Bivalent BET Bromodomain Inhibitors Confer Increased Potency and Selectivity for BRDT via Protein Conformational Plasticity.

Bromodomain and extraterminal domain (BET) proteins are important regulators of gene transcription and chromatin remodeling. BET family members BRD4 and BRDT are validated targets for cancer and male contraceptive drug development, respectively. Due to the high structural similarity of the acetyl-lysine binding sites, most reported inhibitors lack intra-BET selectivity. We surmised that protein-protein interactions induced by bivalent inhibitors may differ between BRD4 and BRDT, conferring an altered selectivity profile. Starting from nonselective monovalent inhibitors, we developed cell-active bivalent BET inhibitors with increased activity and selectivity for BRDT. X-ray crystallographic and solution studies revealed unique structural states of BRDT and BRD4 upon interaction with bivalent inhibitors. Varying spacer lengths and symmetric vs unsymmetric connections resulted in the same dimeric states, whereas different chemotypes induced different dimers. The findings indicate that the increased intra-BET selectivity of bivalent inhibitors is due to the differential plasticity of BET bromodomains upon inhibitor-induced dimerization.

Targeting the BRD4-HOXB13 Coregulated Transcriptional Networks with Bromodomain-Kinase Inhibitors to Suppress Metastatic Castration-Resistant Prostate Cancer.

Resistance to androgen receptor (AR) antagonists is a significant problem in the treatment of castration-resistant prostate cancers (CRPC). Identification of the mechanisms by which CRPCs evade androgen deprivation therapies (ADT) is critical to develop novel therapeutics. We uncovered that CRPCs rely on BRD4-HOXB13 epigenetic reprogramming for androgen-independent cell proliferation. Mechanistically, BRD4, a member of the BET bromodomain family, epigenetically promotes HOXB13 expression. Consistently, genetic disruption of HOXB13 or pharmacological suppression of its mRNA and protein expression by the novel dual-activity BET bromodomain-kinase inhibitors directly correlates with rapid induction of apoptosis, potent inhibition of tumor cell proliferation and cell migration, and suppression of CRPC growth. Integrative analysis revealed that the BRD4-HOXB13 transcriptome comprises a proliferative gene network implicated in cell-cycle progression, nucleotide metabolism, and chromatin assembly. Notably, although the core HOXB13 target genes responsive to BET inhibitors (HOTBIN10) are overexpressed in metastatic cases, in ADT-treated CRPC cell lines and patient-derived circulating tumor cells (CTC) they are insensitive to AR depletion or blockade. Among the HOTBIN10 genes, AURKB and MELK expression correlates with HOXB13 expression in CTCs of mCRPC patients who did not respond to abiraterone (ABR), suggesting that AURKB inhibitors could be used additionally against high-risk HOXB13-positive metastatic prostate cancers. Combined, our study demonstrates that BRD4-HOXB13-HOTBIN10 regulatory circuit maintains the malignant state of CRPCs and identifies a core proproliferative network driving ADT resistance that is targetable with potent dual-activity bromodomain-kinase inhibitors.

Molecular Basis for the N-Terminal Bromodomain-and-Extra-Terminal-Family Selectivity of a Dual Kinase-Bromodomain Inhibitor.

As regulators of transcription, epigenetic proteins that interpret post-translational modifications to N-terminal histone tails are essential for maintaining cellular homeostasis. When dysregulated, "reader" proteins become drivers of disease. In the case of bromodomains, which recognize N-ε-acetylated lysine, selective inhibition of individual bromodomain-and-extra-terminal (BET)-family bromodomains has proven challenging. We describe the >55-fold N-terminal-BET bromodomain selectivity of 1,4,5-trisubstituted-imidazole dual kinase-bromodomain inhibitors. Selectivity for the BRD4 N-terminal bromodomain (BRD4(1)) over its second bromodomain (BRD4(2)) arises from the displacement of ordered waters and the conformational flexibility of lysine-141 in BRD4(1). Cellular efficacy was demonstrated via reduction of c-Myc expression, inhibition of NF-κB signaling, and suppression of IL-8 production through potential synergistic inhibition of BRD4(1) and p38α. These dual inhibitors provide a new scaffold for domain-selective inhibition of BRD4, the aberrant function of which plays a key role in cancer and inflammatory signaling.

