In This Issue

See article, p. 38

The histone acetyltransferase CREBBP is frequently mutated in diffuse large B-cell lymphomas (DLBCL) and follicular lymphomas (FL), which arise from germinal center (GC) B cells. CREBBP-mediated acetylation of histone H3 lysine 27 (H3K27) is required for enhancer activation, but it is unclear how CREBBP mutations contribute to lymphomagenesis. Jiang and colleagues found that, in a mouse model of lymphoma, CREBBP depletion reduced acetylation of H3K27 at enhancers and accelerated lymphomagenesis, indicating a tumor-suppressive role for CREBBP. In mouse and human lymphoma cells, CREBBP depletion resulted in loss of H3K27ac at genes silenced in GC B cells including enhancers involved in B-cell differentiation and antigen presentation such as MHC class II genes, suggesting that CREBBP loss of function might aberrantly maintain the germinal center transcriptional program and prevent immune recognition. Moreover, CREBBP mutations enhanced transcriptional repression of these same genes in patients with FL and DLBCL. Mechanistically, the BCL6/SMRT/HDAC3 repressor complex mediated deacetylation of enhancers that are normally activated by CREBBP. Accordingly, CREBBP loss of function resulted in unopposed repression of these genes by HDAC3, and HDAC3 depletion reduced tumor growth in mice with CREBBP-mutant lymphoma. In addition, HDAC3 inhibitors rescued the repression of MHC class II genes, and the diminished T-cell alloreactivity caused CREBBP loss, further supporting a model whereby BCL6/SMRT/HDAC3 oppose the activity of CREBBP, resulting in increased transcriptional repression when CREBBP loss-of-function mutations are present. Altogether, these findings describe a tumor-suppressive role for CREBBP, and suggest HDAC3 as a potential therapeutic target in CREBBP-mutant tumors.

See article, p. 54

Patients with prostate cancer who receive androgen deprivation therapies (ADT), such as the androgen receptor (AR) inhibitor enzalutamide, often develop castration-resistant prostate cancer (CRPC). However, resistance to such therapies can rapidly develop and trigger differentiation into neuroendocrine prostate cancer (NEPC), a highly aggressive subtype of CPRC characterized by loss of AR signalling and expression of neuroendocrine lineage markers. To elucidate the mechanism underlying neuroendocrine differentiation of CRPC, Bishop and colleagues developed a preclinical mouse model of enzalutamide-resistant CRPC that exhibits heterogeneous mechanisms of resistance to enzalutamide. Expression of the neural transcription factor (TF) POU class 3 homeobox 2 (BRN2, encoded by POU3F2) was upregulated in both neuroendocrine-like enzalutamide-resistant CRPC and human NEPC and was inversely correlated with AR expression and activity. Binding of AR to the androgen response element in the BRN2 promoter ablated BRN2 expression and BRN2-mediated transcription of SRY-box 2 (SOX2), a TF that both is a target of and interacts with BRN2 to drive transcription of neuroendocrine genes. Consistent with these findings, BRN2 ablation caused a decrease in the expression of neuroendocrine markers and growth of enzalutamide-resistant CRPC in vitro and in vivo. Taken together, these findings identify BRN2 as a driver of neuroendocrine differentiation in ADT-resistant CRPC and suggest that BRN2 is a potential therapeutic target for patients with NEPC.

