Growth-dependent podocyte failure causes glomerulosclerosis

Abstract

Podocyte depletion leads to glomerulosclerosis, but whether an impaired capacity of podocytes to respond to hypertrophic stress also causes glomerulosclerosis is unknown. We generated transgenic Fischer 344 rats that express a dominant negative AA-4E-BP1 transgene driven by the podocin promoter; a member of the mammalian target of rapamycin complex 1 (mTORC1) pathway, 4E-BP1 modulates cap-dependent translation, which is a key determinant of a cell's hypertrophic response to nutrients and growth factors. AA-4E-BP1 rat podocytes expressed the transgene and had normal kidney histology and protein excretion at 100 g of body weight but developed ESRD by 12 months. Proteinuria and glomerulosclerosis were linearly related to both increasing body weight and transgene dose. Uni-nephrectomy reduced the body weight at which proteinuria first developed by 40%-50%. The initial histologic manifestation of disease was the appearance of bare areas of glomerular basement membrane from the pulling apart of podocyte foot processes, followed by adhesions to the Bowman capsule. Morphometric analysis confirmed the mismatch between glomerular tuft volume and total podocyte volume (number × size) per tuft in relation to weight gain and nephrectomy. Proteinuria and glomerulosclerosis did not develop if dietary calorie restriction prevented weight gain and glomerular enlargement. In summary, failure of podocytes to match glomerular tuft growth in response to growth signaling through the mTORC1 pathway can trigger proteinuria, glomerulosclerosis, and progression to ESRD. Reducing body weight and glomerular growth may be useful adjunctive therapies to slow or prevent progression to ESRD.

Figure 1.

Relationships between body weight and glomerular tuft volume (closed triangles) and body weight and total podocyte volume as measured by the GLEPP1-positive tuft volume (open squares). In wild-type Fischer 344 intact rats, glomerular tuft volume increased exponentially in relation to body weight gain in rats kept on an ad libitum diet. Podocyte volume also increased in relation to body weight gain, but not at the same rate as the glomerular tuft volume, representing an apparent mismatch developing between these two variables as body weight increases.

Figure 2.

Characterization of AA-4E-BP1 transgenic rats. (A) Expression of human 4E-BP1 by immunofluorescence (red) in podocytes as identified by WT1-positive nuclear immunofluorescence (green) in homozygous transgenic rats (upper panels) and wild-type rats (lower panels). The right two panels show DAPI-stained nuclei for the same images. n=3. Ab, antibody. Original magnification, ×160 (B) Western blots of protein extracts from isolated glomeruli of wild-type (WT), heterozygous (Het), and homozygous (Homo) rats. Note that the heterozygous and homozygous glomeruli contain hypophosphorylated lower-molecular-weight bands not seen in wild-type glomeruli (upper blot), and antihuman 4E-BP1 antibodies recognize bands in the heterozygous and homozygous rat glomerular extracts not present in wild-type extracts (middle blot). β-actin is shown as a reference loading marker. n=2 per experiment, with two experiments performed. (C) SEM (top panel) and TEM (bottom panel) images of homozygous AA-4E-BP1 rats, illustrating normal podocyte structure at 100-g body weight. n=4 for SEM and n=2 for TEM. Original magnification ×2500. (D) Western blots of wild-type, heterozygous, and homozygous isolated glomerular extracts demonstrating increased S6 protein and phospho-S6 in heterozygous and homozygous rats compared with wild-type rats; there is no obvious difference in expression of podocyte products (nephrin, podocin, GLEPP1, and vascular endothelial growth factor [VEGF]). β-actin is shown as a reference loading marker. n=2 per experiment, with two experiments performed. (E) Glomerulus from a 100-g homozygous AA-4EBP1 rat developed for phospho-S6 immunofluorescence demonstrating predominant expression of phospho-S6 in podocytes as confirmed by WT1-positive (green) nuclei. The right panel shows DAPI-stained nuclei in the same image. n=3. Original magnification ×160. (F) Hematoxylin and eosin (H and E)–stained sections of rat renal cortex from an 11-month-old heterozygous AA-4EBP1 rat (left panel) versus the cortex from an 11-month-old wild-type Fischer 344 rat (right panel), demonstrating that ESRD had developed in the transgenic rat. n=7. Original magnification ×50.

Figure 3.

