In This Issue

See article, p. 36.

Rearrangements involving the neurotrophic tyrosine kinase receptor type 1 (NTRK1), NTRK2, and NTRK3 genes, which encode the TRK family of tyrosine kinase receptors, have been implicated in the progression of various cancers, including colorectal cancer. TRK kinase inhibitors such as the pan-TRK inhibitor entrectinib are currently being tested in phase I clinical trials in patients with solid tumors; however, potential acquired resistance mechanisms that may limit the efficacy of entrectinib have not yet been identified. To address this question, Russo, Misale, Wei, and colleagues analyzed circulating tumor DNA (ctDNA) from a patient with metastatic colorectal cancer harboring an LMNA–NTRK1 fusion who experienced a partial response to entrectinib treatment followed by the onset of resistance. Longitudinal analysis of ctDNA isolated before treatment and at relapse revealed the emergence of two NTRK1 point mutations during entrectinib treatment, resulting in the G595R and G667C substitutions in the TRKA kinase domain. The same mutations in NTRK1 were detected following development of entrectinib resistance in a xenopatient and in colorectal cancer cell line models harboring NTRK1 rearrangements, and were sufficient to drive acquired resistance to TRK inhibitors. Structural analysis revealed that the G595R mutation resulted in complete abrogation of entrectinib binding to the TRKA catalytic domain and that the G667C mutation reduced entrectinib binding affinity. These results identify NTRK1 mutation as a mechanism of acquired resistance to entrectinib in colorectal cancer and support development of next-generation TRKA inhibitors.

See article, p. 45.

Many recurrent amplifications in cancer are not associated with known oncogenes, suggesting that additional oncogenes remain to be discovered. To identify previously uncharacterized oncogenic drivers in human cancer, Neupane and colleagues performed an unbiased genome-scale screen using a model system in which primary epithelial cells are transformed by the combination of the SV40 early region, hTERT, and active RAS, and substituted an expression library in place of RAS to screen for library members that could transform cells. Of the genes that were sufficient to promote anchorage-independent growth, MECP2 was selected due to its frequent amplification in various tumor types in The Cancer Genome Atlas. MECP2 was shown to promote tumor formation in xenograft models. Mechanistically, the transforming potential of MECP2 required its DNA-binding ability, and specifically binding to the activating epigenetic mark 5-hydroxymethylcytosine, suggesting that MECP2 functions through an epigenetic mechanism to activate gene expression. Across multiple tumor types, tumors tended to have either MECP2 amplification or activating RAS alterations, but not both. Consistent with these findings, MECP2 activated the MAPK and PI3K pathways, which are both induced by RAS, and rescued the growth of mutant KRAS–addicted cells following KRAS downregulation. Additionally, tumors with MECP2 overexpression exhibited MECP2 oncogene addiction, and the growth defect induced by MECP2 loss could be rescued by active RAS expression. These findings indicate that MECP2 is an important driving oncogene that functions through an epigenetic mechanism, suggesting the potential for epigenetic inhibitors to treat MECP2-amplified tumors.

See article, p. 59.

MYC expression or activity is elevated in more than 50% of cancers, but treatment options are limited in these tumors. MYC-driven B-cell lymphomas are initially sensitive to PI3K/AKT/mTOR inhibitors, but eventually relapse occurs, suggesting that targeting multiple steps of the ribosome signaling network may improve therapeutic outcomes and prevent acquired resistance to therapy. To test this hypothesis, Devlin and colleagues evaluated the efficacy of the mTOR complex 1 (mTORC1) inhibitor everolimus and the inhibitor of RNA polymerase I transcription, CX-5461, in Eμ-Myc B-lymphoma cells. CX-5461 induced apoptosis through activation of nucleolar stress and p53-mediated apoptosis, whereas although PI3K/AKT/mTORC1 inhibition suppressed ribosome biogenesis and function, it induced apoptosis independent of p53 pathway activation. In contrast to CX-5461, PI3K/AKT/mTOR inhibition resulted in upregulation of the pro-apoptotic BH3-only protein BCL2-modifying factor (BMF). Loss of BMF did not affect the sensitivity of B-lymphoma cells to CX-5461, but did reduce AKT inhibitor–induced apoptosis, indicating that PI3K/AKT/mTORC1 inhibitors and CX-5461 promote apoptosis via distinct mechanisms. Consistent with these findings, combined treatment with everolimus and CX-5461 enhanced apoptosis, more effectively suppressed rDNA gene transcription, and resulted in prolonged survival of tumor-bearing mice compared with treatment with either agent alone. Taken together, these results demonstrate that PI3K/AKT/mTOR pathway inhibition and CX-5461 cooperatively inhibit rRNA synthesis and induce apoptosis through independent mechanisms requiring functional BMF and p53, respectively, and suggest that combination treatments that target rDNA transcription and mRNA translation at multiple points may enhance therapeutic benefit in MYC-driven cancers.

See article, p. 71.

