TGF-β–Activated Kinase 1 Is Crucial in Podocyte Differentiation and Glomerular Capillary Formation

Abstract

TGF-β–activated kinase 1 (TAK1) is a key intermediate in signal transduction induced by TGF-β or inflammatory cytokines, such as TNF-α and IL-1, which are potent inducers of podocyte injury responses that lead to proteinuria and glomerulosclerosis. Nevertheless, little is known about the physiologic and pathologic roles of TAK1 in podocytes. To examine the in vivo role of TAK1, we generated podocyte-specific Tak1 knockout mice (Nphs2-Cre⁺:Tak1^fx/fx; Tak1^{[INCREMENT]/[INCREMENT]}). Targeted deletion of Tak1 in podocytes resulted in perinatal lethality, with approximately 50% of animals dying soon after birth and 90% of animals dying within 1 week of birth. Tak1^{[INCREMENT]/[INCREMENT]} mice developed proteinuria from P1 and exhibited delayed glomerulogenesis and reduced expression of Wilms’ tumor suppressor 1 and nephrin in podocytes. Compared with Tak1^fx/fx mice, Tak1^{[INCREMENT]/[INCREMENT]} mice exhibited impaired formation of podocyte foot processes that caused disruption of the podocyte architecture with prominent foot process effacement. Intriguingly, Tak1^{[INCREMENT]/[INCREMENT]} mice displayed increased expression of vascular endothelial growth factor within the glomerulus and abnormally enlarged glomerular capillaries. Furthermore, 4- and 7-week-old Tak1^{[INCREMENT]/[INCREMENT]} mice with proteinuria had increased collagen deposition in the mesangium and the adjacent tubulointerstitial area. Thus, loss of Tak1 in podocytes is associated with the development of proteinuria and glomerulosclerosis. Taken together, our data show that TAK1 regulates the expression of Wilms’ tumor suppressor 1, nephrin, and vascular endothelial growth factor and that TAK1 signaling has a crucial role in podocyte differentiation and attainment of normal glomerular microvasculature during kidney development and glomerular filtration barrier homeostasis.

TGF-β–activated kinase 1 (TAK1), a serine/threonine kinase originally identified as a member of the mitogen-activated protein kinase (MAPK) kinase kinase family named MAPK kinase kinase 7, is rapidly activated by TGF-β1.¹ In addition to TGF-β1, TAK1 can also be activated by various stimuli, including other TGF-β family cytokines (such as bone morphogenetic proteins [BMPs]^2,3), proinflammatory cytokines (such as TNF-α⁴ and IL-1⁵), LPS,⁶ environmental stress,⁷ and Ca²⁺/calmodulin-dependent protein kinase II.⁸ Activation of TAK1 occurs through phosphorylation of Thr187 and Ser192 in the activation loop⁹ and subsequently, triggers the activation of several downstream signaling cascades, including MAPK kinase 4/7–c-Jun N-terminal kinase (JNK), MAPK kinase 3/6-p38 MAPK, and NF-κB–inducing kinase–inhibitor of κB kinase.^4,5 In addition, TAK1 is activated by agonists of AMP kinase and ischemia, which in turn, activate the AMP kinase pathway, a key energy-sensor pathway.^10,11 TAK1 has also been shown to be activated by the Wingless/integration 1 signaling pathway,¹² which plays essential roles in tissue development, including in the kidney.¹³ Changes in Wingless/integration 1 ligands and signaling pathway components are associated with podocyte injury^14,15 and various kidney diseases.^13,16 Thus, a significant body of evidence suggests that TAK1 could be implicated in the onset and progression of renal diseases by various inducers.

Podocytes are terminally differentiated glomerular visceral epithelial cells that form multiple foot processes and slit diaphragms covering the outer layer of the glomerular basement membrane (GBM) and play a crucial role in regulation of glomerular function.¹⁷ Proper differentiation of podocytes during kidney development and retention of the differentiated state of podocytes are critical for maintenance of the kidney glomerular filtration barrier and function. Among the various transcription factors, including Wilms’ tumor suppressor gene product (WT1),^18,19 LIM homeobox transcription factor 1-β,^20,21 and Pod1 (also called Tcf21/capsulin/epicardin),^22,23 implicated in podocyte differentiation, WT1 is thought to be the master transcription factor that regulates podocyte differentiation.²⁴ It is widely accepted that podocyte injury is the key early event in initiating glomerulosclerosis, thereby resulting in ESRD in various animal models and humans.^25,26 In response to injurious stimuli, podocytes undergo dedifferentiation accompanied by podocyte effacement, detachment, and apoptosis, depending on the severity and duration of the injury.^27,28 Because of their limited proliferative capacity, podocyte detachment from GBM and apoptosis inevitably leads to reduction of podocyte density and results in impaired glomerular filtration and proteinuria.

Vascular endothelial growth factor-A (VEGF-A) is a multifunctional cytokine with important roles in both vasculogenesis and angiogenesis. By alternative splicing of exons 6 and 7 in the VEGF gene, at least five isoforms of VEGF are expressed. In addition, a novel family of human VEGF splice variants (VEGF_xxxb) has been identified, which is formed by alternative splicing of the conventional exon 8 and shown to exhibit antiangiogenic activity.^29,30 In human kidneys, podocytes are the major cell type in the glomerulus producing VEGF³¹ during kidney development, and VEGF is essential for the development and maintenance of glomerular integrity through its crucial role in cell migration, differentiation, and survival of glomerular endothelial and mesangial cells.^32,33 Podocyte-specific heterozygous deletion of VEGF leads to glomerular capillary endothelial cell dysfunction and proteinuria.³² In addition, normal animals treated with neutralizing antibody to block VEGF actions systemically displayed endothelial cell damage, reduced nephrin expression, and proteinuria.³⁴ Similarly, patients treated with humanized monoclonal VEGF antibody (bevacizumab) developed glomerular disease characterized by thrombotic microangiopathy.^35,36 Conversely, podocyte-specific overexpression of mouse VEGF₁₆₄ leads to collapsing glomerulopathy, capillary enlargement, and postnatal lethality.³² In addition, the induction of podocyte-specific VEGF₁₆₄ overexpression in adult transgenic mice led to proteinuria, GBM thickening, mesangial expansion, loss of slit diaphragms, and podocyte foot process effacement.³⁷ Thus, both deficiency and excess of VEGF seem to be detrimental to the physiologic integrity of the glomerular filtration barrier.