Advances of small molecule targeting of kinases.

Reversible protein phosphorylation regulates virtually all aspects of life in the cell. As a result, dysregulation of protein kinases, the enzymes responsible for transferring phosphate groups from ATP to proteins, are often the cause or consequence of many human diseases including cancer. Almost three dozen protein kinase inhibitors (PKIs) have been approved for clinical applications since 1995, the vast majority of them for the treatment of cancer. According to the NCI, there are more than 100 types of cancer. However, FDA-approved PKIs only target 14 of them. Importantly, of the more than 500 protein kinases encoded by the human genome, only 22 are targets for currently approved PKIs, suggesting that the reservoir of PKIs still has room to grow significantly. In this short review we will discuss the most recent advances, challenges, and alternatives to currently adopted strategies in this burgeoning field.

BET Bromodomain Inhibitors with One-Step Synthesis Discovered from Virtual Screen.

Chemical inhibition of epigenetic regulatory proteins BrdT and Brd4 is emerging as a promising therapeutic strategy in contraception, cancer, and heart disease. We report an easily synthesized dihydropyridopyrimidine pan-BET inhibitor scaffold, which was uncovered via a virtual screen followed by testing in a fluorescence anisotropy assay. Dihydropyridopyimidine 3 was subjected to further characterization and is highly selective for the BET family of bromodomains. Structure-activity relationship data and ligand deconstruction highlight the importance of the substitution of the uracil moiety for potency and selectivity. Compound 3 was also cocrystallized with Brd4 for determining the ligand binding pose and rationalizing subsequent structure-activity data. An additional series of dihydropyridopyrimidines was synthesized to exploit the proximity of a channel near the ZA loop of Brd4, leading to compounds with submicromolar affinity and cellular target engagement. Given these findings, novel and easily synthesized inhibitors are being introduced to the growing field of bromodomain inhibitor development.

Dual Targeting of WEE1 and PLK1 by AZD1775 Elicits Single Agent Cellular Anticancer Activity.

Inhibition of the WEE1 tyrosine kinase enhances anticancer chemotherapy efficacy. Accordingly, the WEE1 inhibitor AZD1775 (previously MK-1775) is currently under evaluation in clinical trials for cancer in combination with chemotherapy. AZD1775 has been reported to display high selectivity and is therefore used in many studies as a probe to interrogate WEE1 biology. However, AZD1775 also exhibits anticancer activity as a single agent although the underlying mechanism is not fully understood. Using a chemical proteomics approach, we here describe a proteome-wide survey of AZD1775 targets in lung cancer cells and identify several previously unknown targets in addition to WEE1. In particular, we observed polo-like kinase 1 (PLK1) as a new target of AZD1775. Importantly, in vitro kinase assays showed PLK1 and WEE1 to be inhibited by AZD1775 with similar potency. Subsequent loss-of-function experiments using RNAi for WEE1 and PLK1 suggested that targeting PLK1 enhances the pro-apoptotic and antiproliferative effects observed with WEE1 knockdown. Combination of RNAi with AZD1775 treatment suggested WEE1 and PLK1 to be the most relevant targets for mediating AZD1775's anticancer effects. Furthermore, disruption of WEE1 by CRISPR-Cas9 sensitized H322 lung cancer cells to AZD1775 to a similar extent as the potent PLK1 inhibitor BI-2536 suggesting a complex crosstalk between PLK1 and WEE1. In summary, we show that AZD1775 is a potent dual WEE1 and PLK1 inhibitor, which limits its use as a specific molecular probe for WEE1. However, PLK1 inhibition makes important contributions to the single agent mechanism of action of AZD1775 and enhances its anticancer effects.

Potent Dual BET Bromodomain-Kinase Inhibitors as Value-Added Multitargeted Chemical Probes and Cancer Therapeutics.

Synergistic action of kinase and BET bromodomain inhibitors in cell killing has been reported for a variety of cancers. Using the chemical scaffold of the JAK2 inhibitor TG101348, we developed and characterized single agents which potently and simultaneously inhibit BRD4 and a specific set of oncogenic tyrosine kinases including JAK2, FLT3, RET, and ROS1. Lead compounds showed on-target inhibition in several blood cancer cell lines and were highly efficacious at inhibiting the growth of hematopoietic progenitor cells from patients with myeloproliferative neoplasm. Screening across 931 cancer cell lines revealed differential growth inhibitory potential with highest activity against bone and blood cancers and greatly enhanced activity over the single BET inhibitor JQ1. Gene drug sensitivity analyses and drug combination studies indicate synergism of BRD4 and kinase inhibition as a plausible reason for the superior potency in cell killing. Combined, our findings indicate promising potential of these agents as novel chemical probes and cancer therapeutics. Mol Cancer Ther; 16(6); 1054-67. ©2017 AACR.

