Chapter1: Renal Tubular Peroxisomes Are Dispensable For Normal Kidney Function

For more info, contact tina.xiang@wecistanche.com

Peroxisomes are specialized cellular organelles involved in a variety of metabolic processes. In humans, mutations leading to complete loss of peroxisomes cause multiorgan failure(Zellweger's spectrum disorders, ZSD), including renal impairment. However, the (patho)physiological role of peroxisomes in the kidney remains unknown. We addressed the role of peroxisomes in renal function in mice with conditional ablation of peroxisomal biogenesis in the renal tubule (cKO mice). Functional analyses did not reveal any overt kidney phenotype in cKO mice. However, infant male cKO mice had lower body and kidney weights, and adult male cKO mice exhibited substantial reductions in kidney weight and kidney weight/body weight ratio. Stereological analysis showed an increase in mitochondria density in proximal tubule cells of cKO mice. Integrated transcriptome and metabolome analyses revealed profound reprogramming of several metabolic pathways, including the metabolism of glutathione and biosynthesis/biotransformation of several major classes of lipids. Although this analysis suggested compensated oxidative stress, the challenge with high-fat feeding did not induce significant renal impairments in cKO mice. We demonstrate that renal tubular peroxisomes are dispensable for normal renal function. Our data also suggest that renal impairments in patients with ZSD are of extrarenal origin.

Click here to learn about cistanche tubulosa extract benefits

Introduction

Peroxisomes are single-membrane-bound organelles that were first found in mouse kidneys by Rhodin in 1954(1). Current estimates suggest that peroxisomes contain more than 50 enzymes involved in a variety of cellular functions, including β-oxidation of very-long-chain fatty acids(VLCFAs), long-chain fatty acids (LCFAs), and long-chain dicarboxylic acid; a-and β-oxidation of branched-chain fatty acids; oxidation of prostaglandins and leukotrienes; metabolism of amino acids; biosynthesis of ether phospholipids (including plasmalogens) and bile acids; redox homeostasis; glyoxylate detoxification; and ferroptosis. Peroxisomes are ubiquitously present in nearly all mammalian cells, with the highest abundance in kidney proximal tubule cells and hepatocytes. In these cells, peroxisomes occupy about 3% of cell volume, but their number, size, and shape vary significantly during embryonic and postembryonic development (2); their number can grow greatly in various stress conditions (3). Although many peroxisomal enzymes have been well characterized, studies addressing their expression and functional role in the kidney are scarce. Moreover, the overall functional relevance of peroxisomes in the kidney remains largely unknown.

Evidence for the role of peroxisomes in renal (patho)physiology has mainly emerged from human genetics studies. Two types of mutations leading to human peroxisomal disorders have been identified: (i)mutations in so-called peroxin (PEX) genes involved in peroxisome biogenesis and (ii)mutations in specific peroxisomal enzymes. The first type of mutation leads to generalized peroxisomal dysfunction causing cerebrohepatorenal Zellweger's spectrum disorders(ZSD)(4). Infants with severe forms of ZSD usually die during the first year of life from multiorgan failure. Even though the presence of cortical renal cysts and/or renal oxalate stones in patients with ZSD has been well documented over years(5, 6), molecular mechanisms leading to these impairments remain so far unknown. Other possible renal abnormalities in ZSD have not been investigated because the disease is dominated by neurological complications. The impact of intermediate or milder forms of ZSD on renal function has not been systemically evaluated. Evidence linking single peroxisomal enzyme deficiency to renal pathophysiology also remains limited.

