Human Molecular Genetics Advance Access originally published online on July 22, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Human Molecular Genetics, 2003, Vol. 12, No. 18 2379-2394
DOI: 10.1093/hmg/ddg240
© 2003 Oxford University Press
Murine Denys–Drash syndrome: evidence of podocyte de-differentiation and systemic mediation of glomerulosclerosis
1Sir Alastair Currie Cancer Research UK Laboratories, Molecular Medicine Centre, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK, 2Department of Pathology, University of Dundee, Ninewells Hospital, Dundee DD1 9SY, UK, 3MRC Human Genetics Unit, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK, 4Division of Pathology, The University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, UK and 5Molecular Physiology Laboratory, Wilkie Building, University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, UK
Received May 28, 2003; Revised June 19, 2003; Accepted July 10, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Denys–Drash syndrome (DDS) is caused by dominant mutations of the Wilms' tumour suppressor gene, WT1, and characterized by a nephropathy involving diffuse mesangial sclerosis, male pseudohermaphroditism and/or Wilms' tumourigenesis. Previously, we reported that heterozygosity for the Wt1tmT396 mutation induces DDS in heterozygous and chimeric (Wt1tmT396/+
+/+) mice. In the present study, the fate of Wt1 mutant cells in chimeric kidneys was assessed by in situ marker analysis, and immunocytochemistry was used to re-examine the claim that glomerulosclerosis (GS) is caused by loss of WT1 and persistent Pax-2 expression by podocytes. Wt1 mutant cells colonized glomeruli efficiently, including podocytes, but some sclerotic glomeruli contained no detectable Wt1 mutant cells. The development of GS was preceded by widespread loss of ZO-1 signal in podocytes (even in kidneys where <5% of glomeruli contained Wt1 mutant podocytes), increased intra-renal renin expression, and de novo podocyte TGF-ß1 expression, but not podocyte Pax-2 expression or loss of WT1, synaptopodin, 
-actinin-4 or nephrin expression. However, podocytes in partially sclerotic glomeruli that still expressed WT1 at high levels showed reduced vimentin expression, cell cycle re-entry, and re-expressed desmin, cytokeratin and Pax-2. The results suggest that: (i) GS is not due to loss of WT1 expression by podocytes; (ii) podocyte Pax-2 expression reflects re-expression rather than persistent expression, and is the consequence of GS; (iii) GS is mediated systemically and the mechanism involves activation of the renin–angiotensin system; and (iv) podocytes undergo typical maturational changes but subsequently de-differentiate and revert to an immature phenotype during disease progression. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The Wilms' tumour suppressor gene, WT1, plays a key role in epithelial differentiation and is expressed in mesodermal tissues that experience mesenchymal–epithelial transformation, which includes the developing kidney. WT1 encodes 24 isoforms due to RNA editing, alternative translation initiation, and alternative RNA splicing of exon 5 and the KTS insert (lysine, threonine and serine) located at the 3' end of exon 9, which may act as transcription factors, transcriptional cofactors and/or post-transcriptional regulators (1).
Kidney development is initiated by reciprocal interaction between the ureteric bud, an epithelial outgrowth of the Wolffian duct, and the pluripotent metanephric mesenchyme located at the caudal extremity of the nephrogenic cord. Following induction, the metanephric blastema undergoes a series of changes commencing with condensation; forming the renal vesicle, which matures further via the comma-shaped, S-shaped body and capillary loop stage into epithelial cells that form the glomerulus and proximal and distal tubules. WT1 is expressed at low levels in uninduced metanephric blastema and peaks in podocyte precursors in late S-shaped body stage, and expression studies, knock-out studies and YAC complementation experiments have established that WT1 is required for induction of the ureteric bud, for survival and epithelial differentation of the metanephric blastema, and has a role throughout the progression of nephrogenesis (2–4). Moreover, the continued expression of WT1 in terminally differentiated podocytes implies it also has a role in maintaining podocyte function. The finding that WT1 can promote epithelial differentiation in mesenchymal fibroblasts (5) is consistent with the view that WT1 plays a prominent role in mesenchymal–epithelial transformation. However, since WT1 is expressed by mesenchymal cells derived from the proliferating coelomic epithelium and by terminally differentiated podocytes which revert from a fully epithelial to partially mesenchymal phenotype, as evident by re-expression of vimentin and loss of cytokeratin (6), it has been proposed that WT1 might also promote epithelial–mesenchymal transformation and serve to switch cells between epithelial and mesenchymal states (4).
Constitutional WT1 mutations are present in WAGR syndrome (Wilms' tumour, aniridia, genitourinary defects and mental retardation), Denys–Drash syndrome (DDS) and Frasier syndrome (7,8). WAGR syndrome is due to hemizygous deletion of a chromosomal segment encompassing WT1 and is associated with a late-onset renal failure (9), and a link between WT1 haploinsufficency and GS has been recently established in mice (10). Frasier syndrome is caused by heterozygous intronic mutations that lead to a reduction in the +KTS/-KTS ratio and is characterized by an adolescent nephropathy involving focal mesangial sclerosis, male-to-female sex reversal and predisposition to gonadoblastoma. In contrast, DDS involves a severe early-onset nephropathy and is due to dominant intragenic WT1 mutations, which can be missense or nonsense, and which primarily affect the C-terminal zinc finger domain (7,8). The invariant feature of DDS is diffuse mesangial sclerosis which is a distinct form of glomerulopathy observed in patients presenting with congenital or infantile nephrotic syndrome, and is characterized by rapid progression of GS with end-stage renal failure before the age of 5 years, hypertension, thickening of the glomerular basement membrane (GBM), and podocyte hypertrophy and vacuolation. DDS can be associated with XY pseudohermaphroditism and/or predisposition to Wilms' tumourigenesis. However, the recent report that a Frasier mutation can induce diffuse mesangial sclerosis further supports the view that these syndromes share a degree of phenotypic overlap (7,8,11).