Marinopyrrole derivatives with sulfide spacers as selective disruptors of Mcl-1 binding to pro-apoptotic protein Bim.

A series of novel marinopyrroles with sulfide and sulphone spacers were designed and synthesized. Their activity to disrupt the binding of the pro-apoptotic protein, Bim, to the pro-survival proteins, Mcl-1 and Bcl-xL, was evaluated using ELISA assays. Fluorescence-quenching (FQ) assays confirmed the direct binding of marinopyrroles to Mcl-1. Benzyl- and benzyl methoxy-containing sulfide derivatives 4 and 5 were highly potent dual Mcl-1/Bim and Bcl-xL/Bim disruptors (IC50 values of 600 and 700 nM), whereas carboxylate-containing sulfide derivative 9 exhibited 16.4-fold more selectivity for disrupting Mcl-1/Bim over Bcl-xL/Bim binding. In addition, a nonsymmetrical marinopyrrole 12 is as equally potent as the parent marinopyrrole A (1) for disrupting both Mcl-1/Bim and Bcl-xL/Bim binding. Some of the derivatives were also active in intact human breast cancer cells where they reduced the levels of Mcl-1, induced programd cell death (apoptosis) and inhibited cell proliferation.

Acetyl-lysine binding site of bromodomain-containing protein 4 (BRD4) interacts with diverse kinase inhibitors.

Members of the bromodomain and extra terminal (BET) family of proteins are essential for the recognition of acetylated lysine (KAc) residues in histones and have emerged as promising drug targets in cancer, inflammation, and contraception research. In co-crystallization screening campaigns using the first bromodomain of BRD4 (BRD4-1) against kinase inhibitor libraries, we identified and characterized 14 kinase inhibitors (10 distinct chemical scaffolds) as ligands of the KAc binding site. Among these, the PLK1 inhibitor BI2536 and the JAK2 inhibitor TG101209 displayed strongest inhibitory potential against BRD4 (IC50=25 nM and 130 nM, respectively) and high selectivity for BET bromodomains. Comparative structural analysis revealed markedly different binding modes of kinase hinge-binding scaffolds in the KAc binding site, suggesting that BET proteins are potential off-targets of diverse kinase inhibitors. Combined, these findings provide a new structural framework for the rational design of next-generation BET-selective and dual-activity BET-kinase inhibitors.

Akt2 and acid ceramidase cooperate to induce cell invasion and resistance to apoptosis.

Both Akt 2 and acid ceramidase (ASAH1) are found aberrantly overexpressed in cancer cells, but whether these two enzymes cooperate to induce malignant transformation is not known. We found that in immortalized, non-transformed cells, ectopic co-expression of Akt2 and ASAH1 is significantly more effective than expression of each gene alone at inducing cell invasion and at conferring resistance to apoptosis. Consistent with these observations, siRNA-mediated depletion of both Akt2 and ASAH1 is much more potent than depleting each alone at inhibiting cell viability/proliferation and cell invasion. Furthermore, pharmacological inhibitors of Akt (TCN or MK-2206) and ASAH1 (B13) synergize to inhibit cell viability/proliferation, and combinations of these drugs are more effective than single-agent treatments at inhibiting cell invasion. Taken together, the results suggest that these two enzymes cooperate to induce malignant transformation and warrant further preclinical studies to evaluate the potential of combining inhibitors of Akt and ASAH1 to treat cancer.

Synthesis and evaluation of substituted hexahydronaphthalenes as novel inhibitors of the Mcl-1/BimBH3 interaction.

