Wilms’ tumor (nephroblastoma) represents a unique example of an aberrant kidney formation that can result from mutations in a tumor suppressor gene, Wilms’ tumor 1 (WT1). Targeted gene inactivation in mice testifies that WT1 is a master switch for the development of the genitourinary system and other organs.
Wilms’ Tumor: The Result of Disrupted Kidney Development
In 1899, Max Wilms made a significant discovery when he first described a young patient with malignant neoplasm of the kidney. Wilms’ tumor (nephroblastoma) affects 1:10,000 children and accounts for ~8% of all pediatric malignancies (14). Although ~80% of these tumors can nowadays be successfully treated, their fascinating biology still attracts scientists from various disciplines. Wilms’ tumor is thought to occur when pluripotent mesenchymal cells in the developing kidney fail to differentiate into glomeruli and tubules but continue to proliferate instead (3, 14). Incomplete kidney formation in the tumors is reflected in their characteristic histology, which seems to mimic the stages of normal renal development (3).
Inactivating mutations in a tumor suppressor gene, Wilms’ tumor 1 (WT1), are responsible for 10–15% of the neoplasms (14). Although nephroblastoma may also develop in response to mutations at other chromosomal sites, WT1 on human chromosome 11p13 is so far the only gene that has been cloned and proven to inhibit tumor growth. WT1’s unique role during organ formation, particularly development of the genitourinary system and mesothelial tissues, sets it apart from other tumor suppressors. In this short review article, we will discuss some of the important functions of WT1 in development. Its main focus is on recently discovered aspects of WT1 in the formation and maintenance of the kidneys and heart. For a broader discussion of the topic, the reader is referred to more comprehensive publications (23, 37).
WT1 Gene and Proteins
The human WT1 gene spans ~50 kb and consists of 10 exons (5, 10) (FIGURE 1⇓). It encodes a protein that shares a high degree of structural homology with the early growth response family of transcription factors (35). The WT1 gene product contains four COOH-terminal C2H2 zinc fingers for nucleic acid binding (34). Its NH2 terminus includes both transcriptional repression and activation domains (FIGURE 1⇓). Additional motifs in the WT1 protein are essential for self-association, nuclear localization, and RNA recognition (FIGURE 1⇓). More than 20 different WT1 gene products with molecular masses of 52–65 kDa are generated by a combination of alternative mRNA splicing (12), initiation of translation at variable start codons (4), and RNA editing (38). Among them, alternatively spliced exon 5 encodes 17 amino acids at a site NH2-terminal of the zinc finger domain (12). A second splicing event, which involves the use of two alternative splice donor sites at the end of exon 9, leads to the insertion/omission of three amino acids (lysine, threonine, and serine; KTS) between zinc fingers 3 and 4 of the WT1 molecule (12) (FIGURE 1⇓). The corresponding proteins, which are designated as WT1(−KTS) and WT1(+KTS), respectively, differ in their DNA binding site selectivity. Computer modeling (18) and in vitro studies (6) revealed a higher affinity for RNA of the +KTS proteins compared with the −KTS forms. Furthermore, the WT1(+KTS) products colocalized with and bound to nuclear splicing factors (7, 9, 22). These findings strongly support the possibility that the WT1(+KTS) proteins play a role in mRNA splicing rather than transcriptional control.
The ratio of the +KTS and −KTS proteins is conserved among tissues (12), and imbalanced expression of both isoforms will lead to developmental abnormalities. In humans, a ~50% reduction of the WT1(+KTS) levels due to heterozygous point mutation in a splice donor site in intron 9 was associated with developmental defects known as Frasier syndrome (2). Frasier syndrome is characterized by severe glomerulopathy of the kidneys and male-to-female sex reversal (female external genitalia, streak gonads, XY karyotype). In an attempt to define the molecular function of the WT1(+KTS) protein, transgenic mice were generated that only expressed the −KTS product. Animals with selective lack of the WT1(+KTS) form displayed a phenotype reminiscent of Frasier syndrome in humans (13). Malformations, i.e., hypoplastic kidneys and streak gonads, were even more severe in the Wt1(−KTS)-deficient mice (13). Hence, the functional complexity of WT1 appears to be determined by the generation of multiple protein forms. It remains a challenging task to elucidate the physiological function of each of the different WT1 gene products and their combined action in development and disease.
