A major aim of our research is to understand the molecular mechanisms that organize the growth and patterning of multicellular embryos. Much of our work has focused on the Hedgehog (Hh) family of secreted signaling proteins. These proteins are expressed in restricted embryonic locations and elicit graded responses in surrounding cells, thereby controlling the patterns of cellular proliferation and differentiation in developing organs. In the past we concentrated on identifying the biological roles of Hedgehog signaling in embryonic patterning, on understanding how the Hedgehog protein signal is produced and how its tissue distribution is regulated, how cells perceive and respond to this signal, and how this and related signaling pathways that pattern multicellular embryos arose in evolution. In our current work we continue to examine mechanistic aspects of the biogenesis, tissue distribution, and transduction of the Hedgehog signal. We have also begun to explore postembryonic functions of the Hh signaling pathway, including its normal role in tissue homeostasis and its pathological role in cancer growth.
The Hedgehog Protein Signal
Our work on the Hedgehog pathway began with isolation of the Drosophila hedgehog gene and the demonstration that it encodes a secreted protein signal expressed in precisely localized patterns within developing tissues. Similar findings subsequently emerged from the isolation of hedgehog genes in higher organisms, including mammals. Our biochemical studies revealed the derivation of the mature, active Hedgehog signal from its protein precursor by a novel autocatalytic processing reaction. This reaction proceeds via a thioester intermediate that replaces a main-chain peptide bond. Cholesterol attack of this thioester then results in cleavage of the precursor and covalent linkage of cholesterol to the carboxyl terminus of the amino-terminal fragment. This amino-terminal fragment is further modified by addition of palmitate at its amino terminus and is then responsible for all signaling activities. The carboxyl-terminal domain initiates processing of the precursor and is required for production of the active signal.
Crystallographic structural studies, in collaboration with Daniel Leahy (Johns Hopkins University), revealed distinct and surprising evolutionary origins for the signaling and processing domains of the Hedgehog protein. The amino-terminal signaling domain displays a striking resemblance to an enzyme involved in shaping the bacterial cell wall. Catalytic activity, however, is no longer required for Hedgehog signaling; instead, this conserved domain appears to function as a structural scaffold for residues that are important in Hedgehog function as a protein ligand. In addition, a portion of the Hedgehog carboxyl-terminal processing domain displays unmistakable structural similarity to the inteins of self-splicing proteins. The inteins constitute a widely disseminated group of genetic parasites whose excision from host proteins also proceeds by an initial step involving a thioester intermediate; the Hedgehog processing domain in addition has acquired a carboxyl-terminal appendage that mediates participation of cholesterol in the second step of the processing reaction. These structural findings provide a remarkable view of the Hedgehog protein as an embryonic patterning signal that evolved by assembly and adaptation of ancient domains for novel functions.
Normal and Pathological Functions of Hedgehog Signaling
Identification of the active Hedgehog signal facilitated studies of its distribution and activities in embryonic tissues, and in turn led to the finding that Hedgehog signaling elicits concentration-dependent responses in a range of developing organs. Expression of the vertebrate Hedgehog family member Sonic hedgehog (Shh), for example, was found to be localized in structures with the ability to organize development of surrounding tissues. Furthermore, embryonic tissue explants respond to the Hedgehog signal in vitro in a concentration-dependent manner. In embryos, localized production of the Hedgehog signal elicits distinct responses from surrounding cells as a function of the length of exposure to or distance from the source. Such graded responses are critical for formation of the normal spatial pattern of structures as diverse as the digits of the limb or neuronal types in the spinal cord.
In mice lacking function of Shh or other pathway components, the organs affected include the brain, spinal cord, axial skeleton and musculature, the limbs, and endodermally derived organs such as the gastrointestinal tract and the lungs. A striking aspect of the Shh mutant phenotype is the occurrence of cyclopia and ventral forebrain-patterning defects, and these malformations suggested a potential connection between embryonic disruption of Hedgehog signaling and human cyclopia. Although rare, human cyclopia is the extreme manifestation of holoprosencephaly, a term that includes a spectrum of birth defects. Genetic or environmental perturbations of Hedgehog signaling are now well recognized as causally linked to human holoprosencephaly.
Cyclopia in Shh mutant mice also led to the discovery that the plant-derived teratogen, cyclopamine, specifically antagonizes Hedgehog signaling through binding and inhibition of the pathway component Smoothened. With the use of cyclopamine as a potent and specific Hedgehog pathway antagonist, our studies have shifted from patterning activities in embryos to postembryonic activities in maintenance of tissue pattern and neoplastic growth. We have thus noted a critical role for continuous Hedgehog pathway activity in driving the growth of a broad group of deadly human cancers that include esophageal, pancreatic, biliary tract, gastric, prostate, and small-cell lung cancers. These cancers arise in the epithelia of endodermally derived organs and in aggregate account for a significant proportion of human cancer deaths. Our studies show that treatment with the Hedgehog pathway antagonist cyclopamine can block cell proliferative effects associated with pathway activation and can cause complete regression of aggressive human and rodent cancers growing in mice. The use of cyclopamine or other pathway antagonists thus may represent a novel, nontoxic approach to therapies for lethal human cancers.
Why is Hedgehog pathway activity a factor in the growth of so many types of cancer? Recent studies point to a requirement for Hedgehog pathway activity in the renewal and maintenance of postembryonic tissues, probably through effects on division and activity of tissue stem cells. These findings are relevant to cancer because of the possible derivation of cancer stem cells—the minority of cells within a cancer that are capable of its propagation—from adult tissue stem cells. We have proposed that the association between Hedgehog pathway activation and cancer growth is related to the normal function of this pathway in regulating stem cell activity. Given the increased cancer incidence associated with chronic tissue injury, we have further proposed that Hh pathway activation and consequent expansion of tissue stem cells may be induced by tissue injury and that inappropriate continuation of pathway activity may initiate cancer growth by transforming these cells into cancer stem cells. Much of our current work is aimed at investigating the normal mechanism and biological role of Hedgehog pathway activation in response to injury, and the pathological mechanism by which pathway activity inappropriately persists and leads to oncogenic transformation.