Mcl-1, an anti-apoptotic member of the Bcl-2 protein family, is overexpressed in a broad range of human cancers and plays a critical role in conferring resistance to chemotherapy. In the course of screening a natural product-like library of sesquiterpenoid analogs, we identified substituted hexahydronaphthalenes that showed activity against the Mcl-1/BimBH3 interaction in vitro. Here, we describe the synthesis of a small library of analogs and their biological evaluation. The most potent inhibitor in the series (19) exhibits an IC(50) of 8.3 µM by ELISA and disrupts the interaction between endogenously expressed Mcl-1 and Bim in cultured MDA-MB-468 breast cancer cells.

Measurement of protein farnesylation and geranylgeranylation in vitro, in cultured cells and in biopsies, and the effects of prenyl transferase inhibitors.

The importance of the post-translational lipid modifications farnesylation and geranylgeranylation in protein localization and function coupled with the critical role of prenylated proteins in malignant transformation has prompted interest in their biology and the development of farnesyl transferase and geranylgeranyl transferase inhibitors (FTIs and GGTIs) as chemical probes and anticancer agents. The ability to measure protein prenylation before and after FTI and GGTI treatment is important to understanding and interpreting the effects of these agents on signal transduction pathways and cellular phenotypes, as well as to the use of prenylation as a biomarker. Here we describe protocols to measure the degree of protein prenylation by farnesyl transferase or geranylgeranyl transferase in vitro, in cultured cells and in tumors from animals and humans. The assays use [(3)H]farnesyl diphosphate and [(3)H]geranylgeranyl diphosphate, electrophoretic mobility shift, membrane association using subcellular fractionation or immunofluorescence of intact cells, [(3)H]mevalonic acid labeling, followed by immunoprecipitation and SDS-PAGE, and in vitro transcription, translation and prenylation in reticulocyte lysates. These protocols require from 1 d (enzyme assays) to up to 3 months (autoradiography of [(3)H]-labeled proteins).

Targeting protein prenylation for cancer therapy.

Protein farnesylation and geranylgeranylation, together referred to as prenylation, are lipid post-translational modifications that are required for the transforming activity of many oncogenic proteins, including some RAS family members. This observation prompted the development of inhibitors of farnesyltransferase (FT) and geranylgeranyl-transferase 1 (GGT1) as potential anticancer drugs. In this Review, we discuss the mechanisms by which FT and GGT1 inhibitors (FTIs and GGTIs, respectively) affect signal transduction pathways, cell cycle progression, proliferation and cell survival. In contrast to their preclinical efficacy, only a small subset of patients responds to FTIs. Identifying tumours that depend on farnesylation for survival remains a challenge, and strategies to overcome this are discussed. One GGTI has recently entered the clinic, and the safety and efficacy of GGTIs await results from clinical trials.

The BH3 alpha-helical mimic BH3-M6 disrupts Bcl-X(L), Bcl-2, and MCL-1 protein-protein interactions with Bax, Bak, Bad, or Bim and induces apoptosis in a Bax- and Bim-dependent manner.

A critical hallmark of cancer cell survival is evasion of apoptosis. This is commonly due to overexpression of anti-apoptotic proteins such as Bcl-2, Bcl-X(L), and Mcl-1, which bind to the BH3 α-helical domain of pro-apoptotic proteins such as Bax, Bak, Bad, and Bim, and inhibit their function. We designed a BH3 α-helical mimetic BH3-M6 that binds to Bcl-X(L) and Mcl-1 and prevents their binding to fluorescently labeled Bak- or Bim-BH3 peptides in vitro. Using several approaches, we demonstrate that BH3-M6 is a pan-Bcl-2 antagonist that inhibits the binding of Bcl-X(L), Bcl-2, and Mcl-1 to multi-domain Bax or Bak, or BH3-only Bim or Bad in cell-free systems and in intact human cancer cells, freeing up pro-apoptotic proteins to induce apoptosis. BH3-M6 disruption of these protein-protein interactions is associated with cytochrome c release from mitochondria, caspase-3 activation and PARP cleavage. Using caspase inhibitors and Bax and Bak siRNAs, we demonstrate that BH3-M6-induced apoptosis is caspase- and Bax-, but not Bak-dependent. Furthermore, BH3-M6 disrupts Bcl-X(L)/Bim, Bcl-2/Bim, and Mcl-1/Bim protein-protein interactions and frees up Bim to induce apoptosis in human cancer cells that depend for tumor survival on the neutralization of Bim with Bcl-X(L), Bcl-2, or Mcl-1. Finally, BH3-M6 sensitizes cells to apoptosis induced by the proteasome inhibitor CEP-1612.