Nephrin, encoded by nephrotic syndrome type 1 (NPHS1), is a major component of the glomerular slit diaphragm³⁸ and functions as a structural protein, and it is a signaling molecule influencing foot process formation and maintenance of podocyte integrity. The NPHS1 is mutated in congenital nephrotic syndrome of the Finnish type, a human disease that leads to proteinuria in utero and early death.³⁹ In addition, reduced expression of nephrin mRNA is often observed in proteinuric glomerular diseases,^40,41 but information is very limited regarding how nephrin expression is regulated during kidney development and under pathologic circumstances. Although it has been previously known that WT1,^42,43 Sp1,⁴⁴ and Snail⁴⁵ are implicated in the regulation of Nphs1 expression, recent studies have shown that NF-κB and WT1 cooperatively regulate Nphs1 transcription.⁴⁶

Here, we show that podocyte-specific deletion of Tak1 in mice resulted in high incidence of perinatal lethality, proteinuria, delayed glomerulogenesis, and reduced expression of WT1 and nephrin in the podocytes. Tak1 deletion in podocytes caused disruption of the podocyte architecture and foot process effacement. Loss of Tak1 also led to increased VEGF expression and abnormally enlarged glomerular capillaries within the glomerulus. Furthermore, in 4- and 7-week-old Tak1^{[INCREMENT]/[INCREMENT]} mice with proteinuria, increased collagen deposition was seen in the mesangium and the adjacent tubular interstitial area. Taken together, we show that TAK1 regulates the expression of WT1, nephrin, and VEGF and that TAK1 is critical in podocyte differentiation and glomerular microvasculature during kidney development and formation of the kidney glomerular filtration barrier.

Results

Podocyte-Specific Tak1 Deletion Results in Perinatal Death and Proteinuria

TGF-β family cytokines, such as TGF-β and BMP, have been shown to play critical roles in kidney development,⁴⁷ and TAK1 is an important mediator of TGF-β and BMP signaling transduction. We have confirmed that TAK1 is activated by TGF-β1 and regulates the activation of JNK and p38 MAPK in human podocytes (Supplemental Figure 1). To examine the in vivo role of TAK1, podocyte-specific Tak1 knockout mice were generated by crossing Tak1^fx/fx mice with Nphs2-Cre⁺ transgenic mice, in which Cre recombinase is expressed in podocytes from capillary loop stage during glomerular development.⁴⁸ Genotypes of the mice were confirmed by the PCR-based genotyping with tail DNA (Figure 1A). Heterozygous mice (Nphs2-Cre⁺:Tak1^fx/+; referred to as Tak1^{[INCREMENT]/+}) were born at the expected Mendelian ratio and survived to adulthood with no apparent defects. However, homozygous mice (Nphs2-Cre⁺:Tak1^fx/fx; referred to as Tak1^{[INCREMENT]/[INCREMENT]}) produced by the crossing of the heterozygous mice (Tak1^{[INCREMENT]/+}) showed high incidence of perinatal lethality, with approximately 10% of Tak1^{[INCREMENT]/[INCREMENT]} mice surviving to adulthood. Homozygous Tak1^{[INCREMENT]/[INCREMENT]} mice used for the production of progenies had a similar death rate of Tak1^{[INCREMENT]/[INCREMENT]} mice after birth as heterozygous breeding. Approximately 50% of newborn Tak1^{[INCREMENT]/[INCREMENT]} mice died within 1 day, and another 40% of mice within 1 week of birth (Figure 1B). By contrast, mice surviving beyond this perinatal period continued to grow, and death rarely occurred. The different lifespans of the newborn Tak1^{[INCREMENT]/[INCREMENT]} mice might be caused by the mosaicism of Tak1 deletion by Cre-mediated recombination in podocytes, which is known to occur.

Figure 1: Podocyte-specific loss of Tak1 results in perinatal lethality and proteinuria. (A) PCR-based genotyping of Tak1^fx/fx, Tak1^fx/+, Tak1^+/+, and Cre with tail DNA. Molecular weight (M) standards are indicated as kilobases (Kb). (B) Survival curve of Nphs2-Cre⁺:Tak1^fx/fx(Tak1^Δ/Δ) mice and control mice (Tak1^fx/fx). (C) Analysis of urinary albumin excretion at P1 of four randomly selected Tak1^Δ/Δ mice and control mice (Tak1^fx/fx) by Coomassie-stained SDS-PAGE. Al represents 2 μg BSA used as a control. (D) Quantitation of urinary albumin from samples in C expressed as micrograms per milligram urine creatinine (Ucr). *P<0.01 versus control Tak1^fx/fx mice. (E) PCR-based genotyping to determine Tak1 deletion in the kidneys of corresponding P1 mice in C.

Podocyte-specific deletion of Tak1 in mice also resulted in a significant increase in urinary albumin excretion in the newborn postnatal day 1 (P1) (Figure 1C). Quantitation of urinary albumin based on urine creatinine used as a correction factor-confirmed increase in albumin in urine from Tak1^{[INCREMENT]/[INCREMENT]} mice (Figure 1D). To confirm whether the albuminuria has a correlation with Tak1 deletion, one fourth of frozen kidney tissues from each of the corresponding mice were subjected to PCR with the primer set 2. As shown in Figure 1E, only the kidney tissues from Tak1^{[INCREMENT]/[INCREMENT]} mice yielded the expected 1.2-kb PCR products amplified from the deleted Tak1 by Cre recombinase, indicating correlation of Tak1 deletion with albuminuria. Thus, these data suggest that TAK1 deficiency in podocytes may cause a defect in glomerular function and perinatal death.