The investigation of DDS provides an important route to examine the role of WT1 in kidney function and disease and, to date, analyses of podocytes suggest that DDS is linked with altered expression of transcription factors, and changes in the protein composition of the slit diaphragm and glomerular basement membrane (GBM). For example, Yang et al. (12) reported that DDS and IDMS (isolated diffuse mesangial sclerosis) patients show a reduced level of the heparan sulphate chain of heparan sulfate proteoglycan (HSPG) in the GBM and speculated that this could be involved in the pathogenic mechanism leading to proteinuria in DDS since this proteoglycan has a role in the GBM charge-selective barrier. However, since loss of HSPG occurs in unrelated diseases, including adriamycin nephropathy, and can be prevented by angiotensin-converting enzyme (ACE) inhibitors (13) the reduced level of HSPG in DDS could be a consequence of kidney dysfunction. Podocalyxin is a major component of the glycocalyx and has a role in maintaining foot process architecture and glomerular permselectivity. Palmer et al. (14) reported that WT1 promotes podocalyxin expression in a rat kidney cell line, and that podocalyxin is expressed at low levels by podocytes in human DDS. However, it was subsequently reported that a DDS transgene (R362X) does not affect endogenous podocalyxin expression by mouse podocytes (15). Pax-2 is a member of the paired-box class of transcription factors, and plays a key role during the early stage of renal epithelial differentiation. Pax-2 expression is repressed in the proximal loop of the S-shaped body, where podocytes originate, and this is linked with high expression of WT1 (16). Yang et al. (17) reported that podocytes in DDS and IDMS patients show loss or reduced levels of WT1 and inappropriate Pax-2 expression, and suggested that GS may result from persistent Pax-2 expression due to failure of WT1 to repress its expression. While this is consistent with reports WT1 can repress Pax-2 promoter activity (18) and that Pax-2 over-expression in transgenic mice causes nephrotic syndrome (19), the changes in WT1 and Pax-2 were not representative of all patients examined and, as with podocalyxin and HSPG, it remains to be determined whether they precede the development of GS.
Since DDS patients can harbour WT1 mutations that truncate within the zinc finger (ZF) domain, including ZF3 (1,7,8), we previously used gene targeting in embryonic stem (ES) cells to generate a heterozygous Wt1tmT396 mutation that disrupts ZF3 and deletes the KTS insert and ZF4, and found it induces features characteristic of DDS in heterozygote (Wt1tmT396/+) and chimeric (Wt1tmT396/++/+) mice, including nephropathy, male genital defects and Wilms' tumourigenesis (20). The heterozygote suffered renal failure at 8 months and displayed diffuse mesangial sclerosis, interstitial fibrosis, hypertensive damage, tubule dilation and microcyst formation, and podocyte hypertrophy and loss of foot processes. Heterozygous and chimeric mice exhibited proteinuria (protein casts), and chimeras exhibited a focal and segmental mesangial sclerosis that developed between 1 and 6 months. In the present study these same mice were examined further to gain insight concerning the pathogenesis of DDS. Given that the Wt1tmT396 mutation is proven to induce nephropathy, chimeras were examined to address whether Wt1 mutant cells efficiently colonize the kidney, whether all podocytes in sclerotic glomeruli, or indeed all sclerotic glomeruli, harbour the Wt1 mutation, and whether the absence of GS in some adult chimeric DDS kidneys reflects the failure of Wt1 mutant cells to colonize glomeruli, and specifically podocytes. These issues were investigated by isozyme analysis and by in situ marking to identify the ES cell-derived Wt1 mutant cells. Using immunocytochemistry the study examined whether the development of GS is preceded by reduced WT1 signal in podocytes and whether inappropriate podocyte Pax-2 expression in DDS reflects re-expression or persistent expression, and is linked specifically with dominant Wt1 mutations. The study also addressed whether the dominant Wt1 mutation affects expression of components of the podocyte filtration barrier, including nephrin, ZO-1, -actinin-4 and synaptopodin, and whether GS is preceded by changes in intra-renal renin expression and in glomerular expression of growth factors and pro-fibrogenic cytokines, including IGF-II, EGFR, TGF-ß1 and PDGF-A, which are encoded by putative target genes regulated by WT1 (1).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Glucose phosphate isomerase (GPI) isozyme analysis
Chimeric mice (Wt1tmT396/+
+/+) were analysed by GPI analysis to determine whether Wt1 mutant (Wt1tmT396/+) cells colonize the kidney as efficiently as do wild-type cells. For comparison, testis and a range of non-urogenital tissues were analysed in parallel. Since the Wt1 mutant embryonic stem (ES) cells (strain 129/Ola) are homozygous for the Gpi-1a allele and express GPI-1A, and wild-type host blastocysts [F2 (C57BL/6LacxCBA/CaLac)] are homozygous for the Gpi-1b allele and express GPI-1B (21), the contribution of the two lineages to the tissues of chimeras was investigated by determining the relative proportions of the two isozymes (Fig. 1). Analyses of 42 adult chimeras in which the ES cell component was either heterozygous for the Wt1tmT396 mutation (n=34) or wild-type (n=8) found that Wt1 mutant cells colonize kidney cortex, testis, heart, epididymis, spleen, lung, liver, brain and gut, as efficiently as wild-type cells. However, Wt1 mutant cells less efficiently colonize skeletal muscle (P<0.01), but preferentially colonize the pancreas (P<0.001).
|
DNA–DNA in situ hybridization
DNA–DNA in situ hybridization was used to determine whether DDS mutant cells colonize all renal lineages efficiently. Since the chimeras were generated by injection of male ES cells into unsexed host blastocysts the XY component of XX
XY chimeric kidneys is ES cell-derived, and therefore Wt1 mutant cells can be positively identified using a Y chromosome probe. Cryostat sections were examined since unlike wax-embedded tissues their analysis does not require proteinase-K treatment that can compromise the identification of sclerotic glomeruli. Female chimeras were examined since they should be XXXY (21) and this was subsequently confirmed by DNA–DNA in situ hybridization: in XXXY kidneys only a proportion of cells (<50%) show characteristic pericentric Y chromosome signal compared with >85% seen routinely in control male kidney sections (Fig. 2A). The failure to see signal in every nucleus is not unexpected and reflects exclusion of all, or part of, the Y chromosome from the section. No signal was detected in wild-type female kidneys.
|
A total of 14 chimeric kidneys were examined and included sclerotic and non-sclerotic kidneys from 10 chimeras necropsied between 4 and 22 months. Wt1 mutant cells colonized all glomerular lineages in sclerotic glomeruli, including mesangial cells, podocytes and parietal epithelial cells of the Bowman's capsule (Fig. 2B). However, Y chromosome signal was detected in only a small proportion of podocytes and mesangial cells in some sclerotic glomeruli (Fig. 2C), indicative that these cell types are polyclonal in origin in individual glomeruli, and that not all podocytes need to harbour the Wt1 mutation for sclerosis to develop. Wt1 mutant cells were detected routinely in the interstitium and in proximal and distal tubular epithelium (Fig. 2D). Wt1 mutant cells were detected frequently in endothelial cells of the interlobular arteries but were only rarely present in the smooth muscle cells (Fig. 2E–G). While Wt1 mutant cells were detected routinely in sclerotic glomeruli some anomalies were evident. Firstly, no Y chromosome signal was detected in 5% of sclerotic glomeruli (Fig. 2H and I). While the presence of Wt1 mutant cells in these glomeruli outside the plane of section cannot be excluded, all glomeruli in wild-type male kidneys routinely showed Y chromosome signal. Secondly, non-sclerotic adult kidneys were identified with similar high levels of chimerism (22–43% GPI-1A) as sclerotic kidneys, and where >40% of glomeruli contained Wt1 mutant cells, including podocytes (Fig. 2J). Wt1 mutant cells were also detected in all cell lineages in non-sclerotic glomeruli from affected chimeric kidneys (Fig. 2K). Importantly, PCR analysis using primers that characterized the Wt1 mutation (20) confirmed its presence in four non-sclerotic chimeric kidneys at 17–22 months (data not shown).