WT1 is required for the development of the kidneys
The critical role of WT1 in kidney development was revealed by the characteristic phenotype of mice with targeted inactivation of the Wt1 gene. Depending on the genetic background, mice with a homozygous Wt1−/− defect were lethal between embryonic day 12 and the end of gestation (15, 20). A predominant feature of the mutant embryos was their lack of kidneys and gonads (20). Further defects were found in mesothelial tissues (i.e., diaphragm, peritoneum, epicardium), heart (20, 27), adrenal glands (27), and spleen (15). Recent studies showed that WT1 is also necessary for the development of neuronal tissues, specifically of the vertebrate retina (43). Important insights into the function of WT1 in the central nervous system will come from future investigations aimed at identifying molecular WT1 downstream target genes in neurons.
The striking phenotype of Wt1-deficient mice has directed the attention of many scientists to the role of WT1 in renal development. The mammalian kidney is formed by the reciprocal interaction of two cellular compartments, the mesodermal mesenchyme and the ureteric bud, which is an outgrowth of the Wolffian duct (reviewed in Ref. 21) (FIGURE 2⇓). The growth and invasion of the ureteric bud is induced by signals, i.e., the glial-derived neurotrophic factor and others, which originate from the metanephric mesenchyme (reviewed in Ref. 21). Once it has invaded the mesenchyme, the ureteric bud branches into the collecting duct tree and triggers the condensation of mesenchymal cells (FIGURE 2⇓). Subsequently, the metanephric condensates transform to the comma- and S-shaped bodies, which connect to the collecting ducts and eventually give rise to nephrons, the functional excretory units of the kidneys. WT1 expression rises in the metanephric mesenchyme as it undergoes epithelial transition (1, 33) (FIGURE 2⇓). Interestingly, the mesenchyme becomes apoptotic and the outgrowth of the ureteric bud does not occur in Wt1−/− embryos (20). Failure of branching of the ureteric bud and enhanced mesenchymal cell death was also reported in a recent study using short interfering RNA technology to knock down WT1 expression in renal organ cultures (8). In the same experimental setting it was shown that repression of WT1 at later phases mimicked certain aspects of Wilms’ tumor formation in that it prevented nephron development and caused abnormal proliferation of mesenchymal cells (8). Thus WT1 seems to fulfill important functions during different phases of kidney development. Initially it may rescue the metanephric mesenchyme from apoptosis, and it may inhibit mesenchymal cell proliferation at later stages.
The molecular pathways through which WT1 exerts its effects in the developing kidney are not completely understood. Among the genes that are transcriptionally activated by WT1(–KTS) is Amphiregulin. Amphiregulin encodes a secreted member of the EGF family, which stimulates branching of the ureteric bud in renal organ cultures (24). The close relationship between amphiregulin and WT1 is reflected in the overlapping expression pattern of both proteins in the developing kidney (24). However, targeted inactivation of the Amphiregulin gene in mice did not impair normal kidney formation, indicating that lack of amphiregulin can be overcome by other, yet-unidentified factors (24).
In addition to locally secreted growth factors, condensation of the metanephric mesenchyme requires the transfer of signals from the extracellular space into the cytosol. The cell adhesion protein E-cadherin is an important molecule in this process. Expression of E-cadherin, like that of WT1, is induced in the condensing metanephric mesenchyme (16). Furthermore, WT1 enhanced transcription from the E-cadherin promoter in cotransfection experiments (16). These findings raise the interesting possibility that the E-cadherin gene belongs to the growing number of transcriptional targets of WT1. Mutations in the E-cadherin gene were found in malignant tumors from various tissues, including breast, kidney, lung, prostate, and colon cancer (reviewed in Ref. 45). In addition to its role in development, E-cadherin is thought to reduce tumor invasiveness through increasing cell adhesion and inhibition of cell proliferation (reviewed in Ref. 45). It is therefore conceivable that E-cadherin could mediate at least in part the tumor-suppressing actions of WT1.
