Accelerated Nephropathy in Diabetic Apolipoprotein E-Knockout Mouse: Role of Advanced Glycation End Products

Assessment of Renal Function

At week 20, the animals were housed in metabolic cages for 24 h for collection of urine samples for the later measurements of albumin concentration by RIA. This assay is a modification of a previously established method for measuring rat albumin in our laboratory (18). In brief, a rabbit anti-mouse albumin antibody (Cappel, Aurora, OH) was used followed by a second antibody, a sheep anti-rabbit antibody (ProSearch, Melbourne, Australia). Plasma urea and creatinine concentrations at the end of the study were measured by autoanalyzer (Beckman Instruments, Fullerton, CA).

Evaluation of Morphologic Changes

Two-micrometer-thick cross-sections of the kidneys were prepared and stained with periodic acid-Schiff to evaluate renal pathology. Assessment of the amount of glomerular and tubulointerstitial injury was performed in a blinded manner using a semiquantitative method (19). Glomerular injury included mesangial matrix expansion and/or hyalinosis with focal adhesions, capillary dilation, and true glomerular tuft occlusion and sclerosis. The assessment of tubulointerstitial injury included tubular dilation, interstitial fibrosis, and tubular cell atrophy. In brief, 40 glomeruli in each kidney were graded according to the severity and amount of the glomerular injury: grade 0, intact glomerulus; grade 1, glomerular injury involving <25% of the glomerulus; grade 2, glomerular injury affecting 25 to 50% of the glomerulus; grade 3, glomerular injury affecting ∼50 to 75% of the glomerulus; grade 4, severe damage or sclerosis involving >75% of the glomerulus. For tubulointerstitial injury, 40 tubulointerstitial fields were graded according to the severity and amount of the tubular and interstitial damage: grade 0, no tubular injury, no interstitial fibrosis; grade 1, <25% of the field involved; grade 2, 25 to 50% of the field involved; grade 3, 50 to 75% of the field involved; grade 4, >75% of the tubules and interstitium per field involved. The indices for glomerular and tubulointerstitial injury were calculated using the following formula:

GSI/TI = (1 × n₁) + (2 × n₂) + (3 × n₃) + (4 × n₄)/(n₀ + n₁+ n₂ + n₃ + n₄)

where n_x = number of glomeruli/tubulointerstitial fields in each grade of glomerular/tubulointerstitial injury. These indices of glomerular/tubulointerstitial injury were calculated by averaging the grades assigned to all glomeruli/tubulointerstitial fields (20). For assessing renal lipid accumulation, frozen renal sections (6 μm) were stained for lipids with Sudan IV-Herxheimer’s solution (Sigma Chemical Co., St. Louis, MO) (16).

Immunohistochemistry

Frozen renal sections (6 μm) were used to immunostain for macrophages with a rat anti-mouse F4/80 antibody (Serotec, Oxford, UK; diluted 1:50) (21). In brief, frozen sections were fixed with cold acetone and blocked with normal rabbit serum (1/5 in Tris-buffered saline [TBS]). The sections were incubated with the primary antibody overnight at 4°C. Endogenous nonspecific binding for biotin was blocked using an avidin/biotin blocking kit (Vector Laboratories, Burlingame, CA), and endogenous peroxidase was inactivated using 0.6% H₂O₂ in TBS. Biotinylated rabbit anti-rat Ig (Vector Laboratories), diluted 1:200 in TBS, was used as the secondary antibody for 60 min, followed by Vectastain ABC reagent (Vector Laboratories) for 30 min. Peroxidase activity was identified by reaction with 3,3′-diaminobenzidine tetrahydrochloride (Sigma Chemical Co) as the chromogen. The slides were then counterstained with hematoxylin, dehydrated, and mounted.