Immunocytochemistry
Since only one heterozygote was generated, which was a sterile male (20), sclerotic and non-sclerotic chimeric DDS kidneys were analysed in parallel to confirm changes in protein expression in the heterozygote, and identify changes in protein expression that precede and may therefore promote sclerotic damage. Importantly, since chimeras display a focal GS their analysis enables direct quantitative comparisons of protein expression between adjacent sclerotic and non-sclerotic glomeruli in the same tissue section. The chimeric kidneys examined contained a similar high proportion (30–58%) of Wt1 mutant cells (as judged by GPI analysis of kidney cortex) and >40% of glomeruli contained Wt1 mutant podocytes (as determined by DNA–DNA in situ hybridization), and one kidney was included from at least four affected (with some sclerotic glomeruli) and four unaffected (no sclerotic glomeruli) chimeras aged 6–11 months. Age-matched wild-type kidneys and chimeric kidneys (24–42% GPI-1A) generated using wild-type ES cells served as controls. Since only weak immunostaining was generally detected in globally sclerotic glomeruli the staining patterns were recorded for partially sclerotic glomeruli (where <50% of the glomerular tuft was sclerotic) and for each antibody a minimum of 10 sections was scored in each kidney (Table 1).
|
WT1 and Pax-2 expression in murine DDS
The C-terminal (C-19) polyclonal antibody was used routinely since it recognizes wild-type WT1 protein that accounts for 95% of WT1 protein in cells heterozygous for the Wt1tmT396 mutation (20). In wild-type kidneys strong WT1 signal was detected routinely in podocytes with weaker signal in some parietal epithelial cells (Fig. 3A; see also Fig. 4D, E and G). While staining in parietal cells appeared to depend on the batch of C-19 antibody, studies with the N-terminal monoclonal antibody (6F-H2) confirmed they express WT1, albeit focally (data not shown). While no podocyte WT1 signal was detected in some globally sclerotic glomeruli (Fig. 3B), strong signal was routinely seen in podocytes from partially sclerotic glomeruli from chimeric (Fig. 3C) and heterozygous kidneys (Fig. 3D), and podocytes in adjacent partially sclerotic and non-sclerotic glomeruli routinely contained similar high levels of WT1 (Fig. 3E). Moreover, in cases of segmental sclerosis WT1 was still expressed strongly by podocytes in the affected segment (Fig. 3F), and all podocytes in non-sclerotic chimeric kidneys contained similar high levels of WT1 as in wild-type kidneys (data not shown) even though >40% of glomeruli contained some Wt1 mutant podocytes. This was also the case with the N-terminal antibody (6F-H2) that recognizes both wild-type and mutant WT1 proteins (Table 1).
|
|
In wild-type kidneys strong focal Pax-2 signal was present in parietal epithelial cells and in epithelial cells lining the distal and collecting tubules (Fig. 3G). No difference in Pax-2 expression, as regards the number of cells expressing Pax-2 or the signal intensity, was evident in cystic versus healthy tubules. Inappropriate podocyte Pax-2 expression was detected in 25–40% of partially sclerotic glomeruli from heterozygous and chimeric kidneys (Fig. 3H). However, only a proportion of podocytes expressed Pax-2 in partially sclerotic glomeruli from heterozygous kidneys (Fig. 3I and J), suggesting that Pax-2 is not expressed persistently. In contrast, podocyte Pax-2 expression was often absent in globally sclerotic glomeruli (Fig. 3K) and present in 3% of non-sclerotic glomeruli from affected chimeric kidneys (data not shown). Importantly, podocyte Pax-2 expression was detected in only a small minority of glomeruli in non-sclerotic kidneys (Fig. 3L, Table 1). Analysis of serial sections from chimeric DDS kidneys found that WT1 and Pax-2 were expressed strongly by podocytes in the same glomerulus (Fig. 3M and N).
WT1 and Pax-2 expression in hypertensive rat and heterozygous Wt1 null mouse kidneys
Since podocyte Pax-2 re-expression was prominent in sclerotic glomeruli it could be a consequence of sclerosis and unrelated to WT1. This possibility was addressed by examination of models of GS that harbour, respectively, no dominant Wt1 mutations (the heterozygous Wt1 null mouse) (10,22) and no Wt1 mutations at all (the TGR(mREN2)27 transgenic rat line that displays malignant hypertension due to overexpression of mouse Ren-2) (23). The partial penetrance of the heterozygous Wt1 null mutation enabled comparative analyses between 3–4-month-old mice that displayed either no proteinuria (n=3), proteinuria but no GS (n=3), or proteinuria and GS (n=3). Age-matched wild-type rat and mouse kidneys served as controls and these showed focal Pax-2 signal in parietal and tubular epithelial cells only (data not shown). Podocyte Pax-2 signal was detected in 20–30% of sclerotic glomeruli from hypertensive rat kidneys (Fig. 4A) and in 3% of glomeruli in kidneys from proteinuric heterozygous Wt1 null mice (Fig. 4B and C). The podocytes in these same kidneys expressed similar high levels of WT1 as in wild-type kidneys (Fig. 4D and E). No podocyte Pax-2 expression was detected in kidneys from non-proteinuric heterozygous Wt1 null mice (data not shown). The analysis of sclerotic kidneys from heterozygous Wt1 null mice was confounded by severe global sclerosis with advanced hyalinosis and loss of cellularity. Nevertheless, strong WT1 and Pax-2 signal was still detected in podocytes (Fig. 4F and G).