In addition to the developing kidney, programmed cell death occurs in most tissues of Wt1-deficient mice that would normally express Wt1, i.e., the gonads (20), spleen (15), adrenal glands (27), and retina (43). Thus WT1 may generally prevent apoptotic cell loss in newly formed tissues. In support of this idea, WT1 activated transcription of the Bcl-2 gene, which encodes a major antiapoptotic protein (26). WT1-expressing cells with elevated Bcl-2 levels were resistant to a variety of proapoptotic stimuli (26). Additional evidence that Bcl-2 may act as a downstream mediator of WT1 came from in vivo studies with Bcl-2 knockout mice. These animals had a reduced kidney mass with fewer nephrons, and their metanephric tissue was more susceptible to apoptosis than that of wild-type mice (29, 41). Thus Bcl-2 appears to function as one of the signals that are induced by WT1 and mediate its antiapoptotic action in the developing kidney and other organs.
WT1 is Necessary for Normal Kidney Function
In the adult kidney, WT1 expression is restricted to the glomerular podocytes (28). Podocytes are highly differentiated cells on the outer surface of the basal membrane of the glomerular capillaries. The cells lend mechanical support to the capillary network and participate in the control of glomerular ultrafiltration (reviewed in Ref. 32). Podocyte damage likely is a key event during progression to chronic glomerular disease. Several lines of evidence suggest that WT1 is important for normal podocyte function. Firstly, mutations in the WT1 gene occur in >90% of patients with Denys-Drash syndrome (DDS), who consistently develop glomerulosclerosis. Secondly, introducing a zinc finger truncation into the murine Wt1 gene caused severe nephropathy similar to that seen in DDS patients (31). Thirdly, WT1 mutations have been identified in patients with nephrotic syndrome and single cases of glomerulosclerosis (17). Finally, reduced expression levels of Wt1 in renal podocytes of transgenic mice caused crescentic glomer-ulonephritis and mesangial sclerosis (11). Interestingly, the membrane proteins nephrin and podocalyxin were also reduced in the podocytes of the transgenic animals, suggesting coregulation of the three proteins (11). Nephrin is the major component of the slit diaphragm, which spans between the foot processes of the podocytes and forms the molecular barrier for glomerular ultrafiltration. Mutations in the NPHS1 gene, encoding nephrin, are associated with congenital nephrotic syndrome of the Finnish type and other glomerular diseases (19). The integral membrane protein podocalyxin connects to the cytoskeleton of the podocytes and is implicated in maintaining the complex three-dimensional shape of the cells (40). WT1 has been demonstrated to activate transcription of the podocalyxin gene (30). It remains to be established whether NPHS1 is also a physiological downstream target of WT1. Unraveling the molecular signaling pathways of WT1 in renal podocytes will hopefully provide novel insights into the pathophysiology of chronic glomerular disease.
Wt1 is Required for Normal Formation of the Heart
Mouse embryos with a homozygous Wt1 defect presumably die from heart failure (20). Their hearts were markedly reduced in size and displayed extreme thinning of the muscular walls (20, 27). Transmural bleeding into the pericardial cavity due to rupture of the epicardium was observed in the Wt1−/− mutants (20, 27). Importantly, WT1 is absent from the cardiac myocytes but is expressed in the epicardium. The epicardium is formed by a sheet of epithelial cells that cover the outer surface of the heart. Therefore, the failure of the Wt1−/− hearts to develop normally is likely due to disruption of signals derived from the epicardium rather than reflecting a cell-autonomous growth defect of the cardiac myocytes.
The epicardium is a mesothelial tissue that originates from the septum transversum of the proepicardial organ (reviewed in Ref. 25). Its importance for normal cardiac development has long been ignored. Studies with knockout mice showed that formation of the epicardium requires Wt1 (20). During development of the heart, Wt1 is initially detected in the proepicardial mesenchymal villi on the cranial surface of the septum transversum (27). Subsequently, Wt1-positive cells cross the pericardial cavity and spread over the surface of the myocardium to form the epicardial layer. Recent studies performed on chick heart slices testified that trophic signals from the epicardium are required for continued cardiac myocyte proliferation and survival (39). Selective inhibition of retinoic acid and erythropoietin signaling in the epicardium inhibited the proliferation of cardiac myocytes in the cultures (39). Considering these findings in the light of the cardiac phenotype of the Wt1-deficient embryos, one is tempted to speculate that WT1 may activate yet-unknown mitogenic factor(s) in the epicardium to promote the proliferation of cardiac myocytes (FIGURE 3⇓).
