Paper Review — Verify any paper quickly

Paper summary and immediate strengths

The preprint reports that CD155 (PVR) is highly expressed in diffuse midline glioma (DMG) tumor cells and that CD155 has two separable roles: (1) a cell autonomous role supporting proliferation and survival via control of oncogenic programs including FOXM1 and cell cycle pathways, and (2) an immune evasion role where CD155 expression reduces sensitivity to CD8+ T cell mediated killing. Loss of CD155 (shRNA) increases CD8 T cell killing in vitro, causes failure of CD155-deficient KAPP tumors to engraft in immunocompetent mice (rescue by CD8 depletion), and reduces growth even in immunodeficient mice — indicating intrinsic viability effects. Transcriptomics + VIPER highlight FOXM1 downregulation as a major downstream effect; FOXM1 knockdown phenocopies reduced proliferation and Thiostrepton (a FOXM1 inhibitor) reduces DMG viability in vitro and delays subcutaneous tumor growth, but fails to prolong survival in orthotopic intracranial models likely because of BBB penetration issues (Main experimental findings)

Why this matters

• DMG is a pediatric brain tumor with poor outcomes and limited immunotherapy success; identifying a tumor cell surface molecule that both sustains tumor viability and impacts immune sensitivity is potentially high‑value for therapeutic targeting (Clinical motivation and expression evidence

Detailed critique: experimental design and key data

1) Expression and translational relevance

The authors show CD155 is highly expressed across multiple human DMG single cell RNAseq samples and cell lines and in murine models (A-367, KAPP, PKC). They also document low PD-L1/PD-L2 but high B7-H3 expression in DMG, supporting that CD155 could be a more relevant immune checkpoint in this tumor type (Expression profiling in DMG)

2) Immune assays and in vivo immunocompetent models

Key strengths: co-culture with OT-I CD8+ T cells using OVA expressing DMG cells reveals increased T cell killing upon CD155 knockdown at multiple E:T ratios; adoptive transfer and CD8 depletion studies in vivo causally link CD8+ T cells to rejection of CD155-deficient tumors (T cell functional and in vivo depletion data)

Notably, T cells exposed to CD155-deficient targets did not show increased canonical activation markers or IFNγ production, suggesting the mechanism is not simple inhibition of T cell activation but may involve tumor cell susceptibility pathways or modulation of cytotoxic synapse efficiency (T cell activation readouts)

3) Tumor cell intrinsic effects and molecular mechanism

Silencing CD155 reduces proliferation and increases apoptosis in multiple murine and human DMG cell lines in vitro and reduces tumor burden in immunodeficient mice — evidence consistent with a tumor‑intrinsic survival role (Cell intrinsic viability data)

Transcriptomics (DESeq2) and pathway analyses (IPA) show downregulation of cell cycle and oncogenic pathways (eg kinetochore metaphase, Ephrin signaling) and upregulation of interferon pathways; VIPER with a DMG regulatory network flags FOXM1 as a top inactivated master regulator after CD155 knockdown, which was validated by qRT-PCR and protein assays for FoxM1 and some of its targets (Egfr, Itga4, Rock2) (Transcriptomic and VIPER analyses)

4) FOXM1 pharmacologic targeting

Thiostrepton reduces FOXM1 protein and DMG viability in vitro with low micromolar IC50 and delays growth of subcutaneous DMG tumors with survival benefit in NSG mice. However, systemic Thio failed to prolong survival in orthotopic intracranial models — authors attribute this to blood brain barrier (BBB) penetration limitations and suggest brain delivery strategies or alternative BBB‑permeable FOXM1 inhibitors are required (Thiostrepton efficacy and BBB limitation)