TAK1 Plays an Important Role in Glomerulogenesis

Given our findings that Tak1^{[INCREMENT]/[INCREMENT]} mice displayed high occurrence of perinatal lethality and albuminuria, we sought to further understand the role of TAK1 in podocytes during kidney development. To this end, kidneys were harvested from Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx mice at P1, P4, and P7. The kidneys from Tak1^{[INCREMENT]/[INCREMENT]} mice were generally smaller than the kidneys from the control Tak1^fx/fx mice, without significant differences in the apparent body size at P1 (data not shown). To analyze the morphology of kidneys, paraffin-embedded kidney sections from control Tak1^fx/fx and Tak1^{[INCREMENT]/[INCREMENT]} mice (n=4 for each group) were subjected to hematoxylin and eosin (H&E) staining. As shown in Figure 2A, control Tak1^fx/fx mouse kidney showed typical gradient of renal development with normal-appearing nephrogenic zones (NZs) at P1. However, NZs in the kidneys from Tak1^{[INCREMENT]/[INCREMENT]} mice encompassed a wider area and had a reduced number of glomeruli (Figure 2A, brown arrows) compared with the control Tak1^fx/fx mice. Average number of glomeruli in five kidney tissue sections from four mice in each group was obtained as follows: Tak1^fx/fx versus Tak1^{[INCREMENT]/[INCREMENT]}=24.4±1.7 versus 17.6±2.1 (Figure 2A). In addition, larger portions of NZs of Tak1^{[INCREMENT]/[INCREMENT]} kidneys were in the vesicle stages of glomerulogenesis compared with Tak1^fx/fx kidneys (Figure 2B, left panel). Furthermore, mature glomeruli in the sub-NZs of Tak1^{[INCREMENT]/[INCREMENT]} kidneys showed unevenly distributed capillary loops compared with Tak1^fx/fx kidneys, and red blood cells were also detected in the urinary space (arrow in Figure 2B, lower right panel). At P4, more comma/S-shaped bodies were observed in the NZ of Tak1^{[INCREMENT]/[INCREMENT]} kidneys in contrast to mostly capillary loop stage vesicles in the NZ of Tak1^fx/fx kidneys (Figure 3A, left panel). At P7, all vesicular structures had become mature glomeruli, and the NZ had completely disappeared in the control Tak1^fx/fx kidneys (Figure 3A, upper right panel), whereas some vesicular structures, especially capillary loop stage structures and the NZ-like region, were still detectable in the peripheral zone of Tak1^{[INCREMENT]/[INCREMENT]} kidneys (red arrow in Figure 3A, bottom right panel). In addition, tubular structures were developed and easily detectable in the peripheral zone of Tak1^fx/fx kidneys at this stage (asterisks in Figure 3A, upper right panel) but not in Tak1^{[INCREMENT]/[INCREMENT]} kidneys (Figure 3A, bottom right panel). Thus, these histology data suggest that TAK1 in podocyte might play a crucial role in glomerulogenesis and even tubulogenesis during kidney development. Furthermore, as shown in Figure 3B, glomeruli in the inner cortical region of the Tak1^{[INCREMENT]/[INCREMENT]} kidneys showed abnormal capillary development with abnormally wide capillary loops filled with red blood cells, reduced numbers of capillary branches, podocytes that appear to be displaced to the periphery of the glomerulus (Figure 3B, a and b), dilated capillary loops (arrows in Figure 3B, c), and decreased numbers of mesangial cells and mesangial regions (Figure 3B, a, b, and d). These phenotypes were not seen in control Tak1^fx/fx kidneys. Transmission electron microscopy (TEM) of Tak1^{[INCREMENT]/[INCREMENT]} kidneys at P1 and P4 (Figure 3C, lower panel) accordingly showed the ultrastructural abnormalities compared with corresponding P1 and P4 control Tak1^fx/fx kidneys (Figure 3C, upper panel), with simplification of the glomerular network and existence of multiple endothelial cell nuclei within capillary loops (indicated by arrowheads in Figure 3C, lower right panel). Occurrence of Tak1 deletion was confirmed by PCR-based genotyping, and as shown in Figure 3D, only the kidney tissues from Tak1^{[INCREMENT]/[INCREMENT]} mice yielded the expected 1.2-kb PCR products. Thus, these data suggest that the abnormal phenotypes in the glomeruli may be caused by the deficiency of TAK1 activity in podocytes.

Figure 2: Podocyte-specific loss of Tak1 results in the reduction of the number of glomeruli and glomerular abnormality. Morphology of kidneys from Tak1^Δ/Δ mice and control Tak1^fx/fx mice at P1 subjected to H&E staining and analyzed under (A) low magnification (×100) and (B) higher magnification (×400). Representative picture from four kidneys in each group is shown. (A) Brown arrows indicate glomeruli. NZ is indicated by black arrows. Histogram depicts the average number of glomeruli in five tissue sections from four kidneys in each group. Data are presented as the means±SEMs. *P<0.05 versus Tak1^fx/fx mice. (B) H&E staining of cortical regions. Abbreviations indicate different vesicular structures during glomerulogenesis: C, capillary loop stage; g, maturing glomerulus; G, mature glomerulus; S, comma/S-shaped body stage; V, vesicle stage. In the lower right panel, the arrow indicates a glomerulus showing red blood cells in the urinary space.

Figure 3: Podocyte-specific loss of Tak1 results in delayed glomerulogenesis and abnormal capillary formation. (A) H&E staining of the cortical region of kidneys obtained from P4 and P7 mice. A representative picture from four kidneys in each group is shown. Abbreviations are the same as in Figure 2. Asterisks indicate tubules in the peripheral region of the control kidney (Tak1^fx/fx), and the red arrow represents the remaining NZ-like region in the Tak1^Δ/Δ kidney. (B) Various abnormalities of glomeruli found in the P4 Tak1^Δ/Δ kidney. (C) TEM showing enlarged capillaries and multiple endothelial cell nuclei in capillary loops (arrowheads) in glomeruli of P1 and P4 Tak1^Δ/Δ kidneys. Scale bars, 10 μm. (D) Confirming Tak1 deletion in the kidney at P4 by the PCR-based genotyping as shown in Figure 1E.