Mutations in the liver peroxisomal alanine: glyoxylate aminotransferase encoded by the AGXT gene lead to primary hyperoxaluria type I, a disease characterized by deposition of calcium oxalate crystals in the kidney (reviewed in ref. 7). Recently, an autosomal dominant mutation leading to the renal Fanconi syn-drome, or generalized dysfunction of the proximal tubule, has been identified in the peroxisomal enzyme enoyl-CoA hydratase and 3-hydroxy acyl CoA dehydrogenase (EHHADH)(8). This mutation causes mistargeting of EHHADH to mitochondria and impairment of mitochondrial function. Genetic studies performed on congenic rat strains have shown that the peroxisomal enzyme hydroxy acids oxidase 2(HAO2)may play a role in blood pressure control (9,10). However, whether this phenotype is due to the altered HAO2 function in the kidney or other tissues remains unknown. In general, data from animal models with kidney-specific defects of peroxisome biogenesis or kidney-specific inactivation of a single peroxisomal enzyme are still missing. Here, we addressed the role of peroxisomes in renal function in male and female mice with conditional ablation of peroxisomal biogenesis in the renal tubule.

Results

Generation and basic characterization of infant male and female mice with conditional ablation of peroxisomal biogenesis in the renal tubule. Peroxisome biogenesis in the renal tubule was disrupted through conditional inactivation of Pex5 encoding cytosolic peroxisome targeting signal 1 (PTS1)receptor, which is essential for import of PTS1 motif-containing proteins into peroxisomes(11,12). Conditional deletion of Pex5 in the renal tubule was achieved using a doxycycline-inducible (DOX-inducible) system(PexSanlo/Pax8-rtTA/LC1 mice, ref.13). Because the functional importance of peroxisomes may be age-dependent(14), basic characterization of Pex5 deficiency in the kidney was performed both in infant and adult mice. In experiments with infant mice, excision of floxed Pex5 allele was achieved through administration of DOX(2 mg/mL in drinking water) to pregnant females at E14 and maintained until P7. Infant mice were sacrificed at P28. Mice with Pex5oilo genotype were used as controls. As shown in Supplemental Figure 1A (supplemental material available online with this article; https://doi.org/10.1172/jci.insight.155836DS1), DOX treatment resulted in near-complete excision of floxed Pex5 allele in Pexswn/Pax8-GTA/LC1 male and female infant mice. Analysis of renal sections did not reveal any overt histological abnormalities or renal stones in infant mice devoid of Pex.5 in the renal tubule (not shown). Albuminuria was absent in spot urine samples of Pex5-devoid infant mice of both sexes(Supplemental Figure 1B). Body weight (BW) and kidney weight (KW)but not KW/BW ratio were lower in male Pex5-devoid infant mice compared with littermate controls (Supplemental Figure 1C).

Generation and basic characterization of adult male and female mice with conditional ablation of peroxisomal biogenesis in the renal tubule. The Pex5 deletion in the renal tubule of adult mice was induced by 2-week treatment with DOX of 8-week-old Pex5xio/Pax8-ntTA/LC1 male and female mice, hereinafter referred to as cKOm and cKOf mice, respectively (see Methods).In parallel, the same DOX treatment was provided to their littermate controls(Pex5a/lx mice, hereafter referred to as Ctrlm and Ctrlf mice, respectively). All experiments on adult mice were performed 4 weeks after the end of DOX treatment. As shown in Figure 1A, Pex5 mRNA expression was significantly reduced in kidneys of both cKOm and cKOf mice. However, in cKOf mice, some residual full-length Pex5 mRNA was also detectable, suggesting incomplete excision of the floxed allele. Similarly, PEX5 protein was virtually absent in the kidneys of cKOm mice but still detectable in cKOf mice, albeit at a substantially lower level. This difference was visible when kidney extracts from cKOm and cKOf mice were loaded on the same SDS-PAGE gel and immunoblotted together(Figure 1B). Both Kim and cKOf mice were viable, displayed no overt abnormalities, and had normal BW compared to control mice (Supplemental Table 1). The KW and KW/BW ratios were substantially lower in cKOm mice compared with Ctrl mice but not in cKOf mice compared to Ctrlf mice (Figures 1, C, and D, respectively). Analysis of 24-hour urine and plasma samples did not reveal an effect of genotype in mice of both sexes, except for lower plasma potassium levels in cKOm mice compared with Ctrl mice(Supplemental Table 1). Albuminuria was absent in the urine of both cKOm and cKOf mice (Supplemental Figure 2A). No gross morphological changes, calcium oxalate deposits, or lipid accumulation were observed in the kidney of cKOm mice (Supplemental Figure 2, B-D, respectively).