Discovery of a novel proteasome inhibitor selective for cancer cells over non-transformed cells.

Numerous proteins controlling cell cycle progression, apoptosis and angiogenesis are degraded by the ubiquitin/proteasome system, which has become the subject for intense investigations for cancer therapeutics. Therefore, we used in silico and experimental approaches to screen compounds from the NCI chemical libraries for inhibitors against the chymotrypsin-like (CT-L) activity of the proteasome and discovered PI-083. Molecular docking indicates that PI-083 interacts with the Thr21, Gly47 and Ala49 residues of the beta5 subunit and Asp114 of the beta6 subunit of the proteasome. PI-083 inhibits CT-L activity and cell proliferation and induces apoptosis selectively in cancer cells (ovarian T80-Hras, pancreatic C7-Kras and breast MCF-7) as compared to their normal/immortalized counterparts (T80, C7 and MCF-10A, respectively). In contrast, Bortezomib, the only proteasome inhibitor approved by the Food and Drug Administration (FDA), did not exhibit this selectivity for cancer over non-transformed cells. In addition, in all cancer cells tested, including Multiple Myeloma (MM), breast, pancreatic, ovarian, lung, prostate cancer cell lines as well as fresh MM cells from patients, PI-083 required less time than Bortezomib to induce its antitumor effects. Furthermore, in nude mouse xenografts in vivo, PI-083, but not Bortezomib, suppressed the growth of human breast and lung tumors. Finally, following in vivo treatment of mice, PI-083 inhibited tumor, but not hepatic liver CT-L activity, whereas Bortezomib inhibited both tumor and liver CT-L activities. These results suggest that PI-083 is more selective for cancer cells and may have broader antitumor activity and therefore warrants further advanced preclinical studies.

Blockade of protein geranylgeranylation inhibits Cdk2-dependent p27Kip1 phosphorylation on Thr187 and accumulates p27Kip1 in the nucleus: implications for breast cancer therapy.

We describe the design of a potent and selective peptidomimetic inhibitor of geranylgeranyltransferase I (GGTI), GGTI-2418, and its methyl ester GGTI-2417, which increases the levels of the cyclin-dependent kinase (Cdk) inhibitor p27(Kip1) and induces breast tumor regression in vivo. Experiments with p27(Kip1) small interfering RNA in breast cancer cells and p27(Kip1) null murine embryonic fibroblasts demonstrate that the ability of GGTI-2417 to induce cell death requires p27(Kip1). GGTI-2417 inhibits the Cdk2-mediated phosphorylation of p27(Kip1) at Thr187 and accumulates p27(Kip1) in the nucleus. In nude mouse xenografts, GGTI-2418 suppresses the growth of human breast tumors. Furthermore, in ErbB2 transgenic mice, GGTI-2418 increases p27(Kip1) and induces significant regression of breast tumors. We conclude that GGTIs' antitumor activity is, at least in part, due to inhibiting Cdk2-dependent p27(Kip1) phosphorylation at Thr187 and accumulating nuclear p27(Kip1). Thus, GGTI treatment might improve the poor prognosis of breast cancer patients with low nuclear p27(Kip1) levels.

A limited screen for protein interactions reveals new roles for protein phosphatase 1 in cell cycle control and apoptosis.

Protein phosphatase 1 (PP1) catalytic subunits typically combine with other proteins that modulate their activity, direct them to distinct substrates, or serve as substrates for PP1. More than 50 PP1-interacting proteins (PIPs) have been identified so far. Given there are approximately 10 000 phosphoproteins in mammals, many PIPs remain to be discovered. We have used arrays containing 100 carefully selected antibodies to identify novel PIPs that are important in cell proliferation and cell survival in murine fetal lung epithelial cells and human A549 lung cancer cells. The antibody arrays identified 31 potential novel PIPs and 11 of 17 well-known PIPs included as controls, suggesting a sensitivity of at least 65%. A majority of the interactions between PP1 and putative PIPs were isoform- or cell type-specific. We confirmed by co-immunoprecipitation that 9 of these proteins associate with PP1: APAF-1, Bax, E-cadherin, HSP-70, Id2, p19Skp1, p53, PCNA, and PTEN. We examined two of these interactions in greater detail in A549 cells. Exposure to nicotine enhanced association of PP1 with Bax (and Bad), but also induced inhibitory phosphorylation of PP1. In addition to p19Skp1, PP1alpha antibodies also coprecipitated cullin 1, suggesting that PP1alpha is associated with the SCF1 complex. This interaction was only detectable during the G1/S transition and S phase. Forced loss of PP1 function decreased the levels of p27Kip1, a well-known SCF1 substrate, suggesting that PP1 may rescue proteins from ubiquitin/proteasome-mediated destruction. Both of these novel interactions are consistent with PP1 facilitating cell cycle arrest and/or apoptosis.

