THE ANATOMICAL RECORD 291:114–121 (2007) Podocyte Loss in Aging OVE26 Diabetic Mice JENNIFER M. TEIKEN,1 JANICE L. AUDETTEY,1 DONNA I. LATURNUS,1 SHIRONG ZHENG,2 PAUL N. EPSTEIN,2 AND EDWARD C. CARLSON1* 1 Department of Anatomy and Cell Biology, University of North Dakota, Grand Forks, North Dakota 2 Department of Pediatrics, University of Louisville, Louisville, Kentucky ABSTRACT Recent studies show that podocyte nuclear density (NV) and numbers of renal podocytes per glomerulus (N) are altered in experimental and spontaneous diabetes mellitus. NV and N are generally reduced, and it has been hypothesized that these morphological changes may relate to the loss of glomerular permselectivity in diabetic nephropathy (DN). In the current study, OVE26 transgenic diabetic mice and age-matched (FVB) controls (60, 150, or 450 days) were ﬁxed by vascular perfusion and renal cortical tissues were prepared for morphometric analyses. ImageJ software and point counting analyses were carried out on light and transmission electron micrographs to determine glomerular volume (VG), NV, and N. As expected, mean VG in OVE26 mice increased substantially (134%) over the course of the study and was signiﬁcantly increased over FVB mice at all ages. At 60 days, NV and N were not statistically distinguishable in OVE26 and control mice, while at 150 days, NV was signiﬁcantly reduced in diabetics but not N. In 450-day-old OVE26 animals, however, NV and N were both signiﬁcantly decreased (231% and 99%, respectively) relative to age-matched FVB mice. These data suggest that in the OVE26 model of diabetes, signiﬁcant podocyte loss occurs relatively late in the course of the disease. Moreover, it seems possible that these podocytic changes could play a role in sustaining the increased permeability of the blood–urine barrier in the later stages of diabetic renal decompensation. Anat Rec, 291:114–121, 2007. Ó 2007 Wiley-Liss, Inc. Key words: podocytes; diabetic mice; electron microscopy; morphometry Podocytes are structurally important components of the blood–urine barrier, and recent studies have shown that they may be critical elements in renal glomerular permeability (see Shankland, 2006, for review). It has long been appreciated that podocytes function in the ultraﬁltration of blood to form the renal glomerular ﬁltrate, and they are also known to synthesize components of the glomerular basement membrane (GBM). In addition, they provide structural stability to the glomerular tuft (Vogelmann et al., 2003). Although the GBM is believed to be the main ﬁlter in renal permselectivity, podocytic slit diaphragms perform the ﬁnal ultraﬁltration step and collectively form a size-selective barrier maintained by several proteins including nephrin, podocin, CD2AP, neph-1, a-actinin4 (Asanuma and Mundel, 2003) FAT, P-cadherin (Paventstädt et al., 2003), and Ó 2007 WILEY-LISS, INC. Grant sponsor: North Dakota Lions Foundation; Grant sponsor: PHS; Grant number: DK072032; Grant sponsor: JDRF; Grant numbers: 1-2005-88 and 3-2005-932. *Correspondence to: Edward C. Carlson, Department of Anatomy and Cell Biology, University of North Dakota, School of Medicine and Health Sciences, Room 1701, 501 North Columbia Road Stop 9037, Grand Forks, ND 58202-9037. Fax: 701-777-2477. E-mail: email@example.com Received 6 September 2007; Accepted 2 October 2007 DOI 10.1002/ar.20625 Published online in Wiley InterScience (www.interscience.wiley. com). PODOCYTE MORPHOMETRY IN DIABETIC MICE ZO-1 (Rincon-Choles et al., 2006). Slit diaphragms and their components are vital for the maintenance of healthy podocytes and also for regulating the passage of protein (Marshall, 2005). Disruptions in the ﬁltration process by diseases such as diabetes may lead to severe proteinuria and/or the passage of other plasma proteins normally retained by the podocytic slit diaphragm. Morphological hallmarks of diabetic nephropathy (DN) include GBM thickening (Adler, 1994; Viberti et al., 1994; Goode et al., 1995), expansion of the mesangial region (Steffes et al., 1989), and an accumulation of mesangial matrix (Osterby et al., 2001). More recently, however, alterations in podocyte number, structure, and function have been shown to occur with the onset and progression of diabetes, and this cell type has recently received increased attention (Pagtalunan et al., 1997; Kriz et al., 1998; Mifsud et al., 2001; Shirato et al., 2001; White et al., 2002; Dalla Vestra et al., 2003; Vogelmann et