Longing for Ligand: Hedgehog, Patched, and Cell Death

See allHide authors and affiliations

Science  08 Aug 2003:
Vol. 301, Issue 5634, pp. 774-776
DOI: 10.1126/science.1088625

Dependence on developmental signals, like dependence on love, can have catastrophic effects. Being unloved or lost in a multicellular organism can lead to self-inflicted death that people call suicide and cells call apoptosis. But how do lost cells recognize that they lack direction from developmental signals? That job may belong to ligand dependence receptors that induce cells bearing them to undergo apoptosis if the receptor remains unoccupied by ligand. These receptors are thought to ensure the survival of cells that remain close to the source of the appropriate developmental signal (the ligand), and the death of those cells that do not. On page 843 of this issue, Thibert et al. (1) provide evidence suggesting that the Patched1 (Ptc1) receptor is a dependence receptor that induces programmed cell death during chick neural tube development in the absence of its ligand, the signaling molecule Sonic hedgehog (Shh).

During neural development in vertebrates, Shh is produced first by the notochord and later by the floor plate (see the figure). A ventrodorsal gradient of Shh directs ventral patterning and cell differentiation [reviewed in (2)]. The Thibert et al. results provide insight into how the neural tube is shaped during development. In multicellular organisms, cells that are poorly positioned as a result of developmental errors can be eliminated because of failure to receive cues instructing them that they are in the correct location [reviewed in (3)]. Given the widespread importance of hedgehog (Hh) signaling during development, the results of Thibert et al. suggest how Ptc1 and Shh signaling may control tissue sculpting through selective cell death and survival.

Tissue sculpting through cell death and survival.

(A) The action of Ptc1 in the Shh signaling pathway. (Left) In the absence of Shh, Ptc1 may act as a proapoptotic dependence receptor inducing programmed cell death. It is not clear how the apoptotic activity of Ptc1 is regulated, apart from the need for Shh to be absent. (Right) In the presence of Shh, Ptc1 no longer represses Smoothened (Smo), leading to activation of the Gli transcription factor, cell survival, and proliferation. (B) Cleavage of part of the carboxyl-terminal region of membrane-bound Ptc1 by caspase-3 exposes an apoptotic domain. (C) Neural tube development in a generic vertebrate embryo in which Shh is secreted by the notochord and floor plate, inducing patterning of the ventral region. (Left) Note that there is a ventral-to-dorsal gradient of Shh ligand and Ptc1 expression. (Right) In the absence of Shh, for example through early removal of the notochord, the neural tube forms but lacks ventral cell types and shows a high degree of Ptc1-mediated apoptosis (1).


To initiate the cell death program, pro-apoptotic dependence receptors require preliminary cleavage of their intracellular domain (at the DXXD site) by caspase enzymes. Thibert et al. present several lines of evidence indicating that Ptc1 is a proapoptotic dependence receptor. They show that overexpression of Ptc1 in cultured cells induces apoptosis, which is blocked by addition of Shh. In the developing chick neural tube, removal of the ventral source of Shh causes massive cell death, which is rescued by expression of a dominant-negative form of Ptc1 that interferes with the proposed function of wildtype Ptc1 in apoptosis. Cleavage of Ptc1 by caspase-3 exposes a carboxyl-terminal apoptotic domain. Transfecting cultured cells with the carboxyl-terminal region of Ptc1 is sufficient to induce cell death. In this region, there is a conserved aspartic acid residue in human, mouse, and chicken; mutation of this site (D1392N) in mouse Ptc1 prevents apoptosis when this receptor is unoccupied by ligand. Transfection of cultured cells with Ptc1 truncated at the caspase-cleavage site induces apoptosis that cannot be rescued by addition of Shh.

These provocative results raise several questions about the in vivo mode of action of Ptc1 in apoptosis. For example, it remains unclear when Ptc1-induced apoptosis takes place during neural tube formation and which cells are selectively killed. Indeed, Ptc1 is widely expressed in regions of the brain that are located far from sources of Shh. In addition, in the developing neural tube, some cells derived from Shh-responsive precursors such as spinal oligodendrocytes (4) happily migrate away from the local environment where Shh acts. This suggests that the proapoptotic activity of Ptc1 must be tightly regulated; for example, a critical threshold level of Ptc1 may be required to completely inhibit the Shh pathway and initiate apoptosis. Also, there could be other Ptc1 ligands in addition to Shh whose absence leads to Ptc1-mediated apoptosis.

In a previous study in the chick (5), a Ptc1 mutant protein was engineered to lack the second extracellular loop and so was unable to bind to Shh. Introducing this mutant Ptc1 (which had an intact carboxyl terminus) into the chick neural tube resulted in changes in cell type specification without obvious changes in cell number (although apoptosis per se was not measured) (5). Reconciling these findings with those of Thibert et al.—who show that Ptc1 with an intact carboxyl terminus induces apoptosis in the chick neural tube—will require both groups to perform cell death and cell differentiation assays. It remains possible, however, that the second extracellular loop of Ptc1 not only controls Shh binding, but also affects carboxyl-terminal cleavage and exposure of the proapoptotic domain. But this seems inconsistent with the Thibert et al. finding that expression of the entire last intracellular domain of Ptc1 is sufficient to induce cell death (1). It is possible that the ectopic expression of mutant Ptc1 lacking the second extracellular loop still permits a low level of Shh signaling that is sufficient to prevent apoptosis.