Podocyte-Specific Tak1 Deletion Increases VEGF Expression in the Glomeruli

For better visualization of the glomerular capillaries and structure, we performed immunohistochemical staining of CD31, an endothelial cell marker. As shown in Figure 4A, comparison of Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx kidneys shows evidence of dilated capillary loops (Figure 4A, arrows) and fewer cells in the mesangial area in Tak1^{[INCREMENT]/[INCREMENT]} kidneys. Thus, these data suggest that TAK1 in podocytes might play an important role in the normal formation of capillary loops in the glomeruli and the maintenance of mesangial cells. Because it is well known that VEGF production by podocytes plays a crucial role in glomerular endothelial cell migration and capillary formation,³² we next determined the effect of podocyte-specific Tak1 deletion on the expression of VEGF during glomerulogenesis. Immunohistochemical staining showed that VEGF expression was relatively low, and no significant differences were observed in the NZ of both Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx kidneys (Figure 4B, left panel). By contrast, VEGF expression was increased in the maturating glomeruli in the sub-NZ (Figure 4B, lower left panel) and significantly higher in the glomeruli located in the inner cortical region of Tak1^{[INCREMENT]/[INCREMENT]} kidneys (Figure 4B, lower center panel) compared with the expression level in Tak1^fx/fx kidneys (Figure 4B, upper center panel). Under high magnification, we confirmed that podocytes in both Tak1^fx/fx and Tak1^{[INCREMENT]/[INCREMENT]} kidneys expressed VEGF, but the expression of VEGF in the glomeruli of the Tak1^{[INCREMENT]/[INCREMENT]} kidney was significantly enhanced (arrows in Figure 4B, lower right panel). VEGF was also detected along with capillaries in the kidneys from both groups, suggesting that VEGF might be mainly localized on the basal surfaces of podocytes or the GBM. Thus, these data suggest that TAK1 might negatively control VEGF expression in podocytes. In this context, we next examined whether TAK1 directly regulates mRNA expression of VEGF in the podocytes. For this examination, quantitative real-time PCR was used. As shown in Figure 5A, inhibition of TAK1 with 0.2 μM LL-Z1640-2 (LZ) enhanced the expression of VEGF mRNA in a time-dependent manner in human podocytes. Because hypoxia inducible factor 1α (HIF-1α) is well known to be a major transcription factor for VEGF expression, the effect of TAK1 inhibition on the expression of HIF-1α was also evaluated. As shown in Figure 5B, inhibition of TAK1 with 0.2 μM LZ also enhanced the expression of HIF-1α mRNA. Therefore, these data suggest that the enhanced VEGF expression in podocytes caused by TAK1 deficiency might be mediated, at least in part, by the increased HIF-1α expression.

Figure 4: Glomerular expression of CD31 and VEGF is increased by podocyte-specific deletion of Tak1. Paraffin-embedded kidney sections obtained from P4 mice were subjected to immunohistochemical staining of (A) CD31 and (B) VEGF. Representative pictures (original magnification, ×400) from three kidneys in each group are shown. The right panel shows the enlarged images of the rectangular insets in the center panel. The arrows in A indicate the increased expression of CD31 and enlarged capillaries. The arrows in B indicate podocytes expressing VEGF in the glomeruli.

Figure 5: TAK1 inhibition increases the expression of VEGF and HIF-1α in immortalized human podocytes. Cells incubated under nonpermissive conditions for 4 (D4) and 7 (D7) days at 37°C were treated with TAK1 inhibitor LZ (0.2 μM) for 3 and 6 hours (h). Expression of (A) VEGF and (B) HIF-1a was assessed by quantitative real-time PCR. The values are expressed as fold change and represent mean values±SEMs (n=3 in all groups); t test was used for statistical analysis. *P<0.05 versus no LZ treatment; **P<0.10 versus no LZ treatment.

TAK1 Regulates the Differentiation of Podocytes

Given that Tak1^{[INCREMENT]/[INCREMENT]} mice develop albuminuria, we next examined whether TAK1 is required for the differentiation of podocytes. Frozen sections from both Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx kidneys at P4 were subjected to immunofluorescence staining for WT1 and nephrin, which are two representative markers of differentiated podocytes. In Tak1^fx/fx kidneys, WT1-positive cells were seen in immature vesicular structures in the NZ, and WT1-positive podocytes were seen in maturing and mature glomeruli in the sub-NZ and the inner cortical region (Figure 6A, column 2, top panel). Nephrin expression was also observed in the podocytes in maturing and mature glomeruli in the sub-NZ and the inner cortical region but hardly detectable in the NZ (Figure 6A, columns 3 and 4, top panels). In Tak1^{[INCREMENT]/[INCREMENT]} kidneys, similar intensity of WT1 was seen in immature vesicular structures in the NZ (Figure 6A, column 2, middle panel). However, apparent intensity of WT1 staining of maturing and mature glomeruli in the inner cortical region of Tak1^{[INCREMENT]/[INCREMENT]} kidney was significantly weaker (Figure 6A, column panel, middle and bottom panels), compared with the intensity of Tak1^fx/fx kidney. It has been previously shown that Cre recombination occurred during the capillary loop stage in glomerular development; therefore, our data from Tak1 deletion by Nphs2-Cre suggest that WT1 expression was decreased because of the loss of TAK1 activity in the podocytes. Similarly, overall intensity of nephrin staining in glomeruli of Tak1^{[INCREMENT]/[INCREMENT]} kidney was weaker (Figure 6A, column 3, middle panel) compared with Tak1^fx/fx kidney. These data indicate that TAK1 plays a crucial role in the regulation of WT1 and nephrin expression. The analysis of the level of fluorescence intensity for WT1 and nephrin expression in kidney showed that WT1 and nephrin levels were reduced by 0.4- and 0.5-fold, respectively in the sub-NZ of Tak1^{[INCREMENT]/[INCREMENT]} kidney compared with Tak1^fx/fx kidney (Figure 6B). To investigate whether TAK1 regulates WT1 mRNA expression, immortalized human podocytes cultured under nonpermissive conditions for 4 and 7 days at 37°C were treated with TAK1 inhibitor LZ (0.2 μM) for 3 and 6 hours. WT1 mRNA level was assessed by quantitative real-time PCR. As shown in Figure 6C, TAK1 inhibition resulted in suppression of WT1 mRNA expression in a time-dependent manner, suggesting that TAK1 regulates expression of WT1.

Figure 6: Podocyte-specific loss of Tak1 results in the reduction of WT1 and nephrin expression. (A) Immunofluorescence staining of frozen sections for WT1 (green; column 2), nephrin (red; column 3), WT1 and nephrin (merged images; column 4), and 4′,6-diamidino-2-phenylindole (DAPI) (blue; column 1). A representative picture from four kidneys in each group is shown. Bottom panels are the enlarged images of rectangular insets in the middle panels. (B) Quantification of WT1 and nephrin expression in the kidney. Fluorescence intensity of 20 vesicular structures in NZ and 20 glomeruli in sub-NZ of each group was measured and analyzed as described in Concise Methods. Values are expressed as means±SEMs. (C) TAK1 inhibition in immortalized human podocytes reduced WT1 mRNA expression, which was assessed by quantitative real-time PCR. The values are expressed as fold change and represent mean values±SEMs (n=3 in all groups); t test was used for statistical analysis. *P<0.05 versus no LZ treatment.