al., 2003; Petermann et al., 2004; Marshall, 2005; Wolf et al., 2005; Yu et al., 2005; Shankland, 2006; Susztak et al., 2006; Macconi et al., 2006). Podocytes typically are reduced in number in diabetes, and in patients some podocytes are recoverable in the urine as viable cells (Vogelmann et al., 2003). The OVE26 transgenic diabetic mouse model was originally created by Epstein and coworkers (1989, 1992) and displays many of the characteristics of human diabetes. These characteristics include signiﬁcant nephropathic features, which have been well characterized. As examples, OVE26 kidney weights were almost doubled between 2 and 5 months (Zheng et al., 2004), and morphometric analyses have shown an increase in GBM thickness (Carlson et al., 1997; 2003), glomerular volume, and mesangial matrix area (Zheng et al., 2004). Interestingly, signiﬁcant albuminuria was frequently seen by 2 months of age, and urinary albumin excretion (UAE) rates increased progressively with age (Zheng et al., 2004). Despite these compelling changes in renal morphology and physiology, podocyte morphometry has not been investigated in the OVE26 diabetic model. Accordingly, we hypothesized that podocyte density and number are reduced progressively in these mice. In the current study, morphometric analyses were carried out on light (LM) and transmission electron microscopic (TEM) images to determine mean glomerular volume (VG) and mean podocyte nuclear density (NV), from which average number of podocytes per glomerulus (N) was calculated. The results of our study show for the ﬁrst time signiﬁcant podocyte loss in OVE26 mice relative to agematched controls. MATERIALS AND METHODS Experimental Animals Control (FVB) and transgenic diabetic (OVE26) mice or ﬁxed renal tissues were received from the laboratories of Dr. Paul Epstein, Department of Pharmacology and Toxicology, University of North Dakota, Grand Forks, North Dakota, and Department of Pediatrics, University of Louisville, Louisville, Kentucky. Diabetic OVE26 mice on the FVB background have been maintained for 15 years in Dr. Epstein’s laboratory (Epstein et al., 1992). All animals had free access to Harlan Teklab Rodent Diet #8640 and tap water, and no insulin therapy was 115 given. All animal procedures adhered to the guidelines of the NIH Guide for the Care and Use of Laboratory Animals and were approved by the IACUC committees of the University of North Dakota and the University of Louisville. Tissue Preparation and Microscopic Technique Tissues from at least three OVE26 animals (60, 150, and 450 days of age), and at least three age-matched FVB mice (20 total animals) were used in this study. Before killing, all mice were weighed, and blood glucose levels were determined by ‘‘Lifescan’’ glucometer. All animals were killed by vascular perfusion with cold Karnovsky’s (1965) ﬁxative following sodium pentobarbital anesthesia. Kidneys from all animals were removed, weighed, and decapsulated. They were sliced longitudinally and the medulla from each slice was removed and discarded. The remaining cortical strips were cut into 1mm3 tissue blocks, and 10–15 blocks from each kidney were selected randomly by unbiased technical personnel and placed in cold ﬁxative for at least 2 hr. This process was followed by post-ﬁxation in OsO4, dehydration, and embedding in Epon/Araldite. Thick (250 nm) sections were made and stained with toluidine blue for LM observation, and thin (silver–gray interference color) sections were mounted on Formvar-coated copper slot grids and stained with uranyl acetate and lead citrate for conventional TEM. All TEM specimens were observed and photographed in a Hitachi 7500 TEM. For TEM calibration, a standard cross-grating (2,160 lines/inch) was photographed at the same optical conditions as were the tissues. All TEM prints were made at a constant enlarger setting to produce identical ﬁnal magniﬁcations of 1,960. Podocyte Morphometry Mean glomerular volume (VG). All data were derived from LM micrographs of renal cortical thick sections taken at 5-mm intervals. At least 15 and up to 20 glomerular proﬁles from each animal age and type (720 total LM images) showing well-perfused and patent glomeruli were chosen randomly by unbiased observers. Glomerular proﬁles were digitally imaged from each tissue block, and mean glomerular proﬁle areas for each animal were determined using ImageJ software. VG was calculated by the method of Weibel and Gomez (Weibel and Gomez, 