No abstract

Transition between differentiation states in development occurs swift but the mechanisms leading to epigenetic and transcriptional reprogramming are poorly understood. The pediatric cancer neuroblastoma includes adrenergic (ADRN) and mesenchymal (MES) tumor cell types, which differ in phenotype, super-enhancers (SEs) and core regulatory circuitries. These cell types can spontaneously interconvert, but the mechanism remains largely unknown. Here, we unravel how a NOTCH3 intracellular domain reprogrammed the ADRN transcriptional landscape towards a MES state. A transcriptional feed-forward circuitry of NOTCH-family transcription factors amplifies the NOTCH signaling levels, explaining the swift transition between two semi-stable cellular states. This transition induces genome-wide remodeling of the H3K27ac landscape and a switch from ADRN SEs to MES SEs. Once established, the NOTCH feed-forward loop maintains the induced MES state. In vivo reprogramming of ADRN cells shows that MES and ADRN cells are equally oncogenic. Our results elucidate a swift transdifferentiation between two semi-stable epigenetic cellular states.

Neuroblastoma is a pediatric tumor of the peripheral sympathetic nervous system with a highly variable prognosis. Activation of the phosphoinositide 3-kinase (PI3K)/AKT pathway in neuroblastoma is correlated with poor patient prognosis, but the precise downstream effectors mediating this effect have not been determined. Here we identify the forkhead transcription factor FOXO3a as a key target of the PI3K/AKT pathway in neuroblastoma. FOXO3a expression was elevated in low-stage neuroblastoma tumors and normal embryonal neuroblasts, but reduced in late-stage neuroblastoma. Inactivation of FOXO3a by AKT was essential for neuroblastoma cell survival. Treatment of neuroblastoma cells with the dual PI3K/mTOR inhibitor PI-103 activated FOXO3a and triggered apoptosis. This effect was rescued by FOXO3a silencing. Conversely, apoptosis induced by PI-103 or the AKT inhibitor MK-2206 was potentiated by FOXO3a overexpression. Furthermore, levels of total or phosphorylated FOXO3a correlated closely with apoptotic sensitivity to MK-2206. In clinical specimens, there was an inverse relationship between gene expression signatures regulated by PI3K signaling and FOXO3a transcriptional activity. Moreover, high PI3K activity and low FOXO3a activity were each associated with an extremely poor prognosis. Our work indicates that expression of FOXO3a and its targets offer useful prognostic markers as well as biomarkers for PI3K/AKT inhibitor efficacy in neuroblastoma. Cancer Res; 73(7); 2189-98. Ó2013 AACR.

Structural requirements for b1 integrin cytoplasmic domain functions in adhesion, migration and signaling have been studied mainly for ®broblasts in vitro. The relevance for b1-dependent in vivo migration of lymphoid cells has not been assessed. To study this, we transfected b1 mutants into b1-de®cient double knockout (DKO) ESb lymphoma cells, and tested the capacity of the cells to metastasize to liver and spleen. This was compared to a4b1-dependent invasion into cell monolayers in vitro and Mn 2+-induced adhesion to ®bronectin. Deletion of the ®ve C-terminal residues or mutation of both threonines T788 and T789 to alanines blocked invasion and metastasis and greatly reduced adhesion, in line with known in vitro eects. However, mutations of the NPXY motif tyrosines had unexpected consequences. A Y783F mutation had no eect at all, but a Y783,795F double mutation strongly reduced Mn 2+ -induced adhesion, whereas it had limited eects on invasion and metastasis. Furthermore, cells expressing a b1b2 chimeric subunit, which contains phenylalanines in the NPXY/F motifs, adhered poorly but invasion and metastasis was fully restored to the same levels as for cells expressing wildtype b1. We conclude that part of the functions of the b1 cytoplasmic domain that are required for adhesion are not essential for b1-dependent invasion and metastasis.

Cancer therapy frequently fails due to the emergence of resistance. Many tumors include phenotypically immature tumor cells, which have been implicated in therapy resistance. Neuroblastoma cells can adopt a lineage-committed adrenergic (ADRN) or an immature mesenchymal (MES) state. They differ in epigenetic landscape and transcription factors, and MES cells are more resistant to chemotherapy. Here we analyzed the response of MES cells to targeted drugs. Activating anaplastic lymphoma kinase (ALK) mutations are frequently found in neuroblastoma and ALK inhibitors (ALKi) are in clinical trials. ALKi treatment of ADRN neuroblastoma cells with a tumor-driving ALK mutation induced cell death. Conversely, MES cells did not express either mutant or wild-type ALK and were resistant to ALKi, and MES cells formed tumors that progressed under ALKi therapy. In assessing the role of MES cells in relapse development, TRAIL was identified to specifically induce apoptosis in MES cells and to suppress MES tumor growth. Addition of TRAIL to ALKi treatment of neuroblastoma xenografts delayed relapses in a subset of the animals, suggesting a role for MES cells in relapse formation. While ADRN cells resembled normal embryonal neuroblasts, MES cells resembled immature precursor cells, which also lacked ALK expression. Resistance to targeted drugs can therefore be an intrinsic property of immature cancer cells based on their resemblance to developmental precursors. Significance: In neuroblastoma, mesenchymal tumor cells lack expression of the tumor-driving ALK oncogene and are resistant to ALKi, but dual treatment with ALKi and mesenchymal cell–targeting TRAIL delays tumor relapse.