Protein phosphatase 1alpha activity prevents oncogenic transformation.

Cyclin-dependent kinase 2 (Cdk2) phosphorylates Thr320 of protein phosphatase 1alpha (PP1alpha) in late G(1), thereby inhibiting its activity. Phosphorylation-resistant PP1alphaT320A, acting as a constitutively active (CA) mutant, causes a late G(1) arrest by preventing the phosphorylation and inactivation of the retinoblastoma protein (pRb). Both PP1alpha-mediated G(1) arrest and PP1alpha phosphorylation in late G(1) require the presence of pRb, indicating that PP1alpha is a crucial regulator of the pRb pathway, which is almost invariably mutated in human cancer. These findings prompted us to investigate whether PP1alpha interferes with oncogenic transformation. The ability of NIH 3T3 cells to form foci after transformation with ras/cyclin D1 was significantly inhibited by co-transfection with PP1alphaT320A, but not PP1alpha. Likewise, cells expressing PP1alphaT320A or PP1alphaT320A fused to green fluorescent protein (GFP) were unable to form colonies in soft agar, regardless of whether PP1alpha constructs were co-transfected with ras/cyclin D1 or transfected into stably transformed cells. Overexpressed wild-type (Wt) PP1alpha and GFP-PP1alpha were phosphorylated in Thr320, most likely explaining its lack of effect. Expression of GFP-PP1alphaT320A was associated with caspase-cleaved pRb in Western blots (WB) and morphological signs of cell death. These findings demonstrate that PP1alpha activity can override oncogenic signaling by causing cell-cycle arrest and/or apoptosis rather than restoring contact inhibition or anchorage dependence.

Interaction between the retinoblastoma protein and protein phosphatase 1 during the cell cycle.

The functions of the retinoblastoma protein (pRb) are in part regulated by reversible and cell cycle-dependent phosphorylation. While the regulation of pRb by cyclin-dependent kinases (Cdks) has been studied extensively, the role(s) of protein phosphatase 1 (PP1) in controlling pRb are only partially understood. In this chapter, we will describe experimental approaches to investigate the interactions between pRb and PP1. Methods will be presented to study the cell cycle-dependent dephosphorylation of pRb by various PP1 isozymes, the specificity of PP1 isozymes for distinct pRb phosphorylation sites, the dephosphorylation of pRb associated with apoptosis, and the cell cycle- and pRb-dependent phosphorylation of PP1.

Protein phosphatase 1alpha is required for murine lung growth and morphogenesis.

Protein phosphatase 1 (PP1) plays important roles in cell cycle control and apoptosis, two processes that impinge on morphogenesis and differentiation. Following the precedent set by other molecules regulating the cell cycle and apoptosis, we hypothesized that PP1 may have context-specific roles in development. Therefore, we have studied the spatial and temporal expression of PP1alpha during murine lung development and determined the consequences of loss of PP1alpha function on branching morphogenesis. By using an immunohistochemical approach, we show here that PP1alpha was expressed throughout the epithelium and mesenchyme upon the emergence of the lung primordium on embryonic day 10, with immunostaining exclusively extranuclear. During the late pseudoglandular stage, PP1alpha was predominantly expressed in the distal lung epithelium, whereas the mesenchyme contained very little or no PP1alpha protein. Peri- and postnatally, PP1alpha immunostaining was mostly nuclear in apparently differentiated cells, as judged by colocalization with well-known markers for lung differentiation. Exposure of fetal lung explants to antisense oligodeoxynucleotides against PP1alpha, resulted in decreased overall size of the cultured lung, a defect in forming new airways, lack of expression of surfactant protein C, and histologic signs of poor differentiation. These data suggest that PP1alpha is required for branching morphogenesis and differentiation.