Another interesting hypothesis for the possible function of WT1 during cardiac development is based on the finding that the epicardium provides the source for coronary vascular cells (FIGURE 3⇑). Between embryonic days 11.5 and 12.5 in the developing mouse embryo, Wt1-positive cells delaminate from the epicardium into the subepicardial zone, where they form a layer of subepicardial mesenchymal cells (SEMCs) (27). Wt1 is thought to play a crucial role in this process by allowing epicardial cells to undergo epithelium-to-mesenchyme transition. This idea was derived from the characteristic phenotype of the Wt1−/− mutants, which lack most of their SEMCs (27). In wild-type mice around embryonic day 12.5, Wt1-positive cells migrate from the SEMC zone into the myocardium. Here they give rise to coronary vascular smooth muscle cells, perivascular and intermyocardial fibroblasts, and presumably also vascular endothelial cells (reviewed in Ref. 25) (FIGURE 3⇑). Wt1 expression is switched off once these cells have become fully differentiated. Thus any failure of the epicardial cells to undergo transition from epithelium to mesenchyme will ultimately lead to defects in coronary vascular development. That blood vessels were absent from the subepicardial zone in Wt1−/− mouse embryos (27) supports this view. It will be an exciting task to define exactly the developmental steps during which WT1 becomes important for vasculogenesis in the heart. Potential sites of action of WT1 include the detachment of epicardial cells from the epithelial layer, epithelium-to-mesenchyme transition, migration of SEMCs into the myocardium, and the formation of cell-cell contacts between SEMCs (FIGURE 3⇑). Several approaches can be taken to further explore the molecular functions of WT1 during cardiac morphogenesis. For example, the fate of the Wt1-positive cells in the embryonic heart can be followed with the use of recently established transgenic mouse lines, which express a lacZ reporter under control of the human WT1 locus (27). Transgene expression in these lines matched the spatiotemporal distribution of the endogenous Wt1 in the developing heart of wild-type embryos (27). Furthermore, identification of downstream targets of WT1 in the embryonic heart will provide important clues to the molecular signaling pathways through which Wt1 exerts its effects during cardiac morphogenesis.
The formation of new blood vessels is critical not only for normal heart development but also for myocardial tissue repair, i.e., after infarction due to coronary artery occlusion. Recent evidence suggests that the molecular mechanisms for coronary vascular formation are similar during fetal and adult life. As such they rely on the activation of a vasculogenesis program, which is controlled by a variety of genes and secreted growth factors. To explore whether WT1 is also involved in the vasculogenic response to myocardial ischemia, permanent occlusion of the left coronary artery was performed in rats. Remarkably, this manipulation caused de novo expression of Wt1 in vascular endothelial and smooth muscle cells in proximity to the necrosis area (42). Induction of Wt1 in myocardial vessels was associated with upregulation of the vasculogenesis-associated proteins CD31/platelet-endothelial cell adhesion molecule-1 and VEGF (42). Ischemic stimulation of Wt1 in the myocardial vessels could be mimicked by exposure of rats at low ambient oxygen (8% inspiratory O2) through a mechanism that involved hypoxia-inducible factor-1 (44). These findings suggest that WT1 represents a novel member of the family of proteins that control the genetic program for the formation of new blood vessels in both the developing and the ischemic heart.
Over the past few years it has become clear that WT1, in addition to its tumor suppressor action, has multiple roles during development. The functional complexity of WT1 is apparently due at least in part to the variety of proteins that are generated from a single gene. To fully understand the physiology of WT1, an effort must be made to obtain a more complete picture of the target genes that are regulated by each of the different WT1 molecules. This is of particular importance for the WT1(+KTS) isoforms, which have been implicated in mRNA processing and for which bona fide candidate targets are still unknown. Once the downstream mediators of WT1 have been characterized by a combination of biochemistry and cell biology, the physiological relevance of each of the identified candidates needs to be analyzed in vivo. This aim will be accomplished in transgenic animals, preferably with the use of tissue-specific and inducible gene-targeting strategies. Replacing Wt1 with the suspected target gene, and analyzing whether at least some of the phenotypic changes are rescued by this approach, will provide important novel insights into the molecular functions of WT1 during development and disease.
We thank Christian Bauer for his critical reading of the manuscript and Olivia Kaferly for her help with editing.
We apologize to all colleagues whose work could not be cited due to space limitations.
We appreciate the continued support by the Deutsche Forschungsgemeinschaft.
- © 2005 Int. Union Physiol. Sci./Am. Physiol. Soc.