In This Issue

See article, p. 247.

Pancreatic ductal adenocarcinoma (PDAC) is characterized by activating mutations in the KRAS oncogene and a proinflammatory tumor microenvironment that is predominantly immunosuppressive. B cells are recruited to pancreatic parenchyma during the early stages of pancreatic intraepithelial neoplasia (PanIN). To determine whether B-cell recruitment contributes to oncogenic progression, Pylayeva-Gupta and colleagues confirmed the presence of B cells in human PanIN lesions and mutant KRAS–driven neoplastic pancreatic lesions in mice, and showed that stromal fibroblasts were positive for the B-cell chemoattractant CXCL13. Neoplastic progression of KRAS-mutant pancreatic cancer cells was inhibited in mice lacking B cells, whereas tumor formation was rescued upon reintroduction of the CD19⁺CD1d^hiCD5⁺ B cell subpopulation, reinforcing the requirement for B cells in PDAC tumorigenesis. Mechanistically, the protumorigenic effect of this subset of B cells in PanIN was shown to be mediated by the production of IL35, but not IL10. Expression of Il12a and Ebi3, which encode the IL35 subunits p35 and EBI3, respectively, was elevated in CD1d^hiCD5⁺ B cells, and the production of IL35 by B cells promoted the growth of KRAS-mutant pancreatic cancer cells in vitro. In line with this finding, PDAC growth was inhibited in Il12a-deficient mice. Together, these findings highlight an essential role of CD1d^hiCD5⁺ B cells in pancreatic oncogenesis and suggest that therapeutic targeting of the B cell/IL35 axis may limit PanIN progression.

See article, p. 256.

Pancreatic ductal adenocarcinoma (PDAC) is characterized by frequent oncogenic mutations in KRAS and an inflammatory, hypoxic tumor microenvironment. To counteract conditions of low oxygen availability, cells activate adaptive transcriptional responses largely mediated by the hypoxia-inducible factor 1α (HIF1α). However, the role of HIF1α in PDAC remains unclear. To study the functional role of HIF1α in PDAC, Lee and colleagues used a mutant Kras–driven PDAC mouse model and found that HIF1α and hypoxia were increased in early pancreatic intraepithelial neoplasia (PanIN) and remained high throughout disease progression. Surprisingly, deletion of Hif1a in KRAS-mutant pancreatic epithelial tissue increased PanIN number and grade concomitant with enhanced cell proliferation, suggesting that HIF1α inhibits pancreatic tumor initiation and progression. Hif1a-deficient murine PanINs and human pancreatic cancers were characterized by increased B-cell infiltration, and a shift in the B-cell population toward a rare B1b subtype was observed in Kras-mutant mouse pancreata, with a further shift observed upon Hif1a deletion. B-cell depletion led to decreased progression of late-stage PanINs in Kras-mutant mice, irrespective of HIF1α status, and reduced microinvasive lesions in Hif1a-deficient animals. Mechanistically, levels of the B-cell chemoattractant CXCL13 and other B-cell migratory chemokines were increased in Kras-mutant;Hif1a-deficient mice. Together, these data point to a previously underappreciated protective role of HIF1α in PDAC initiation and highlight B cell intrapancreatic recruitment as a driver of PDAC progression.

See article, p. 270.

Lack of early detection methods and ineffective therapies contribute to poor survival rates in patients with pancreatic ductal adenocarcinoma (PDAC). PDAC tumors are characterized by infiltrating leukocytes, which have been implicated in neoplastic progression. To identify tractable immune targets that contribute to PDAC, Gunderson, Kaneda, and colleagues showed that human PDAC samples displayed high levels of the B-cell marker CD20 and increased infiltration of CD45⁺ leukocytes and FcRγ-positive myeloid cell populations compared to normal pancreatic tissue. Analysis of a mutant KRAS–driven mouse model of PDAC demonstrated similar immune cell infiltration and revealed that tumor growth was dependent on B cells and FcRγ-positive myeloid cells. In line with this finding, activation of Bruton tyrosine kinase (BTK) signaling was detected in B cells and myeloid cells from human and murine PDAC tumors, and high levels of the BTK-activating PI3Kγ isoform were observed in myeloid cells. Pharmacologic inhibition of BTK or PI3Kγ suppressed PI3Kγ-dependent integrin activation and macrophage adhesion and impaired protumorigenic T_H2 macrophage polarization. Functionally, co-culture of PDAC-derived B cells with primary macrophages revealed that B cells were responsible for driving BTK-dependent macrophage polarization. Treatment of PDAC-bearing mice with BTK or PI3Kγ inhibitors reduced myeloid cell infiltration and stimulated CD8⁺ T-cell recruitment, resulting in inhibition of both early- and late-stage tumor formation, and enhanced response to gemcitabine chemotherapy. Together, these data suggest that PI3Kγ/BTK signaling drives protumorigenic, immunosuppressive myeloid cell recruitment and highlight BTK as a potential immunotherapeutic target in PDAC.

