Treatment of tubular damage in high-fat-diet-fed obese mice using sodium-glucose co-transporter inhibitors

Abstract

A long-term high-fat diet (HFD) causes obesity and changes in renal lipid metabolism and lysosomal dysfunction in mice, causing renal damage. Sodium-glucose co-transporter inhibitors, including phlorizin, exert nephroprotective effects in patients with chronic kidney disease, but the underlying mechanism remains unclear. A HFD or standard diet was fed to adult C57BL/6J male mice, and phlorizin was administered. Lamellar body components of the proximal tubular epithelial cells (PTECs) were investigated. After phlorizin administration in HFD-fed mice, sphingomyelin and ceramide in urine and tissues were assessed and label-free quantitative proteomics was performed using kidney tissue samples. Mitochondrial elongation by fusion was effective in the PTECs of HFD-fed obese mice under phlorizin administration, and many lamellar bodies were found in the apical portion of the S2 segment of the proximal tubule. Phlorizin functioned as a diuretic, releasing lamellar bodies from the apical membrane of PTECs and clearing the obstruction in nephrons. The main component of the lamellar bodies was sphingomyelin. On the first day of phlorizin administration in HFD-fed obese mice, the diuretic effect was increased, and more sphingomyelin was excreted through urine than in vehicle-treated mice. The expressions of three peroxisomal β-oxidation proteins involved in fatty acid metabolism were downregulated after phlorizin administration in the kidneys of HFD-fed mice. Fatty acid elongation protein levels increased with phlorizin administration, indicating an increase in long-chain fatty acids. Lamellar bodies accumulated in the proximal renal tubule of the S2 segment of the HFD-fed mice, indicating that the urinary excretion of lamellar bodies has nephroprotective effects.

Fig 1. Phlorizin improved increased blood glucose levels in high-fat diet (HFD)-fed mice.

(A) Experimental design: All data were measured under non-fasting conditions. (B) Weight gain in HFD-fed and standard diet (STD)-fed mice (n = 10) over 16 weeks. (C) Blood glucose levels (n = 10). (D) Blood glucose levels pre- and post-vehicle (Veh) administration as well as pre- and post-phlorizin (PLZ) administration in STD- and HFD-fed mice at 20 weeks of age. (E) Subtracted glucose levels from the data shown in (D) before and after vehicle or phlorizin administration. (F) Urinary glucose levels in STD- and HFD-fed mice at 20 weeks of age. (G) Urinary glucose levels pre- and post-Veh as well as pre- and post-PLZ administration in HFD mice at 20 weeks of age. (H) Serum insulin levels in STD- and HFD-fed mice at 12 and 20 weeks of age. (I) Serum triglyceride (TG) and (J) total cholesterol (T-CHO) in the serum of STD- and HFD-fed mice at 20 weeks of age. (K) Serum L-FABP levels in STD- and HFD-fed mice at 20 weeks of age. (L) Urinary L-FABP levels in STD- and HFD-fed mice at 20 weeks of age. Results are presented as the mean ± standard deviation (SD). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Statistical significance was calculated by unpaired two-tailed Student t-test (B, C, E, F, G, H, K, and L), paired two-tailed Student t-test (D), and one-way analysis of variance, followed by the post-hoc Turkey–Kramer tests (I, J).

Fig 2. Phlorizin improved injured proximal tubules in high-fat diet (HFD)-fed mice.

(A–D) Light micrographs of 2-μm paraffin kidney sections stained with hematoxylin and eosin (H&E) in standard diet (STD)-fed and HFD-fed mice. (E–H) Light micrographs of 1-μm thick Epon kidney sections stained with toluidine blue (TB). The TB sections have glomeruli and renal tubules at the outer cortex of four groups: (E) STD-vehicle, (F) STD-phlorizin (PLZ), (G) HFD-vehicle, and (H) HFD-PLZ (G–H). Red arrows indicate damaged proximal tubules in HFD-vehicle and HFD-PLZ mice. (I–L) Electron micrographs of proximal tubules of four groups; (I) STD-vehicle, (J) STD-PLZ, (K) HFD-vehicle, and (L) HFD-PLZ. (K and L) The high intensity of the lamellar body is indicated by red arrows corresponding to the red arrows in (G) and (H). (M–P) Light micrographs of immunohistochemistry for Ki-67 in the renal cortex of four groups; (M) STD-vehicle, (N) STD-PLZ, (O) HFD-vehicle, and (P) HFD-PLZ. Ki-67-positive tubular nuclei are indicated by red arrows. (Q) The number of injured proximal tubules per high-magnified light micrograph of TB sections (E–H). (R) The number of Ki-67-positive nuclei in the proximal tubules per high-power light micrograph of immunohistochemistry. Results are presented as the mean ± standard deviation, *P < 0.05, ****P < 0.0001. Statistical significances were calculated by unpaired two-tailed Student’s t-test (Q) and one-way analysis of variance, followed by the post-hoc Turkey–Kramer tests (R). Bars; (C and D) 200 μm, (E–H) 50 μm, (I–L) 5 μm, (M–P) 50 μm.