Ki67, desmin, cytokeratin and vimentin expression in murine DDS
Since Pax-2 is expressed by podocyte precursors during early nephrogenesis (16,18) its re-expression in sclerotic glomeruli could indicate that podocytes de-differentiate and revert to an immature phenotype during disease progression. This possibility was explored further by examination of vimentin, cytokeratin, desmin and Ki67 expression. Vimentin and desmin are intermediate filament proteins expressed by mesenchymal cells, and cytokeratins are epithelial proteins. During nephrogenesis, podocyte development is marked by loss of DNA synthesis and cytokeratin expression, and re-expression of vimentin (4,6). Desmin is a muscle-associated intermediate filament protein that can be regarded as a marker of early nephrogenesis since it is expressed by murine metanephric ridge cells in vitro (24) and by the blastemal component of human Wilms' tumours (25) that is believed to derive from the induced metanephric mesenchyme (1).
Gut sections served as a positive staining control for Ki67 (which identifies cells between the late-G1 and the G2 phases of the cell cycle), and showed the expected signal in intestinal crypts (Fig. 5A). No Ki67 signal was detected in wild-type adult kidneys but strong signal was detected in podocytes and parietal epithelial cells in 30–45% of partially sclerotic glomeruli from both heterozygous and chimeric kidneys (Fig. 5B). In contrast, podocyte Ki67 signal was rarely seen in non-sclerotic kidneys (Table 1). Ki67 signal was detected routinely in tubular epithelial cells but was only rarely present in mesangial cells of sclerotic glomeruli. In wild-type kidneys, no cytokeratin signal was detected in podocytes (Fig. 5C), whereas strong podocyte staining was detected in 40–55% of partially sclerotic glomeruli from heterozygous and chimeric kidneys (Fig. 5D), but in only a small minority of glomeruli in non-sclerotic kidneys (Table 1). All podocytes in wild-type and non-sclerotic chimeric DDS kidneys showed strong vimentin signal (Fig. 5E). In contrast, podocytes showed a focal loss of vimentin signal in 42–65% of partially sclerotic glomeruli from heterozygous and chimeric kidneys (Fig. 5F). As for human Wilms' tumours (25), desmin was expressed by the blastemal component of a murine Wilms' tumour (Fig. 5G) that developed in a DDS chimera (20). In wild-type kidneys, no desmin signal was detected in podocytes and only weak signal was present in mesangial cells (Fig. 5H). In contrast, prominent signal was detected in podocytes and parietal epithelial cells in 45–60% of partially sclerotic glomeruli from heterozygous and chimeric kidneys (Fig. 5I). However, desmin signal in podocytes was detected in only a small minority of glomeruli in non-sclerotic kidneys (Table 1).
|
Renin expression in murine DDS
The renin–angiotensin system is the principal regulator of intravascular volume and systemic blood pressure, and renin is involved in the generation of angiotensin II (AngII) that promotes salt and fluid retention in concert with aldosterone (26). Increased intra-renal renin expression, as evident by elongation of renin signal along the afferent arteriole adjacent to the juxtaglomerular apparatus (JGA hyperplasia), can be indicative of glomerular dysfunction due to a reduction in the glomerular filtration rate and/or tubule salt loss. In wild-type kidneys strong renin signal was present in the JGA (Fig. 6A). In contrast, JGA hyperplasia was evident in affected chimeric kidneys adjacent to both sclerotic (Fig. 6B) and non-sclerotic (data not shown) glomeruli, and in all four non-sclerotic chimeric kidneys examined (Fig. 6C). However, no renin signal was detected in either kidney from the heterozygous DDS mouse (data not shown) that exhibited malignant hypertension and end stage renal failure (20). The result may reflect renin degranulation due to malignant hypertension (27).
|
ZO-1, synaptopodin,
-actinin-4 and nephrin expression in murine DDSThe effect of the dominant Wt1 mutation on components of the podocyte filtration barrier was examined, including nephrin, ZO-1 (zonula occludens-1), synaptopodin and
-actinin-4, since mutations in these genes and/or altered patterns of expression are linked with the development of proteinuria (28).
In wild-type kidneys, strong synaptopodin, nephrin and -actinin-4 signal was present in podocyes, and each showed a distinct linear staining pattern that bordered the capillary loops (Fig. 7A–C). Weak -actinin-4 stain was also detected in distal tubules and parietal epithelial cells, and weak synaptopodin signal in the walls of glomerular capillaries. While the signal strength of all three proteins was greatly reduced in 30–40% of partially sclerotic glomeruli from heterozygous and chimeric kidneys, some glomeruli in heterozygous kidneys still showed strong, albeit focal, signal (Fig. 7D–F). Thus, their reduced levels in sclerotic glomeruli reflect loss of expression during disease progression rather than failure of expression during nephrogenesis. Importantly, the intensity and pattern of staining of each protein were indistinguishable between wild-type and non-sclerotic chimeric DDS kidneys where >40% of glomeruli contain Wt1 mutant podocytes (Table 1).
|
In wild-type kidneys, strong linear podocyte ZO-1 staining was present along the GBM, and weaker signal between parietal and tubular epithelial cells (Fig. 7G). Greatly reduced podocyte signal was seen in 50–70% of partially sclerotic glomeruli but strong ZO-1 signal persisted in parietal epithelial cells (Fig. 7H). This was also the case for non-sclerotic glomeruli in affected kidneys (data not shown). However, the finding that some glomeruli in heterozyous kidneys continued to show strong focal ZO-1 signal (Fig. 7I) suggests that the reduced ZO-1 signal in sclerotic glomeruli reflects loss of expression with disease progression. Importantly, loss of podocyte ZO-1 signal was also noted in 30–40% of glomeruli from non-sclerotic chimeric kidneys (Fig. 7), including kidneys from four additional DDS chimeras that contained only 5–17% Wt1 mutant cells, of which two were XX
XY and had only 3–5% of glomeruli containing Wt1 mutant cells (Table 1).