To further examine the role of TAK1 in foot process formation, ultrastructure of Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx kidneys was analyzed by TEM, and representative images are shown in Figure 7, A and B. Tak1^fx/fx kidneys (Figure 7A, left panel) showed normal architecture of podocyte foot processes. By contrast, Tak1^{[INCREMENT]/[INCREMENT]} kidneys (Figure 7A, right panel) showed disruption of the normal architecture with podocyte foot process effacement, suggesting that TAK1 deficiency causes impaired podocyte foot process formation. In addition, glomeruli of Tak1^{[INCREMENT]/[INCREMENT]} kidneys (Figure 7A, right panel) displayed irregularity and thickening of the GBM (Figure 7B) (Tak1^fx/fx kidney versus Tak1^{[INCREMENT]/[INCREMENT]} kidney=187±5 versus 313±15 nm, P<0.02), disruption of endothelial fenestration, and presence of multiple endothelial cell nuclei within a capillary loop (Figure 7C) (Tak1^fx/fx kidneys versus Tak1^{[INCREMENT]/[INCREMENT]} kidneys=0.96±0.06 versus 1.53±0.07, P<0.05).

Figure 7: TAK1 deficiency in podocytes leads to abnormal ultrastructure of glomeruli. (A) TEM of kidneys from P1 and P4 mice as indicated. Cap, capillary lumen; En, endothelial cell; P, podocyte foot processes. Arrowheads show endothelial fenestration. Scale bars, 500 nm. (B) GBM thickness of Tak1^fx/fx and Tak1^Δ/Δ kidneys. Arrows show GBMs. The values are expressed in nanometers and represent mean values±SEMs. Scale bars, 500 nm. *P<0.02 versus Tak1^fx/fx kidney. (C) Multiple endothelial cell nuclei present within a capillary loop in the kidney from Tak1^Δ/Δ mice. Endothelial cell nuclei are indicated with arrowheads. Numbers of endothelial cell nuclei are represented as the mean values±SEMs. Scale bar, 2 μm. *P<0.05 versus Tak1^fx/fx kidney.

Podocyte-Specific Tak1 Deletion Increases Collagen Deposition in the Mesangium and the Tubular Interstitial Area

It is well established that podocyte injury is associated with increased proteinuria, leading to development of glomerulosclerosis and interstitial fibrosis and progression of CKD.^25,26 Our data suggest that Tak1 deletion in podocytes might repress podocyte differentiation and glomerular capillary morphogenesis and cause development of proteinuria in the newborn mice. Thus, we sought to determine whether the mice had progressively increased collagen deposition with time. To this end, paraffin-embedded Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx kidney sections obtained from 2-, 4-, and 7-week-old mice were subjected to Masson’s trichrome staining. As shown in Figure 8, A and B, collagen deposition in Tak1^{[INCREMENT]/[INCREMENT]} kidneys was increased in the mesangium and the adjacent tubulointerstitial area from 4- (Figure 8A, right panel) and 7-week-old mice (Figure 8B, right panel) but not Tak1^fx/fx kidneys (Figure 8, A, left panel, and B, left panel) or the kidneys from 2-week-old Tak1^{[INCREMENT]/[INCREMENT]} mice (data not shown). Moreover, the increased collagen deposition in the mesangium of 7-week-old Tak1^{[INCREMENT]/[INCREMENT]} mice correlated with increased albuminuria (Figure 8, C, lanes 1 and 2, and D).

Figure 8: TAK1 deficiency in podocytes leads to collagen deposition and proteinuria. Kidney sections obtained from (A) 4- (4W) and (B) 7-week-old (7W) Tak1^fx/fx and Tak1^Δ/Δ mice were subjected to Masson’s trichrome staining. (C) Analysis of urinary albumin excretion of four 7W Tak1^Δ/Δ mice and control mice (Tak1^fx/fx) by Coomassie-stained SDS-PAGE. Asterisks indicate the corresponding 7W mice shown in B. (D) Quantitation of urinary albumin expressed as micrograms per milligram Ucr. The number in each bar indicates the lane number in C.

Discussion

Our present findings support an important role of TAK1 signaling in the development and integrity of the kidney glomerular filtration barrier. There is ample evidence in the literature that TGF-β1 and BMPs play crucial roles in podocyte differentiation, glomerulogenesis, and nephrogenesis during kidney development and podocyte injury responses. Their inhibition has been shown to cause, among others, abnormal glomerular capillary formation and development. For instance, inhibition of TGF-β1 actions by using neutralizing TGF-β1 antibodies administered in neonatal rats markedly reduced the invasion of endothelial cells into the comma/S-shaped bodies and delayed glomerular capillary formation.⁴⁹ In addition, blockade of BMP4 function in the transgenic Nephrin-Noggin mice with podocyte-selective expression of BMP inhibitor Noggin caused enlargement of glomerular capillaries or microaneurysms and dilated Bowman's capsules with collapsed glomerular capillary tufts, which were similar to those capillaries and capsules observed in Bmp4^+/− mice.⁵⁰Nephrin-Noggin mice also showed a decrease in the number of mesangial cells in the glomeruli. Similarly, in the Tak1^{[INCREMENT]/[INCREMENT]} kidney, a major morphologic change found was the enlargement of glomerular capillaries and abnormal capillary development. Thus, it is suggested that survival and differentiation of mesangial cell might be dependent on successful establishment of a glomerular capillary system. However, the dilated Bowman’s capsule with collapsed glomerular capillary tufts seen in the kidney from transgenic Nephrin-Noggin mice was a rare finding in the Tak1^{[INCREMENT]/[INCREMENT]} kidney. Indeed, it is well known that BMP4 induces the activation of TAK1 as well as other signaling pathway, including the Smad signaling pathway. Therefore, it is plausible that the mechanism leading to abnormal formation and development of glomerular capillaries might be caused by the blockade of BMP4-induced TAK1 activation in the podocytes. By contrast, the dilation of Bowman’s capsule with collapse of glomerular capillary tufts observed in the kidneys from Nephrin-Noggin mice may be mediated primarily by other signaling pathways induced by BMP.

