An in vivo chemical genetic screen identifies phosphodiesterase 4 as a pharmacological target for hedgehog signaling inhibition.

Hedgehog (Hh) signaling plays an integral role in vertebrate development, and its dysregulation has been accepted widely as a driver of numerous malignancies. While a variety of small molecules target Smoothened (Smo) as a strategy for Hh inhibition, Smo gain-of-function mutations have limited their clinical implementation. Modulation of targets downstream of Smo could define a paradigm for treatment of Hh-dependent cancers. Here, we describe eggmanone, a small molecule identified from a chemical genetic zebrafish screen, which induced an Hh-null phenotype. Eggmanone exerts its Hh-inhibitory effects through selective antagonism of phosphodiesterase 4 (PDE4), leading to protein kinase A activation and subsequent Hh blockade. Our study implicates PDE4 as a target for Hh inhibition, suggests an improved strategy for Hh-dependent cancer therapy, and identifies a unique probe of downstream-of-Smo Hh modulation.

Selective small molecule targeting ß-catenin function discovered by in vivo chemical genetic screen.

The canonical Wnt signaling pathway, mediated by the transcription factor ß-catenin, plays critical roles in embryonic development and represents an important therapeutic target. In a zebrafish-based in vivo screen for small molecules that specifically perturb embryonic dorsoventral patterning, we discovered a compound named windorphen that selectively blocks the Wnt signal required for ventral development. Windorphen exhibits remarkable specificity toward ß-catenin-1 function, indicating that the two ß-catenin isoforms found in zebrafish are not functionally redundant. We show that windorphen is a selective inhibitor of p300 histone acetyltransferase, a coactivator that associates with ß-catenin. Finally, windorphen robustly and selectively kills cancer cells that harbor Wnt-activating mutations, supporting the therapeutic potential of this Wnt inhibitor class.

Synthesis and structure-activity relationships of a novel and selective bone morphogenetic protein receptor (BMP) inhibitor derived from the pyrazolo[1.5-a]pyrimidine scaffold of dorsomorphin: the discovery of ML347 as an ALK2 versus ALK3 selective MLPCN probe.

A structure-activity relationship of the 3- and 6-positions of the pyrazolo[1,5-a]pyrimidine scaffold of the known BMP inhibitors dorsomorphin, 1, LDN-193189, 2, and DMH1, 3, led to the identification of a potent and selective compound for ALK2 versus ALK3. The potency contributions of several 3-position substituents were evaluated with subtle structural changes leading to significant changes in potency. From these studies, a novel 5-quinoline molecule was identified and designated an MLPCN probe molecule, ML347, which shows >300-fold selectivity for ALK2 and presents the community with a selective molecular probe for further biological evaluation.

Enteric nervous system specific deletion of Foxd3 disrupts glial cell differentiation and activates compensatory enteric progenitors.

The enteric nervous system (ENS) arises from the coordinated migration, expansion and differentiation of vagal and sacral neural crest progenitor cells. During development, vagal neural crest cells enter the foregut and migrate in a rostro-to-caudal direction, colonizing the entire gastrointestinal tract and generating the majority of the ENS. Sacral neural crest contributes to a subset of enteric ganglia in the hindgut, colonizing the colon in a caudal-to-rostral wave. During this process, enteric neural crest-derived progenitors (ENPs) self-renew and begin expressing markers of neural and glial lineages as they populate the intestine. Our earlier work demonstrated that the transcription factor Foxd3 is required early in neural crest-derived progenitors for self-renewal, multipotency and establishment of multiple neural crest-derived cells and structures including the ENS. Here, we describe Foxd3 expression within the fetal and postnatal intestine: Foxd3 was strongly expressed in ENPs as they colonize the gastrointestinal tract and was progressively restricted to enteric glial cells. Using a novel Ednrb-iCre transgene to delete Foxd3 after vagal neural crest cells migrate into the midgut, we demonstrated a late temporal requirement for Foxd3 during ENS development. Lineage labeling of Ednrb-iCre expressing cells in Foxd3 mutant embryos revealed a reduction of ENPs throughout the gut and loss of Ednrb-iCre lineage cells in the distal colon. Although mutant mice were viable, defects in patterning and distribution of ENPs were associated with reduced proliferation and severe reduction of glial cells derived from the Ednrb-iCre lineage. Analyses of ENS-lineage and differentiation in mutant embryos suggested activation of a compensatory population of Foxd3-positive ENPs that did not express the Ednrb-iCre transgene. Our findings highlight the crucial roles played by Foxd3 during ENS development including progenitor proliferation, neural patterning, and glial differentiation and may help delineate distinct molecular programs controlling vagal versus sacral neural crest development.

