Apical proximal tubule fatty acid uptake–generated ceramides cause endoplasmic reticulum stress from altered membrane fluidity
Zhiyu Liu, Robert J. Gaivin, Shenaz Khan, Vincent Li, Amal Chaba, Fraser J. Moss, Usman Sabir, Takhar Kasumov, Tingwei Mu, Jeffrey R. Schelling
Zhiyu Liu, Robert J. Gaivin, Shenaz Khan, Vincent Li, Amal Chaba, Fraser J. Moss, Usman Sabir, Takhar Kasumov, Tingwei Mu, Jeffrey R. Schelling
View: Text | PDF
Research Article Cell biology Nephrology

Apical proximal tubule fatty acid uptake–generated ceramides cause endoplasmic reticulum stress from altered membrane fluidity

Abstract

Circulating fatty acids (FAs) are constitutively taken up by basolateral kidney proximal tubule transporters and are the preferred metabolic substrate. In many chronic kidney diseases, the damaged glomerular filtration barrier permits passage of albumin-bound FAs, which are reabsorbed by apical FA transport protein-2 (FATP2). Bilateral FA uptake leads to lipotoxicity and progressive renal function decline, but the relative apical versus basolateral contribution and intracellular mechanisms are not established. Apical or bilateral (but not basolateral) palmitate incubation with human proximal tubule cells stimulated endoplasmic reticulum (ER) stress gene expression, ER stress pathway activation, and ER fragmentation. Apical or bilateral palmitate was associated with reduced lipid droplets and decreased expression of ER-localized lipid droplet biogenesis transcripts. Inhibition of lipid droplet formation also precipitated ER stress, suggesting diminished sequestration of FA metabolites as the cause. Indeed, C16:0 ceramide was increased in bilateral palmitate-treated cells and in kidneys from mice that phenocopy progressive diabetic kidney disease. Ceramide synthesis inhibition abrogated ER stress, and transfection with C16:0 ceramide decreased ER membrane fluidity and caused ER stress. We conclude that aberrant filtration and uptake of FAs by apical FATP2 exceeded the capacity for lipid droplet incorporation and led to cytotoxicity from ceramide-induced ER lipid bilayer stress.

Authors

Zhiyu Liu, Robert J. Gaivin, Shenaz Khan, Vincent Li, Amal Chaba, Fraser J. Moss, Usman Sabir, Takhar Kasumov, Tingwei Mu, Jeffrey R. Schelling

×

Figure 5

In vivo and in vitro models are associated with decreased LD biogenesis.

Options: View larger image (or click on image) Download as PowerPoint
In vivo and in vitro models are associated with decreased LD biogenesis....
(A) Representative kidney cortex section from db/db mouse stained for LDs with LipidSpot, as described in Methods. Original magnification, ×400. (B) Quantification by ImageJ of LDs from kidney cortex of mice (4 non-consecutive sections from 4 mice per genotype). Data are percentage tissue area that contained LDs from an individual mouse and are expressed as mean ± SEM, and were analyzed by 1-way ANOVA and Tukey’s post hoc test for multiple comparisons. (C) BL and/or AP surfaces of human proximal tubule cells on permeable supports were incubated with albumin (0.2%) or albumin (0.2%) complexed with palmitate (100 μM) for 16 hours at 37°C. Cells were fixed, labeled for LDs with LipidSpot as described in Methods, and quantified using ImageJ. Each symbol represents mean LD number in 110 to 289 individual cells per group from 3 separate experiments. Data are expressed as mean ± SD. Data from multiple groups were analyzed by 1-way ANOVA and Tukey’s post hoc test for multiple comparisons. (D) LD biogenesis gene transcript expression in human proximal tubule cells treated with AP + BL versus BL palmitate (100 μM complexed with 0.2% BSA, 16 hours, 37°C). Data are mean ± SD from 3 experiments. P values reflect AP + BL versus BL palmitate comparisons by 2-tailed t test. FPKM, fragments per kilobase transcript per million mapped reads; BSCL2, seipin; FITM2, fat storage–inducing transmembrane protein-2; DGAT, diacylglycerol acyltransferase; PLIN3, perilipin-3; AGPAT1, 1-acyl-sn-glycerol-3-phosphate acyltransferase-α; ACSL5, long-chain acyl-CoA synthetase-5. (E) Human proximal tubule cells were pretreated for 1 hour with the DGAT2 inhibitor PF-06424439 at indicated concentrations, then treated with AP + BL or BL palmitate (100 μM complexed with 0.2% BSA, 16 hours, 37°C), and probed for CHOP or β-actin by immunoblot analysis, as described in Methods (N = 3 biologically independent samples). Immunoblot data were quantitated using ImageJ software and are shown in histograms below representative blots. Data are depicted as mean ± SEM and were analyzed by 1-way ANOVA and Tukey’s post hoc test for multiple comparisons.