sign in sign up
<<
Home , Nuclear receptor, Transcription factor, WT1

J Am Soc Nephrol 11: S106–S115, 2000 Wilms’ Tumor Suppressor WT1: From Structure to Renal Pathophysiologic Features

CHRISTIAN MROWKA*† and ANDREAS SCHEDL* *Max Delbru¨ck Center for Molecular Medicine and †Franz Volhard Clinic, Humboldt University of Berlin, Berlin-Buch, Germany.

Abstract. Normal development of the is a highly com- current understanding of the structural features of WT1, how plex process that requires precise orchestration of proliferation, they modulate the transcriptional and post-transcriptional ac- differentiation, and . In the past few years, a number tivities of the , and how affecting individual of that regulate these processes, and hence play pivotal isoforms can lead to diseased kidneys is summarized. In addi- roles in kidney development, have been identified. The Wilms’ tion, results from transgenic experiments, which have yielded WT1 has been shown to be one of these important findings regarding the function of WT1 in vivo, are essential regulators of kidney development, and mutations in discussed. Finally, data on the unusual feature of RNA editing this gene result in the formation of tumors and developmental of WT1 transcripts are presented, and the relevance of RNA abnormalities such as the Denys-Drash and Frasier syndromes. editing for the normal functioning of the WT1 protein in the A fascinating aspect of the WT1 gene is the multitude of kidney is discussed. isoforms produced from its genomic . In this review, our

The kidney is crucial for survival. The molecular mechanisms Structural Features underlying the development of this organ, however, are only The human WT1 gene spans approximately 50 kb and in- beginning to be elucidated. Three sets of kidneys develop cludes 10 exons, which generate a 3-kb mRNA (3,4). Four during embryogenesis, i.e., the pronephros, the mesonephros, finger motifs in the carboxyl-terminal portion form the DNA- and the metanephros (1). The precursors of the permanent binding domain and share with the early growth kidney in the metanephros result from reciprocal interactions response gene 1 family, suggesting a role for the WT1 protein between the epithelial ureteric bud and the pluripotent meta- as a factor (Figure 1A). Up to 24 different iso- nephric mesenchyme. The mesenchyme induces the bud to forms may result from a combination of alternative transla- grow and branch, thus forming the ureter, renal pelvis, and tional start sites, alternative RNA splicing, and RNA editing. collecting ducts. The ureteric bud induces the mesenchyme to One alternative translational start site (CTG) is located 204 bp undergo an epithelial transformation, to condense and form upstream of the major ATG site, creating an isoform with 68 (via the comma- and S-shaped bodies) the mature nephron with additional amino acids at its amino terminus (5). Recently, the glomerulus, proximal convoluted tubule, loop of Henle, shorter versions of WT1, resulting from initiation at and distal convoluted tubule. In the past decade, several genes the second in-frame ATG, have also been detected (6). Two have been identified that demonstrate spatially and temporally alternatively spliced exons are known, one inserting or exclud- distinct expression patterns during kidney development and ing exon 5, which encodes 17 amino acids, and the other lead to abnormal organ development when disrupted by gene- affecting exon 9 via insertion or exclusion of three amino acids targeting experiments (for review, see reference (2). One of [lysine-threonine-serine (KTS)] between the third and forth these genes is the Wilms’ tumor suppressor gene WT1. This zinc fingers. The ratio of splice variants is highly conserved in review focuses on structural features of and how they relate to normal fetal kidney and is maintained throughout development (7). Finally, an RNA-editing site has been identified in tran- WT1 function during kidney development. This is particularly scripts of rat and human WT1 genes, leading to the replacement relevant because we want to understand the basis of WT1 of an in exon 6 (8). A more detailed description of mutations related to human renal diseases, such as the Denys- RNA editing is provided below. Drash syndrome (DDS) and the Frasier syndrome (FS). It is now well established that WT1 can bind, via its zinc fingers, to the regions of Ͼ20 putative downstream target genes (for review, see reference (9). WT1 effects on transcription can be either repressing or activating, depending Correspondence to Dr. Andreas Schedl, Max Delbru¨ck Center for Molecular on the type and the target gene with which WT1 interacts. Medicine, Robert-Ro¨ssle Str. 10, D-13092 Berlin-Buch, Germany. Phone: 49-30-9406-2337; Fax: 49-30-9406-2110; E-mail: [email protected] Most target genes have been identified in transfection assays, 1046-6673/1111-0106 and data should be interpreted with care, because the effects of Journal of the American Society of Nephrology or other interacting are usually neglected in Copyright © 2000 by the American Society of Nephrology vitro. However, in vitro experiments have been corroborated J Am Soc Nephrol 11: S106–S115, 2000 WT1, Glomerulopathy, and RNA Editing S107

Figure 1. (A) Structural features of the Wilms’ tumor suppressor gene WT1. The schematic drawing depicts the main features of the WT1 protein and its various isoforms (numbers in parentheses refer to references). (B) Frasier syndrome (FS). The schematic drawing indicates the most common intronic mutations affecting lysine-threonine-serine (KTS) splicing. aa, amino acid; PU, proteinuria; NS, nephrotic syndrome; ESRF, end-stage renal failure. by in vivo data for an increasing number of target genes that are no-terminally shorter versions seem to lead to reduced repres- either activated or repressed by WT1. These genes include sion but increased activation potential (6). How important those for the cell surface epidermal these minor products are for WT1 function in vivo remains receptor (10), the paired-box-containing unclear. However, alternatively spliced exon 5 has been sug- gene 2 (PAX2) (11), and the growth factor -like growth gested to increase the repressing effect of WT1 on some factor-2 (12) as targets downregulated by WT1. Amphiregulin, promoters (16). Interestingly, the presence of the alternative a member of the epidermal growth factor family, has been exon 5 has been found only in mammalian species (14,15), shown to be a target that can be activated by WT1 (13). indicating that it has been adopted by WT1 at a later stage in Because WT1 is expressed in so many different isoforms, . which are highly conserved throughout evolution (14,15), what Increasing amounts of compelling data suggest a role for properties do the different isoforms have? The addition of a WT1 not only as a transcription factor but also as a potent 68-amino acid amino-terminal region from an alternative trans- modulator at the post-transcriptional level. WT1 is able to bind lational start site has been demonstrated to have little effect on to RNA of the insulin-like growth factor-2 gene (17). How the the transcriptional activity of the protein (5). In contrast, ami- WT1 protein interacts with RNA is still under discussion. S108 Journal of the American Society of Nephrology J Am Soc Nephrol 11: S106–S115, 2000