Limitations, confounders, and open questions

1. shRNA specificity and knockdown artifacts — most key phenotypes rely on shRNA knockdown. Although multiple shRNAs were used, full CRISPR knockout, rescue (re‑expression) experiments, or orthogonal perturbations strengthen causality and help rule out off-target or interferon‑related responses to RNAi (Use of shRNA and limitations)
2. Mechanistic gap between CD155 and FOXM1 — the authors show correlated transcriptional changes and VIPER-inferred FOXM1 inactivation after CD155 loss but the direct signaling pathway linking a membrane adhesion/checkpoint receptor to a transcription factor program is not fully delineated (eg which kinases/adaptors transduce CD155 loss to FOXM1 downregulation?) (Mechanistic gap CD155 to FOXM1)
3. Tumor microenvironment complexity — most immune experiments use OT-I/OVA model antigen and adoptive OT-I cells; while clean for mechanism, the OVA model does not represent endogenous tumor antigen repertoires or human TCR diversity. The role of innate cells (NK) and TIGIT/DNAM1/CD96 interactions deserves deeper exploration, especially since CD155 has multiple immune receptors and prior studies emphasize CD155-TIGIT in NK evasion (Immune receptor complexity and OVA model)
4. Delivery for therapeutic translation — Thiostrepton is large and not BBB‑permeable; the promising subcutaneous data do not prove efficacy in orthotopic brain tumors (Therapeutic delivery limitation)
5. Species/model differences and clinical heterogeneity — murine DMG lines carry engineered driver lesions and are p53-deficient; human DMG heterogeneity and immune microenvironment differ from murine models, so translational predictiveness is uncertain (Model limitations and heterogeneity)

Recommendations to strengthen and extend the study

1. Perform CRISPR knockout and re‑expression rescue of CD155 (wild type and point mutants) to verify specificity and map the functional domains required for cell intrinsic and immune effects (Need for orthogonal genetic validation)
2. Map signaling between CD155 and FOXM1 via phospho-proteomics and small‑molecule kinase inhibitor screens to identify intermediates (eg Src family kinases, MAPK, PI3K/AKT) and test whether exogenous activation of those pathways rescues FOXM1 after CD155 loss.
3. Test whether blocking CD155 interactions with TIGIT vs DNAM1 vs CD96 recapitulates phenotypes via receptor‑specific antibodies or receptor knockout on immune cells; examine NK contributions using NK‑competent models given CD155 biology (CD155 immune receptor interactions)
4. Prioritize development of BBB permeant FOXM1 inhibitors, CD155 targeted modalities amenable to CNS (eg bispecifics, BBB-shuttle conjugates, ASOs against PVR) or adoptive cell strategies (CAR-T targeting CD155) while carefully interrogating expression in normal brain to avoid on‑target toxicity.

Bottom line

Overall the study provides strong preclinical evidence that CD155 has both tumor‑intrinsic and immune‑modulatory roles in DMG and identifies FOXM1 as a key downstream master regulator. The work is convincing at the level of phenotype, transcriptomics, and VIPER inference, but key mechanistic links (membrane receptor to FOXM1 regulation), orthogonal genetic validation, and brain‑delivery demonstration are needed before translational claims can be advanced. The preclinical therapeutic concept (target CD155 and/or FOXM1) is attractive but requires brain‑capable modalities and deeper immune receptor dissection for safe clinical development (Conclusion and translational caution)

Actionable next experiments (concise)

1. CRISPR KO plus CD155 re-expression (WT and domain mutants) with proliferation, apoptosis and T cell killing assays.
2. Phospho‑proteomic time course after CD155 perturbation to identify signaling nodes linking CD155 to FOXM1 regulation.
3. Receptor axis dissection: block TIGIT CD226 CD96 separately in vitro and in vivo; include NK depletion studies.
4. Test BBB‑permeable FOXM1 inhibitors or local intracerebroventricular/ convection enhanced delivery of Thiostrepton or ASO against FOXM1 and CD155 in orthotopic DMG models.

Reference (primary source)

Primary preprint: CD155 in DMG

Author Review buttons