In This Issue

See article, p. 2158.

The main side effect of EGFR inhibitors is induced by inhibition of the MAPK pathway in the skin, while BRAF inhibitors paradoxically activate MAPK downstream of EGFR in BRAF wild-type cells. Lacouture and colleagues conducted a phase I clinical trial administering the topical BRAF inhibitor LUT014 to improve skin toxicities induced by EGFR inhibitors. LUT014 gel was well tolerated in 10 patients with colorectal cancer with acneiform rash induced by cetuximab or panitumumab. The rash improved in the six patients who started with grade 2 rash. Therefore, topical LUT014 is safe and efficacious in improving rash from EGFR inhibitors.

See article, p. 2168.

Microsatellite instability and elevated tumor mutational burden (TMB) are observed in approximately 20% of gastric cancers and are good predictors for clinical response to anti–PD-1 antibodies. However, only 50% of microsatellite instability–high (MSI-H) gastric cancers are responsive to PD-1 therapies, and the underlying mechanisms are largely unknown. Kwon, An, Klempner, Lee, Kim, and colleagues conducted a phase II trial of pembrolizumab in patients with advanced MSI-H gastric cancer and found through analysis of serial and multiregion tissue samples and serial peripheral blood analyses that TMB, diverse T-cell receptor repertoire, TCR clonal diversity, expansion of circulating CD8⁺ lymphocytes, and less exhausted T cells were associated with favorable responses to anti–PD-1 treatment. On-treatment tumor characterization may help distinguish patients with MSI-H gastric cancer who would benefit from anti–PD-1 monotherapy versus those who would benefit from combination therapies.

See article, p. 2186.

The adoptive transfer of chimeric antigen receptor (CAR) T cells represents a breakthrough in clinical oncology, yet both between- and within-patient differences in autologously derived T cells are major contributors to therapy failure. Chen, Chen, Das, and colleagues performed RNA sequencing on sorted T-cell subsets from 71 patients with B-cell malignancies on trial treated with anti-CD19 CAR T-cell therapy, followed by paired CITE-Seq and single-cell ATAC-Seq on T cells from 6 of these patients, identifying transcriptomic and gene regulatory factors associated with long- and short-term CAR T-cell persistence. These findings provide key insights into the underlying molecular determinants of clinical CAR T-cell function.

See article, p. 2200.

Arabzade and colleagues developed an in utero electroporation mouse model of ZFTA–RELA-driven ependymoma, mirroring the most common genetic lesion occurring in children with cortical ependymoma. ZFTA–RELA was shown to directly engage DNA elements within active chromatin, thus leading to increased oncogenic gene expression. Addition of the ZFTA protein fragment conferred de novo binding of the ZFTA–RELA fusion protein to transcription factor motifs shared with the PLAGL family, which play important roles in brain development and gene imprinting. Finally, ZFTA–RELA invades a network of core transcriptional circuitry transcription factors to regulate a substantial portion of the ependymoma tumor transcriptome.

See article, p. 2216.

Ependymomas are hard-to-treat brain tumors driven by translocations between ZFTA (also known as C11orf95) and a variety of transcriptional coactivators including RELA and YAP1. The function of ZFTA is not known. Kupp and colleagues interrogated the function of deletion-mutant ZFTA fusion genes using transcriptomic, chromatin immunoprecipitation sequencing, proteomic and mouse modeling approaches. They reveal the widespread chromatin binding properties of ZFTA, which tethers ZFTA-containing fusions across the genome, modifying chromatin to an active state and enabling its partner transcriptional coactivators to promote promiscuous gene expression. These studies unmask the molecular basis of ependymoma tumorigenesis and vulnerabilities for therapeutic targeting.

See article, p. 2230.

The majority of supratentorial ependymomas (ST-EPN) are driven by fusion genes involving RELA and ZFTA, previously designated C11orf95. Zheng, Ghasemi, Okonechnikov, and colleagues identified ZFTA as a major effector in ST-EPN that mediates oncogenicity when fused to alternative transcriptional activators. Generation of ZFTA fusion-driven mouse models followed by cross-species comparison uncovered potential therapeutic vulnerabilities in fusion-induced tumors. These tumors are now classified as ST-EPN, ZFTA fusion–positive in the upcoming 5th edition of the WHO Classification of Tumors of the Central Nervous System.