Although one study reported that the region (in type has not yet been observed) and early bilateral nephrec- particular, zinc finger 1) is necessary for these protein-RNA tomy before the development of Wilms’ tumor are -saving interactions (17), computer modeling suggests that an RNA treatments. However, a late manifestation of Wilms’ tumor recognition motif in the amino-terminal region is important after renal transplantation in a child with retrospectively sus- (18). These two regions may be equally important, with the pected DDS has been observed (26). amino-terminal region serving as a nonspecific binding domain More than 60 germline mutations (both familial and de and the zinc finger region providing sequence specificity. novo) have been described, which essentially leave the WT1 Interestingly, the distinct isoforms of WT1 seem to behave protein intact but affect the DNA-binding zinc finger domain differently, in terms of their nucleic acid-binding capacity. (27) (for review, see reference (28). Most of these mutations Whereas variants lacking the KTS sequence (ϪKTS) bind to are missense mutations within exons 8 and 9, which code for DNA sequences and can act as potent transcriptional regula- zinger finger domains 2 and 3, respectively. A hot tors, KTS-containing isoforms (ϩKTS) demonstrate higher spot seems to be present at nucleotide 1180 in codon 394, affinities for RNA. The actual function of the interaction of leading to the replacement of arginine with tryptophan WT1 with RNA is still unknown; it could stabilize or destabi- (R394W). Only a few deletions, insertions, and nonsense mu- lize certain RNA forms or could be involved in post-transcrip- tations result in truncated proteins. All mutations, however, tional modifications such as RNA editing or splicing. Evidence alter the structure of the DNA-binding domain, thus changing for the latter is derived from immunocolocalization and immu- its ability to bind to both DNA and RNA. The severe pheno- noprecipitation experiments. In transfection studies, the ϩKTS type produced by only one mutated allele raises several ques- isoforms seem to preferentially colocalize with splicing fac- tions. Does the mutation act in a dominant way with new tors, whereas ϪKTS forms exhibit more diffuse nuclear local- devastating effects? Or does the mutated protein act in a ization similar to that of transcription factors (19). Further- dominant-negative manner, actively suppressing and inactivat- more, direct interaction of WT1 with the splicing machinery of ing the influence of the wild-type allele? Evidence for the latter the nucleus and particularly the splicing factor U2AF65 has hypothesis has been obtained from both in vitro and in vivo recently been demonstrated by two-hybrid analysis and bio- analyses. First, WT1 is able to homodimerize, and a nonfunc- chemical studies (20,21). Despite these exciting and provoca- tional allele may thus render some functional WT1 products tive data, the function of these isoforms in vivo is still unclear. (from the wild-type allele) nonfunctional (29). This could explain the much more severe defects observed for patients WT1 Mutations and Implications for Human with DDS, compared with patients with heterozygous deletions Renal Disease and Tumorigenesis of WT1. Second, gene-targeting experiments with a truncation of WT1 in the zinc finger region demonstrated a The identification of WT1 mutations in human patients has with some of the characteristics of DDS (30). Interestingly, provided us with important clues regarding the cellular and expression from the mutated allele accounted for only 5% of developmental functions of this gene. A variety of WT1 mu- the total amount of WT1 protein, suggesting that even minor tations, which either affect development or induce tumor for- amounts of mutant protein can have devastating effects on mation, have been identified. Developmental defects include urogenital development. the dominant Wilms’ tumor/aniridia/genital anomaly/mental retardation syndrome, DDS, and FS, whereas tumors include nephroblastoma (Wilms’ tumor), , breast , Frasier Syndrome aggressive desmoplastic small round cell tumor, and . FS includes a slowly progressing nephrotic syndrome, at- For reasons of space, we restrict our discussion to diseases tributable to minimal glomerular changes or focal segmental affecting the urogenital system. glomerulosclerosis, and complete male to female gender rever- sal in 46,XY patients but no signs of Wilms’ tumor (31). Denys-Drash Syndrome Gonadal development in XX female patients is normal. In The DDS is a rare congenital childhood syndrome that contrast to DDS, end-stage renal disease develops more slowly includes diffuse mesangial sclerosis, severe hypertension, a and at a later stage in life (for review, see reference (28). The -resistant nephrotic syndrome that rapidly progresses to late onset of renal disease in some patients with FS suggests end-stage renal disease before the age of 5 yr, male pseudoherm- that WT1 is also important for normal functioning of the aphroditism, and a high risk of developing Wilms’ tumor kidney after development has been completed. (22,23). In addition to these symptoms, incomplete forms with Four similar point mutations downstream of the second renal involvement and either varying genital anomalies or splice donor site in intron 9 have been detected in FS (32) Wilms’ tumor have been described (24). Podocytes exhibit (Figure 1B). In vitro experiments have shown that the intronic decreased or absent WT1 expression but persistent