1962; Weibel, 1979) from the formula: VG 5 Area1.5 3 b /K, or VG 5 Area1.5 3 1.38/1.01, (equation 1) where b (shape coefﬁcient for a sphere) is 1.38, and K (size distribution coefﬁcient assuming a 10% coefﬁcient of variation) is 1.01. Mean podocyte nuclear density (NV). All data were derived from TEM images of glomerular proﬁles. Assuming one nucleus per cell, podocyte nuclear density (i.e., the number of podocyte nuclei per unit of glomerular tissue, NV) and podocyte density are the same. Therefore, determining NV in TEM micrographs provides the desired numerical data and also conveniently obviates the difﬁculty of quantifying variable and complex podocytic cell processes, which may belong to the same cell. Also, unequivocal identiﬁcation of podocyte nuclei is facilitated at the level of ultrastructure. In an effort to avoid bias produced by variable nuclear shape and size distribution, the Weibel and Gomez 116 TEIKEN ET AL. Fig. 1. A,B: Low-power transmission electron micrographs (TEMs) of glomerular proﬁles in FVB (A) and OVE26 (B) mice at 450 days of age. TEM resolution is sufﬁcient for clear recognition of podocyte nuclei (arrows) in both genotypes. 3800. (Weibel and Gomez, 1962; Weibel, 1979) method as described by White and Bilous (2004) was used in the current study. By this technique, point counting measurements were carried out on TEM micrographs of randomly chosen sections of entire glomerular proﬁles similar to those shown in Figure 1. Measurements were performed on 9–14 TEM sectional proﬁles from animals of each age and type (65 total TEM images). Podocyte NV was determined using the formula: NV 5 K/bH (NA3/VV) (equation 2). A transparent plastic grid with ﬁne points 11.5 mm apart was superimposed on each TEM glomerular proﬁle, and the number of podocyte nuclear proﬁles (n) and ﬁne points hitting the glomerular tuft (Ptuft) were counted to estimate podocyte nuclear proﬁle area (NA): NA 5 Sn / [S Ptuft 3 area per point] (equation 3). Fine points hitting the glomerular tuft at all ages averaged 125.3 in FVB mice and 196.7 in OVE26 animals. The number of ﬁne points hitting podocyte nuclear proﬁles (Pnucleus) were then counted to calculate the volume fraction of podocyte nuclei in the glomerular tuft (VV): VV 5 SPnucleus /SPtuft (equation 4). Fine points hitting podocyte nuclei at all ages averaged 4.3 in both FVB and OVE26 mice. Podocyte NV was then calculated from equation 2, with a size distribution (K) of 1.01 and a shape constant (b) of 1.55. Mean podocyte number per glomerulus (N). For determination of N, we used the general formula for numerical density: N 5 VG 3 NV, using VG from equation 1 and NV from equation 2. used. Statistical analyses were carried out using SigmaStat 3.0 (SPSS, Inc.). RESULTS Blood Glucose Blood glucose levels in nonfasted FVB control animals at all ages ranged from 84–230 mg/dl and averaged 134.5 mg/dl. As early as 50 days of age, blood glucose values in OVE26 mice were more than twice that of controls, and by 150 days, OVE26 values were approximately 435 mg/dl. Frequent testing of older diabetic animals conﬁrmed that blood glucose levels remained high throughout their lifespan (average 443.5 mg/dl) and occasionally exceeded 500 mg/dl. At each age, blood glucose was signiﬁcantly higher in OVE26 mice than in age-matched controls. Body Weight Between 60 and 150 days average body weights of FVB mice increased from 23.8 g to 32.7 g (Table 1). However, by 450 days, their weights decreased slightly to 30.9 g (total weight gain 30%). Similarly, body weights in the OVE26 model increased between 60 and 150 days from 21.7 g to 29.5 g, and then decreased at 450 days to 28.5 g (total weight gain 31%). Mean OVE26 body weights were slightly lower than FVB body weights in all age groups. Kidney Weight Statistics Data were analyzed using the unpaired t-test. Level of signiﬁcance for all tests was set at P < 0.05 and power was set at 0.8. In those cases where data did not meet assumptions of normal distribution and equal variance, the nonparametric Mann-Whitney rank-sum test was Mean kidney weights of FVB mice increased 24% between 60 and 150 days of age (Table 1). Substantial gains also occurred between 150 and 450 days. In OVE26 mice, kidney weight increases were even more dramatic (135%) between 60 days and 150 days, and the increase was maintained at 450 days. Although 117 PODOCYTE MORPHOMETRY IN DIABETIC MICE TABLE 1. Mean body and kidney weights of FVB and OVE26 mice 6 SEMa