in these controls. This points to the importance of VPS34’s initially discovered role: endolysosomal trafficking, which involves how the cell sorts and routes material to different destinations, including the cell membrane.

Combination therapy improves survival

In mouse models, tumors grew poorly when researchers targeted ATG14 and other genes associated with autophagy. Wang added that the two functions of VPS34—autophagy and endolysosomal trafficking—likely work together to help cancer cells survive. When the team paired VPS34 inhibition with monoclonal antibodies in mice and cells, the combination outperformed either approach alone.

Zhang explained that the resulting "brighter" cancer cells with elevated GD2 levels became easier targets not only for antibodies but also for natural killer (NK) cells. These NK cells bind to the antibodies and release perforin, a protein that creates holes in the cancer cell membrane, along with granzymes, enzymes that enter through those holes to trigger cell death. In both cell experiments and mice, tumor growth was significantly suppressed, and survival rates improved in the animal models.

"Combining the two therapies gave us better results," Zhang said. According to Wang, high levels of GD2 are also found in melanoma and osteosarcoma, but anti-GD2 immunotherapy is currently only clinically approved as standard care for neuroblastoma. "This could potentially be tested in other tumor types that express high GD2 on their surface, too," he noted.

Path forward for clinical application

Wang highlighted that VPS34 inhibitors are still mostly restricted to pre-clinical studies, with off-target effects often complicating research in live subjects. He expressed hope that this work encourages the scientific community to view VPS34 as a potentially important target, accelerating drug discovery and testing efforts to complement anti-GD2 immunotherapy.

"Right now, what's missing is a clinical-grade VPS34 drug," Wang said. "We're providing the pre-clinical evidence and rationale that, once better VPS34 inhibitors are developed, combining them with anti-GD2 could be an important neuroblastoma treatment strategy."

Other authors from Penn State College of Medicine include Longgui Chen, Todd D. Schell, Giselle Saulnier Sholler, Vladimir Spiegelman, and Yoshinori Takahashi. Xiaoming Liu, who conducted the research as a doctoral student at Penn State and has since graduated, also contributed to the study.

Reference: Zhang, J., et al. (2026). Inhibiting PIK3C3/VPS34 enhances anti-GD2 immunotherapy in neuroblastoma. Autophagy. https://doi.org/10.1080/15548627.2026.2717948