Ptc1 may be a dependence receptor that controls the size and shape of the chick neural tube by inducing some ventral cells to undergo apoptosis (6). Indeed, its ligand, Shh, is critical for regulating precursor cell numbers in different regions of the central nervous system, such as the cerebellum [reviewed in (7)]. However, as Thibert et al. show in Shh-deprived embryos, blocking Ptc1-induced apoptosis with a dominant-negative Ptc1 protein lacking proapoptotic activity does not fully rescue the loss-of-Shh phenotype; the neural tube is still small and poorly shaped. Therefore, the overall action of Shh signaling in ventral neural tube patterning could be a combination of mitogenic, morphogenetic, and cell survival activities.

The relationship of Ptc1-induced apoptosis to the Shh signaling pathway remains to be clarified. Expression of a Ptc1 mutant lacking part of its carboxyl-terminal domain still induces cell death even when Shh is added to the cells (1). Because there is no evidence that cells in which the Shh signaling pathway is active die, these results suggest that expression of a truncated Ptc1 receptor does not activate the Shh pathway. However, several PTCH1 mutations that appear to cause inappropriate activation of the SHH pathway in tumors do map to the carboxyl terminus [reviewed in (8)]. Moreover, in the fruit fly Drosophila, removal of the carboxyl-terminal region of the Ptc1 receptor induces activation of the Hh pathway (9). These results raise interesting questions. Do fly Ptc and vertebrate Ptc1 behave similarly? Does the related Ptc2 receptor behave like Ptc1? What is the relationship between Ptc1 and Gli3 (10), a downstream transcription factor in the Hh pathway that is known to affect apoptosis? Direct in vivo testing of the activation of Shh-Gli targets under various conditions, including rescue experiments with the dominant-negative form of Ptc1, should help to clarify these issues.

Most of our knowledge of the Hh signaling pathway originally came from studies in flies, where amazing conservation of Hh function enabled extrapolation of results to larger organisms. In flies, however, there is no evidence that Ptc acts as a dependence receptor. Cells that cease to receive an Hh signal, for example, in smoothened (Smo) mutant clones (11), or that express high levels of Ptc (5, 12) in areas outside the organizing region of the wing imaginal disc (anterior-posterior compartment border), survive and proliferate. In the organizing region, cells that do not receive an Hh signal do not die, but a complete block in Hh signaling in the wing primordium prevents the activation of other morphogenetic signals such as Decapentaplegic (Dpp) (13), leading to inhibition of wing development (14). In the Drosophila Ptc receptor, the consensus caspase recognition site is missing. This is not entirely unexpected because in both nematodes and flies, extrinsic death receptor activation of caspases seems to be absent (15). It is possible that Ptc-induced apoptosis could be a late evolutionary acquisition that was never present in insects, or that insects eliminated this function over the course of evolution.

Beyond normal development, the finding that Ptc1 may be a proapoptotic dependence receptor could have important implications for our understanding of human cancer. PTCH1 is a tumor suppressor protein that is mutated in patients with basal cell nevus syndrome (16, 17) and in cells of various types of sporadic tumors, including those of the skin and brain [reviewed in (18)]. Mice carrying Ptc mutations also develop tumors of the cerebellum (medulloblastomas) (19). Mutations in PTCH1 may be one of the many ways in which the HH pathway is switched on, leading to activation of GLI transcription factors and the initiation of tumor formation [reviewed in (18)]. Indeed, expression of GLI1, a marker of HH pathway activation, is the hallmark of sporadic tumors such as basal cell carcinomas and medulloblastomas that arise from inappropriate HH pathway activity (20, 21). Misexpression of GLI1 is sufficient to induce basal cell carcinoma-like tumors (20, 21); and medulloblastomas and basal cell carcinomas require an active SHH-GLI pathway for maintained proliferation (21, 23, 24). In this context, the results of Thibert et al. suggest an additional twist. Is it possible that null mutations in PTCH1 leading to the absence of caspase-mediated cell death and pathway activation allow certain cells to survive and initiate tumorigenesis? Such a scenario could provide a possible reason for why PTCH1 loss-of-function mutations (versus those in other HH pathway components) are common in sporadic tumors: two effects for the price of one component. Nevertheless, given that the absence of Shh can induce PTCH1-mediated apoptosis, why does the activation of SHH pathway components downstream of PTCH1 [including SMOH (25, 26)] in the apparent absence of SHH result in tumorigenesis? Can these components talk back to PTCH1 or prevent PTCH1-mediated apoptosis? Or is there a tonic level of SHH required to inhibit PTCH1-induced apoptosis and allow cancer growth?


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
View Abstract

Stay Connected to Science

Navigate This Article