Mean body weight (g) Age (days) 60 150 450 Mean kidney weight (g) FVB OVE26 FVB OVE26 23.8 6 1.4 32.7 6 0.5 30.9 6 1.3 21.7 6 1.2 29.5 6 1.1 28.5 6 1.1 0.21 6 0.01 0.26 6 0.01 0.33 6 0.01 0.20 6 0.01 0.47 6 0.02* 0.46 6 0.01* a Asterisks indicate that comparisons between FVB and OVE26 show statistical signiﬁcance (P < 0.05). Fig. 2. A,B: Transmission electron micrographs of podocytes (P) in 450-day-old FVB (A) and OVE26 (B) mice. 36,800. OVE26 mean kidney weights were somewhat lower (5.0%) than controls at 60 days of age, at 150 and 450 days, mean kidney weights in the diabetic mice were 81% and 39% greater, respectively, than age-matched normal mice. Renal Glomerular Morphology At 60 days of age, renal glomeruli from FVB and OVE26 mice were indistinguishable by LM and TEM. However, by 150 days, they were easily differentiated. At that time, glomeruli in OVE26 mice appeared larger in diameter than those of FVB controls. Diabetic animals also exhibited thicker GBMs and expanded mesangial regions owing primarily to increased mesangial matrix. By 450 days, the morphological differences were further exaggerated (Fig. 1). In FVB mice, podocytes exhibited typical ultrastructural features at all ages (Fig. 2A). Primary processes extended from cell bodies and secondary processes with associated foot processes showed features considered typical for this cell type. Although at 60 days podocytes in OVE26 and control mice were virtually indistinguishable, by 150 days occasional foot process broadening, or effacement was noted in the diabetics and at 450 days signiﬁcant effacement was common in the diabetic mice. In the 450-day-old diabetics, slits between foot processes were much farther apart and the narrow cross-sectional foot process proﬁles seen in controls were replaced by club-like extensions that covered large areas of the GBM (Fig. 2B). Podocyte Morphometry Mean glomerular volume (VG). Mean glomerular proﬁle areas were determined in mm2 from digital images using ImageJ software as described in the Materials and Methods section. These were used to calculate mean VG for FVB and OVE26 mice. In FVB mice, VG increased slightly (9%) between 60 and 150 days of age (Table 2; Fig. 3A). However, the increase was much greater between 150 and 450 days (41%) resulting in a 53% overall increase between ages 60 and 450 days. As expected, however, changes were much greater in aging diabetic mice where substantial increases in VG occurred between 60 days and 150 days (111%) followed by a slight rise (11%) between 150 and 450 days. Between 60 and 450 days, mean VG increased 134% in the OVE26 strain. At 60 days, VG differences between OVE26 and FVB mice were small but signiﬁcant (19%; P < 0.001). Signiﬁcant differences occurred also at; both 150 and 450 days (131%; P 5 0.001, and 81%, P 5 0.019, respectively) when VG increased dramatically in the diabetics (Table 2; Fig. 3A). Mean podocyte nuclear density (NV). In control FVB mice, mean NV increased slightly between 60 and 150 days (7.29 3 1024/mm3 vs. 7.84 3 1024/mm3), but then decreased to 5.19 3 1024/mm3 at 450 days, resulting in an overall decrease in NV with age (Table 2; Fig. 3B). In OVE26 mice, a substantial mean NV decrease (86%) occurred from 5.92 3 1024/mm3 at 60 days, to 3.19 3 1024/mm3 at 150 days and further decreased to 1.57 3 1024/mm3 by 450 days of age (27% of 60-day value). At 60 days, mean podocyte NV in OVE26 mice was slightly less (23%) than in FVB mice. However, at 150 days NV was signiﬁcantly less (146%; P < 0.001) in OVE26 mice. Moreover, at 450 days OVE26 mice showed 118 TEIKEN ET AL. TABLE 2. Mean glomerular volume, podocyte nuclear density, and podocyte number per glomerulus in FVB and OVE26 mice 6 SEMa VG: mean glomerular volume 3105mm3 Age (days) 60 150 450 a NV: mean podocyte nuclear density 31024/mm3 N: mean podocyte number per glomerulus FVB OVE26 FVB OVE26 FVB OVE26 2.09 6 0.16 2.27 6 0.21 3.20 6 0.50 2.48 6 0.13* 5.24 6 0.47* 5.80 6 0.47* 7.29 6 0.94 7.84 6 1.33 5.19 6 1.14 5.92 6 1.26 3.19 6 0.31* 1.57 6 0.26* 151.33 6 19.79 172.85 6 23.77 164.15 6 26.52 147.09 6 30.68 165.35 6 14.78 82.56 6 12.98* Asterisks indicate that comparisons between FVB and OVE26 mice show statistical signiﬁcance (P < 0.05). Fig. 3. A–C: Mean glomerular volume (A), mean podocyte nuclear density (B), and mean podocyte number per glomerulus (C) 6 SEM in normal (FVB) and transgenic diabetic (OVE26) mice. Normal mice are represented by open bars; diabetic mice are represented by ﬁlled bars. Asterisks indicate that comparisons between FVB and OVE26 mice show statistical signiﬁcance (P < 0.05). 