PAX2 mutations interfere with recognition of the second splice donor expression (25). Genital anomalies in 46,XY patients with site and result in loss of the ϩKTS isoform from the mutated DDS vary from the total absence of epithelial structures, with allele. Interestingly, FS mutations are dominant, and the wild- “streak” gonads, to the presence of both the wolffian (male) type allele still produces both the ϩKTS and ϪKTS isoforms. and mullerian (female) ducts or female gonads and genitalia. Therefore, a mere change of the ratio of the two isoforms must Kidney transplantation (recurrence of the glomerular pheno- be responsible for the severe developmental defects, empha- J Am Soc Nephrol 11: S106–S115, 2000 WT1, Glomerulopathy, and RNA Editing S109 sizing the importance of the ϩKTS isoform for urogenital mental abnormalities, including aniridia, genital anomalies, development. and mental retardation (40,41). Although abnormalities of the Recently, a mutation in exon 9 that did not alter the isoform urogenital system can be attributed to the of WT1, ratio in two patients classified as having FS was described (33). other abnormalities are attributable to the deletion of additional In addition, intronic mutations characteristic of FS have been genes mapping to this region, including PAX6 and reticulocal- observed in patients with DDS. Furthermore, a patient with the bindin (42,43). A 25-bp intragenic deletion in a sporadic classic FS mutation but an unexpected Wilms’ tumor has been Wilms’ tumor (44) and a germline intragenic deletion in a described (34). These recent data fuelled a controversy regard- patient with bilateral nephroblastoma (45) provided the first ing the correct classification of DDS and FS. The majority of solid evidence that WT1 is indeed the long-sought tumor sup- published data clearly distinguish between the two syndromes, pressor gene. WT1 mutations, including deletions, truncations, on both clinical and molecular biologic grounds. However, FS translocations, and missense mutations, were observed in ap- could also be an atypical subtype of DDS, because of the proximately 20% of Wilms’ tumors. The cause in the other variety of incomplete syndromes and overlapping mutational 80% of cases is unclear but may involve mutations in genes characteristics. The very similar in FS and DDS, acting upstream or downstream of WT1. together with the fact that patients with FS do not produce abnormal WT1 proteins, support the hypothesis that DDS is Role of WT1 during Kidney Development attributable to dominant-negative mutations, rather than a new The expression pattern of WT1 during embryogenesis is function of the mutated protein. Classification of the two highly complex (46,47). The initial differentiation of the meta- syndromes should therefore be performed at the molecular nephric mesenchyme seems to be independent of WT1 (48), level, rather than on the basis of the observed phenotype. and the gene is only weakly expressed in the uncondensed metanephric blastema (49). Expression dramatically increases Isolated Diffuse Mesangial Sclerosis (IDMS) and during the mesenchyme-to- conversion in the con- Idiopathic Persistent Nephrotic Syndrome densing mesenchyme, when the renal vesicles and comma- In rare familial cases, only the renal phenotype of either shaped bodies are formed (Figure 2, A and B). Similar mes- DDS (diffuse mesangial sclerosis) or FS (focal segmental enchyme-epithelium interactions have been observed in a glomerulosclerosis) is observed. Additional symptoms of the variety of different developing organs (50). is diseases, including abnormalities of the gonads, are not found highest in the proximal part of the S-shaped body, which (35). IDMS has been shown to be associated with de novo flattens to form the glomerular podocytes. It has been reported mutations of both exon 8 and exon 9 (24), whereas the most that WT1 activity is essential for the switching of cells between frequent DDS mutation (R394W) has not been detected. In mesenchymal and epithelial cell states. Terminally differenti- addition, some patients with IDMS do not present with WT1 ated cells, such as epithelial tubular cells, exhibit no WT1 mutations at all (36). Although the description of FS has been expression, whereas cells with epithelial/mesenchymal switch- restricted to XY female patients, the FS mutation can also be ing potential, such as podocytes, continue to produce WT1 detected in 46,XX patients. These female patients exhibit nor- (51,52). In the fully developed kidney, WT1 expression persists mal development of the genital system, whereas kidney biop- in the podocytes and, at a much lower level, in epithelial cells sies reveal segmental glomerular sclerosis (37,38). From a of the Bowman’s capsule, suggesting that the gene may play a nephrologic point of view, one should search for both intronic role in more than the initial stages of kidney development. In and exonic mutations of WT1, to assess the risk of inherited addition to the kidney, high levels of WT1 expression are found diseases of the kidney or gonads among the children of pa- in the spleen (53), the , and the genital ridges that tients. This is particularly important for female patients with develop into testes or , depending on the presence or focal segmental glomerulosclerosis or karyotypic 46,XY male absence of the Y (54). patients with a family history of therapy-resistant nephrotic The complex pattern of WT1 expression suggests that the syndrome. The classification of these renal diseases is difficult, function of WT1 is required at multiple stages during kidney because of the heterogeneity of observed phenotypes and the development, and data obtained using genetically modified different mutations of the WT1 gene. mice are beginning to support this hypothesis. In knockout mice, the ureteric bud fails to grow out and the metanephric Wilms’ Tumor blastema undergoes apoptosis (55). Interestingly, apoptosis Wilms’ tumor (nephroblastoma) is a childhood tumor of the occurs even when blastema from knockout is recom- kidney (1:10,000 live births) originating from the metanephric bined with ureteric buds from wild-type animals in organ blastema. Nephrogenic rests, caused by the developmental culture experiments. These data suggest that WT1 has at least arrest of nephrogenesis, are thought to be precursors of Wilms’ two functions during this first stage of kidney development. tumor formation (39). The WT1 gene has been proven to play First, it may be required for the inductive signaling that induces an essential role in the pathogenesis of this tumor. Heterozy- the outgrowth of the ureter from the mesonephros. Second, it gous deletions on chromosome 11p13 in human subjects are seems to be involved in either survival or the reception of the associated with the congenital Wilms’ tumor/aniridia/genital survival signal from the ureteric bud. anomaly/mental retardation syndrome, which consists of a high Recent transgenic experiments from our own laboratory risk of developing Wilms’ tumor and a combination of develop- have characterized a second phase of WT1 action (56). We S110 Journal of the American Society of Nephrology J Am Soc Nephrol 11: S106–S115, 2000

