Prolonged exposure of mouse and human podocytes to insulin induces insulin resistance through lysosomal and proteasomal degradation of the insulin receptor

Abstract

Aims/hypothesis: Podocytes are insulin-responsive cells of the glomerular filtration barrier and are key in preventing albuminuria, a hallmark feature of diabetic nephropathy. While there is evidence that a loss of insulin signalling to podocytes is detrimental, the molecular mechanisms underpinning the development of podocyte insulin resistance in diabetes remain unclear. Thus, we aimed to further investigate podocyte insulin responses early in the context of diabetic nephropathy.

Methods: Conditionally immortalised human and mouse podocyte cell lines and glomeruli isolated from db/db DBA/2J mice were studied. Podocyte insulin responses were investigated with western blotting, cellular glucose uptake assays and automated fluorescent imaging of the actin cytoskeleton. Quantitative (q)RT-PCR was employed to investigate changes in mRNA. Human cell lines stably overproducing the insulin receptor (IR) and nephrin were also generated, using lentiviral constructs.

Results: Podocytes exposed to a diabetic environment (high glucose, high insulin and the proinflammatory cytokines TNF-α and IL-6) become insulin resistant with respect to glucose uptake and activation of phosphoinositide 3-kinase (PI3K) and mitogen-activated protein kinase (MAPK) signalling. These podocytes lose expression of the IR as a direct consequence of prolonged exposure to high insulin concentrations, which causes an increase in IR protein degradation via a proteasome-dependent and bafilomycin-sensitive pathway. Reintroducing the IR into insulin-resistant human podocytes rescues upstream phosphorylation events, but not glucose uptake. Stable expression of nephrin is also required for the insulin-stimulated glucose uptake response in podocytes and for efficient insulin-stimulated remodelling of the actin cytoskeleton.

Conclusions/interpretation: Together, these results suggest that IR degradation, caused by high levels of insulin, drives early podocyte insulin resistance, and that both the IR and nephrin are required for full insulin sensitivity of this cell. This could be highly relevant for the development of nephropathy in individuals with type 2 diabetes, who are commonly hyperinsulinaemic in the early phases of their disease.

Keywords: Albuminuria; Diabetic nephropathy; Genetic background; Insulin resistance; Kidney injury.

Fig. 1

Expression of podocyte markers and key insulin signalling proteins in WT mouse podocytes. Representative western blots and densitometry demonstrating (a) levels of the podocyte markers nephrin, CD2-associated protein (CD2AP), synaptopodin and the heat-sensitive SV40 transgene, and (b) insulin signalling proteins IRβ, IRS-1 and IRS-2, normalised to β-actin levels, under conditions of proliferation (33°C) and differentiation (37°C). *p < 0.05, **p < 0.01, unpaired t test, n = 3

Fig. 2

Loss of IRs in diabetic mouse podocytes in vitro and in glomeruli from db/db mice. WT mouse podocytes were treated for 10 days with 1 ng/ml TNF-α, 1 ng/ml IL-6, 100 nmol/l insulin and 25 mmol/l glucose (labelled Diabetic, D), prior to insulin stimulation (100 nmol/l). (a) dpm counts representing cellular uptake of [³H]2-deoxy-d-glucose following exposure of podocytes to the diabetic factors (D); n = 4. (b) Representative western blots and densitometry of IRS-1, IRβ, IRS-2 and IGF-IRβ protein following exposure of podocytes to the diabetic factors (D), or with mannitol (in parallel with insulin and inflammatory cytokines) included in place of glucose as an osmotic control (O); n = 4. (c) Representative western blots and densitometry of insulin-stimulated phosphorylation of Akt (T308, S473), mitogen-activated ERK-activating kinase (MEK1/2) and p44/42 MAPK (Thr202/Tyr204); n = 3. (d) IR protein in glomeruli isolated from db/db mice at 8 (n = 4, two males and two females) and 12 weeks (n = 5, three males and two females) of age. (e) IRβ protein in podocyte cultures isolated from 3-month-old db/db mice and WT littermate controls; n = 6. *p < 0.05, **p < 0.001, ***p < 0.001, one-way ANOVA, Tukey’s multiple comparison

Fig. 3

Chronic insulin exposure is responsible for the loss of IR protein in mouse podocytes. Total IRβ levels following the exposure of podocytes to either (a) 25 mmol/l glucose (G) or mannitol (M), or (b) TNF-α and IL-6, at the stated concentrations, for 10 days. Representative western blot (top) and matched densitometry (bottom). (c) Total IRβ, IGF-IRβ, IRS-1 and IRS-2 protein following growth of podocytes in the presence of insulin in vitro. ***p < 0.001, one-way ANOVA, Tukey’s multiple comparison; n = 4