We identified that kidneys from Tak1^{[INCREMENT]/[INCREMENT]} mice had increased expression of VEGF in the glomeruli along the capillaries as well as the podocytes. Previous studies have shown that podocyte-specific VEGF (nephrin-VEGF₁₆₄)-overexpressing mice developed a collapsing glomerulopathy featuring collapsed glomerular tufts, the few visible patent capillary loops were larger in diameter, and the mice were clinically unwell and exhibited albuminuria and death at 5 days of age.³² In the Tak1^{[INCREMENT]/[INCREMENT]} mice, although the underlying molecular mechanism for the enhanced VEGF expression in the podocytes is not yet understood, it has been previously reported that HIF-1α positively regulates the transcription of the VEGF gene during glomerulogenesis.⁵¹ More recently, Cook et al.⁵² have reported evidence that activator protein-1 transcription factor JunD controls VEGF expression in podocytes. The study showed that, in both cultured glomeruli from Jund knockout (Jund^−/−) mice and small interfering RNA–mediated JunD knockdown in immortalized human podocytes, the expressions of VEGF and HIF-1α were increased.⁵² JunD is one of the Jun family transcription factors regulated by JNK MAPK, a downstream target of TAK1 signaling cascade. In addition, our data show that the inhibition of TAK1 signaling in immortalized human podocytes using TAK1 inhibitor LZ resulted in increased mRNA expressions of VEGF and HIF-1α. Therefore, it is possible that VEGF expression might be increased by the elimination of JunD function in the TAK1-deficient podocytes in Tak1^{[INCREMENT]/[INCREMENT]} mice. Furthermore, it is notable that the nephrin-VEGF₁₆₄ transgenic mice developed proteinuria soon after birth,³² like in our findings in Tak1^{[INCREMENT]/[INCREMENT]} mice, suggesting that the increased expression of VEGF, at least in part, might be a cause of albuminuria in the Tak1^{[INCREMENT]/[INCREMENT]} mice.

WT1 has been shown to play crucial roles in embryogenesis. During kidney development, WT1 is weakly expressed in the metanephric blastema but increased in the comma/S-shaped bodies, and subsequently, WT1 expression progressively decreases and becomes restricted primarily to podocytes.⁵³ Immunofluorescence staining of WT1 in the control Tak1^fx/fx kidney represented the typical expression pattern in the normal developing kidney. We show that WT1 expression was similar in the Tak1^{[INCREMENT]/[INCREMENT]} and Tak1^fx/fx mouse kidneys until the capillary loop stage, but it was dramatically reduced in the maturing and mature glomeruli of the Tak1^{[INCREMENT]/[INCREMENT]} kidney. Given that Cre recombination occurs during the capillary loop stage in glomerular development, the reduction of WT1 expression beyond this stage is temporally correlated with Tak1 deletion by the expression of Nphs2-Cre in the Tak1^{[INCREMENT]/[INCREMENT]} mice, and these results corroborate the tenet that WT1 expression was repressed because of the loss of TAK1 activity in the podocytes. In functional studies in vitro using immortalized human podocytes, we showed that inhibition of TAK1 signaling suppressed WT1 mRNA expression, supporting evidence for positive regulation of WT1 by TAK1.

Interestingly, in studies using transgenic mice expressing a dominant negative mutant form of WT1 in podocytes, abnormal capillary development was observed with unusually wide and enlarged capillaries within the glomeruli,⁵⁴ similar to our findings in the Tak1^{[INCREMENT]/[INCREMENT]} mice, indicating that the reduction of WT1 function also affects glomerular capillary development. Hence, TAK1 signaling regulates both VEGF and WT1 expressions in podocytes and represents a mechanism that controls glomerular capillary development. Although retinoic acid⁵⁵ and several transcription factors,^56,57 including NF-κB,⁵⁸ are known to modulate Wt1 gene expression, the molecular mechanism regulating Wt1 gene expression in the podocytes during differentiation is poorly understood. Two NF-κB binding sites are found and conserved in the human and murine Wt1 promoter, and the transactivation of the murine Wt1 promoter is mediated by NF-κB.⁵⁹ Given that NF-κB is activated by TAK1, it is plausible that the reduction of WT1 expression in the podocytes of Tak1^{[INCREMENT]/[INCREMENT]} kidney might be caused by the blockade of NF-κB activation by TAK1 deletion.

Various target genes of WT1 are known, such as nphs1, the gene encoding nephrin. Recent evidence indicates that WT1 and NF-κB physically interact with the promoters for two homologous proteins, nephrin and Neph3 (filtrin), expressed in the podocyte slit diaphragm, and they cooperatively regulate the expression of nephrin and Neph3.⁴⁶ Thus, these results present a plausible explanation that the reduction of nephrin expression occurring in the Tak1^{[INCREMENT]/[INCREMENT]} kidney might be caused by decreased WT1 expression and failed NF-κB activation resulting from Tak1 deletion. Null mutations in the nephrin gene nphs1 cause congenital nephrotic syndrome of the Finnish type characterized by podocyte foot process effacement and absence of slit diaphragms.⁵⁹ Moreover, nephrin-deficient mice exhibit massive proteinuria and death within 24 hours of birth.⁶⁰ However, it has been noted that nephrin is involved in podocyte maturation but not survival of podocytes during glomerular development.⁶¹ Our H&E and TEM ultrastructure data confirm the existence of podocytes within the glomeruli of the Tak1^{[INCREMENT]/[INCREMENT]} kidney; however, notably, there were disruptions of the podocyte foot process architecture and effacement of foot process. Thus, these data suggest that TAK1 deficiency causes impaired podocyte foot process formation, which might be associated with downregulation of WT1 and nephrin expression. Nephrin is also downregulated in acquired glomerular diseases associated with proteinuria and foot process effacement.^62,63

In our studies, podocyte-specific Tak1 deletion also led to glomerular endothelial defects and dramatic simplification of the glomerular tufts with a reduced number of capillary branches and abnormally enlarged capillaries, irregularity and thickening of the GBM, disruption of endothelial fenestration, and presence of multiple endothelial cell nuclei within a capillary loop. Although a similar phenotype is seen in podocyte-specific VEGF-A–overexpressing kidneys,³⁴ the mechanism for capillary loop abnormalities with multiple endothelial cell nuclei in Tak1^{[INCREMENT]/[INCREMENT]} kidneys is unknown. Recent evidence indicated that Semaphorin3a (sema3a) deletion resulted in defects in renal vascular patterning, excess endothelial cells within glomerular capillaries, podocyte foot process effacement, and albuminuria, which was seen in Tak1^{[INCREMENT]/[INCREMENT]} kidneys.⁶⁴ Furthermore, it has been suggested that there is functional competition between sema3a and VEGF.⁶⁵ Therefore, it is possible that Tak1 deletion-induced overexpression of VEGF suppresses the function of sema3a that leads to multiple endothelial cell defects.