The antitumor response of CTLs is generally weak, but can be promoted by immunotherapy with immunostimulatory mAbs targeting programmed cell death 1 (PD-1) or CD137. The activation of CTLs requires cross-presentation of tumor antigens by dendritic cells (DC), which is impaired by loss of the basic leucine zipper transcription factor, ATF-like 3 (BATF3). Sánchez-Paulete and colleagues used a Batf3^−/−mouse model to demonstrate that BATF3-dependent DCs are required to cross-prime CTLs and upregulate PD-1 and CD137 on CTLs, making these lymphocytes sensitive to immunotherapeutic stimulation. Treatment with anti–PD-1 or anti-CD137 mAbs reduced tumor growth in wild-type mice, but not Batf3^−/− mice, indicating that BATF3-dependent DCs are necessary for the baseline immune response that is enhanced by immunotherapy with mAbs. Coculture of T cells with DCs from Batf3^−/− mice impaired T-cell proliferation and production of IFNγ. In vivo, the anti-CD137 mAb–associated increase in tumor antigen–specific CD8⁺ T cells was abrogated by Batf3 loss, further supporting an essential role for BATF3-dependent DCs in cross-priming CTLs. Administration of soluble FLT3 ligand (sFLT3L) and poly-ICLC to enhance BATF3-dependent DC expansion and activation increased the surface expression of PD-1 and CD137 on antigen-specific CD8⁺ tumor-infiltrating lymphocytes and promoted the antitumor effects of anti-CD137 and anti-PD-1 mAbs in vivo. These results suggest that cross-priming BATF3-dependent DCs act synergistically with anti-CD137 and anti–PD-1 immunotherapy and that augmenting DC-mediated cross-priming may enhance the efficacy of tumor immunotherapy.

See article, p. 80.

Cross-talk between tumor and stromal cells has been implicated in cancer progression. For example, activated myeloid-derived suppressor cells (MDSC) have been shown to suppress T-cell function and are enriched in the prostate tumor microenvironment. To better understand the mechanisms by which MDSCs are recruited and contribute to prostate carcinogenesis, Wang, Lu, and colleagues used a Pten/Smad4-deficient mouse model that rapidly develops metastatic prostate cancer. These tumors showed extensive infiltration of CD11b⁺Gr1⁺ polymorphonuclear MDSCs, which suppressed T-cell proliferation and activation. Immunodepletion of MDSCs using an anti-Gr1 neutralizing antibody or MDSC-specific peptide-Fc fusion protein inhibited prostate tumor growth and extended the overall survival of tumor-bearing mice. Mechanistically, transcriptional profiling revealed a significant upregulation of the MDSC-recruiting cytokine chemokine (C-X-C motif) ligand 5 (CXCL5) in Pten/Smad4-deleted prostate tumor cells and a concurrent increase in the expression of its cognate receptor, CXCR2, in tumor-associated stromal cells. Pharmacologic inhibition of CXCL5–CXCR2 signaling impaired MDSC migration and limited tumor progression in Pten/Smad4-deleted mice. Furthermore, induction of Cxcl5 was dependent on YAP1/TEAD–driven transcription, and suppression of YAP1 inhibited MDSC infiltration and prostate tumor progression in vivo. In line with these findings, human prostate tumor samples with an elevated MDSC-related gene signature were characterized by increased expression of YAP1 signature genes. Together, these data suggest that hyperactivation of YAP1 in prostate cancer cells drives disease progression by promoting MDSC infiltration via the CXCL5–CXCR2 axis and suggest potential therapeutic strategies for advanced prostate cancer.

See article, p. 96.

Anaplastic lymphoma kinase (ALK) is activated by point mutations or amplification in approximately 8% of neuroblastomas. A recent phase I trial of the ATP-competitive ALK inhibitor crizotinib in pediatric patients with ALK-driven malignancies showed that crizotinib exhibited greater efficacy against ALK-rearranged tumors than ALK-mutated neuroblastomas, highlighting the need to identify more potent inhibitors of full-length ALK mutants. Infarinato, Park, and colleagues compared the potency and efficacy of PF-06463922, a next-generation selective ALK/ROS1 inhibitor, with that of crizotinib in neuroblastoma cell lines and xenografts harboring mutations in the ALK tyrosine kinase domain (TKD) that are either resistant or sensitive to crizotinib. In vitro, PF-06463922 was more broadly potent than crizotinib against clinically relevant full-length ALK mutants and exhibited greater affinity than crizotinib for the mutated ALK TKD at lower concentrations. Single-agent treatment with PF-06463922 induced rapid and sustained tumor regression of both patient-derived and cell line–derived xenografts harboring hotspot ALK mutants, including both crizotinib-sensitive and crizotinib-resistant ALK mutants, whereas crizotinib exhibited limited inhibition of crizotinib-sensitive ALK-mutant xenograft growth and had no effect on the growth of crizotinib-resistant ALK-mutant xenografts. Mechanistically, PF-06463922 inhibited ALK phosphorylation and mutated ALK variant–mediated transformation more effectively than crizotinib. In summary, these findings suggest that the potent antitumor activity of PF-06463922 against known clinically acquired ALK mutations makes it a lead candidate for rapid clinical development for the treatment of pediatric patients with ALK-driven neuroblastoma.

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