Figure 2. Immunohistochemical analysis of Wilms’ tumor suppressor gene 1 (WT1) (A and B) and paired-box-containing gene 2 (PAX2) (C and D) (red) and synaptopodin (A to D) (green) during kidney development (embryonic day 17.5). (A) Evidence that WT1 is expressed in the condensing blastema (arrowhead), the proximal part of S-shaped bodies (arrows), and the podocyte layer of functional glomeruli (asterisks). (B) High-power view of a glomerulus, showing nuclear staining for WT1 in podocytes and cytoplasmic staining for podocyte-specific synaptopodin. (C) Evidence of PAX2 expression in the collecting ducts (arrows) and condensing blastema (arrowheads) but not in mature glomeruli (asterisks). (D) High-power view of the reciprocal interactions between the branching ureteric bud (UB) and the condensing metanephric blastema (arrowhead), both of which express PAX2. Sections in A, C, and D were counterstained with diaminophenylindole (blue). Magnifications: ϫ200 in A and C; ϫ630 in B and D. examined whether the human WT1 locus is able to complement analyses revealed varying degrees of complementation, i.e.,a the knockout mutation in mice, which are known to die at complete absence of ureteric budding, outgrowth of the ureter embryonic day 13.5 because of multiple organ defects, with without further branching, or normal development of the neph- renal and gonadal agenesis. To ensure proper expression of all rogenic zone with nephrogenesis to the stage of comma-shaped isoforms as well as correct transcriptional regulation of the bodies. Mature glomeruli, however, were never formed. Be- WT1 gene, we decided to introduce the entire genomic locus on cause the experiments were not performed with a defined a 280-kb yeast artificial chromosome (YAC). Interestingly, the genetic background, the varying degrees of kidney develop- transgene only partially rescued the knockout phenotype when ment may be attributable to modifier genes. The defects in it was crossed into the WT1 knockout background. The trans- nephron formation led to increased apoptosis and overall re- gene complementation, however, led to unexpected postnatal duced kidney size (Figure 3). Interestingly, after the ureteric death, within 48 h, because of kidney failure. Histologic bud invasion, condensation of the metanephric blastema J Am Soc Nephrol 11: S106–S115, 2000 WT1, Glomerulopathy, and RNA Editing S111

Figure 3. Complementation analysis of the WT1 knockout phenotype. WT1 knockout mice complemented with a human, 280-kb, yeast artificial chromosome (C to E) have smaller kidneys, with increased amounts of stroma (S), compared with kidneys of wild-type animals (A and B). Branching of the ureteric bud (UB) and condensation (arrowheads) occur normally (E), but no functional glomeruli (asterisk; compare with B) are observed. Magnifications: ϫ40 in A and C; ϫ200 in B and D; ϫ400 in E. seemed to occur normally (Figure 3E), suggesting that WT1 is cytes) and mature glomeruli (Figure 2C), simultaneously with not required for this first step of nephron formation. Taken increased WT1 expression, suggesting that PAX2 may be a target together, these data indicate a continuous requirement for WT1 of WT1. Indeed, binding assays and cotransfection studies dem- during nephrogenesis. onstrated that WT1 can repress PAX2 transcription by binding to As illustrated in our rescue experiments, a multitude of regulatory sequences in the promoter region of PAX2 (61). Inter- positive and mechanisms operate during the estingly, the converse, i.e., regulation of the WT1 promoter by formation of an organ such as the kidney. For a deeper under- PAX2, also seems to occur. Two PAX2-binding sites have been standing of development, it is important to examine these identified in the upstream region of WT1 (62). Indeed, transfection interactions on a molecular level. With respect to WT1, the experiments result in high levels of transcriptional activation of transcription factor PAX2 is particularly interesting. PAX2 is an WT1 by PAX2 (63). Taken together, these data suggest that the evolutionarily highly conserved member of the paired box gene two genes are targets of each other. It is presently unclear whether family (57) that plays a pivotal role during the development of the down-regulation of PAX2 is necessary for proper kidney the kidney, central nervous system, and sensory organs. Ectopic development and is exclusively mediated by WT1. From our expression of PAX2 is associated with deregulated proliferation rescue experiments, however, it seems unlikely that PAX2 is the during nephrogenesis (58), whereas PAX2 null mutants lack kid- only target of WT1 during nephron formation, because a lack of neys, ureters, and gonads (59,60). PAX2 is down-regulated in PAX2 repression (resulting in ectopic expression) is unlikely to precursor cells of the visceral glomerular epithelium (prepodo- cause the observed complete absence of nephrons. S112 Journal of the American Society of Nephrology J Am Soc Nephrol 11: S106–S115, 2000

The incomplete kidney development in our rescued animals is somewhat difficult to understand but could be explained by insufficient levels of WT1 expression. Indeed, RNA analysis indicated three- to fivefold lower levels of transgene expres- sion, compared with the endogenous WT1 gene. In addition, deregulation of post-transcriptional modifications may contrib- ute to the insufficient action of the WT1 transgene. The ratio of the four different mRNA splice forms, however, was identical to the wild-type situation. This prompted us to investigate whether the partial rescue of the urogenital system in WT1 null mice with the WT1 YAC transgenic lines could be attributed to deregulated RNA editing.