231% fewer podocyte nuclei/mm3 of glomerular tissue than age-matched controls (Table 2; Fig. 3B). Mean podocyte number per glomerulus (N). For aging FVB mice, the mean N remained fairly constant (151 at 60 days, 172 at 150 days, and 164 at 450 days; Table 2; Fig. 3C). Of interest, N values in OVE26 animals were similar to FVB mice at both 60 days of age and 150 days of age. However, a substantial reduction (100%) occurred between 150 days and 450 days when the mean N in the oldest OVE26 mice was reduced to 83 (56% of 60 day values). When average N was compared in FVB and OVE26 mice, controls and diabetics were similar at 60 days and 150 days. However, at 450 days, N was signiﬁcantly less (99%; P 5 0.005) in diabetic mice than in age-matched controls (Table 2; Fig. 3C). DISCUSSION The OVE26 transgenic diabetic mouse is particularly interesting because, despite substantial pancreatic bcell–speciﬁc damage (Epstein et al., 1989, 1992) and blood glucose levels frequently exceeding 600 mg/dl (Zheng et al., 2004), insulin secretion is sufﬁcient to allow these mice to live as long as 450 days without treatment. Signiﬁcantly, these mice show numerous PODOCYTE MORPHOMETRY IN DIABETIC MICE morphological features that parallel chronic complications of DN in humans (Carlson et al., 1997, 2003; Zheng et al., 2004). Moreover, they are prone to early and frequently severe microalbuminuria and the major blood pressure effect is hypertension (Zheng et al., 2004). On the contrary, throughout their life-span, FVB mice show no signiﬁcant functional or morphological complications considered characteristic of DN (Carlson et al., 1997, 2003; Zheng et al., 2004), despite their tendency toward blood glucose levels that could be considered hyperglycemic in human (Zheng et al., 2004). Numerous reports and reviews suggest that podocytes may be major players in the development of DN (Marshall, 2005; Shankland, 2006) and that podocyte injury or loss may be directly or indirectly related to proteinuria. In the current study, we attempted to take advantage of the unique nephropathic features of diabetic OVE26 mice in an effort to compare their mean values of VG, NV and N, with those in age-matched FVB control mice. In patients, DN is characterized by renal and glomerular hypertrophy (Viberti et al., 1994). In the current study, renal hypertrophy was evident in OVE26 mice by 150 days of age when kidney weights exceeded controls by approximately 81%. Similar data were reported by Zheng and coworkers (2004), who showed that at 5 months, kidney weights in OVE26 mice were 97% greater than those from controls. Both studies showed that body weights in diabetics and age-matched controls were not signiﬁcantly different. In the current study, visual inspection of LM sections of renal cortex in OVE26 animals showed glomerular proﬁles that were larger in diameter than in agematched controls. This ﬁnding was conﬁrmed by morphometric analyses carried out by the method of Weibel and Gomez (Weibel and Gomez, 1962; Weibel, 1979) as described by Lane and coworkers (1992) who demonstrated that when at least 15 glomerular proﬁles were analyzed and adequate correction factors were applied, reliable estimates of VG could be determined. Using this method, we showed that at 60, 150, and 450 days, VG was signiﬁcantly increased in OVE26 mice over agematched controls. This ﬁnding was not unexpected as glomerular hypertrophy is a consistent feature of both early and late diabetic glomerulopathy (Osterby and Gundersen, 1975, 1977). Moreover, using similar techniques on periodic acid-Schiff–stained parafﬁn sections, Zheng and associates (2004) showed that glomerular volumes in OVE26 mice at 2.5, 6, and 10 months of age were signiﬁcantly enlarged relative to FVB control mice. Because number of podocytes per glomerulus (N) was a goal of the current studies and LM sections could not be resolved sufﬁciently to quantify podocytes unequivocally, low-magniﬁcation TEM sections of entire glomerular proﬁles were used to determine NV, and ultimately N was calculated for each glomerulus studied. Several methods were available for determining NV, including the disector method (Sterio, 1984), which some consider to be the gold standard technique. However, this procedure required exhaustive sectioning of each glomerulus and was not practical for our