Dual expression of WT1 and nephrin in murine DDS
In wild-type kidneys the podocytes showed characteristic nuclear WT1 stain and cytoplasmic nephrin stain (Fig. 8A). In contrast, in heterozygous kidneys, only nephrin staining was detected in some globally sclerotic glomeruli (Fig. 8B). Thus, absence of WT1 signal in DDS does not necessarily reflect podocyte loss.
|
TGF-ß1, PDGF-A, IGF-II and EGFR expression in murine DDS
In wild-type kidneys TGF-ß1 staining was present in the endothelial cells of the glomerular capillaries (Fig. 9A). In contrast, strong TGF-ß1 stain was present in podocytes in 30–50% of partially sclerotic glomeruli from heterozygous and chimeric kidneys, and in tubular epithelial cells and the lumens (Fig. 9B–D). De novo TGF-ß1 expression by podocytes was also detected in 20–30% of glomeruli from non-sclerotic chimeric kidneys (Fig. 9E, Table 1). In wild-type kidneys weak PDGF-A signal was present in podocytes and endothelial cells of the glomerular (and peritubular) capillaries (Fig. 9F). Strong podocyte signal was evident in partially sclerotic glomeruli from chimeric and heterozygous kidneys, and was particularly marked when comparing adjacent glomeruli with different levels of sclerosis (Fig. 9G). Strong PDGF-A signal was also detected in tubule lumens and in interstitial macrophages (data not shown). Weak EGFR and IGF-II signal was present in tubular epithelial cells in wild-type kidneys, but no de novo expression was detected in glomeruli from DDS kidneys (Table 1).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
While experimental studies have reported that aberrant expression of WT1 can induce GS, the underlying mechanism(s) is unclear (10,11,20,22). Moreover, while WT1 transgenes can affect the expression of endogenous genes in transfection studies and in mouse kidneys (1,15), the significance of the findings is complicated by effects due to promoter and cellular contexts, and the inability to examine the combined effect of all WT1 isoforms. The DDS mouse (Wt1tmT396/+) enables examination of the effect of an endogenous Wt1 mutation on the expression of endogenous genes, and thereby has the potential to identify bona fide changes in gene expression that promote glomerular damage. The present study found that Wt1 mutant cells colonize glomeruli efficiently, the development of GS is preceded by increased intra-renal renin expression, and de novo TGF-ß1 expression and loss of ZO-1 in podocytes, and that podocytes in partially sclerotic glomeruli, that express WT1 at high levels, show inappropriate Pax-2, cytokeratin and desmin expression, cell cycle re-entry, and reduced expression of vimentin, nephrin,
-actinin-4 and synaptopodin. The findings challenge earlier claims that the dominant WT1 mutation may prevent podocyte maturation, and that GS is initiated by alterations in WT1 and Pax-2 expression by podocytes, and shed important insight concerning possible systemic mechanisms that promote glomerular scarring.
Wt1tmT396/+ mutant cells colonize glomeruli efficiently and DDS podocytes show typical maturational changes
Since WT1 is indispensable for kidney development (2) and DDS involves a nephropathy with early onset, which can be congenital, WT1 mutations may affect podocyte development, and reports that podocytes are often under-developed in DDS patients (29) have prompted the suggestion that WT1 is required for proper podocyte differentiation (1). However, GPI analysis and in situ marking found that Wt1 mutant cells colonize all glomerular lineages efficiently, including podocytes, which express WT1 and where the mutation could act. The results suggest that Wt1 mutant podocyte precursor cells behave like wild-type cells in that they can respond to the complex cascade of regulatory signals that govern podocyte determination and differentiation (1,6), and is consistent with finding that most (65%) adult DDS chimeras with bilaterally chimeric kidneys exhibit GS (20). Further, examination of non-sclerotic DDS kidneys found that podocytes undergo typical maturational changes in that they express high levels of WT1, nephrin, -actinin-4 and synaptopodin, express vimentin but not Pax-2 or cytokeratin, and show a cessation of mitotic activity. Thus, the dominant DDS mutation does not affect the partial epithelial–mesenchymal transformation that accompanies the terminal differentiation of podocytes (4,6). Further, the similar patterns of nephrin, synaptopodin, -actinin-4, IGF-II and EGFR expression in wild-type and non-sclerotic DDS kidneys, and the finding that podocytes (in heterozygous kidneys) can express nephrin in the absence of WT1, suggests that the genes encoding these proteins are unlikely to be targets regulated by WT1. The results are consistent with reports that WT1 does not affect the endogenous expression of these genes in rat embryonic kidney cell precursors derived from undifferentiated renal mesenchyme (14), that DDS is not linked with faulty suppression of EGFR (30), and that a DDS transgene (R362X) does not affect nephrin or synaptopodin expression in mouse kidneys (15).
Wt1tmT396/+ mutant cells colonize pancreas efficiently but not muscle
GPI analysis found that mutant Wt1 cells were selected against in skeletal muscle, and in situ marking found that Wt1 mutant cells only rarely colonize smooth muscle cells of the interlobular renal arteries. The results suggest that WT1 may have a role in muscle development, and is consistent with reports that WT1 protein is present in skeletal muscle, in smooth muscle cells of the renal arteries, and in abdominal wall musculature (1,31), and that Wt1 is expressed in heart and vertebrate somites (3,32). In contrast, it is claimed that WT1 may suppress muscle development since aberrant myogenic differentiation in Wilms' tumour is linked with LOH for WT1, and WT1 inhibits myogenesis in vitro (33). However, this relationship is clearly not invariant since no muscle development occurred in a murine Wilms' tumour that formed in a DDS chimera (20) or in chemically induced rat Wilms' tumours (34), all of which express only mutant WT1 protein. GPI analysis found selection against wild-type cells in pancreas. Since WT1 is not expressed in pancreas (31,35) analysis of WT1 expression in pancreatic precursor cells, and of pancreatic development in Wt1 null mice, may help to explain this selection.
Glomerulosclerosis is not preceded by changes in WT1 and Pax-2 expression
Yang et al. (17) reported that podocytes from DDS and IDMS patients contain low levels of WT1 and show inappropriate Pax-2 expression, and suggested that the loss of WT1-dependent transcriptional repression may cause persistent Pax-2 expression (beyond the S-shaped body stage) that may participate in the pathological mechanism leading to glomerular dysfunction. It was also suggested that mutant DDS proteins prevent the accumulation of wild-type WT1 in the nucleus since they show a reduced capacity to bind DNA and may act in a dominant-negative manner by self-association (1). While loss of WT1 and inappropriate podocyte Pax-2 expression was a feature of murine DDS we found no evidence of persistent Pax-2 expression or that loss of WT1 preceded the development of GS. Using the C-19 antibody as used by Yang et al. (17), which in the present study detects wild-type but not mutant WT1 (20), it was found that podocytes in partially sclerotic glomeruli contained high levels of wild-type WT1, including those in scarred regions in cases of segmental sclerosis, and all podocytes in non-sclerotic chimeric kidneys contained high levels of WT1 protein despite the presence of Wt1 mutant podocytes in >40% of glomeruli and altered patterns of renin, ZO-1 and TGF-ß1 expression. The results suggest that the dominant Wt1 mutation does not prevent the accumulation of wild-type WT1 in the nucleus, and is consistent with the report that DNA binding by wild-type WT1 is not inhibited by an excess of mutant WT1 protein (36). Thus, the loss of podocyte WT1 signal in globally sclerotic glomeruli most likely reflects loss of expression as a result of disease progression.