New Evaluation of RNA Editing of WT1 RNA editing is a post-transcriptional processing event that generates a new transcript with nucleotides that do not match the bases present in the original genomic sequence. In mam- mals, direct nucleotide modifications via nucleotide deamina- tion (-to- or adenosine-to-inosine conversion) Figure 4. Single-strand conformation polymorphism analysis to detect are the most frequently observed events (for review, see ref- RNA editing. Total RNA was isolated from snap-frozen kidneys erence (64). A different editing mechanism has been described [kidneys from adult female (MK) and fetal (f MK) (embryonic day for WT1 RNA, in which a uridine is converted to a cytidine at 9.5) NMRI mice and a human cadaveric kidney from a hypertensive adult patient (HK)]. cDNA was prepared using 1 ␮g of RNA, 25 units nucleotide 839, resulting in the replacement of leucine by of avian myeloblastosis virus reverse transcriptase (Boehringer Mann- (leucine/proline dimorphism). This process is thought heim) in 10ϫ reverse transcription buffer, 20 units of RNAsin, 100 to be developmentally regulated and occurs in adults rather ␮M dNTP, and 100 ␮M random hexanucleotide primers, in a total than neonates (8). On a functional level, the proline-containing volume of 20 ␮l. The reaction was performed at 42°C for 1 h and at edited isoform seems to be a less repressive transcriptional 95°C for 7 min. The forward primer for PCR was placed at the end of molecule for growth-promoting genes, compared with the non- exon 5 (5Ј-CCACGGTATAGGGTACGAGA-3Ј). We chose a reverse edited isoform. The physiologic consequences, however, have primer located at the beginning of exon 7 (5Ј-CAGATACACGCCG- not yet been clarified. CACATC-3Ј). Although identical in length (116 bp), the PCR prod- Because RNA editing has not yet been described in mice, we ucts varied at single nucleotides between the different species, result- first attempted to detect edited forms of WT1 mRNA in wild- ing in differently moving bands in single-strand conformation type mice, using tissues from rats and a human patient as polymorphism analyses. The PCR products were diluted (1:1) with stop solution (deionized 95% formamide, 0.05% bromphenol blue, 20 positive control samples. We used single-strand conformation mM ethylenediaminetetraacetate), denatured at 95°C for 5 min, kept polymorphism analysis to detect the edited RNA, which should on ice for 3 min, and immediately loaded on a 0.5ϫ Hydrolink-MDE alter the mobility of single-stranded DNA. In the case of polyacrylamide gel (BioWhittaker Molecular Application, Rockland, editing, the PCR product would contain two slowly moving ME), at 3-W constant wattage, for 12 h at 4°C. The gel was stained Ϯ-strands of nonedited T/A839 DNA and two strands of the using a silver staining . Only two bands of the nonedited form were faster moving edited C/G839 DNA (see the legend to Figure 4 detected in all samples. for experimental details). Surprisingly, kidneys from mice, rats, and a human patient revealed only the nonedited RNA form and not the four strands expected for edited RNA samples lung, we detected edited RNA in 2 of 30 clones, whereas all (Figure 4) (data for Sprague-Dawley, Wistar-Kyoto, and spon- other organs exhibited only nonedited forms (Figure 5C). taneously hypertensive rats are not shown). The results of our As expected from the sequence conservation among differ- analysis were in marked contrast to those of earlier studies, in ent species, we were able to demonstrate RNA editing also in which RNA editing was observed in 30% (adult or tumorous the kidney of mice, suggesting that RNA editing is relevant for kidney) to 90% (Rat 2 cell line) of mRNA (8). Could it be that biologic processes. However, the low prevalence in adult mice, our experimental design overlooks a small but relevant amount as well as in “control-species” rats and hypertensive patients, is of edited RNA? To exclude this possibility, we changed our in striking contrast to the published data. The biologic effects strategy to a PCR-based subcloning method. As shown in of RNA editing of the WT1 gene might be more variable than Figure 5, most colonies harbored the nonedited form, with expected. RNA editing has not been observed in 15 primary three bands after MnlI digestion (Figure 5, A and B). Only Wilms’ tumors (65) or in the partial rescue of the kidney 0.4% of the clones from adult mice (1 of 220 colonies) and phenotype with the WT1-YAC transgenic line in WT1 knock- 0.7% of those from rats (1 of 126 colonies) exhibited the edited out mice. To explain these conflicting results, subtle differ- form in the kidney (Figure 5C). In addition, we examined 30 ences in the genetic background, nutritional and gender influ- colonies from four different adult Sprague-Dawley rats (kidney ences, or circadian effects (day/night course) might be for all four rats, liver for two rats, and lung for one rat). In rat considered. J Am Soc Nephrol 11: S106–S115, 2000 WT1, Glomerulopathy, and RNA Editing S113