Two-micrometer paraffin sections of kidneys were used to stain for α-smooth muscle actin [α-SMA], collagen type I and IV, TGF-β1, AGE, and RAGE. The primary antibodies that were used included a monoclonal mouse anti-human α-SMA antibody (Dako A/S, Copenhagen, Denmark; diluted 1:50), a polyclonal goat anti-collagen type IV antibody (diluted 1:1600; Southern Biotechnology, Birmingham, AL), a polyclonal goat anti-collagen type I antibody (diluted 1:200; Southern Biotechnology), a polyclonal rabbit anti–TGF-β1 antibody (diluted 1:400; Santa Cruz Biotechnology, Santa Cruz, CA), a monoclonal mouse anti-mouse AGE antibody (AGE 4G9, which recognizes the AGE carboxymethyllysine [CML]; diluted 1:100; a gift from Dr. H. Founds, Alteon, Ramsey, NJ), and a polyclonal goat anti-RAGE antibody (diluted 1:250; Chemicon, Temecula, CA). In brief, sections for α-SMA and AGE were dewaxed, hydrated, and quenched with 3% H₂O₂ in PBS (pH 7.6) to inhibit endogenous peroxidase activity. This was followed by incubation in Protein Blocking Agent (Lipshaw-Immunon, Pittsburgh, PA) for 30 min at room temperature. The sections were then incubated with anti–α-SMA, anti-AGE, or goat anti-RAGE antibodies overnight at 4°C. Sections for TGF-β1 were first heated, dewaxed, and then underwent pressure-cooker microwaving in 0.01 M citrate buffer at pH 6.0. This was followed by incubation with 3% H₂O₂ in PBS (pH 7.6) and Protein Blocking Agent and incubation with the primary antibody for TGF-β1 for 1 h at room temperature. For type I and type IV collagen, dewaxed and hydrated sections were quenched with 3% H₂O₂ in PBS. In addition, sections were digested with 0.4% pepsin (Sigma Chemical Co.) in 0.01 M HCl at 37°C. This was followed by incubation with 1% normal horse serum diluted in TBS to block nonspecific binding. Subsequently, sections were incubated with the primary antibody for either type I or type IV collagen overnight at 4°C in a humid atmosphere followed by avidin/biotin blocking. Thereafter, biotinylated anti-mouse Ig (Vector Laboratories, Burlingame, CA), diluted 1:250 for α-SMA and AGE, biotinylated anti-rabbit Ig (Vector Laboratories), diluted 1:500 for TGFβ1, or anti-goat, diluted 1:500 for collagen type I and IV and RAGE, was applied as the secondary antibody, followed by horseradish peroxidase–conjugated streptavidin (VECTASTAIN Elite ABC Staining Kit, Vector Laboratories). Peroxidase conjugates were subsequently visualized using 3,3′-diaminobenzidine tetrahydrochloride (Sigma Chemical Co.) in 0.08% H₂O₂/PBS as the chromogen. Finally, sections were counterstained with Mayer’s hematoxylin, dehydrated, and mounted.

All sections were examined under light microscopy (Olympus BX-50, Olympus Optical) and digitized with a high-resolution camera. For the quantification of proportional area of staining, ∼20 views were randomly located in the renal cortex and corticomedullary junction of each slide (Optimas 6.2-Video Pro-32; Bedford Park, SA, Australia). For some parameters with obvious differences detected with respect to glomerular, tubular, and interstitial expression, separate analyses were performed for these individual renal compartments. All assessments were performed in a blinded manner.

AGE Peptides in Serum

Fluorescence AGE peptides in serum were assayed using HPLC with on-line spectrofluorometric detection (22,23⇓). Spectrofluorometric detection of AGE peptides was measured at 440 nm with excitation at 247 nm. Samples were run in triplicate, and the area under the curve was used for signal measurements. The assay was calibrated against AGE peptide obtained from enzymatic hydrolysis of AGE-BSA (10 g/L). Total peptide content was estimated using an on-line spectrophotometric detector set to a wavelength of 280 nm.

Reverse Transcription–PCR

Six micrograms of total RNA extracted from a whole kidney using Trizol method were used to synthesize cDNA with Superscript First Strand synthesis system for reverse transcription–PCR (Life Technologies BRL, Grand Island, NY) (24). RAGE, AGE receptor-2 (AGE R-2), and AGE R-3 gene expression were analyzed by real-time quantitative reverse transcription–PCR using the TaqMan system based on real-time detection of accumulated fluorescence (ABI Prism 7700, Perkin-Elmer, Foster City, CA) (24). Fluorescence for each cycle was quantitatively analyzed by an ABI Prism 7700 Sequence Detection System (Perkin-Elmer, PE Biosystems). For controlling for variation in the amount of DNA available for PCR in the different samples, gene expression of the target sequence was normalized in relation to the expression of an endogenous control, 18S ribosomal RNA (rRNA) (18S rRNA TaqMan Control Reagent kit; ABI Prism 7700, Perkin-Elmer). Primers and Taqman probes for the target sequences and the endogenous reference 18S rRNA were constructed with the help of Primer Express (ABI Prism 7700, Perkin-Elmer).

