Introduction

In the famous screen of Nüsslein-Volhard and Wieschaus for genes governing Drosophila embryonic patterning, hedgehog (hh) and patched (ptc) were initially, shown to be involved in setting up the segmental boundary pattern of the fly embryo1. The hh gene encodes a secreted ligand that can bind to its receptor complex composed of the transmembrane proteins on the surface of the target cells encoded by ptc and smoothened (smo) genes2. This binding releases the suppression of PTC towards a transcription factor encoded by cubitus interruptus (ci, a homolog of the mammalian Gli1 gene) through an intracellular signal transduction cascade, thus allowing several important genes including wingless (wg, a homolog of the mammalian Wnt genes), decapentaplegic (dpp, a member of the TGFb family), and ptc itself to be transcribed. The activation of these genes then specifies the fate of the target cells.

Studies in the past several years have revealed that the hh/ptc signaling pathway is conserved in a number of postembryonic developmental processes in Drosophila and embryonic development in vertebrates. For example, during Drosophila metamorphosis, hh and ptc act to control cell fate decisions for patterning the central nervous system3, legs, wings, (reviewed in4, 5) and eyes6. In vertebrates, the hh/ptc pathway is involved in neurogenesis and limb pattering7, 8. More recently, the hh/ptc pathway has also been shown to play a crucial role in regulating stem cell division and differentiation in adult Drosophila and mammalian epithelia9, 10, 11, 12, even though they appear to act on a different set of downstream genes. The hh/ptc signaling cascade in non-stem cells, especially with regard to components downstream of ptc inside the target cells has been reviewed in detail by Ruiz i Altaba13. This paper reviews the novel function of hh/ptc in controlling stem cell divisions, with emphasis on the unique features of the hh/ptc pathway in this process.

hh/ptc pathway controls epithelial cell proliferation and differentiation in Drosophila oogenesis

Drosophila oogenesis begins at the apical tip of the ovary in 15-17 functionally equivalent structures called the ovarioles (Fig 1a). Within each ovariole, oogenesis starts at the apical tip in a specialized structure called the germarium in which germline stem cells produce germline cysts. Each cyst then becomes enveloped by a layer of somatic epithelial cells called the follicle cells to form an egg chamber, which leaves the germarium and moves towards the oviduct as it develops further. To encapsulate new germline cysts, the somatic stem cells in the middle region of the germarium must undergo active divisions to generate a continuous supply of follicle cells14 (Fig 1b). Recent work by Forbes et al. 9, 10 revealed that the division and differentiation of the somatic stem cells and their daughter cells are controlled by a group of non-mitotic somatic cells at the very tip of the germarium called the terminal filament cells through the hh/ptc signaling pathway.

Fig. 1
figure 1

a. A schematic drawing of the Drosophila ovariole, which is composed of a string of egg chambers (EC) produced by the germarium (G).

b. A magnified view of the apical half of the germarium as boxed in panel a. hh is expressed in terminal filament cells (TF), while ptc is expressed in inner sheath cells (IC), somatic stem cells (SSC), and follicle cells (FC). GSC: germline stem cells; Cyst: germline cyst. The germline is shaded.

Full size image

Forbes et al. 9, 10 first demonstrated a requirement for hh in regulating somatic cell proliferation and differentiation during oogenesis by showing that hh is specifically expressed in the terminal filament cells. A hh enhancer trap line specifically stains the terminal filament; so did anti-HH antibodies. Functional evidence for hh requirement comes from examining the effect of removing or ectopically expressing hh in the germarium. When females homozygous for a temperature-sensitive hh mutation were raised at a permissive temperature to adulthood and then shifted to a restrictive temperature to remove hh activity in the ovary, this shift caused the severe underproliferation of follicle cells. In contrast, when hh was ectopically expressed as driven by a hsp70 promoter, this overexpression resulted in an increase in the number of follicle cells. These experiments were the first to establish that the hh signaling pathway plays a role in oogenesis and epithelial cell proliferation.

Given that the epithelium is a stem cell-derived tissue, hh may control the process by regulating stem cell division. This appears to be true since timed ectopic hh expression experiments revealed that hh acts on somatic stem cells and their immediate daughters9.

Since hh was known to interact with ptc in other Drosophila developmental processes, the function of ptc and other segment polarity genes in oogenesis were also tested10. Both the ptc-LacZ reporter and the anti-PTC antibody staining experiments showed that ptc indeed is expressed in a pattern complementary to the hh expression pattern in the ovary. Loss of ptc function in clones of somatic stem cells and/or their daughter cells caused these cells to overproliferate, mimicking the effects of hh-overexpression. This suggests that ptc acts as a receptor molecule for the hh signal in controlling somatic stem cell division during oogenesis (Fig 2). Consistent with this, ectopic expression of ptc in the ovary showed no discernible effect on oogenesis, since the PTC protein is already present in the somatic stem cells and their immediate daughter cells without ectopic expression. These experiments provide substantial evidence that the functional relationship between hh and ptc is conserved in the epithelial stem cell system in the Drosophila ovary. These results also suggest that downstream targets of the ptc signal are insensitive to the concentration of PTC protein as long as it is above the threshold level, yet complete lack of PTC would fail to repress downstream target genes.

Fig. 2
figure 2

A simplified diagram of the hh/ptc signaling pathway in regulating epithelial stem cell division in the Drosophila ovary. This pathway appears to function also in vertebrate epithelia.