investigation. Because TEM images were required and several investigators (Steffes et al., 2001; White and Bilous, 2004) indicated that the results using the disector method were not statistically different than those generated by the single- 119 section Weibel and Gomez (Weibel and Gomez, 1962; Weibel, 1979) technique, provided the data were generated from electron micrographs, the latter method was chosen. As a caveat, however, it should be pointed out that a recent elegant light microscopic cell counting study (Basgen et al., 2006) showed that the Weibel and Gomez (Weibel and Gomez, 1962; Weibel, 1979) counting method overestimated total glomerular cell counts by 10%. Our data showed that, at 150 days and 450 days of age, podocyte NV in OVE26 mice was signiﬁcantly reduced (146% and 231%, respectively) relative to age-matched controls. Similar studies have not been carried out previously on OVE26 mice, so no direct comparisons can be made. Nevertheless, it has been well documented that podocyte numbers are reduced in human (Pagtalunan et al., 1997; Shirato et al., 2001; Steffes et al., 2001; White et al., 2002; Dalla Vestra et al., 2003) and experimental (Petermann et al., 2004; Susztak et al., 2006) DN. Although the precise mechanisms by which podocytes are lost have not been elucidated, it has been shown that glucose-induced apoptosis plays a role in the process (Susztak et al., 2006). Of interest, however, it is likely that not all lost podocytes die, as viable podocytes have been recovered in the urine of streptozotocin-induced diabetic rats (Petermann et al., 2004) and human subjects with kidney disease (Vogelmann et al., 2003). Regardless of the pathogenetic mechanisms it is clear that, in human DN, podocyte loss is signiﬁcant and on average approximately 20% of podocytes are lost within the ﬁrst few years of the disease (Steffes et al., 2001). In the current study, podocyte NV was signiﬁcantly reduced in diabetic mice at 150 and 450 days of age. However, NV was calculated as one step in determining average N, and, although at 450 days N was signiﬁcantly reduced (99%) in OVE26 animals compared with normal mice, at 60 and 150 days, N was nearly identical in diabetic and controls. These results were somewhat unexpected as it has been shown that, in DN patients (Pagtalunan et al., 1997; Shirato et al., 2001; Steffes et al., 2001; White et al., 2002; Dalla Vestra et al., 2003) and in several diabetic animal models (Petermann et al., 2004; Susztak et al., 2006), N is generally reduced. It should be pointed out, however, that in the study by White and coworkers (2002), there was no reduction in podocyte numbers at baseline, and only after 3 yr was there a signiﬁcant podocyte loss, suggesting that this change can occur later in the disease. We are intrigued by the signiﬁcant reductions in NV in 150- and 450-day-old diabetic mice and its variable affect on N, which is necessarily directly related to concomitant increases in VG in these animals. In fact, because N is dependent upon VG and NV, which are inversely related in OVE26 and control mice, it is not surprising that diabetics and age-matched controls show relatively small differences in N. An important question is thus raised: Is podocyte NV or N the more functionally relevant factor? In this regard, it must be pointed out that a decrease in NV can be a result of (1) increased VG, (2) an absolute decrease in N, or (3) a combination of both factors. Because a decrease in N is not necessary in order for a decrease in NV to occur, it is possible that NV may be the most relevant (and independent) data point in determining 120 TEIKEN ET AL. glomerular function. Similar conclusions are expressed by Dalla Vestra and coworkers (2003) who emphasize the signiﬁcance of NV and point out that N does not provide important information on glomerular architecture such as how many cells may be available to cover the peripheral GBM in a large or small glomerular tuft. Further support for this position is derived from human studies in which decreased podocyte NV correlates with increased albumin excretion rates, suggesting a link between reduced NV and permeability to albumin (Dalla Vestra et al., 2003). We recently showed that OVE26 mice exhibit pronounced UAE early in life and that it increases progressively with age (Zheng et al., 2004). Moreover, the data indicate that, although the proteinuria is variable within animal groups, by 9 months, average UAE exceeds 