Podocyte Pax-2 expression was detected in only a small minority of glomeruli in non-sclerotic chimeric kidneys but in a substantially higher proportion of sclerotic kidneys. The result suggests that podocyte Pax-2 expression in DDS reflects re-expression with disease progression rather than persistent expression from early nephrogenesis. Importantly, the finding that podocyte Pax-2 expression was a rare event in non-sclerotic kidneys shows that Pax-2 expression is not due to the dominant Wt1 mutation per se, and is supported by the finding here of inappropriate podocyte Pax-2 expression in other models of GS, including mice that do not harbour a dominant Wt1 mutation (the heterozygous Wt1 null mouse) and, indeed, transgenic rats [line TGR(mREN2)27] that have no Wt1 mutation at all. In agreement, podocyte Pax-2 expression has been reported in diseases that rarely, if at all, harbour WT1 mutations, including idiopathic collapsing glomerulopathy (ICG), HIV-associated nephropathy (HIV-AN) and primary focal segmental GS (FSGS) (37,38). Further, we have detected podocyte Pax-2 expression in focal necrotising glomerulo- nephritis and HIV collapsing nephropathy, which have not previously been linked with WT1 mutations (unpublished data). Moreover, podocyte Pax-2 expression was not due to loss of WT1 since Pax-2 was detected in heterozygous Wt1 null mice, DDS mice and hypertensive rats, where podocytes still contained high levels of WT1, and analysis of serial sections found that Pax-2 and WT1 were co-expressed strongly in DDS glomeruli. Overall, the data do not support the claims by Yang et al. (17) that DDS is caused by reduced levels of WT1 and that the low levels are due to the WT1 mutation per se. Furthermore, these claims are difficult to reconcile with their finding that podocytes from one of six DDS patients and two of three IDMS patients, with defined WT1 mutations, contain normal levels of WT1, as was also the case in asymptomatic non-tumoural tissue in 2/3 Wilms' tumour patients with constitutional WT1 mutations (17). The loss of WT1 staining in some DDS patients may reflect the severity of GS since the transition from partial to global sclerosis in murine DDS was marked by loss of WT1 in podocytes. While podocyte Pax-2 expression has also been linked with loss of WT1 in FSGS, ICG and HIV-AN, the cellular lesions in FSGS showed strong Pax-2 signal and low levels of WT1, but Pax-2 and WT1 were co-expressed strongly in the more advanced monolayer lesions (38), and Pax-2 was detected in only some ICG and HIV-AN patients where podocytes contained reduced levels of WT1, and in an ICG patient where the podocytes still expressed WT1 at high levels (37).
Overall, the present results based on three different models of GS (the hypertensive rat, and DDS and heterozygous Wt1 null mice) do not support the view that WT1 represses Pax-2 expression by podocytes which is based on the inverse correlation between WT1 and Pax-2 expression in podocyte precursors, and evidence that WT1 can repress Pax-2 promoter activity in transient transfection assays (16,18). Importantly, the report that WT1 does not affect endogenous Pax-2 expression in rat embryonic kidney precursor cells (14) also suggests that WT1 does not repress Pax-2 expression. Moreover, the finding that podocyte Pax-2 expression in murine DDS was prominent in sclerotic kidneys only, does not support the view that Pax-2 initiates glomerular scarring (17), which was based on the report that constitutive overexpression of Pax-2 in transgenic mice causes nephrotic syndrome (19). However, Pax-2 overexpression did not promote GS, and the more severe phenotype in these mice (neonatal death, lack of foot processes, glomerular atrophy, hypertrophic Bowman's capsules and capillary abnormalities) could reflect a number of factors, including persistent Pax-2 expression, the high level of Pax-2 expressed, the expression of a dominant gain of function Pax-2 mutation and/or expression of the Pax-2b isoform only. Thus, it remains to be determined whether re-expression of endogenous Pax-2, as occurs in DDS, can promote sclerotic damage.
Although we found no evidence that the development of GS in DDS was preceded by loss or reduced levels of WT1 in podocytes, immunoperoxidase staining may not detect minor reductions in the level of WT1 that might be sufficient to promote GS since it develops in kidneys that express 5–30% less WT1 mRNA than normal (22). It is also possible that the mutation can promote GS even in the absence of an overall change in the level of WT1 in podocytes, since should it act in a dominant-negative manner, this is likely to involve sequestration of the wild-type protein into inactive multimeric complexes rather than a reduction in its level as assessed by immunocytochemistry (1).
Evidence that podocytes de-differentiate with disease progression
Since Pax-2 expression by DDS podocytes was prominent in sclerotic kidneys, it could be the consequence of GS, and since Pax-2 is normally expressed by podocyte precursors (until the S-shaped body stage when expression attenuates in the proximal region from where podocytes originate) (16), the re-expression in murine DDS could reflect podocyte de-differentiation and their reversion to an immature phenotype. This view is further supported by the finding that DDS podocytes in sclerotic glomeruli exhibit other features of early nephrogenesis, including cell cycle re-entry, cytokeratin re-expression and loss of vimentin expression (4,6). However, it has been proposed that podocytes adopt a ‘dysregulated’ phenotype in renal disease rather than revert to an earlier developmental stage. For example, Barisoni et al. (39) reported that podocytes in a murine model of HIV-AN express desmin but not WT1 and suggested they are ‘dysregulated’ since WT1 is normally expressed from the very beginning of podocyte ontogeny, whereas desmin is not expressed by podocytes at any stage of glomerular development. However, desmin expression could reflect the reversion of podocytes to an immature fetal phenotype since desmin is expressed by metanephric mesenchymal cells (24) and by the immature blastemal component of Wilms' tumours (25, present data) and therefore can be regarded as a marker of metanephric blastema stem cells. The loss of podocyte WT1 signal in murine HIV-AN, as also reported in human HIV-AN and idiopathic FSGS (40), together with loss of other ‘differentiated’ podocyte markers, including synaptopodin, which has been suggested to imply a ‘dysregulated’ phenotype (37,40–42), could reflect the severity of GS since in murine DDS we found that, while podocytes in globally sclerotic glomeruli show loss or greatly reduced levels of WT1, as well as nephrin, synaptopodin and -actinin-4, these proteins were routinely detected at high levels in partially sclerotic glomeruli. It is also claimed that podocytes undergo ‘transdifferentiation’ since they express macrophage markers in ICG and post-transplantation relapse of primary FSGS (41,43). However, the possibility this reflects macrophage infiltration was not excluded. Importantly, the view that DDS podocytes may revert to an immature fetal phenotype during disease progression could apply to renal diseases in general that culminate in GS: in addition to Pax-2 re-expression in FSGS, ICG and HIV-AN (37,38), podocytes re-express cytokeratin in primary and recurrent FSGS (38,41,44), express desmin in murine HIV-AN, the Zucker diabetic rat, and in the rat remnant kidney model (39,45,46), and show cell cycle re-entry in HIV-AN, ICG, collapsing glomerulopathy and in cellular lesions of FSGS (39,40,43,47), and loss of vimentin expression in ICG, HIV-AN and recurrent FSGS (37,41,43). Thus, the progression of nephropathy in DDS may share a common pathogenic pathway with renal diseases characterized by irreversible glomerular damage.