Full size image

Perhaps the most unexpected aspect of the hh/ptc pathway in the Drosophila epithelial stem cell system is the finding that dpp and wg are not the downstream effectors of the pathway as is the case in other developmental processes examined so far 10, Fig 2. Although the ci expression suggests that it is under the control of the hh/ptc pathway, its target genes no longer include wg and dpp. It has been well documented in the Drosophila embryo and imaginal disc systems that wg and dpp are the effector genes of the hh/ptc signaling pathway, serving as the direct targets of the CI transcription factor. The finding that the epithelial stem cells do not require wg and dpp [10, H.L., unpublished results] suggests that there must be other effector genes more specific or relevant to the stem cell system. Thus, studies by Forbes et al. have shown not only the involvement of the hh/ptc pathway in controlling stem cell proliferation but also the divergence of downstream effector genes in this process. This role of the hh/ptc pathway in Drosophila epithelial stem cell division parallels its role in mammalian epithelial development as revealed by recent findings in vertebrate systems.

hh/patched as regulators of cell proliferation in mammalian epithelia

The hh/ptc pathway has been characterized in vertebrate systems by isolating homologs, examining their expression profiles and manipulating their expression experimentally to examine the developmental consequences. For example, ectopic expression of Sonic hedgehog (Shh), a chick homolog of hh, caused mirror-like duplications in limb structures that paralleled the mutant phenotype in Drosophila8. For another example, ptc in mice was shown to be expressed in tissues tightly linked to those expressing Shh7. Mice deficient in Shh had severe defects in embryonic development such as neurogenesis, axial determination, skeleton formation and limb development15. These results, together with the apparent similarity at the amino acid sequence level, suggest structural and functional conservation of the hh/ptc pathway in governing differentiation and patterning during mammalian embryonic development.

Recently, a growing body of evidence shows that the Shh/ptc pathway also controls cell proliferation in vertebrate epithelia, a role similar to its function in controlling epithelial stem cell division during Drosophila oogenesis. Two recent reports11, 12 have shown that the human homolog of ptc, 96% and 40% amino acid identity to the mouse and Drosophila Ptc proteins respectively, is mutated in individuals with an autosomal dominant disorder called Nevoid Basal Cell Carcinoma Syndrome (NBCCS). Consistent with the multiple developmental phenotypes of Drosophila mutants, defects in human ptc include a number of developmental anomalies which define its role in embryonic patterning. In addition, NBCCS individuals are prone to develop basal cell carcinomas (BCCs), as well as the induction of several other classes of tumors, suggesting ptc plays an important role in controlling cell proliferation in the affected tissues. The BCCs originate from pluripotent basal skin cells and tumors are associated with loss of heterozygosity at the human ptc locus. Since pluripotent basal skin cells by definition should be epidermal stem cells, the overproliferation of these cells in BCCs in ptc-deficient patients implies a role for ptc as a negative regulator of epidermal stem cell division as shown in Drosophila.

Further support for the hh/ptc pathway in regulating basal cell division was provided by overexpressing Shh in transgenic mice bearing a Shh transgene driven by the keratin 14 promoter. This ectopic expression was sufficient to induce BCClike skin lesions in mice, thereby mimicking the loss of ptc function16.

It is not surprising that several downstream targets of the hh/ptc signaling pathway in Drosophila are themselves agents in tumorigenesis when mutated in vertebrates. Perhaps the most illustrative case is the role Gli1, a vertebrate homolog of ci, in incidences of various cancers including BCCs (reviewed in13). In all Drosophila developmental systems investigated, the transcription factor ci has been shown to be the downstream target of ptc signaling. In human, Glil was identified as a protooncogene causing incidences of glioblastomas (reviewed in17). Ectopic expression of Gli1 in frog embryos has recently been shown to cause the induction of BCCs, suggesting that Gli1 is potentially under the control of the hh/ptc signal in regulating basal stem cell division. This hypothesis is consistent with the observation that in every case of human skin tumors with Shh expression, Gli1 is always expressed18.

In addition to its role in BCCs, ptc has been shown to be mutated in 2 out of 14 sporadic medulloblastomas and in 2 out of 7 incidences of breast carcinomas examined19. Also, activation of Wnt-2 and Wnt-3 has been correlated with tumor formation in the mouse mammary gland as shown by expression profiles and by mutant analysis20. Given that breast carcinomas and mammary gland tumors are of epithelial origin, it is likely that ptc and Wnt genes are involved in controlling epithelial stem cell divisions in this organ. Particularly, the involvement of Wnt-2 and Wnt-3 genes in this process suggests that other wg homologs in Drosophila, instead of wg itself, may be involved in regulating ovarian epithelial stem cell division. One future direction may be to examine which of these Wnt genes are actual targets of ptc regulation in Drosophila and mammals.

Although the use of Drosophila to dissect the molecular mechanisms of tumorigenesis is hindered by the short life cycle of the fly, screens for recessive mutations that cause overproliferation have been successfully conducted21. Genetic methods have been developed to screen specifically for tumor suppressors in the fly using Flp/FRT mediated genetic mosaics22. Given the known conservation between Drosophila and mammalian hh/ptc pathways as well as many other developmental pathways, the fly remains an effective model system for exploring the nature of stem cell division and overproliferation.