Loss of Foxd3 results in decreased ß-cell proliferation and glucose intolerance during pregnancy.

A complete molecular understanding of ß-cell mass expansion will be useful for the improvement of therapies to treat diabetic patients. During normal periods of metabolic challenges, such as pregnancy, ß-cells proliferate, or self-renew, to meet the new physiological demands. The transcription factor Forkhead box D3 (Foxd3) is required for maintenance and self-renewal of several diverse progenitor cell lineages, and Foxd3 is expressed in the pancreatic primordium beginning at 10.5 d postcoitum, becoming localized predominantly to ß-cells after birth. Here, we show that mice carrying a pancreas-specific deletion of Foxd3 have impaired glucose tolerance, decreased ß-cell mass, decreased ß-cell proliferation, and decreased ß-cell size during pregnancy. In addition, several genes known to regulate proliferation, Foxm1, Skp2, Ezh2, Akt2, and Cdkn1a, are misregulated in islets isolated from these Foxd3 mutant mice. Together, these data place Foxd3 upstream of several pathways critical for ß-cell mass expansion in vivo.

Influence and timing of arrival of murine neural crest on pancreatic beta cell development and maturation.

Interactions between cells from the ectoderm and mesoderm influence development of the endodermally-derived pancreas. While much is known about how mesoderm regulates pancreatic development, relatively little is understood about how and when the ectodermally-derived neural crest regulates pancreatic development and specifically, beta cell maturation. A previous study demonstrated that signals from the neural crest regulate beta cell proliferation and ultimately, beta cell mass. Here, we expand on that work to describe timing of neural crest arrival at the developing pancreatic bud and extend our knowledge of the non-cell autonomous role for neural crest derivatives in the process of beta cell maturation. We demonstrated that murine neural crest entered the pancreatic mesenchyme between the 26 and 27 somite stages (approximately 10.0 dpc) and became intermingled with pancreatic progenitors as the epithelium branched into the surrounding mesenchyme. Using a neural crest-specific deletion of the Forkhead transcription factor Foxd3, we ablated neural crest cells that migrate to the pancreatic primordium. Consistent with previous data, in the absence of Foxd3, and therefore the absence of neural crest cells, proliferation of insulin-expressing cells and insulin-positive area are increased. Analysis of endocrine cell gene expression in the absence of neural crest demonstrated that, although the number of insulin-expressing cells was increased, beta cell maturation was significantly impaired. Decreased MafA and Pdx1 expression illustrated the defect in beta cell maturation; we discovered that without neural crest, there was a reduction in the percentage of insulin-positive cells that co-expressed Glut2 and Pdx1 compared to controls. In addition, transmission electron microscopy analyses revealed decreased numbers of characteristic insulin granules and the presence of abnormal granules in insulin-expressing cells from mutant embryos. Together, these data demonstrate that the neural crest is a critical regulator of beta cell development on two levels: by negatively regulating beta cell proliferation and by promoting beta cell maturation.

Requirement for Foxd3 in the maintenance of neural crest progenitors.

Understanding the molecular mechanisms of stem cell maintenance is crucial for the ultimate goal of manipulating stem cells for the treatment of disease. Foxd3 is required early in mouse embryogenesis; Foxd3(-/-) embryos fail around the time of implantation, cells of the inner cell mass cannot be maintained in vitro, and blastocyst-derived stem cell lines cannot be established. Here, we report that Foxd3 is required for maintenance of the multipotent mammalian neural crest. Using tissue-specific deletion of Foxd3 in the neural crest, we show that Foxd3(flox/-); Wnt1-Cre mice die perinatally with a catastrophic loss of neural crest-derived structures. Cranial neural crest tissues are either missing or severely reduced in size, the peripheral nervous system consists of reduced dorsal root ganglia and cranial nerves, and the entire gastrointestinal tract is devoid of neural crest derivatives. These results demonstrate a global role for this transcriptional repressor in all aspects of neural crest maintenance along the anterior-posterior axis, and establish an unprecedented molecular link between multiple divergent progenitor lineages of the mammalian embryo.