While the results suggest that DDS podocytes undergo a phenotypic change reminiscent of epithelial–mesenchymal transformation in the later stages of disease, the significance and mechanism(s) driving this change are unclear. It could perhaps be a regenerative pathway to combat further glomerular damage since de-differentiation has recently been implicated as a mechanism of repair of tissue injury in liver and renal tubules (48). While it has been proposed that WT1 may serve to switch cells between epithelial and mesenchymal states (4) it is unlikely that the mutation or changes in the level of WT1 initiate podocyte de-differentiation in DDS since WT1 was expressed at high levels in partially sclerotic glomeruli that showed aberrant changes in Pax-2, desmin, cytokeratin and vimentin expression, and de-differentiation was prominent in sclerotic kidneys only. Interestingly, recent studies suggest that tubulointerstitial fibrosis involves epithelial–mesenchymal transformation, and that tubular epithelial–myofibroblast transformation can be induced by TGF-ß1 and collagen type 1 (49,50). The report that sclerotic glomeruli in human DDS contain increased levels of extracellular matrix (ECM) proteins, including collagen type I, and that podocytes show de novo TGF-ß1 expression (12), raises the possibility that podocyte de-differentiation in DDS could be triggered by increased accumulation of collagen type 1 in the adjacent mesangium and/or intrinsically by de novo TGF-ß1 expression which, in murine DDS, was shown to precede podocyte de-differentiation. Importantly, the finding that parietal epithelial cells in sclerotic glomeruli also show cell cycle re-entry and express desmin suggests that glomerular epithelial cells in general are subject to phenotypic change during disease progression in DDS.
Mechanistic aspects of glomerulosclerosis
The development of GS in murine DDS was not linked with changes in WT1 or Pax-2 expression but was preceded by JGA hyperplasia, and widespread loss of ZO-1 staining and de novo TGF-ß1 expression by podocytes. It is believed that following the loss of a critical number of functional nephrons due to sclerosis, systemic factors affect the remaining nephrons and so promote progressive scarring (51). Importantly, the analysis of chimeric DDS kidneys where only some cells harbour the mutant Wt1 allele suggests that a systemic mechanism does indeed operate in disease pathogenesis since (1) some sclerotic glomeruli contained no detectable Wt1 mutant cells, and many contained <20% Wt1 mutant podocytes, and (2) kidneys showed widespread loss of ZO-1 even in cases with low levels of chimerism and where <5% of glomeruli contained Wt1 mutant podocytes. ZO-1 is a tight junction protein that plays a critical role in maintaining the functional properties of the epithelial slit diaphragm, and the development of proteinuria has been linked with reduced ZO-1 signal due to loss of expression, or its re-distribution from the slit diaphragm into the cytoplasm (52,53). Irrespective of which mechanism operates in DDS the early loss of ZO-1 would contribute greatly to dysfunctional filtration and the development of sclerosis. AngII plays a central role in promoting renal injury by both haemodynamic and non-haemodynamic effects, and can promote the accumulation of ECM proteins, cell proliferation and hypertrophy by inducing cytokine and chemokine expression (26). Since renin is a rate-limiting step in AngII synthesis, the detection of JGA hyperplasia in DDS kidneys suggests that AngII is part of the systemic mechanism that promotes sclerotic damage, and the expected increase in AngII could account for the widespread loss of ZO-1 staining in DDS podocytes given that ACE inhibitors prevent the loss of ZO-1, proteinuria, and further sclerosis development, in the MFW rat (53).
TGF-ß1 and PDGF-A are potent pro-fibrotic cytokines (54,55). However, their high expression in DDS podocytes is unlikely to reflect loss of WT1-dependent transcriptional repression (56) since WT1 does not affect endogenous PDGF-A expression in rat embryonic kidney precursor cells (14), expression of a DDS mutant allele does not affect TGF-ß1 or PDGF-A expression in the M15 mouse mesonephric cell line (57), and increased glomerular TGF-ß1 and PDGF-A expression is a general feature of diseases characterized by an increase in ECM proteins (54,55). However, the latter is complicated by few descriptions regarding the exact immunocytochemical localization of TGF-ß1, and reports that TGF-ß1 is expressed either by podocytes (58) or mesangial cells (59). Since podocytes have AngII receptors (60) and ACE inhibitors can suppress glomerular TGF-ß1 levels (54), it is possible that AngII induces TGF-ß1 expression in DDS podocytes. While AngII can promote accumulation of ECM proteins in mesangial cells by upregulation of TGF-ß1 expression (61) such a direct mechanism is unlikely to operate in DDS since de novo TGF-ß1 expression (and increased PDGF-A expression) occurred in podocytes only. While these cytokines are unlikely to traverse the GBM to promote sclerosis in the adjacent mesangium, a systemic action is possible since the development of GS in transgenic mice that over-express TGF-ß1 is linked with high plasma levels of TGF-ß1 (62).
