Influence of Different Levels of Lipoic Acid Synthase Gene Expression on Diabetic Nephropathy

Abstract

Oxidative stress is implicated in the pathogenesis of diabetic nephropathy (DN) but outcomes of many clinical trials are controversial. To define the role of antioxidants in kidney protection during the development of diabetic nephropathy, we have generated a novel genetic antioxidant mouse model with over- or under-expression of lipoic acid synthase gene (Lias). These models have been mated with Ins2Akita/+ mice, a type I diabetic mouse model. We compare the major pathologic changes and oxidative stress status in two new strains of the mice with controls. Our results show that Ins2Akita/+ mice with under-expressed Lias gene, exhibit higher oxidative stress and more severe DN features (albuminuria, glomerular basement membrane thickening and mesangial matrix expansion). In contrast, Ins2Akita/+ mice with highly-expressed Lias gene display lower oxidative stress and less DN pathologic changes. Our study demonstrates that strengthening endogenous antioxidant capacity could be an effective strategy for prevention and treatment of DN.

Fig 1. Characterizations of the new mouse model.

(A) Generation of mice with genetically graded expression of lipoic acid synthase (Lias). Wild type (top line): Endogenous Lias gene 3’-UTR (white column) is located after exon 11 of Lias gene. The targeting construct (second line): consisted of the 3’-UTR sequences of bovine growth hormone (bGH) gene and a Neo gene, two lox P sites flanking the two fragments, and followed by the 3’-UTR of cFos gene and thymidine kinase gene (TK). H/H (third line): The locus after homologous recombination. Lias expression is now controlled by the 3’-UTR of bGH, which stabilizes Lias mRNA. L/L (bottom line): The locus after Cre-lox P recombination. Lias expression is controlled by the 3-UTR of cFos, which destabilizes the Lias mRNA. (B) The kidney mRNA levels of Lias in 12-week-old L/L, H/H and WT male mice and in 28-week-old diabetic L/L, H/H and WT male mice. Lias gene expression in the non-diabetic WT mice as a reference for the all six groups of mice. (C) Lipoic acid synthase (LIAS) concentrations of kidney cortex mitochondria, measured by Western blot, and VDAC1 as loading control, in 12-week-old L/L, H/H and WT male mice. n = 5, in each group. (D) The amounts of lipoic acid in non-diabetic Lias^High/High and Lias^Low/Low kidney, detected by Western blot, were quantified by Image Quant software. Data are expressed as the mean ± SE.

Fig 2. Changes of urine albumin in the mice.

Urine albumin/creatinine ratio in Lias^Low/LowIns2^Akita//+ mice is significantly elevated (P<0.01) whereas the ratio in Lias^High/HighIns2^Akita/+ mice is reduced but does not reach significance, compared with Lias^+/+Ins2^Akita/+ mice at 28 weeks of age. The numbers inside the bars indicate the number of animals.

Fig 3. Changes of Lias gene expression influences mesangial matrix expansion (MME).

Representative PAS staining of glomeruli in diabetic Ins2^Akita//+ mice at 28 weeks of age. (A) Lias^Low/LowIns2^Akita//+ mice. (B) Lias^+/+Ins2^Akita/+ mice. (C) Lias^High/HighIns2^Akita/+ mice. Original magnification x400, Bars = 50μm. (D) MME score, quantified as the region of positive PAS staining, is expressed as a function of total glomerular tuft area. The numbers inside the bars indicate the number of animals. Lias^Low/LowIns2^Akita//+ and Lias^High/HighIns2^Akita/+ mice are compared with Lias^+/+Ins2^Akita/+ mice. Values are expressed as the mean ± SEM.

Fig 4. Representative electron micrographs from diabetic Ins2^Akita//+ mice at 28 weeks of age.

(A) Segmentally thickened glomerular basement membrane (GBM) with minor irregularities along the lamina rara externa (“undulations”) and segmental podocyte foot process effacement in Lias^Low/LowIns2^Akita//+ mice. Original magnification, x8,000. (B) Lias^High/HighIns2^Akita/+ mice. Original magnification, x8,000. (C) Lias^+/+Ins2^Akita/+ mice. Original magnification, x8,000. (D) Illustrated is the basal part of a tubular epithelial cell containing distorted mitochondria. Cristae are disrupted and mitochondria filled presumably with lipid products vaguely resembling zebra-bodies in Lias^Low/LowIns2^Akita//+ mice. Original Magnification, x5,000. (E) A small number of damaged mitochondria in proximal tubular cells of Lias^High/HighIns2^Akita/+ mice. Original magnification, x8,000. (F) Damaged mitochondria in proximal tubular cells of Lias^+/+Ins2^Akita/+ mice. Magnification, x8,000. (G) Electron microscopic quantitative examination using Image J showed thickening of the GBM significant increase in Lias^Low/LowIns2^Akita/+ mice compared to Lias^+/+Ins2^Akita/+ mice. (H) Ratio of average damaged mitochondrion over entire counted mitochondria in proximal tubules of Lias^Low/LowIns2^Akita//+, Lias^High/HighIns2^Akita/+ and Lias^+/+Ins2^Akita/+ mice, using one-way ANOVA for the comparison.

Fig 5. Alternated oxidative stress.

(A). Oxidative stress in kidney cortex is associated with plasma concentration of lipid perioxidation marker, 4-hydroxynonenal (4-HNE), in diabetic Ins2^Akita//+ mice and non-diabetic mice with differential Lias gene expression. Lias^+/+ mice as a control. Data were analyzed using two-way ANOVA. (B). Urine 8-isoprostane varies with Lias expression in diabetic mice. Data were analyzed using one-way ANOVA. The numbers inside the bars indicate the number of animals in each group. Results are expressed as mean ±SEM.

Fig 6. Kideny inflammation.

Urinary MCP-1 levels in Lias^Low/LowIns2^Akita/+, Lias^+/+Ins2^Akita/+ and Lias^High/HighIns2^Akita/+ mice. The numbers inside the bars indicate the number of animals in each group. Results are expressed as mean ±SEM.