For amplification of the RAGE cDNA, the forward primer was 5′-GGGAAGGAGGTCAAGTCCAACTA-3′ and the reverse primer was 5′-TGAGTTCAGAGGCAGGATCCA-3′. The probe specific to RAGE was 5′-TCTACCAGATTCCTGGGAAGCCAGAAATTG-3′. For the AGE R-2 cDNA, the forward primer was 5′-ACGAGCTCACCACCAATGAGT-3′ and the reverse primer was 5′-TTTGGGTTTCTGGGAGACCA-3′. The probe specific to AGE R-2 was 5′-CGTCTACCGGCTTTGCCCCTTCA-3′. For the AGE R-3 cDNA, the forward primer was 5′-CAGACAGCTTTTCGCTTAACGA-3′ and the reverse primer was 5′-CCATGCACCCGGATATCC-3′. The probe specific to AGE R-3 was FAM-5′-TTAGCTGGCTCTGGAAACCCAAACCCT-3′-TAMRA. Every amplification was performed with the following time course: 50°C, 2 min and 10 min at 95°C and 40 cycles of 94°C, 20 s, 60°C, 1 min. Each sample was tested in triplicate. Results are expressed relative to values in the control apo E-KO group, which were arbitrarily assigned a value of 1.

Statistical Analyses

Data were analyzed by ANOVA using Statview V (Brainpower, Calabasas, CA). Comparisons of group means were performed by Fisher least significant difference method. Data are shown as mean ± SEM, unless otherwise specified. P < 0.05 was viewed as statistically significant.

Metabolic Parameters and Systolic BP

There was no difference in body weight gain over the 20-wk period between nondiabetic C57BL/6 and apo E-KO mice, but diabetic mice of both strains gained less weight than nondiabetic mice (Table 1). In both mouse strains, diabetes was associated with an increase in HbA_1c levels (Table 1). Plasma cholesterol levels were increased in nondiabetic apo E-KO mice when compared with the C57BL/6 mice, whereas triglyceride levels were similar in nondiabetic animals of both strains (Table 1). Diabetes was associated with a further increase in plasma cholesterol and triglyceride levels in diabetic apo E-KO mice but not in C57BL/6 mice. Systolic BP was modestly increased in apo E-KO mice when compared with C57BL/6 mice, but the induction of diabetes did not alter systolic BP significantly in either strain (Table 1).

Aminoguanidine or ALT-711 treatments had no significant effect on body weight, HbA_1c, triglyceride levels, or systolic BP when compared with untreated diabetic apo E-KO mice (Table 1). Treatment with ALT-711 but not aminoguanidine was associated with a modest decrease in plasma cholesterol, albeit the levels remained markedly elevated.

Renal Parameters

Nondiabetic apo E-KO mice had higher kidney/body weight ratios than C57BL/6 mice. Diabetes in both strains was associated with an increase in the kidney/body weight ratio, but no difference in this parameter was detected between diabetic apo E-KO and diabetic C57BL/6 mice. Plasma urea and creatinine concentrations were similar in nondiabetic mice of either strain (Table 1). Diabetes was associated with increased serum urea concentrations in both strains, but serum creatinine concentrations were increased by diabetes in apo E-KO mice only. There was no difference in urinary albumin excretion rate between nondiabetic apo E-KO mice and C57BL/6 mice (Figure 1A). In both strains, induction of diabetes significantly increased urinary albumin excretion. The highest albumin excretion rate was seen in diabetic apo E-KO mice, which showed increased albuminuria when compared with both diabetic C57BL/6 mice and nondiabetic apo E-KO mice (Figure 1A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Albuminuria (A), glomerular injury index (B), and tubulointerstitial injury (C) in mice after the 20-wk treatment period. Representative pictures of glomerular (D) and tubulointerstitial (E) injury in nondiabetic apo E-KO, diabetic C57BL/6, and diabetic apo E-KO mice. Data are shown as mean ± SEM. **P < 0.01 versus appropriate nondiabetic control group; †P < 0.05, ‡P < 0.01 versus appropriate C57BL/6 group; #P < 0.05, ##P < 0.01 versus diabetic apo E-KO mice. Magnifications: ×400 in D, ×200 in E.

Treatment with aminoguanidine or ALT-711 did not significantly affect kidney/body weight ratio, plasma urea, or creatinine concentrations when compared with untreated diabetic apo E-KO mice (Table 1). However, the increase in albuminuria was attenuated by treatment with either aminoguanidine or ALT-711 (Figure 1A).