Animal models can contribute greatly to understanding the mechanisms involved in initiation and progression of human disease, and thereby influence the design of diagnostic and therapeutic strategies. The DDS mouse meets these criteria since the urogenital pathology mimics closely the human disease (20) and podocytes show similar changes, including inappropriate Pax-2 expression, de novo TGF-ß1 expression, increased PDGF-A expression, and cell cycle re-entry (12,17). However, the function of WT1 in terminally differentiated podocytes, the site of action of dominant WT1 mutations, and the pathogenesis of proteinuria in DDS are presently unknown. Importantly, the present findings that DDS podocytes undergo typical maturational changes, and that Wt1 mutant podocytes colonize both non-sclerotic adult kidneys and non-sclerotic glomeruli in affected kidneys, raise the possibility that the development of sclerosis is not tightly linked with the presence of mutant Wt1 allele in podocytes, and therefore, DDS mutations may not affect podocytes directly. Indeed, the presence of nephrogenic rests and atrophic glomeruli in DDS kidneys (1,7,8,29) is indicative of defective nephrogenesis. Further, the finding that Wt1 mutant cells colonize the podocyte lineage efficiently, and evidence that GS is mediated systemically, suggests that DDS nephropathy is not strictly dependent on a constitutional WT1 mutation, and therefore mosaicism for WT1 mutations, as so far reported in non-DDS cases (63,64), could contribute to the wide phenotypic variation in DDS (7,8). Indeed, the first DDS patient reported was XX/XY mosaic (65). While treatment of DDS necessitates ultimately kidney transplantation, the use of renal replacement is controversial (66). However, the finding that JGA hyperplasia is an early event in murine DDS indicates that further study of the role of the renin–angiotensin system is merited to address the possibility that the symptoms can be ameliorated in DDS patients by using ACE inhibitors until suitable donors become available.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Tissues
The heterozygous (Wt1tmT396/+) and chimeric (Wt1tmT396/+
+/+) mice used were described previously (20). At necropsy chimeric kidneys were divided into four parts which were respectively fixed in buffered formalin overnight at 4°C and embedded in paraffin wax (for immunohistology), used for DNA isolation (for PCR analysis), used for GPI analysis (to assess chimerism), and embedded in OCT (Miles Scientific; for DNA–DNA in situ hybridization).
DNA–DNA in situ hybridization
Five-micrometre kidney cryostat sections were fixed for 1 h in 4% paraformaldehyde solution in phosphate buffered saline (PBS) and examined by DNA–DNA in situ hybridization using an 3H-labelled Y chromosome probe as described previously (21).
Glucose phosphate isomerase (GPI) isozyme analysis
Extraction, separation and detection of GPI isozymes were performed as previously described (21). Samples were analysed in triplicate on separate gels and the relative proportions of the two GPI isoforms, GPI-1A and GPI-1B, were quantified by densitometry. In the case of kidneys, 2 mm3 cortex samples were analysed. Estimates of the proportion of ES cells in each tissue of chimeras produced from heterozygous and wild-type ES cells, standardized for variation in overall extent of chimerism between mice, were determined by logistic regression analysis (67). Data were fitted using the GLIM statistical package to a model ln{x/(1-x)}=A+BM+CT, where x is the proportion of GPI-1A, A is a constant, BM takes a different value for each mouse and CT takes a different value for each tissue. CT was allowed to take different values for the same tissue from heterozygous and wild-type chimeras except in the case of brain where the two values were constrained to be identical. The estimates graphed are calculated using a mean value of BM. The statistical significance of differences between estimates for the two genotypes for a given tissue were assessed by computing an F statistic from the scaled deviances of a complete regression model and a reduced model in which these estimates were constrained to be identical, and using it to determine a P-value at the level of the whole study by the Dunn–
idák method for multiple planned comparisons (68).
Immunohistology
The antibodies used are listed in Table 2. Immunoperoxidase detection was performed on 4 µm formalin-fixed, paraffin-embedded kidney sections using the following antigen retrieval methods: (i) microwave pre-treatment (4x5 min) in 10 mM sodium citrate buffer pH 6.0—for synaptopodin, vimentin, Ki67, renin, desmin and EGFR; (ii) microwave pre-treatment (4x5 min) in 0.8 M urea buffer pH 6.4—for WT1 (C-19), nephrin, TGF-ß1, PDGF-A and Pax-2; (iii) protease digestion (0.1% in PBS) at 37°C for 5 min—for ZO-1; (iv) microwave pre-treatment (4x5 min) in 10 mM citrate buffer pH 6.0 followed by trypsin digestion (0.25% trypsin in PBS) at 37°C for 45 s—for IGFII, pan-cytokeratin and -actinin-4; and (v) pepsin digestion (0.4% in 0.2N HCl) for 3 min at 37°C followed by incubation in saponin (0.0004% in PBS) for 30 min—for WT1 (6F-H2). In the case of rat kidney sections a pressure cooker (10–13 psi for 4 min; A Menarini Diagnostics) was used for antigen retrieval. Peroxidase staining was performed using either Vectastain Elite ABC kits for rabbit and goat polyclonal antibodies, or the Vector M.O.M. Immunodetection kit for mouse monoclonal antibodies (Vector Laboratories Inc.). Since sclerotic kidneys contained high levels of endogenous protein, when using the M.O.M. kit the concentration of blocking reagent was increased x10-fold, sections were blocked overnight, and 50% less secondary antibody was used. Kidney sections were incubated for 2–16 h with the appropriate dilution of primary antibody in PBS containing 0.1% BSA. Sections were then washed in PBS and incubated with biotinylated secondary antibody for 30 min. For quenching of endogenous peroxidase, sections were treated with 3% hydrogen peroxide in distilled water for 5 min, then washed in PBS for 20 min. Sections were incubated for 30 min with ABC reagent. After washing with PBS, sections were stained with 3,3'-diaminobenzidine (DAB) for 5 min, and counter-stained with 50% haematoxylin and 10% eosin stain, or methyl green. In dual antibody staining of cytoplasmic and nuclear proteins, cytoplasmic signal was visualized using the VIP substrate kit (Vector Laboratories Inc.). Negative controls consisted of substitution of the primary antibody with 1% BSA in PBS, equivalent concentrations of an irrelevant murine monoclonal antibody, or normal rabbit or goat IgG. No signal was detected in negative control sections.
|
|
ACKNOWLEDGEMENTS |
|---|
We thank Deborah Bruce and Rachel Berry for technical assistance, Dr Pekka Kilpelainen (Karolinska Institute, Stockholm) for the nephrin antibody and Dr Setsuo Hirohashi (National Cancer Center Research Institute, Tokyo) for the antibody against
-actinin-4. The study was supported by the Medical Research Council, the National Kidney Research Fund, and the Cancer Research Campaign (CRC; now Cancer Research UK).
|
FOOTNOTES |
|---|
* To whom correspondence should be addressed. Tel: +44 1316511078; Fax: +44 1316511072; Email: m.hooper{at}ed.ac.uk
Present address: Institute of Human Genetics, International Centre for Life, Central Parkway, Newcastle-upon-Tyne NE1 3B7, UK.
Present address: Centre for Research in Biomedicine, Faculty of Applied Sciences, University of the West of England, Bristol BS16 1QY, UK.
|
REFERENCES |
|---|
|
TOP |
|---|