Effects of adenovirus-mediated expression of p27Kip1, p21Waf1 and p16INK4A in cell lines derived from t(2;5) anaplastic large cell lymphoma and Hodgkin's disease.

We investigated the response of SUDHL-1 and L428 cells, derived from t(2;5)-anaplastic large cell lymphoma (ALCL) and Hodgkin's disease (HD), respectively, to recombinant adenoviruses expressing cyclin-dependent kinase inhibitors (CDKIs) p27Kip1 (Adp27), p21Waf1 (Adp21) and p16INK4A (Adp16). Cell cycle analysis of SUDHL-1 cells after 24 h of infection with 200 multiplicity of infection (MOI) of Adp27, Adp21, and Adp16, showed very high levels of cell debris in the subG1 area. The magnitude of cell debris-events was Adp27/Adp21 > Adp16. Cell cycle analysis of L428 cells revealed absence of cell debris and increased G2 phase in all the groups of cells tested as compared to the controls (mock and AdNull). A minimal increase in G1 phase was also evident in cells infected with Adp27 (52%) compared to uninfected cells (43%), AdNull (45%) and to cells infected with Adp21 (37%) and Adp16 (31%). The presence of significant levels of Coxsackie-adenovirus receptor (CAR) on the cell surface of L428 cells excluded the cell membrane-barrier as responsible for the differences in cell observed in response to the recombinant adenovirus-mediated CDKIs expression as compared to SUDHL-1. We also showed that the recombinant adenovirus-mediated cytotoxicity measured as apoptosis was MOI- and vector-dependent in SUDHL-1 cells at lower MOI (100). In conclusion, the therapeutic effect induced by recombinant adenoviruses expressing p27Kip1, p21Waf1 and p16INK4A is cell-dependent in cells derived from selected lymphoid malignancies. Biochemical cellular differences more than cell surface barriers seem to be responsible for differences in response to recombinant adenovirus-mediated expression of cytotoxic genes. Moreover, the cytotoxicity of recombinant adenoviruses expressing p27Kip1, p21Waf1 and p16INK4A may be further explored as a tool for gene therapy of t(2;5)-derived ALCL.

Model of inhibition of the NPM-ALK kinase activity by herbimycin A.

Anaplastic large cell lymphoma (ALCL) exhibiting the t(2;5) translocation is characterized by the resulting expression of the oncogenic fusion protein nucleophosmin-anaplastic lymphoma kinase (NPM-ALK) gene product. The ALK domain of NPM-ALK contains kinase activity, which is responsible for the autophosphorylation of tyrosine residues of the oncogenic protein and phosphorylation of SH2-protein substrates. Herbimycin A is a general protein tyrosine kinase inhibitor active as an antiproliferative compound against different types of mammalian cells. Herbimycin A inhibited the NPM-ALK-associated autophosphorylating activity in an in vitro cell-free kinase assay. The inhibition was specific when tested against other kinase inhibitors and extended to other cell lines derived from t(2;5)-ALCL. SUDHL-1 cells showed increasing percentage of cells in G(1) after 18 h of incubation with a dose of herbimycin A. NPM-ALK, Akt, and pAkt were down-regulated after 24 h of incubation with herbimycin A. Apoptosis was observed only if the dose of inhibitor was given every 12 h for prolonged time. Our results show that herbimycin A interferes with NPM-ALK and Akt pathways in SUDHL-1 cells. It seems that prolonged inhibition of these biochemical pathways may lead to cell cycle arrest and apoptosis. This study supports the idea of investigating protein kinase inhibitors as therapeutic compounds for t(2;5)-ALCL.