See article, p. 2248.

Chimeric antigen receptor (CAR) T-cell therapy for solid tumors faces many challenges, including the suppressive tumor microenvironment (TME). While the direct tumor-targeting activity of CAR T cells has been well elucidated, the interplay with endogenous immune cells is not well understood. Alizadeh and colleagues demonstrate that glioma-targeted CAR T cells can activate host intratumoral myeloid and T cells to promote anti-glioma immunity via a mechanism that require CAR T-cell production and immune cell responsiveness to IFNγ. These studies highlight the critical role for IFNγ-dependent remodeling of the TME by CAR T cells for productive therapy against solid tumors.

See article, p. 2266.

Promoter hypermethylation may drive cancer through repressing tumor suppressor genes, but it is difficult to differentiate driver DNA methylation changes from passenger events. Pan and colleagues present MethSig, a novel inference framework for DNA methylation driver discovery to chart the role of aberrant DNA methylation in cancer. MethSig resulted in reproducible inference of putative DNA methylation drivers with increased accuracy across multiple cancer types compared with benchmarked methods. As a proof-of-concept, knockout of MethSig-identified chronic lymphocytic leukemia DNA methylation drivers provided fitness advantages with and without therapeutic intervention, and DNA methylation driver risk score was closely associated with adverse outcomes in independent CLL cohorts. MethSig may represent a broadly useful tool to interrogate the role of aberrant DNA methylation in cancer.

See article, p. 2282.

CRISPR-based cancer dependency maps have identified a multitude of potential disease-selective targets. Malone and colleagues performed a series of in vitro and in vivo functional genetic screens to prioritize candidate genetic dependencies in neuroblastoma and identified the nuclear transport factor NXT1 as a lethal, neuroblastoma-selective target. CRISPR knockouts and an inducible degradation system revealed that loss of NXT1 leads to destabilization of its binding partner NXF1, which is essential for cell survival. However, loss of NXT1 only leads to loss of NXF1 in the setting of low expression of NXT2, a paralog of NXT1, which is common in neuroblastoma, suggesting that selective targeting of essential proteins may be possible by exploiting context-specific synthetic lethal relationships.

See article, p. 2300.

Epigenetic silencing associated with DNA hypermethylation is a pervasive yet poorly understood phenomenon in acute myeloid leukemia (AML). Yang and colleagues used a CRISPR screen to discover that a subset of AML becomes addicted to Fanconi anemia DNA repair proteins, which is caused by the epigenetic silencing of the aldehyde detoxifying enzyme ALDH2. ALDH2 silencing occurs in a recurrent manner in AML, and the function of this enzyme is compensated for by Fanconi anemia proteins. This study demonstrates how the epigenetic reprogramming in cancer leads to molecular vulnerabilities through a synthetic lethality mechanism.

See article, p. 2316.

Loss or reduction of NKX3.1 is a common early event in prostate cancer and is associated with increased cancer risk, particularly among African American men. Papachristodoulou and colleagues show that NKX3.1 suppresses prostate cancer initiation by localizing to mitochondria, where it protects against aberrant oxidative stress and restores mitochondrial function. Polymorphisms of NKX3.1 associated with increased risk of aggressive disease are inactive for protecting mitochondria. Localization of NKX3.1 to mitochondria is associated with favorable disease outcomes in patients. These findings suggest that monitoring expression levels and localization of NKX3.1 in prostate cancer may improve risk assessment for men undergoing active surveillance.

See article, p. 2334.

There is increasing evidence in the clinical and preclinical setting that the RB tumor suppressor prevents tumor development and progression through mechanisms beyond cell cycle control. Mandigo and colleagues utilized isogenic model systems to interrogate RB function across cancer progression using transcriptomic, cistromic, and metabolomic approaches, revealing stage-specific, E2F1-dependent deregulation of glutathione synthesis after RB loss. Elevated glutathione production resulting from RB loss afforded protective advantage against reactive oxygen species induced by cytotoxic therapy. RB-deficient models also showed increased sensitivity to glutathione synthesis inhibition, providing a novel target for therapeutic intervention in advanced disease.