15 mg/24 hr. It seems possible, therefore, that the changes in NV and N shown in the current study at 5 and 15 months may reﬂect podocyte losses that correlate with the degree of increased proteinuria. Additional support for this concept is offered in a recent transgenic rat study that demonstrated a quantitative relationship between the degree of podocyte depletion and the development of proteinuria and glomerulosclerosis (Wharram et al., 2005). It should be pointed out that the effects of blood pressure changes in OVE26 mice might be related to podocyte damage and loss. In this regard, it has been shown that podocytes respond to injury by increased production of extracellular matrix, detachment, and hypertrophy (Kriz, 1996; Barisoni et al., 2000), and the latter is noted speciﬁcally in response to stretch and mechanical stress (Petermann et al., 2005). Because the primary blood pressure effect in OVE26 mice is hypertension (Zheng et al., 2004), it seems possible that the increased vasoactivity in this model may be linked to podocyte hypertrophy and reduced NV in late stages of life. In conclusion, our data show that, in OVE26 transgenic diabetic mice, podocyte density and number decrease signiﬁcantly with age. Because these changes occur relatively late in the course of the disease, it seems possible that they may be a consequence of early molecular diabetic nephropathic sequelae. Moreover, because podocytes apparently do not divide in vivo (Kriz, 1996), they have little or no chance for cell proliferation in response to injury. Accordingly, the observed reduction in podocyte numbers may be largely irreversible and may play a role in sustaining the well-known increased permeability of the blood–urine barrier in late stages of diabetic renal decompensation. ACKNOWLEDGMENTS The authors thank Dr. Kap Lee for superb help with animal health care and surgery, and Dr. Patrick Carr for expert assistance with the statistical methods. E.C.C. was funded by the North Dakota Lions Foundation; P.N.E. was funded by a PHS grant and a JDRF grant; and S.Z. was funded by a JDRF grant. LITERATURE CITED Adler S. 1994. Structure-function relationships associated with extracellular matrix alterations in diabetic glomerulopathy. J Am Soc Nephrol 5:1165–1172. Asanuma K, Mundel P. 2003. The role of podocytes in glomerular pathology. Clin Exp Nephrol 7:255–259. Barisoni L, Mokrzycki M, Sablay L, Nagata M, Yamase H, Mundel P. 2000. Podocyte cell regulation and proliferation in collapsing glomerulopathies. Kidney Int 58:137–143. Basgen JM, Nicholas SB, Mauer M, Rozen S, Nyengaard JR. 2006. Comparison of methods for counting cells in the mouse glomerulus. Nephron Exp Nephrol 103:E139–E148. Carlson EC, Audette JL, Klevay LM, Nguyen H, Epstein PN. 1997. Ultrastructural and functional analyses of nephropathy in calmodulin-induced diabetic transgenic mice. Anat Rec 247:9–19. Carlson EC, Audette JL, Veitenheimer NJ, Risan JA, Laturnus DI, Epstein PN. 2003. Ultrastructural morphometry of capillary basement membrane thickness in normal and transgenic diabetic mice. Anat Rec 271A:332–341. Dalla Vestra M, Masiero A, Roiter AM, Saller A, Crepaldi G, Fioretto P. 2003. Is podocyte injury relevant in diabetic nephropathy? Studies in patients with type 2 diabetes. Diabetes 52:1031–1035. Epstein PN, Overbeek PA, Means AR. 1989. Calmodulin-induced early-onset diabetes in transgenic mice. Cell 58:1067–1073. Epstein PN, Ribar TJ, Decker GL, Yaney G, Means AR. 1992. Elevated b-cell calmodulin produces a unique insulin secretory defect in transgenic mice. Endocrinology 130:1387–1393. Goode NP, Shires M, Crellin DM, Aparicio SR, Davison AM. 1995. Alterations of glomerular basement membrane charge and structure in diabetic nephropathy. Diabetologia 38:1455–1465. Karnovsky MJ. 1965. A formaldehyde-glutaraldehyde ﬁxative of high osmolality for use in electron microscopy. J Cell Biol 27:137A–138A. Kriz W. 1996. Progressive renal failure—inability of podocytes to replicate and the consequences for development of glomerulosclerosis. Nephrol Dial Transplant 11:1738–1742. Kriz W, Gretz N, Lemley KV. 1998. Progression of glomerular diseases: is the podocyte the culprit? Kidney Int 54:687–697. Lane PH, Steffes MW, Mauer M. 1992. Estimation of glomerular volume: a comparison of four methods. Kidney Int 41:1085–1089. Macconi D, Bonomelli M, Benigni A, Plati T, Sangalli F, Longaretti L, Conti S, Kawachi H, Hill P, Remuzzi G, Remuzzi A. 2006. Pathophysiologic