This will enable us to investigate the effects of altered isoform ratios or the developmental and physiologic functions of indi- vidual isoforms, by crossing animals with the heterozygous mutation to homozygosity. This will allow us to study the association of individual isoforms with components of the nuclear machinery, such as transcription and splicing factors, in a natural environment. Is an altered WT1 isoform ratio responsible for gender determination, and could decreased gene expression be associated with abnormal genital develop- ment? Does a DDS mutation in exon 8 affect genital develop- ment more severely than does a mutation in exon 9, as recently suggested (24)? More insight into molecular mechanisms will certainly permit better classification of the different syn- dromes. WT1 has been shown to control cellular proliferation and mesenchyme-to-epithelium transitions. It is tempting to spec- ulate on whether WT1 makes cells of mesodermal origin susceptible to inductive signals. It will be exciting to further investigate the function of deregulated WT1 activity by ectopi- cally expressing the gene under the control of a renal tissue- specific promoter. What will happen if WT1 is constantly expressed in cells that normally do not exhibit gene activity after differentiation? Will these cells return to their origins, undergo apoptosis, or proliferate into tumor cells? What is the possible role of WT1 in fully differentiated cells, such as Sertoli or granulosa cells? Expression of PAX2 under the control of WT1 will allow us to gain insight into the regulation and function of podocytes. It is fascinating to speculate on whether PAX2 overexpression can inactivate WT1 function or may be associated with the loss of WT1 activity and cell at certain stages of nephrogenesis. These questions are awaiting answers from studies with transgenic models, as expected soon. Such studies can ultimately Figure 5. Colony PCR and MnlI digestion of PCR products. The PCR help us determine the appropriate conditions for gene therapy product of the reverse-transcribed RNA (Figure 4) was subcloned for patients with deregulated WT1 gene function. (TOPO TA cloning kit; Invitrogen, San Diego, CA). (A) The 116-bp PCR product contains either two MnlI restriction sites (nonedited CTC form, three resulting fragments) or one MnlI site (edited CCC References form, two fragments) because of the T839/C839 conversion, resulting 1. Saxen L: Organogenesis of the kidney. In: Developmental and in loss of one restriction site. (B) MnlI digestion of the PCR product Cell Series, 19th Ed., edited by Barlow PW, Green PB, from the majority of colonies revealed the nonedited form. (C) The White CC, Cambridge, Cambridge University Press, 1987, pp prevalence of RNA editing in a and four different rats 1–71 (Sprague-Dawley strain; organ numbers refer to the same animals) is 2. Lipschutz JH: Molecular development of the kidney: A review of shown. the results of gene disruption studies. Am J Kidney Dis 31: 383–397, 1998 3. Call KM, Glaser T, Ito CY, Buckler AJ, Pelletier J, Haber DA, Rose EA, Kral A, Yeger H, Lewis WH, Jones C, Housman DE: Summary and Perspectives Isolation and characterization of a zinc finger polypeptide gene at Research in the past decade has proven that WT1 has a the human Wilms’ tumor locus. Cell 60: 509– critical role in proper kidney development. Disrupted gene 520, 1990 expression can lead to developmental abnormalities, as well as 4. Gessler M, Poustka A, Cavenee W, Neve RL, Orkin SH, Bruns tumor induction. Target genes for ϪKTS isoforms have been GAP: Homozygous deletion in Wilms tumours of a zinc-finger gene identified by chromosome jumping. Nature (Lond) 343: identified, but what is the role of other WT1 variants? The 774–778, 1990 functions of the various splice variants, the regulation of tran- 5. Bruening W, Pelletier J: A non-AUG translational initiation scription, and splicing of the gene or RNA editing are far from event generates novel WT1 isoforms. J Biol Chem 271: 8646– being understood. It will be fascinating to generate mouse 8654, 1996 models for DDS by inserting the most common point mutation 6. Scharnhorst V, Dekker P, Van Der Eb AJ, Jochemsen AG: or to mimic the intronic KTS splice mutation observed in FS. Internal translation initiation generates novel WT1 protein iso- S114 Journal of the American Society of Nephrology J Am Soc Nephrol 11: S106–S115, 2000

forms with distinct biological properties. J Biol Chem 274: nephropathie parenchymateuse et d’un mosaicisme XX/XY. 23456–23462, 1999 Arch Fr Pedriatr 24: 729–739, 1967 7. Haber DA, Sohn RL, Buckler AJ, Pelletier J, Call KM, Housman 23. Drash A, Sherman F, Hartmann W, Blizzard RMA: A syndrome DE: and genomic structure of the Wilms of pseudohermaphroditism, Wilms’ tumor, hypertension, and tumor gene WT. Proc Natl Acad Sci USA 88: 9618–9622, 1991 degenerative renal disease. J Pediatr 76: 585–593, 1970 8. Sharma PM, Bowman M, Madden SL, Rauscher FJ III, Sukumar 24. Jeanpierre C, Denamur E, Henry I, Cabanis MO, Luce S, Cecille S: RNA editing in the Wilms’ tumor susceptibility gene, WT1. A, Elion J, Peuchmaur M, Loirat C, Niaudet P, Gubler MC, Genes Dev 8: 720–731, 1994 Junien C: Identification of constitutional WT1 mutations, in 9. Menke A, McInnes L, Hastie ND, Schedl A: The Wilms’ tumor patients with isolated diffuse mesangial sclerosis, and analysis of suppressor WT1: Approaches to gene function. Kidney Int 53: genotype/phenotype correlations by use of a computerized mu- 1512–1518, 1998 tation database. Am J Hum Genet 62: 824–833, 1998 10. Englert C, Hou X, Maheswaran S, Bennett P, Ngwu C, Re GG, 25. Yang Y, Jeanpierre C, Dressler GR, Lacoste M, Niaudet P, Garvin AJ, Rosner MR, Haber DA: WT1 suppresses synthesis of Gubler MC: WT1 and PAX-2 podocyte expression in Denys- the epidermal growth factor receptor and induces apoptosis. Drash syndrome and isolated diffuse mesangial sclerosis. Am J EMBO J 14: 4662–4675, 1995 Pathol 154: 181–192, 1999 11. Dressler GR, Deutsch U, Chowdhury K, Nornes HO, Gruss P: 26. Cleper R, Davidovitz M, Krause I, Bar Nathan N, Ash S, Pax2, a new murine paired-box-containing gene and its expres- Schwarz M, Mor