Electron microscopy assessment of proximal tubule cells in adult cKO mice. Electron microscopy assessment of the kidney cortex revealed a high number of peroxisomes in proximal tubule cells of Ctrlm and Ctrlf mice (Figure 2, A, and B, respectively). Peroxisomes were virtually absent in proximal tubules of cKOm and cKOf mice (Figure2, Cand D, respectively). Stereological tools were employed for quantitative analysis of cellular

organelles and cellular dimensions in proximal tubules of Ctrlm and cKOm mice. Both peroxisomes'volume density（number/μm² cytoplasm） and the percentage of cytoplasm occupied by peroxisomes in proximal tubule cells were dramatically decreased in cKOm mice (Figure 3, A and B, respectively), Conversely, mitochondrial volume density, as well as the percentage of cytoplasm occupied by mitochondria in proximal tubule cells, was increased in cKOm mice as compared with Ctrl mice (Figure 3, C and D, respectively). Also-comes volume density and percentage of cytoplasm occupied by lysosomes in proximal tubule cells were not different between Kim and Ctrlm mice (Figure 3, E and F, respectively). As shown in Figure 3G, proximal tubule cells from cKOm mice tended reduced cell width (P= 0.083).

Integrated transcriptomic and metabolomic analyses suggest profound reprogramming in metabolic, antioxidant, and lipid synthesis pathways in the kidneys of cKO mice. Integrated deep transcriptome-sequencing and metabolome analysis were performed to identify molecular alterations in the kidneys of cKO mice(GSE179202). Comparisons of transcriptomes revealed 1350 transcripts differentially expressed in kidneys of Ctrlm and cKOm mice (Figure 4A and Supplemental Table 2, FDR< 5%)and 121 transcripts differentially expressed in kidneys of Ctrlf and cKOf mice (Figure 4B and Supplemental Table 3, FDR< 5%). 61 differentially expressed transcripts were present in both sexes(Figure 4C and Supplemental Figure 3). Enrichment analysis of 100 genes encoding proteins related to peroxisomal function(Kyoto Encyclopedia of Genes and Genomes [KEGG] pathway hsa04146)performed on transcriptomes of Ctrlm and cKOm mice revealed 52 transcripts either up-or downregulated in kidneys of cKOm mice (Figure 4D and Supplemental Figure 4). Gene set enrichment analysis(GSEA) performed on the KEGG Pathway database (release on December 11, 2020)showed downregulation of pathways related to the metabolism of pyruvate, glyoxylate and dicarboxylate, amino acids (arginine, proline, cysteine, methionine, tryptophan, alanine, and histidine), or glutathione(Figure 4E and Supplemental Figure 5) and upregulation of pathways related to fatty acid synthesis and degradation, PPAR signaling, and ABC transporters (Figure 4F and Supplemental Figure 6). No enrichment was found in pathways linked to inflammation or fibrosis. Targeted analysis of several groups of functionally related genes that are not represented in KEGG pathways revealed substantial changes in the expression of genes encoding enzymes involved in the peroxisomal metabolism of fatty acids (ACSL1/3/4/6, ACSVL1, ACOT3/4, ACNAT1, ACOX3, ABCD3, EHHADH, and ACAA1B), plasmalogen biosynthesis (FAR1 and AGPS), bile acid synthesis(AMACR and HSD17B4), and reactive oxygen species (ROS) detoxification(CAT and SOD1) (Supplemental Figure 3). Alterations were also noted in the expression of a large number of genes critical for membrane transport processes in the proximal tubule, thick ascending limb (TAL), and distal nephron(Supplemental Figure 7 and Supplemental Table 4). For instance, in the proximal tubule, we found a reduction in the expression of an aquaporin-1 water channel(Agp1); megalin (Lrp2); phosphate transporter Napi-2a (Slc34al); urate transporters Uratl (Slc22a12), Npt1/4 (SIc17al/3), and Oat1 (Slc22a6); and numerous other organic anion and amino acid transporters. In the TAL and distal nephron, there was a substantial increase in the expression of genes encoding proteins involved in sodium reabsorption: NKCC2 (SLC12A1), CLC-Kb (CLCNKB), βENAC (SCNN1B) (FDR<0.05), and NCC (SLC12A3)(FDR= 0.05).