In nondiabetic C57BL/6 mice, renal structure was preserved. In nondiabetic apo E-KO mice, there was a slight but significant increase in the glomerular injury index as well as in the tubulointerstitial injury index (Figure 1, B and C). Glomerular changes were characterized by deposits of lipids in the mesangium and in the capillaries, slight glomerular hyperplasia, deposition of extracellular matrix (ECM) in the mesangium associated with mesangial expansion, and mesangial hypercellularity. Tubulointerstitial changes included macrophage infiltration and a modest increase in interstitial space as a result of deposition of ECM. The induction of diabetes was associated with an increase in the degree and severity of the glomerular and tubular changes compared with nondiabetic apo E-KO mice. These changes included glomerular capillary dilation, increased capillary and mesangial lipid deposits, and increased mesangial expansion with increased infiltration by inflammatory cells and true glomerulosclerosis with glomerular tuft occlusion (Figure 1D). There was not only an increase in tubulointerstitial lesions but also an increase in their severity. These changes included increased infiltration of mononuclear cells, tubulointerstitial ECM deposition, and dilation of proximal and distal tubules and occasional tubular cell atrophy (Figure 1E). The indices for glomerular and tubulointerstitial injury were greater in diabetic apo E-KO mice when compared with diabetic C57BL/6 mice (Figure 1, B and C). Treatment with aminoguanidine or ALT-711 was associated with amelioration of these renal pathologic changes with reduced glomerular and tubulointerstitial injury indices in both treatment groups (Figure 1, B and C).

Glomerular Lipid Accumulation

In nondiabetic or diabetic C57BL/6 mice, no staining for Sudan IV was observed. In contrast, nondiabetic apo E-KO mouse kidneys showed clear lipid accumulation (0.74 ± 0.20%). In diabetic apo E-KO mice, there was a significant increase in glomerular lipid accumulation (1.72 ± 0.46%; P < 0.05 versus nondiabetic apo E-KO mice). Aminoguanidine (1.63 ± 0.40%) or ALT-711 (1.53 ± 0.46%) had no effect on glomerular lipid accumulation.

Macrophage Infiltration

In the nondiabetic animals, macrophage infiltration, as assessed by F4/80-positive cells, was observed in the interstitium with virtually no macrophages detected in glomeruli (Figure 2). Diabetes was associated with increased interstitial and glomerular infiltration of macrophages in both mouse strains (Figure 2). Treatment with aminoguanidine or ALT–711 was associated with a significant decrease in macrophage infiltration in both the glomerulus and the tubulointerstitium (Figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. F4/80-positive macrophage staining in frozen section of the kidneys. (A) Quantification. (B) Representative pictures. Diabetic apo E-KO mice show increased macrophage infiltration in the glomerulus and the interstitium. Treatment with aminoguanidine or ALT-711 reduced macrophage infiltration. Data are shown as mean ± SEM. **P < 0.01 versus appropriate nondiabetic control group; ##P < 0.01 versus diabetic apo E-KO mice. Magnification, ×200 in B.

Expression of α-SMA–Positive Cells

In the nondiabetic mouse, renal immunostaining for α-SMA was essentially confined to renal arterioles and arteries, with only a little staining observed in the peritubular interstitium (Figure 3). Nondiabetic apo E-KO mice showed a slight increase in α-SMA expression in glomeruli and the tubulointerstitium when compared with C57BL/6 mice (Figure 3). In kidneys of diabetic mice, α-SMA expression was increased in glomeruli and in the tubulointerstitium when compared with nondiabetic mice. There was no significant difference between diabetic apo E-KO mice and diabetic C57BL/6 mice. In the diabetic kidneys, α-SMA–positive cells were detected in virtually all glomeruli, particularly in the mesangial cells as well as in the interstitium and occasionally in tubular cells. Treatment with ALT-711 prevented the increase in α-SMA–positive cell expression in diabetic apo E-KO mice (Figure 3). By contrast, aminoguanidine treatment had no effect on this parameter.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Quantification (A) and representative pictures (B) of immunohistochemistry for α-smooth muscle actin in the kidneys. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01 versus appropriate nondiabetic control group; †P < 0.05 versus appropriate C57BL/6 group; ##P < 0.01 versus diabetic apo E-KO mice; §§P < 0.01 versus diabetic apo E-KO mice treated with aminoguanidine. Magnification, ×400 in B.

Type I and IV Collagen Expression

There was no difference in collagen type I expression between nondiabetic C57BL/6 and apo E-KO mice (Figure 4, A and B), whereas collagen type IV was increased in nondiabetic apo E-KO when compared with nondiabetic C57BL/6 mice (Figure 4, C and D). The induction of diabetes was associated with a similar increase in collagen type I and type IV protein expression in both mouse strains (Figure 4). The increase in collagen expression in diabetic kidneys was predominantly detected in the glomeruli as well as in the peritubular interstitium (Figure 4). Treatment with aminoguanidine or ALT-711 reduced accumulation of collagen type I and type IV, with ALT-711 treatment being associated with a more prominent reduction in collagen type I and IV protein expression than aminoguanidine treatment (Figure 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. Immunohistochemistry for type I collagen (quantification [A] and representative pictures [B]) as well as type IV collagen (quantification [C] and representative pictures [D]) expression in the kidneys. Data are shown as mean ± SEM. **P < 0.01 versus appropriate nondiabetic control group; †P < 0.05 versus appropriate C57BL/6 group; ##P < 0.01 versus diabetic apo E-KO mice; §§P < 0.01 versus diabetic apo E-KO mice treated with aminoguanidine. Magnification, ×400 in B and D.