C, Eisentstein B: Unexpected Wilm’s tumor in sion in the developing excretory system. Development 109: 787– a pediatric renal transplant recipient: Suspected Denys-Drash 795, 1990 syndrome. Transplant Proc 31: 1907–1909, 1999 12. Drummond IA, Madden SL, Rohwer Nutter P, Bell GI, 27. Pelletier J, Bruening W, Kashtan CE, Mauer SM, Manivel JC, Sukhatme VP, Rauscher FJ: Repression of the insulin-like Striegel JE, Houghton DC, Junien C, Habib R, Fouser L, Fine growth factor II gene by the Wilms tumor suppressor WT1. RN, Silvermann BL, Haber DA, Housmann D: Germline muta- Science (Washington DC) 257: 674–678, 1992 tions in the Wilms’ tumor suppressor gene are associated with 13. Lee SB, Huang K, Palmer R, Truong VB, Herzlinger D, Kolquist abnormal urogenital development in Denys-Drash syndrome. KA, Wong J, Paulding C, Yoon SK, Gerald W, Oliner JD, Haber Cell 67: 437–447, 1991 28. Gubler MC, Yang Y, Jeanpierre C, Barbaux S, Niaudet P: WT1: DA: The Wilms tumor suppressor WT1 encodes a transcriptional Renal development and glomerulopathies. Adv Nephrol Necker of amphiregulin. Cell 98: 663–673, 1999 Hosp 29: 299–315, 1999 14. Kent J, Coriat AM, Sharpe PT, Hastie ND, van Heyningen V: 29. Moffett P, Bruening W, Nakagama H, Bardeesy N, Housman D, The evolution of WT1 sequence and expression pattern in the Housman DE, Pelletier J: Antagonism of WT1 activity by protein vertebrates. 11: 1781–1792, 1995 self-association. Proc Natl Acad Sci USA 92: 11105–11109, 15. Miles C, Elgar G, Coles E, Kleinjan DJ, van Heyningen V, 1995 Hastie ND: Complete of the Fugu WAGR region 30. Patek CE, Little MH, Fleming S, Miles C, Charlieu JP, Clarke from WT1 to PAX6: Dramatic compaction and conservation of AR, Miyagawa K, Christie S, Doig J, Harrison DJ, Porteus DJ, synteny with human chromosome 11p13. Proc Natl Acad Sci Brookes AJ, Hooper ML, Hastie ND: A zinc finger truncation of USA 95: 13068–13072, 1998 murine WT1 results in the characteristic urogenital abnormalities 16. Hewitt SM, Fraizer GC, Wu YJ, Rauscher FJ, Saunders GF: of Denys-Drash syndrome. Proc Natl Acad Sci USA 96: 2931– Differential function of Wilms’ tumor gene WT1 splice isoforms 2936, 1999 in transcriptional regulation. J Biol Chem 271: 8588–8592, 1996 31. Frasier S, Bashore RA, Mosier HD: Gonadoblastoma associated 17. Caricasole A, Duarte A, Larsson SH, Hastie ND, Little M, with pure in monozygotic twins. J Pediatr Holmes G, Todorov I, Ward A: RNA binding by the Wilms 64: 740–745, 1964 tumor suppressor zinc finger proteins. Proc Natl Acad Sci USA 32. Barbaux S, Niaudet P, Gubler MC, Grunfeld JP, Jaubert F, 93: 7562–7566, 1996 Kuttenn F, Fekete CN, Souleyreau Therville N, Thibaud E, 18. Kennedy D, Ramsdale T, Mattick J, Little M: An RNA recog- Fellous M, McElreavey K: Donor splice-site mutations in WT1 nition motif in Wilms’ tumour protein (WT1) revealed by struc- are responsible for Frasier syndrome. Nat Genet 17: 467–470, tural modelling. Nat Genet 12: 329–331, 1996 1997 19. Larsson SH, Charlieu JP, Miyagawa K, Engelkamp D, Rassou- 33. Kohsaka T, Tagawa M, Takekoshi Y, Yanagisawa H, Tadokoro lzadegan M, Ross A, Cuzin F, van Heyningen V, Hastie ND: K, Yamada M: Exon 9 mutations in the WT1 gene, without Subnuclear localization of WT1 in splicing or transcription factor influencing KTS splice isoforms, are also responsible for Frasier domains is regulated by alternative splicing. Cell 81: 391–401, syndrome. Hum Mutat 14: 466–470, 1999 1995 34. Barbosa AS, Hadjiathanasiou CG, Theodoridis C, Papathanasiou 20. Davies RC, Calvio C, Bratt E, Larsson SH, Lamond AI, Hastie A, Tar A, Merksz M, Gyorvari B, Sultan C, Dumas R, Jaubert F, ND: WT1 interacts with the splicing factor U2AF65 in an iso- Niaudet P, Moreira CA, Cotinot C, Fellous M: The same muta- form-dependent manner and can be incorporated into spliceo- tion affecting the splicing of WT1 gene is present in Frasier somes. Genes Dev 12: 3217–3225, 1998 syndrome patients with or without Wilms’ tumor. Hum Mutat 13: 21. Ladomery MR, Slight J, McGhee S, Hastie ND: Presence of 146–153, 1999 WT1, the Wilms’ tumor suppressor , in nuclear 35. Habib R, Gubler MC, Antignac C: Diffuse mesangial sclerosis: ϩ poly(A) ribonucleoprotein. J Biol Chem 274: 36520–36526, A congenital glomerulopathy with nephrotic syndrome. Adv 1999 Nephrol Necker Hosp 22: 43–56, 1993 22. Denys P, Malvaux P, van den Berghe H, Tanghe W, Proesmans 36. Schumacher V, Scha¨rer K, Wu¨hl E, Altrogge H, Bonzel KE, W: Association d’un syndrome anatomo-pathologique de pseu- Guschmann M, Neuhaus TJ, Pollastro RM, Kuwert-Broking E, do-hermaphrodisme masculin, d’une tumeur de Wilms, d’une Billa M, Tondera AM, Mundel P, Helmchen U, Waldherr P, J Am Soc Nephrol 11: S106–S115, 2000 WT1, Glomerulopathy, and RNA Editing S115

Weirich A, Royer-Pokora B: Spectrum of early onset nephrotic mentally regulated transcription factor in the kidney that func- syndrome associated with WT1 missense mutations. Kidney Int tions as a tumor suppressor. FASEB J 7: 896–903, 1993 53: 1594–1600, 1998 53. Herzer U, Crocoll A, Barton D, Howells N, Englert C: The 37. Demmer L, Primack W, Loik V, Brown R, Therville N, McEl- Wilms tumor suppressor gene WT1 is required for development reavey K: Frasier syndrome: A cause of segmental glomerulo- of the spleen. Curr Biol 9: 837–840, 1999 sclerosis in a 46,XX female. J Am Soc Nephrol 10: 2215–2218, 54. Nachtigal MW, Hirokawa Y, Enyeart VanHouten DL, Flanagan 1999 JN, Hammer GD, Ingraham HA: Wilms’ tumor 1 and Dax-1 38. Klamt B, Koziell A, Poulat F: Frasier syndrome is caused by modulate the orphan SF-1 in sex-specific gene defective alternative splicing of WT1 leading