Podocyte-specific loss of Tak1 results in the reduction of the number of glomeruli, and at 4 and 7 weeks of age, the mice displayed severe proteinuria and histologic evidence of increased collagen deposition within the glomerular mesangium and adjacent tubulointerstitial area. The impact of the initial degree of podocyte depletion may represent a major mechanism driving glomerulosclerosis. Recent evidence suggests that podocyte damage and podocyte loss can induce further injury by triggering secondary damage of other podocytes and form a vicious cycle that leads to podocyte depletion and development of glomerulosclerosis.^66,67 However, these studies were done in otherwise normally developed adult glomeruli, and the impact of Tak1 deletion under these conditions will need to be investigated in the future.

It is well established that TGF-β family cytokines play important roles in glomerulogenesis and glomerulosclerosis. However, the cell signaling pathways involved in these events are poorly understood. To our knowledge, the current study is the first to show the physiologic role of TAK1 in podocytes. Our data provide evidence that TAK1, a major mediator of Smad-independent signaling pathways induced by TGF-β family cytokines, plays critical roles in the podocyte differentiation and glomerular capillary formation during kidney development and the integrity of the glomerular filtration barrier, which may be mediated through the regulation of the expression of VEGF, WT1, and nephrin in podocytes.

Concise Methods

Reagents

Recombinant human TGF-β1 and goat anti-mouse nephrin antibody (AF3159) were obtained from R&D Systems (Minneapolis, MN). Polyclonal antibodies against TAK1 (sc-7162), VEGF-A (sc-507), WT-1 antibody (sc-192), and horseradish peroxidase-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against phospho–Thr-187-TAK1 (p¹⁸⁷-TAK1), phospho-JNK, JNK, phospho-p38, and p38 were obtained from Cell Signaling Technologies (Beverly, MA). Anti-CD31 monoclonal antibody (14–0311) was obtained from eBioscience (San Diego, CA). Texas Red-conjugated donkey anti-goat IgG and Alexa Fluor 488-conjugated goat anti-rabbit IgG were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA) and Invitrogen (Grand Island, NY), respectively. Pharmacological inhibitor of TAK1, LZ (3S,5Z,8S,9S,11E)-3,4,9,10-tetrahydro-8,9,16-trihydroxy-14-methoxy-3-methyl-1H-2-benzoxacyclotetradecin-1,7(8H)-dione (LL-Z1640-2; herein referred to as LZ), was obtained from Alexis Biochemicals.

Knockout Animal

Podocin (Nphs2) promoter-driven Cre recombinase transgenic mice (Nphs2-Cre)⁶⁸ with mixed genetic background were backcrossed eight times with C57BL/6J mice to obtain Nphs2-Cre mice carrying nearly pure C57BL/6J genetic background. The new C57BL/6J Nphs2-Cre mice were crossed with C57BL/6J mice carrying the floxed allele of Tak1 (Tak1^fx/fx mice).^69,70 Heterozygous Nphs2-Cre:Tak1^fx/+ mice were obtained by crossing Tak1^fx/fx mice with Nphs2-Cre mice in the expected Mendelian ratio and survived to adulthood with no apparent defects. Homozygous podocyte-specific Tak1-deficient mice (Tak1^{[INCREMENT]/[INCREMENT]}) were produced by breeding heterozygous mice (Nphs2-Cre:Tak1^fx/+). The genotype of the newborn mice was confirmed by PCR with tail DNA and primer set 1 (forward: 5′-GGCTTTCATTGTGGAGGTAAGCTGAGA-3′; reverse: 5′-GGAACCCGTGGATAAG TGCACTTGAAT-3′). To confirm the deletion of the part of Tak1 by Cre recombinase, other PCR primers (primer set 2: forward: 5′-GCAACTTCGACAACTTGCTTCCTGTG-3′; reverse: 5′-GCACTTGAATTAGCGGCCGCAAGCTTATAACT-3′) were used; only deleted Tak1 produces apparent 1.2-kb DNA after PCR. The binding sites of the PCR primer sets and PCR product sizes are shown in Supplemental Figure 2. Mice received humane care according to the National Institutes of Health recommendations outlined in their guide for the care and use of laboratory animals. All animal experiments were approved by the Institutional Animal Care and Use Committee at Harvard Medical School.

Quantitative Real-Time PCR

Human podocytes were grown under nonpermissive conditions for 4 and 7 days at 37°C with replacement of fresh media containing 10% FBS every other day, and they were treated with 0.2 μM TAK1 inhibitor LZ for 3 and 6 hours. Cells were washed one time with cold PBS, and total RNA was isolated using Trizol Reagent (Invitrogen). After DNAse digestion, the first cDNA was synthesized using the AffinityScript multiple temperature cDNA Synthesis Kit (Agilent Technologies). For real-time PCR, CFX96 Touch Real-Time PCR Detection System (Bio-Rad) with Quantifast SYBR green PCR kit (Qiagen) was used according to the manufacturer’s instruction. Samples were run in duplicate, and experiments were repeated a minimum of three times with appropriate primers for VEGF, HIF-1α, WT1, and nephrin (Supplemental Table 1). The cycle threshold of amplified products was detected using CFX Manager software. The expression level of target mRNAs was normalized to β-actin and represented as a fold change relative to untreated sample.