The integrin cytoplasmic domain-associated protein-1 (ICAP-1) binds via its C-terminal PTB (phosphotyrosine-binding) domain to the cytoplasmic tails of beta1 but not other integrins. Using the yeast two-hybrid assay, we found that ICAP-1 binds the ROCK-I kinase, an effector of the RhoA GTPase. By coimmunoprecipitation we show that ICAP-1 and ROCK form complexes in cells and that ICAP-1 contains two binding sites for ROCK. In cells transfected with both ICAP-1 and ROCK, the proteins colocalized at the cell membrane predominantly in lamellipodia and membrane ruffles, but also in retraction fibers. ROCK was not found at these sites when ICAP-1 was not co-transfected, indicating that ICAP-1 translocated ROCK. In lamellipodia ICAP-1 and ROCK colocalized with endogenous beta1 integrins and this colocalization was also observed with the isolated ICAP-1 PTB domain. The plasma membrane localization of ROCK did not depend on beta1 integrin ligation or ROCK kinase activity, and in truncated ROCK proteins it required the presence of the ICAP-1-binding domain. To show that the interaction was direct, we measured fluorescence resonance energy transfer (FRET) between cyan fluorescent protein (CFP) fused to ICAP-1 and yellow fluorescent protein (YFP) fused to ROCK. FRET was observed in lamellipodia in cells that were induced to spread. These results indicate that ICAP-1-mediated binding of ROCK to beta1 integrin serves to localize the ROCK-I kinase to both the leading edge and the trailing edge where ROCK affects cell migration.

Abstract. CD44 splice variants have been shown to be involved in metastasis of carcinomas. In addition, the standard form of CD44 has been implicated in metastasis, particularly of melanomas and lymphomas. To investigate this, we have generated a CD44-negative mutant of the highly metastatic murine MDAY-D2 lymphosarcoma. The two CD44 alleles of this diploid cell line were sequentially disrupted by homologous recombination, using isogenic CD44 genomic constructs interrupted by a neomycin or hygromycin resistanceconferring gene. The resulting double knockout (DKO) cells had completely lost the capacity to bind to immobilized hyaluronic acid, but did not differ from MDAY-D2 cells in integrin expression or in vitro growth.Subcutaneous (s.c.) growth potential and metastatic capacity of MDAY-D2 and DKO cells were assessed by s.c. and i.v. injection of the lowest cell dose (103 or 104 , respectively) that gave rise to tumor formation by MDAY-D2 cells in N100% of the mice. Quite unexpectedly, we observed no difference at all in either s.c. growth rate or local invasion into surrounding tissues between MDAY-D2 cells and the CD44-negative DKO cells. Also hematogenous metastasis formation upon i.v. injection was similar: both parental and DKO cells metastasized extensively to the spleen, liver, and bone marrow. We conclude that, at least for these MDAY-D2 lymphosarcoma cells, the standard form of CD44 is dispensable for tumor growth and metastasis. Our resuits show that targeted disruption of genes in tumor cells is a feasible approach to study their role in tumorigenesis and metastasis.C D44 is a cell surface protein expressed by many different cell types (for reviews see references 21 and 42) that can act as a receptor for the extracellular matrix component hyaluronic acid (HA) 1 (2) and the proteoglycan serglycin (41). CD44 is required for in vitro lymphopoiesis, particularly of B cells (24), that use CD44 to bind to the abundant HA. In addition, CD44 can act as costimulator in lymphocyte activation (32) and the induction of lymphocyte effector functions (5,11,28). In addition to the standard 85-95-kD form, various larger CD44 variants can be generated by alternative splicing of at least 10 exons (40).CD44 has been proposed to play a major role in metastasis of different types of tumor cells (for review see reference 31). The most compelling evidence concerns a splice variant of CD44 containing exon v6 that upon transfection into a rat pancreatic carcinoma cell line conferred metastatic potential (12). In these cells, the standard form of CD44 is

Integrin Cytoplasmic domain-Associated Protein-1 (ICAP-1) binds specifically to the beta1 integrin subunit cytoplasmic domain. We observed that RNAi-induced knockdown of ICAP-1 reduced migration of C2C12 myoblasts on the beta1 integrin ligand laminin and that overexpression of ICAP-1 increased this migration. In contrast, migration on the beta3 integrin ligand vitronectin was not affected. ICAP-1 knockdown also greatly diminished migration of microvascular endothelial cells on collagen. The number of central focal adhesions in C2C12 cells on laminin was reduced by ICAP-1 knockdown and increased by ICAP-1 overexpression. Previously, we demonstrated that ICAP-1 binds to the ROCK-I kinase and translocates ROCK-I to the plasma membrane. We show here that the ROCK kinase inhibitor Y27362 reduces migration on laminin and causes a loss of central focal adhesions, similarly as ICAP-1 knockdown. ICAP-1 and ROCK were co-immune-precipitated from C2C12 cells, and in cells that overexpressed ICAP-1, YFP-ROCK was translocated to membrane ruffles. These results indicate that ICAP-1 regulates beta1 integrin-dependent cell migration by affecting the pattern of focal adhesion formation. This is likely due to ICAP-1-induced translocation of ROCK to beta1 integrin attachment sites.