to an altered ratio expression. Cell 93: 445–454, 1998 of WT1ϩ/ϪKTS splice isoforms. Hum Mol Genet 7: 709–714, 55. Kreidberg JA, Sariola H, Loring JM, Maeda M, Pelletier J, 1998 Housman D, Jaenisch R: WT-1 is required for early kidney 39. Park S, Bernard A, Bove KE, Sens DA, Hazen-Martin DJ, development. Cell 74: 679–691, 1993 Garvin AJ, Haber DA: Inactivation of WT1 in nephrogenic rests, 56. Moore AW, McInnes L, Kreidberg J, Hastie ND, Schedl A: YAC genetic precursors to Wilms’ tumour. Nat Genet 5: 363–367, complementation shows a requirement for Wt1 in the develop- 1993 ment of epicardium, and throughout nephrogen- 40. Riccardi VM, Sujansky E, Smith AC, Francke U: Chromosomal esis. Development 126: 1845–1857, 1999 imbalance in the aniridia-Wilms’ tumor association: 11p inter- 57. Gruss P, Walther C: Pax in development. Cell 69: 719–722, stitial deletion. Pediatrics 61: 604–610, 1978 1992 41. Rose EA, Glaser T, Jones C, Smith CL, Lewis WH, Call KM, 58. Dressler GR, Wilkinson JE, Rothenpieler UW, Patterson LT, Minden M, Champagne E, Bonetta L, Yeger H, Housman DE: Williams Simons L, Westphal H: Deregulation of Pax-2 expres- Complete physical map of the WAGR region of 11p13 localizes sion in transgenic mice generates severe kidney abnormalities. a candidate Wilms’ tumor gene. Cell 60: 495–508, 1990 Nature (Lond) 362: 65–67, 1993 42. Ton CC, Hirvonen H, Miwa H, Weil MM, Monaghan P, Jordan 59. Torres M, Gomez Pardo E, Dressler GR, Gruss P: Pax-2 controls T, van Heyningen V, Hastie ND, Meijers-Heijboer H, Drechsler multiple steps of urogenital development. Development 121: M, Royer-Pokora B, Collins F, Swaroop A, Strong LC, Saunders 4057–4065, 1995 GF: Positional cloning and characterisation of a paired box- and 60. Favor J, Sandulache R, Neuhauser Klaus A, Pretsch W, Chatter- -containing gene from the aniridia region. Cell 67: jee B, Senft E, Wurst W, Blanquet V, Grimes P, Sporle R, 1059–1074, 1991 Schughart K: The mouse Pax2 (1Neu) mutation is identical to a 43. Kent J, Lee M, Schedl A, Boyle S, Fantes J, Powell M, Rushmere human PAX2 mutation in a family with renal-coloboma syn- N, Abbott C, van Heyningen V, Bickmore WA: The reticulocal- drome and results in developmental defects of the , ear, eye, bindin gene maps to the WAGR region in human and to the small and kidney. Proc Natl Acad Sci USA 93: 13870–13875, 1996 eye Harwell deletion in mouse. 42: 260–267, 1997 61. Ryan G, Steele Perkins V, Morris JF, Rauscher FJ, Dressler GR: 44. Haber DA, Buckler AJ, Glaser T, Call KM, Pelletier J, Sohn RL, Repression of Pax-2 by WT1 during normal kidney development. Douglass EC, Housman DE: An internal deletion within an Development 121: 867–875, 1995 11p13 zinc finger gene contributes to the development of Wilms’ 62. McConnell MJ, Cunliffe HE, Chua LJ, Ward TA, Eccles MR: tumor. Cell 61: 1257–1269, 1990 Differential regulation of the human Wilms tumour suppressor 45. Huff V, Miwa H, Haber DA, Call KM, Housman D, Strong LC, gene (WT1) promoter by two isoforms of PAX2. Oncogene 14: Saunders GF: Evidence for WT1 as a Wilms tumor (WT) gene: 2689–2700, 1997 Intragenic germinal deletion in bilateral WT. Am J Hum Genet 63. Dehbi M, Ghahremani M, Lechner M, Dressler G, Pelletier J: 48: 997–1003, 1991 The paired-box transcription factor, PAX2, positively modulates 46. Pritchard Jones K, Fleming S, Davidson D, Bickmore W, Por- expression of the Wilms’ tumor suppressor gene (WT1). Onco- teous D, Gosden C, Bard J, Buckler A, Pelletier J, Housman D: gene 13: 447–453, 1996 The candidate Wilms’ tumour gene is involved in genitourinary 64. Niswender CM: Recent advances in mammalian RNA editing. development. Nature (Lond) 346: 194–197, 1990 Cell Mol Life Sci 54: 946–964, 1998 47. Buckler AJ, Pelletier J, Haber DA, Glaser T, Housman DE: 65. Gunning KB, Cohn SL, Tomlinson GE, Strong LC, Huff V: Isolation, characterization, and expression of the murine Wilms’ Analysis of possible WT1 RNA processing in primary Wilms tumor gene (WT1) during kidney development. Mol Cell Biol 11: tumors. Oncogene 13: 1179–1185, 1996 1707–1712, 1991 66. Bickmore WA, Oghene K, Little MH, Seawright A, van Hey- 48. Donovan MJ, Natoli TA, Sainio K, Amstutz A, Jaenisch R, ningen V, Hastie ND: Modulation of DNA binding specificity by Sariola H, Kreidberg JA: Initial differentiation of the metaneph- alternative splicing of the Wilms tumor gene transcript. ric mesenchyme is independent of WT1 and the ureteric bud. Dev Science (Washington DC) 257: 235–237, 1992 Genet 24: 252–262, 1999 67. Wang ZY, Qiu QQ, Gurrieri M, Huang J, Deuel TF: WT1, the 49. Armstrong JF, Pritchard Jones K, Bickmore WA, Hastie ND, Wilms’ tumor suppressor gene product, represses transcription Bard JB: The expression of the Wilms’ tumour gene, WT1, in the through an interactive nuclear protein. Oncogene 10: 1243–1247, developing mammalian embryo. Mech Dev 40: 85–97, 1993 1995 50. Birchmeier C, Meyer D, Riethmacher D: Factors controlling 68. Madden SL, Cook DM, Rauscher FJ: A structure-function anal- growth, motility, and of normal and malignant ysis of transcriptional repression mediated by the WT1, Wilms’ epithelial cells. Int Rev Cytol 160: 221–266, 1995 tumor suppressor protein. Oncogene 8: 1713–1720, 1993 51. Grubb GR, Yun K, Williams BR, Eccles MR, Reeve AE: Ex- 69. Reddy JC, Morris JC, Wang J, English MA, Haber DA, Shi Y, pression of WT1 protein in fetal kidneys and Wilms tumors. Lab Licht JD: WT1-mediated transcriptional activation is inhibited Invest 71: 472–479, 1994 by dominant negative mutant proteins. J Biol Chem 270: 10878– 52. Rauscher FJ: The WT1 Wilms tumor gene product: A develop- 10884, 1995