Tissue Analyses

Kidney tissues were fixed with 4% paraformaldehyde and embedded in paraffin, and 5-μm-thick sections were collected for histology. H&E staining for histology and Masson’s trichrome staining for the detection of collagen deposition were performed. The paraffin sections were also subjected to immunohistochemistry with the Vectastain ABC kit (Vector Laboratories) and 3,3′-diaminobenzidine according to the manufacturer’s instructions. Primary antibodies and dilutions were used: anti-CD31 (14–0311; 1:300; eBioscience) and anti–VEGF-A (sc-507; 1:500; Santa Cruz Biotechnology). For the immunofluorescence staining, freshly harvested kidneys were embedded in optimal cutting temperature compound, frozen in liquid nitrogen, and stored at −80°C. Cryosections were fixed in acetone precooled at −20°C, permeabilized with 0.3% Triton-X, and blocked in 2%–5% donkey serum, 1% BSA, 0.1% gelatin, 0.1% Triton-X, 0.05% Tween, and 0.05% sodium azide. Rabbit anti-WT1 antibody (sc-192; 1:10; Santa Cruz Biotechnology) and goat anti-mouse nephrin antibody (AF3159; 1:200; R&D) were used as the primary antibodies, and Texas Red-conjugated donkey anti-goat IgG (1:500; Jackson ImmunoResearch Laboratories) or Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:500; Invitrogen) was used to detect the corresponding primary antibodies. Images were obtained using a Nikon E-1000 inverted laser microscope with digital camera (Nikon Coolpix 990).

Quantitative Immunofluorescence

The analysis of the level of fluorescence intensity for WT1 and nephrin expression in kidney was performed using the ImageJ program (National Institutes of Health Image) and calculated as mean fluorescence intensity per glomerulus or vesicular structure in each image. Samples from each experiment were stained on the same occasion and measured together. The total pixel intensity for each glomerulus of 20 glomeruli per section was measured, and mean pixel intensity of 20 glomeruli for each mouse was obtained.

TEM and Morphology

Kidneys obtained at P1 and P4 were cut in half just after harvest and fixed in fixative solution (2.5% glutaraldehyde, 1.25% paraformaldehyde, and 0.03% picric acid in 0.1 M sodium cacodylate buffer [pH 7.4]) for at least 2 hours at room temperature. The fixed kidney was washed with 0.1 M cacodylate buffer, postfixed with 1% osmium tetroxide (OsO4)/1.5% potassium ferrocyanide (KFeCN6) for 1 hour, washed in water three times, and incubated in 1% aqueous uranyl acetate for 1 hour followed by two washes in water and subsequent dehydration in grades of ethanol (10 minutes each with 50%, 70%, and 90% and 10 minutes each for two times with 100%). The samples were then put in propylene oxide for 1 hour and infiltrated overnight in a 1:1 mixture of propylene oxide and TAAB Epon (Marivac Canada Inc., St. Laurent, Canada). The next day, the samples were embedded in TAAB Epon and polymerized at 60°C for 48 hours. Ultrathin sections (approximately 60 nm) were cut on a Reichert Ultracut-S microtome, picked up on copper grids stained with lead citrate, and examined in a TecnaiG² Spirit BioTWIN, and images were recorded with an AMT 2k CCD camera. Basement membrane thickness was established with the ImageJ program. Measurements of distance between the endothelial cytoplasmic membrane and the outer lining underneath the cytoplasmic membrane of the podocyte foot processes were made perpendicular to the axis of the GBM every 500 nm for 70- to 80-µm GBM length on 4800× TEM images. To quantitate glomerular endothelial cells in Tak1^fx/fx and Tak1^{[INCREMENT]/[INCREMENT]} kidneys, endothelial cell nuclei were counted in 10 capillary loops per kidney section (n=3 in each group) on 1900× TEM images.

Urine Analyses

Urine samples of newborn mice were obtained from the bladder without blood contamination during euthanization. In the case of adult mice (4- and 7-week-old mice), 24-hour urine samples were collected using metabolic cage. Urine samples after collection were centrifuged at 14,000 rpm for 3 minutes. Supernatant was transferred into a new tube, and 2 μl supernatant was applied to 10% SDS-PAGE followed by Coomassie blue staining for urinary albumin analysis. As a control, 2 μg BSA was loaded. Remaining urine was stored at −70°C for the quantitation of albumin and creatinine. The frozen urine samples were thawed and centrifuged at 14,000 rpm for 15 minutes. Quantitation of urinary albumin and urine creatinine was carried out using mouse albumin-specific ELISA kits (ab108792; Abcam, Inc.) and creatinine assay kits (500701; Cayman Chemicals) following the manufacturer’s instructions.

Western Blot Analyses

Cells were washed one time with ice-cold PBS and lysed in buffer containing 1% NP-40, 20 mM Tris (pH 8.0), 150 mM NaCl, 12.5 mM β-glycerophosphate, 1.5 mM MgCl₂, 2 mM EGTA, 1 mM NaF, 2 mM dithiothreitol, 1 mM Na₃VO₄, 1 mM PMSF, and 20 µM aprotinin. Cells were disrupted using sonication and then centrifuged for 15 minutes at 14,000×g at 4°C to remove cellular debris. The protein concentration of cell lysates was determined by BCA protein assay reagent kit (Pierce). For Western blotting, protein samples (20 μg) were subjected to 10% SDS-PAGE and then transferred to polyvinylidene difluoride membranes. The membranes were blocked with either 5% nonfat milk or 5% BSA for 1 hour and then incubated with primary antibodies overnight on a rocker at 4°C. The membranes were washed three times (15 minutes each) with buffer of 10 mM Tris (pH 7.5), 50 mM NaCl, and 0.05% Tween 20 and then incubated with horseradish peroxidase-conjugated secondary antibodies for 30 minutes at room temperature. The target proteins were detected with LumiGLO (Cell Signaling Technology).

Statistical Analyses

Results are presented as means±SEMs; t test was used to analyze the difference between the two groups. Values were regarded as significant at P<0.05; all error bars represent SEM.

Disclosures

None.

The authors thank Dr. Shizuo Akira (Osaka University) and Dr. Jordan A. Kreidberg (Children’s Hospital, Harvard Medical School) for sharing Tak1^fx/fx and Nphs2-Cre mice, respectively.

This work was supported, in part, by the Carl W. Gottschalk Research Scholar Grant from the American Society of Nephrology (to S.I.K.), Beginning Grant-in-Aid 0665379U from the American Heart Association (to S.I.K.), National Institutes of Health Grant R01-DK57661 from the National Institute of Diabetes and Digestive and Kidney Diseases (to M.E.C.), and the M. James Scherbenske Grant from the American Society of Nephrology (to M.E.C.).