From the total of 852 detected metabolites, 207 showed differential abundance in kidneys of Ctrlm and cKOm mice, and 118 showed differential abundance in kidneys of Ctrlf and cKOf mice (Supplemental Table 5, FDR<5%). Seventy-nine metabolites demonstrating differential abundance were common in both comparisons(Figure 5A and Supplemental Table 6).Among metabolites showing the most significant differences were plasmalogens and sphingomyelins(decreased abundance)and glutathione-related metabolites and dicarboxylic acids (increased abundance) in kidneys of both cKOm and cKOf mice (Figure 5B and Supplemental Figure 8, respectively). Global analysis of metabolome confirmed that plasmalogens and sphingomyelins constituted a majority of metabolites showing a decreased abundance in kidneys of both cKOm and cKOf mice (Figure 5C).LCFA dicarboxylates, very long-chain fatty acyl-carnitines, phosphatidylcholines, and phosphatidylethanolamines were present among metabolites with increased abundance, in addition to glutathione-related metabolites and dicarboxylic acids(Figure 5D).

Joint pathway analysis (MetaboAnalyst 5.0)of transcripts and metabolites exhibiting increased or decreased abundance in kidneys of cKOm mice compared with Ctrl mice identified 22 downregulated and 7upregulated pathways(Figure 6, A and B, respectively; FDR<0.1; Supplemental Table 7). Nitrogen metabolism, pyruvate metabolism, glycolysis, and gluconeogenesis were identified among the downregulated pathways while fatty acid degradation was identified among the upregulated pathways (Figure 6, A and B, respectively). Glutathione metabolism and retinol metabolism were present among both up-and downregulated pathways (Figure 6, A and B). Detailed analysis of transcripts and metabolites related to glutathione metabolism identified 16 transcripts and 3 metabolites with decreased abundance in kidneys of cKOm mice (Figure 6, C and D, respectively), along with 6 transcripts and 8 metabolites exhibiting increased abundance(Figure 6, E and F, respectively). The oxidized form of glutathione(GSSG)was present in cKOm but not in Ctrlm mice, thereby suggesting oxidative stress in cKOm mice(Figure 6F).

High-fat feeding challenge does not lead to additional oxidative stress in ckKOmice. Alterations in glutathione-related pathways, depletion of plasmalogens, and decreased expression of catalase suggested oxidative stress and/or rearrangement of different antioxidant defense systems in kidneys of cKOm mice. To test these hypotheses, we challenged cKOm mice with a high-fat diet (HFD)for 4 weeks. This challenge did not result in albuminuria but increased urinary volume and urinary excretion of calcium and urate. No difference was observed in tested plasma parameters, including calcemia and uricemia (Supplemental Table 8). No gross morphological changes or lipid accumulation were detected in the kidneys of HFD-challenged Ctrlm and cKOm mice (Figure 7A). Total and nonenzymatic antioxidant capacity as assessed by Trolox assay (Figure 7B)as well as tissue levels of malondialdehyde, a marker of polyunsaturated fatty acid peroxidation(Figure 7C), were not different between treatments and genotypes. The abundance of lipid peroxidation product 4-hydroxynonenal (4-HNE)was increased in Ctrlm mice treated with HFD as compared with Ctrl mice on the control diet, but no difference was observed in cKOm mice treated with the 2 diets (Figure 7D).