Fig 3. Scanning transmission electron microscope-energy-dispersive X-ray spectroscopy (STEM)-EDX analysis of lamellar bodies for elementary mapping.

(A–B) STEM-high-angle annular dark-field (HAADF) images show lamellar bodies in the proximal tubule of high-fat diet (HFD)-fed vehicle mice at 20 weeks. (C–I) EDX elementary mapping images. Each image has the atomic number in the upper left corner and the element name in the lower-left corner. (C) Phosphorus; atomic number 15. (D) Chlorine; atomic number 17. (E) Potassium; atomic number 19. (F) Calcium; atomic number 20. (G) Osmium; atomic number 76. (H) Lead; atomic number 82. (I) Ulan; atomic number 92.

Fig 4. Electron microscopic 3D reconstruction showing the proximal tubule as a characteristic mitochondrial shape in three segments (S1, S2, and S3 segments).

(A) For scanning electron microscopy (SEM), a diamond knife (yellow) was used to cut ultrathin slices from the top of the block-faces of the Epon block (a). Serial block-face SEM (SBF-SEM) can produce a series of hundreds of serial electron micrographs of the block surfaces similar to conventional transmission electron microscopy (TEM) images (b). After the ultrathin sections are cut, the SBF-SEM data are acquired for backscattered electrons using a backscatter detector. (B) Three-dimensional reconstruction of serial SBF-SEM images of proximal tubule segment 1. (C) Lower magnification SBF-SEM image of the outer cortex in the kidney of an STD-vehicle mouse. (D) Higher magnification image of proximal tubule segment 1 in the red box in (C). (E) Lower magnification SBF-SEM image of the outer medulla in the kidney of an STD-vehicle mouse. (F) Higher magnification image of proximal tubule segments 2 and 3 in the blue box in (E). (G–I) Reconstructed 3D image of mitochondria (various colors other than magenta) and lysosomes (magenta). The yellow line along the basement membrane is shown in steps of 500 nm (0.5 um) for every 10 slices. (G) Image of SBF-SEM data corresponding to the S1 segment shown in (D). (H and I) Image of SBF-SEM data corresponding to the S2 and S3 segments shown in (F). (J–L) Reconstructed 3D image of vacuoles (green) and lysosomes (magenta). The yellow line along the basement membrane is shown in steps of 500 nm (0.5 μm) for every 10 slices. The white line along the apical surface of tubular cells (brush border-bottom line) is shown in steps of 500 nm (0.5 μm) for every 10 slices. (J) Image of SBF-SEM data corresponding to the S1 segment shown in (D). (K and L) Image of SBF-SEM data corresponding to the S2 and S3 segments shown in (F). Bars; (C and E) 100 μm, (D and F) 50 μm.

Fig 5. Electron microscopic 3D reconstruction of the HFD phlorizin assay model revealed mitochondrial volume changes.

(A–C) Segmentation of mitochondria (various colors other than magenta) and lysosomes (magenta) on serial block-face (SBF)-SEM 2D images of the proximal tubule S1 segment. (D–F) Three-dimensional reconstruction of serial SBF-SEM images of the proximal tubule S1 segment. (A, D) Standard diet (STD) vehicle, (B, E) high-fat diet (HFD) vehicle, and (C, F) HFD phlorizin. (G–I) Segmentation of mitochondria (various colors other than magenta), lysosomes (magenta), and lamellar bodies (magenta; yellow arrows) on SBF-SEM 2D images of the proximal tubule S2 segment. (J–L) Three-dimensional reconstruction of serial SBF-SEM images of proximal tubule S2 segment. (G, J) STD-vehicle, (H, K) HFD-vehicle, and (I, L) HFD phlorizin. (M–O) Statistical analyses of mitochondrial volume in SBF-SEM reconstruction data. Results are presented as median with an interquartile range (M–O). *P < 0.05, ***P < 0.001, ****P < 0.0001. Statistical significance was calculated by the Kruskal–Wallis test with the post-hoc test, followed by Dunn’s multiple tests. PLZ, phlorizin.