implications of reduced podocyte number in a rat model of progressive glomerular injury. Am J Pathol 168:42–54. Marshall SM. 2005. The podocyte: a major player in the development of diabetic nephropathy? Horm Metab Res 37:9–16. Mifsud SA, Allen TJ, Bertram JF, Hulthen UL, Kelly DJ, Cooper ME, Wilkinson-Berka JL, Gilbert RE. 2001. Podocyte foot process broadening in experimental diabetic nephropathy: amelioration with renin-angiotensin blockade. Diabetologia 44:878–882. Osterby R, Gundersen HJ. 1975. Glomerular size and structure in diabetes. I. Early abnormalities. Diabetologia 11:225–229. Osterby R, Gundersen HJ. 1977. Glomerular size and structure in diabetes. I. Late abnormalities. Diabetologia 13:43–48. Osterby R, Bangstad H-J, Nyberg G, Rudberg S. 2001. On glomerular structural alterations in type-1 diabetes. Virchows Arch 438:129–135. Pagtalunan ME, Miller PL, Jumping-Eagle S, Nelson RG, Myers BD, Rennke HG, Coplon NS, Sun L, Meyer TW. 1997. Podocyte loss and progressive glomerular injury in type II diabetes. J Clin Invest 99:342–348. Pavenstädt H, Kriz W, Kretzler M. 2003. Cell biology of the glomerular podocyte. Physiol Rev 83:253–307. Petermann AT, Pippin J, Krofft R, Blonski M, Grifﬁn S, Durvasula R, Shankland SJ. 2004. Viable podocytes detach in experimental diabetic nephropathy: potential mechanism underlying glomerulosclerosis. Neph Exp Nephrol 98:E144–E123. Petermann AT, Pippin J, Durvasula R, Pichler R, Hiromura K, Monkawa T, Couser WG, Shankland SJ. 2005. Mechanical stretch induces podocyte hypertrophy in vitro. Kidney Int 67:157–166. Rincon-Choles H, Vasylyeva Tl, Pergola PE, Bhandari B, Bhandari K, Zhang JH, Wang W, Gorin Y, Barnes JL, Abboud HE. 2006. ZO-1 expression and phosphorylation in diabetic nephropathy. Diabetes 55:894–900. Shankland SJ. 2006. The podoctye’s response to injury: role in proteinuria and glomerulosclerosis. Kidney Int 69:2131–2147. PODOCYTE MORPHOMETRY IN DIABETIC MICE Shirato I, Hishiki T, Tomino Y. 2001. Podocyte loss and progression of diabetic nephropathy. Contrib Nephrol 134:69–73. Steffes MW, Osterby R, Chavers B, Mauer SM. 1989. Mesangial expansion as a central mechanism for loss of kidney function in diabetic patients. Diabetes 38:1077–1081. Steffes MW, Schmidt D, McCrery R, Basgen JM. 2001. Glomerular cell number in normal subjects and in type 1 diabetic patients. Kidney Int 59:2104–2113. Sterio DC. 1984. The unbiased estimation of number and sizes of arbitrary particles using the disector. J Microscopy 134:127–136. Susztak K, Raff AC, Schiffer M, Bottinger EP. 2006. Glucoseinduced reactive oxygen species cause apoptosis of podocytes and podocyte depletion at the onset of diabetic nephropathy. Diabetes 55:225–233. Viberti G, Wiseman MJ, Pinto JR, Messent J. 1994. Diabetic nephropathy. In: Kahn CR, Weir GC, editors. Joslin’s diabetes mellitus. 13th ed. Philadelphia: Lea and Febiger. p 691–737. Vogelmann SU, Nelson WJ, Myers BD, Lemley KV. 2003. Urinary excretion of viable podocytes in health and renal disease. Am J Physiol Renal Physiol 285:F40–F48. Weibel ER. 1979. Stereological methods. Vol. 1. Practical methods for biological morphometry. London: Academic Press. p 44–45. Weibel ER, Gomez DM. 1962. A principle for counting tissue structures on random sections. J Appl Physiol 17:343–348. 121 Wharram BL, Goyal M, Wiggins JE, Sanden SK, Hussain S, Filipiak WE, Saunders TL, Dysko RC, Kohno K, Holzman LB, Wiggins RC. 2005. Podocyte depletion causes glomerulosclerosis: diphtheria toxin-induced podocyte depletion in rats expressing human diphtheria toxin receptor transgene. J Am Soc Nephrol 16:2941–2952. White KE, Bilous RW. 2004. Estimation of podocyte number: a comparison of methods. Kidney Int 66:663–667. White KE, Bilous RW, Marshall SM, El Nahas M, Remuzzi G, Piras G, De Cosmo S, Viberti G. 2002. Podocyte number in normotensive type I diabetic patients with albuminuria. Diabetes 51:3083– 3089. Wolf G, Chen S, Ziyadeh FN. 2005. From the periphery of the glomerular capillary wall toward the center of disease. Podocyte injury comes of age in diabetic nephropathy. Diabetes 54:1626– 1634. Yu D, Petermann A, Kunter U, Rong S, Shankland SJ, Floege J. 2005. Urinary podocyte loss is a more speciﬁc marker of ongoing glomerular damage than proteinuria. J Am Soc Nephrol 16:1733– 1741. Zheng S, Noonan WT, Metreveli NS, Coventry S, Kralik PM, Carlson EC, Epstein PN. 2004. Development of late-stage diabetic nephropathy in OVE26 diabetic mice. Diabetes 53:3248– 3257.