Fluorescence AGE in Serum

Fluorescence-labeled serum AGE levels were higher in nondiabetic apo E-KO than in nondiabetic C57BL/6 mice (Figure 5A). Diabetes was associated with increased serum AGE levels in both mouse strains, with the highest levels detected in the diabetic apo E-KO group. Aminoguanidine and ALT-711 treatment decreased serum AGE to levels similar to those observed in nondiabetic apo E-KO mice (Figure 5A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. Quantification of fluorescence advanced glycation end products (AGE) in serum (A) and AGE immunostaining (carboxymethyllysine) in the kidneys (glomerulus quantification [B]; tubulus quantification [C]; representative pictures [D]). Data are shown as mean ± SEM. **P < 0.01 versus appropriate nondiabetic control group; †P < 0.05, ‡P < 0.01 versus appropriate C57BL/6 group; ##P < 0.01 versus diabetic apo E-KO mice; §§P < 0.01 versus diabetic apo E-KO mice treated with aminoguanidine. Magnification, ×200 in D.

AGE Immunostaining in the Kidney

AGE immunostaining (CML) was detected in the kidneys of C57BL/6 mice and was mainly observed in proximal and distal tubuli (Figure 5, B through D). In apo E-KO mice, there was in increase in AGE staining in tubuli, and some AGE staining was also detected in glomeruli (Figure 5, B through D). Induction of diabetes in both strains was associated with a marked increase in AGE immunostaining in both the glomerulus and the tubulointerstitium. Renal AGE immunostaining was significantly higher in diabetic apo E-KO mice than in diabetic C57BL/6 mice. Renal AGE expression was reduced to control apo E-KO levels by treatment with either aminoguanidine or ALT-711 (Figure 5, B and C).

Renal AGE Receptor Expression

Gene expression of the AGE receptors RAGE and AGE R-3 was increased in diabetic kidneys in both mouse strains to a similar degree (Table 2). No changes in renal AGE R-2 gene expression in diabetic kidneys were detected (Table 2). Immunohistochemical analysis showed that RAGE protein was increased in glomerular and tubular cells of nondiabetic apo E-KO mice. Induction of diabetes caused a further increase in RAGE immunostaining, but there was no difference between the diabetic groups of both strains (Table 2).

Treatment with aminoguanidine or ALT-711 was associated with decreased renal RAGE and AGE R-2 gene expression when compared with the untreated diabetic apo E-KO mice (Table 2). No significant effect on AGE R-3 gene expression was detected with either treatment (Table 2). Both treatments decreased RAGE protein expression when compared with untreated diabetic apo E-KO mice (Table 2).

Renal TGF-β1 Expression

In the kidneys of nondiabetic C57BL/6 mice, there was minimal staining for TGF-β1 (Figure 6). In nondiabetic apo E-KO mice, positive staining for TGF-β1 was seen in the glomeruli as well as in the tubulointerstitial areas (Figure 6). Diabetes was associated with a prominent increase in renal TGF-β1 immunostaining, which was seen in the glomerular epithelial and mesangial cells as well as in the tubulointerstitium. Tubulointerstitial TGF-β1 staining was significantly higher in diabetic apo E-KO mice than in diabetic C57BL/6 mice, whereas no difference in glomerular TGF-β1 staining was observed between these two diabetic groups (Figure 6). Treatment with ALT-711 or aminoguanidine reduced TGF-β1 expression in the glomeruli as well as in the tubulointerstitium (Figure 6) to levels similar to that observed in nondiabetic apo E-KO mice.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6. Quantification of TGF-β1 protein expression in the glomeruli (A) and the tubulointerstitium (B) as well as representative pictures of the kidneys (C; top, glomeruli; bottom, tubulointerstitium) after TGF-β1 immunostaining. Data are shown as mean ± SEM. **P < 0.01 versus appropriate nondiabetic control group; ‡P < 0.01 versus appropriate C57BL/6 group; ##P < 0.01 versus diabetic apo E-KO mice; §§P < 0.01 versus diabetic apo E-KO mice treated with aminoguanidine. Magnification, ×400.