Fig 6. Raman spectrometry showed that lamellar bodies (spherical particles) in the proximal tubule in HFD-vehicle mice are sphingomyelin.

(A) Hematoxylin and eosin (H&E) staining of the frozen sections. Spherical particles (blue arrow) correspond to lamellar bodies in EM images. (B–D) Differential interference contrast (DIC) images of Raman microscopy. (B) Glomerulus (red arrow) in the outer cortex, (C) spherical particle (blue arrow) of the proximal tubule (PT) in the outer cortex, (D) PT segment 3 (S3) (white arrow) in the outer medulla. (E) Schematic diagram of the detection system of a Raman microscope. (F) Name of each part of the Raman spectrum. (G) Raman spectra of PT S3 in the outer medulla (blue), glomerulus (green), and the spherical particle of PT in the outer cortex (red). (H) Five-point variation of Raman spectra of spherical particles in PT in the outer cortex. (I) Five-point variation of Raman spectra of PT S3 in the outer medulla. (J) The spectra of the spherical particles and segment S3 were compared, and the spectrum at 2800–3000 cm⁻¹ derived from the methyl/methylene group differed (H, I). A prominent peak was observed at 2800–3000 cm⁻¹, suggesting that the subtracted Raman peak has more chain hydrocarbon structures. In addition, differential peaks were observed below 1700 cm^−1, and weak P = O bonds were observed at 1063 cm⁻¹ and 1296 cm⁻¹, suggesting that spherical particles (lamellar body) might contain phosphate. Two main types of hydrocarbons were seen: saturated hydrocarbons (such as palmitic acid) and unsaturated hydrocarbons (such as oleic acid). The shift (H2–I3) of the subtracted Raman peak was similar to that of the Raman spectrum of the purified sphingomyelin.

Fig 7. Protocols for urine and tissue collection for sphingomyelin and ceramide analysis and changes in urine volume and proteinuria.

(A) Diagram showing urine collection and tissue sampling scheduled after administration of phlorizin (PLZ) or vehicle (Veh) to C57/BL6J mice fed with a standard diet (STD) or high-fat diet (HFD) at 20 weeks. Urine was collected in the absence of dosing (Fed); 3 days after administration of Veh; after the second administration of Veh (Veh day 1, Veh day 2); and after administration of PLZ (PLZ day 1, PLZ day 2). (B) The urinary volume of PLZ assay. (C) The urinary volume of Veh assay. (D) The urinary protein of PLZ assay. (E) The urinary protein of Veh assay. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Statistical significances were calculated by two-way analysis of variance followed by the post-hoc Turkey–Kramer tests.

Fig 8. Sphingomyelin levels in urinary extracts and those in tissues from the kidneys of standard diet (STD)- and high-fat diet (HFD)-fed mice in the phlorizin assay.

Liquid chromatography-tandem mass spectrometry was used to quantify sphingomyelin levels in urinary extracts from (A) total sphingomyelin (d18:0), sphingomyelin (d18:1), (B) total ceramide (d18:0), and ceramide (d18:1) in the HFD obese mouse model in the phlorizin assay. (C) Total sphingomyelin (d18:0), sphingomyelin (d18:1), (D) total ceramide (d18:0), and ceramide (d18:1) of tissues in the HFD obese mouse model from the phlorizin assay. Results are presented as the mean ± standard deviation. *P < 0.05, **P < 0.01. Statistical significances were calculated by two-way analysis of variance followed by the post-hoc Turkey–Kramer tests and Sidak’s multiple comparison test. Urine was obtained from C57/BL6J mice fed with a STD (STD-Fed) or HFD (HFD-Fed) prior to vehicle administration, vehicle-treated mice (Veh) prior to phlorizin or vehicle administration, phlorizin- or vehicle-treated mice on day one (PLZ-D1 or Veh-D1), and phlorizin- or vehicle-treated mice on day two (PLZ-D2 or Veh-D2).