See article, p. 286.

Targeted therapies and immunotherapies for malignant melanoma have greatly improved patient care and survival; however, the lack of biomarkers makes it difficult to determine which patients will benefit from these treatments and to select second-line therapies. Girotti, Gremel, and colleagues used next-generation sequencing (NGS) and xenografts to develop a personalized medicine platform utilizing 364 tumor samples from 214 patients with melanoma. Analysis of circulating tumor DNA (ctDNA) from patients with melanoma revealed that mutant BRAF and NRAS expression predicted clinical response to therapies. NGS of ctDNA also predicted relapse due to common mechanisms of resistance several weeks before it was detectable by CT scanning. To complement this approach and guide clinical care, patient-derived xenografts (PDX) from patients with melanoma were used to test hypothesis-driven personalized therapies identified by whole exome sequencing and to validate combination therapies that may result in clinical benefit, such as dual treatment with MEK inhibitor and paclitaxel in a non-BRAF^V600E and RB-mutant tumor. As an alternative in cases where late-stage tumors are not accessible and PDXs cannot be generated, xenografts were derived from circulating tumor cells (CDX) in 13% of cases. These CDX models recapitulated the genetic and histologic features of patient tumors and could predict response to therapy. Taken together, these findings demonstrate the efficacy of a personalized medicine approach whereby NGS of ctDNA can identify mutations for rational drug targeting, which can be tested using PDX or CDX models to optimize treatment for patients with melanoma.

See article, p. 300.

BRAF mutations, the most prominent of which is BRAF^V600E, activate MAPK signaling to drive the growth of numerous tumor types. However, drugs that target BRAF^V600E, which functions as a monomer, paradoxically activate MAPK signaling by inducing RAF dimerization in BRAF wild-type cells and are inactive against RAF dimers. LY3009120 is a pan-RAF and RAF dimer inhibitor currently in clinical trials. In this study, Chen and colleagues treated a panel of tumor cell lines harboring different genetic mutations. Some tumor cell lines which exhibited sensitivity to LY3009120 were found to harbor novel BRAF in-frame deletions within or adjacent to the residues from leucine 485 to proline 490 (L485-P490) in the αC-helix region. Consistent with these findings, examination of publicly available sequencing data revealed that patients with pancreatic or thyroid cancers harbored similar BRAF in-frame deletions that were mutually exclusive of KRAS or BRAF^V600 mutations. The in-frame BRAF deletion variants functioned as homodimers and promoted the oncogenic transformation of NIH3T3 cells. Mechanistically, the previously unidentified BRAF in-frame deletions resulted in conformational changes that favored the formation of homodimers that activated MAPK signaling, which was abrogated by LY3009120 but was resistant to vemurafenib. Similarly, tumor cells harboring BRAF in-frame deletions were inhibited by LY3009120 but not by vemurafenib in vitro and in vivo. Together, these results identify the activation mechanism of BRAF in-frame deletions, which may potentially act as a common activation mechanism of other kinases.

See article, p. 316.

Gain-of-function mutations in the KRAS oncogene are implicated in 30% of all human cancers, but no targeted therapies have been successfully developed for cancers with KRAS mutations. Building on the recent discovery of an allosteric pocket under the Switch II loop region exposed in the GDP-bound, inactive state of KRAS, Patricelli, Janes, and colleagues identified a selective, covalent inhibitor of mutant KRAS^G12C activation. Structural evaluation revealed that the small molecule ARS-853 reacted specifically with GDP-bound KRAS^G12C and, once bound, prevented formation of the GTP-bound, active state. In KRAS^G12C-mutant cells, binding of ARS-853 blocked the interaction between the RAF–RAS binding domain and KRAS, and inhibited KRAS downstream signaling via the MAPK and PI3K pathways. Binding of ARS-853 also drove apoptosis and inhibited cell growth in culture. In contrast to the classic view of mutant KRAS as a statically active enzyme, it was discovered that KRAS^G12C rapidly cycles its nucleotide state, allowing ARS-853 to achieve complete binding and functional inhibition. Furthermore, similar to wild-type RAS, the activity of KRAS^G12C was found to be modulated by upstream signaling factors, suggesting that combined inhibition of KRAS and its upstream activators might provide enhanced benefit over direct KRAS^G12C inhibition alone. Taken together, these findings lay the groundwork for future efforts to optimize ARS-853 into a compound suitable for in vivo studies and offer promise for the development of therapeutic strategies for cancers harboring the KRAS^G12C mutation.

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