Relationship between proteinuria, age, and weight. (A) Time course (starting at 100-g body weight) for development of proteinuria of wild-type, wild-type uni-nephrectomized (NX), heterozygous, heterozygous uni-nephrectomized, homozygous, and homozygous uni-nephrectomized rats. Note that the time course of proteinuria varied depending on transgene dose and whether the rats had been uni-nephrectomized. (B) Same data plotted against body weight on the x-axis, demonstrating that as weight increases a weight proteinuria threshold is crossed; beyond this weight threshold, the proteinuria increases linearly in relation to weight gain. Nephrectomy reduced this threshold by 40%–50% for each rat strain (wild-type, heterozygous, or homozygous). (C) For the uni-nephrectomized rat strains, the slope of proteinuria versus weight gain in homozygous rats is approximately twice that for heterozygous rats under both nephrectomy and intact conditions, thereby demonstrating that the gene dose plays a critical role in determining both the weight proteinuria threshold and the relationship between weight gain and proteinuria. Data are shown as the mean ± SEM.

Figure 4.

Relationships between histologic features, proteinuria, and urine podocyte biomarkers in wild-type, heterozygous, and homozygous nephrectomized AA-4E-BP1 rats. Original magnification ×60. (A) Masson trichrome–stained histologic sections at 7 weeks and 14 weeks after nephrectomy (at 100-g body weight). Homozygous rats were euthanized at 9 weeks because they developed uremic symptoms. (B) Proportion of glomeruli that contained adhesions at 7 and 14 weeks (9 weeks for homozygous nephrectomized rats). Het, heterozygous; Homo, homozygous; Wt, wild-type. *P<0.05 and **P<0.01, as assessed by Kruskal-Wallis test and then Scheffe's test. (C) There was a direct relationship between the urine protein-to-creatinine ratio and the proportion of glomeruli that contained adhesions. (D) SEM images from heterozygous nephrectomized rats demonstrating that foot processes were mostly intact (upper left panel) but that glomeruli contained patches of bare GBM where the foot processes had apparently been pulled apart (three examples shown). Foot processes abutting these areas remained relatively intact. Original magnification ×4000. (E) Upper panel shows a TEM image demonstrating two adhesions (arrows) distant from the pole of the glomerulus (P), and podocytes adjacent to these areas contained granules and cysts (arrowheads). Most capillary loops were lined by intact foot processes, although occasional areas of effaced foot processes were present (not shown). Also shown are plump parietal epithelial cells adjacent to the adhesions. Original magnification ×260. Glomerular tufts could not be visualized by SEM because capsules remained attached to the tuft, presumably because of adhesions (lower panel). Original magnification ×70. (F) Time course of urine protein-to-creatinine ratio (upper panel), urine podocin-to-nephrin mRNA ratio as a measure of podocyte stress (middle panel), and urine podocin-to-aquaporin2 mRNA ratio per as a measure of the rate of podocyte loss into the urine. Note that there was no measurable increase in the rate of podocyte loss in wild-type rats, although podocyte stress was increased throughout the 14-week observation period. In contrast, podocyte stress was markedly increased in both heterozygous and homozygous rats, and the rate of podocyte loss was higher in homozygous than in heterozygous rats. As homozygous rats reached ESRD, glomeruli became depleted of podocytes, as reflected by a reduced rate of podocyte loss. Data are shown as the mean ± SEM.

Figure 5.

Effect of calorie restriction in preventing podocyte stress, podocyte loss, and adhesion formation in 300-g nephrectomized heterozygous AA-4EBP1 rats. (A) Forty percent calorie restriction resulted in loss of about 70 g in weight, whereas ad libitum–fed rats continued to gain weight over the 16-week period of observation. (B) Urine protein-to-creatinine ratio remained at baseline in calorie-restricted rats and increased in ad libitum–fed rats. (C) Representative Masson trichrome–stained sections illustrating the development of adhesions present in ad libitum–fed rats, which were not present in calorie-restricted rats. Original magnification ×110. (D) Quantitation of the adhesions. *P<0.05, as assessed by t test. (E) Both the urine podocin-to-nephrin mRNA ratio as a measure of podocyte stress and the urine podocin-to-aquaporin2 ratio as a measure of the rate of podocyte loss remained at baseline in calorie-restricted rats but increased in ad libitum–fed rats. Data are shown as the mean ± SEM.