See article, p. 72

Patients with estrogen receptor (ER)–positive tumors are often treated with tamoxifen, which has mixed antagonistic and agonistic effects on ERs depending on the cell type. ERs are also expressed on most immune cells, but the effects of tumor cell–independent estrogen signaling on antitumor immune responses has not been determined. Svoronos and colleagues found that in a subset of serous ovarian carcinomas ERα expression was detected in the tumor-infiltrating leukocytes independent of tumor cell ERα expression. Further, in a mouse model of aggressive ovarian cancer that recapitulates the inflammatory microenvironment of human tumors, ERα was detected in the tumor-associated myeloid cells, but not the tumor cells. Although estradiol supplementation accelerated tumor progression in these mice, isolated tumor cells did not respond to estradiol treatment in vitro, suggesting that ERα expressed on hematopoietic cells in the tumor microenvironment can accelerate progression independent of tumor cell ERα. ERα signaling in hematopoietic cells increased the mobilization and activity of myeloid-derived suppressor cells (MDSC), enhancing their immunosuppressive activity and accelerating tumor progression. Mechanistically, ERα signaling promoted activation of the JAK2 and SRC kinases, resulting in increased STAT3 phosphorylation that drives MDSC expansion. Although estradiol signaling also suppressed T cell–dependent antitumor immunity, it was not sufficient to drive progression, and the estradiol-mediated acceleration of tumor progression is likely mediated by enhanced activity of immunosuppressive MDSCs. The finding that tumor cell–independent estradiol signaling promotes immunosuppression suggests that estrogen antagonists may enhance immunotherapy even in ER-negative tumors.

See article, p. 86

Lung squamous cell carcinomas (LSCC) are characterized by the homozygous loss of TP53 in over 80% of tumors and mutations in the Kelch-like ECH associated protein 1 (KEAP1)/nuclear factor 2 erythroid like 2 (NF2EL2, also known as NRF2) antioxidant response pathway in over 33% of tumors. To further elucidate the role of the KEAP1/NRF2 pathway in LSCC and identify the LSCC cell of origin, Jeong and colleagues developed mouse models of LSCC driven by Trp53 loss with or without Keap1 loss. Although deletion of Trp53 and codeletion of Trp53 and Keap1 in peripheral lung cells or in tracheal epithelial cells gave rise, respectively, to lung adenocarcinomas (LUAD) or LSCCs, Trp53^−/−;Keap1^−/− LSCCs were more aggressive than Trp53^−/− LSCCs. Codeletion of Trp53 and Keap1 resulted in the development of LSCCs from airway basal stem cells (ABSC), but not luminal tracheal cells, and promoted ABSC self-renewal in vitro and ABSC expansion in vivo. Further, compared with Trp53^−/− LSCCs, Trp53^−/−;Keap1^−/− LSCCs showed enhanced proliferation, resistance to radiation and oxidative stress, and NRF2 signaling, and displayed decreased intracellular reactive oxygen species. Moreover, the mutation status of KEAP1 in LSCCs and LUADs could be detected in circulating tumor DNA and significantly predicted patient response to radiotherapy. These findings demonstrate how the establishment of genetic mouse models of LSCC enabled the identification of the LSCC cell of origin and the characterization of the role of the KEAP1/NRF2 pathway in LSCC.

See article, p. 102

AKT signaling is frequently activated in human tumors as a result of multiple mechanisms including mutations in AKT1, AKT2, AKT3, or PIK3CA, or loss of PTEN, suggesting the potential for therapeutic targeting of AKT. However, allosteric AKT inhibitors that bind the inactive state cause serious toxicities in patients. ATP-competitive inhibitors, such as ipatasertib, inhibit the phosphorylated active conformation of AKT, and might therefore spare normal cells and exhibit reduced toxicity. Saura, Roda, and colleagues evaluated ipatasertib in a phase I dose-escalation study of 52 patients with advanced solid tumors. The primary objectives were to evaluate the safety and tolerability of ipatasertib and determine the maximum tolerated dose (MTD). Ipatasertib was generally well tolerated, and most adverse events were grade 1 to 2. As expected, ipatasertib treatment resulted in a dose-dependent downregulation of multiple AKT target genes, indicating on-target effects. At the determined MTD of ipatasertib, 11 of 25 patients (44%) achieved stable disease or an incomplete response, including patients with colon, ovarian, prostate, breast, and lung cancers, and chondrosarcoma. Stable disease was observed in 6 of 9 patients (67%) who exhibited PTEN loss or mutations in PIK3CA or AKT, whereas only 10 of 32 patients (31%) without alterations achieved stable disease. Further, stable disease was not seen in patients with known KRAS mutations. Collectively, the results of this phase I study suggest that AKT can be effectively targeted with ATP-competitive inhibitors without significant toxicity, and support further clinical investigation of ipatasertib.

Note: In This Issue is written by Cancer Discovery editorial staff. Readers are encouraged to consult the original articles for full details.