Fig. 4

Chronic insulin exposure is sufficient to attenuate podocyte insulin signalling. Mouse podocytes were grown in the presence of insulin (10 and 100 nmol/l) prior to further insulin stimulation (black bars). (a) Representative western blots and matched densitometry of insulin-stimulated phosphorylation cascades; n = 4. (b, c) Insulin-stimulated glucose uptake assays; n = 3 in duplicate. ^† p = 0.041, unpaired t test; *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA, Tukey’s multiple comparison

Fig. 5

High glucose is the cause of IRS-1 loss in podocytes. Mouse podocytes were grown in the presence of 25 mmol/l glucose (G) or mannitol (M) for 10 days prior to insulin stimulation (100 nmol/l, 10 min, black bars). (a) Representative western blot and densitometry of IRS-1 protein in podocytes. n = 3; *p < 0.05; ^† p = 0.09 vs control cells. (b) Representative western blots and densitometry of insulin-stimulated signalling in podocytes following chronic glucose exposure; n = 3. (c, d) Insulin-stimulated glucose uptake assays (n = 4), in duplicate. *p < 0.05, **p < 0.01, ***p < 0.001; no significant difference between any of the insulin-stimulated groups, one-way ANOVA, Tukey’s multiple comparison

Fig. 6

No significant changes in Ir mRNA in podocytes following chronic insulin exposure. (a) Endpoint RT-PCR and (b) qRT-PCR of Ir mRNA following chronic exposure of mouse podocytes to insulin; n = 4. (c) Representative qRT-PCR and (d) endpoint RT-PCR demonstrating the relative abundance of podocyte Ir-A/B mRNA following growth in chronic insulin. **p < 0.01. Individual changes in (e) Ir-A and (f) Ir-B mRNA relative to basal condition. No significant differences, unpaired t test; n = 4 in triplicate

Fig. 7

Proteasome and lysosome inhibition blocks insulin-induced IR degradation in podocytes. Podocytes were treated with (a) 50 nmol/l bafilomycin, (b) 10 μmol/l MG132, or (c) both bafilomycin and MG132 for 8 h, alone or in combination with insulin for 8 or 24 h. IRβ levels were determined by western blotting. Levels of p62 and ubiquitin (Ub) were determined as positive controls for successful lysosomal and proteasomal inhibition, respectively. (d) Representative western blots and matched densitometry showing IRβ levels after mTOR inhibition (24 h, 10 nmol/l rapamycin) alone or in combination with chronic insulin stimulation; phosphorylation of mTOR (S2448) was determined as a positive control for successful mTOR inhibition; n = 4–6 experiments. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA, Tukey’s multiple comparison; ^† p = 0.0151, two-tailed t test vs 24 h

Fig. 8

Stable IR and nephrin expression rescues insulin signalling in insulin-resistant human podocytes. Human (WT) podocytes were stably transfected with lentiviral particles containing human IR (WT-IR), human nephrin (WT-Neph) or both (WT-IR-Neph), and insulin responses were investigated. (a) Representative western blots and matched densitometry demonstrating increased phosphorylation of Akt (S473, T308) in WT-IR podocytes. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA, Tukey’s multiple comparison; n = 4. Insulin-stimulated glucose uptake experiments showed no significant increase in this response in either (b) WT or (c) WT-IR podocytes. (d) Representative western blots demonstrating stable expression of nephrin and IR in appropriate cell lines, and insulin-stimulated phosphorylation (nmol/l, 10 min) of Akt (S473); n = 3. (e) WT-IR-Neph cells showed a significantly increased percentage glucose uptake following insulin stimulation (100 nmol/l), compared with unstimulated cells. *p = 0.0142 Mann–Whitney vs unstimulated cells. (f) WT-Neph cells have no significant increase in insulin-stimulated glucose uptake; n = 4

Fig. 9

IR and nephrin expression are both required for effective insulin-stimulated actin remodelling in podocytes. Human (WT) podocytes were stably transfected with lentiviral particles containing human nephrin (WT-Neph), human IR (WT-IR) or both (WT-IR-Neph). Cell lines were stimulated with insulin at the stated doses and times, prior to F-actin and nuclear staining. The percentage of cells positive for actin reorganisation was calculated as described in the Methods. Modest changes in brightness and contrast were uniformly applied to all images for visual purposes; unmodified images were used for quantification. (a) Representative fluorescent images of each cell line under basal conditions and following insulin stimulation (100 nmol/l, 10 min). Scale bar, 50 μm. (b) Percentage of WT-IR-Neph cells displaying evidence of F-actin remodelling when stimulated with insulin at 1 nmol/l (light grey line), 10 nmol/l (dark grey line) and 100 nmol/l (black line). (c) Comparison of responses between stable cell lines (grey line, WT; grey dotted line, WT-Neph; black dotted line, WT-IR; black line, WT-IR-Neph) stimulated with 100 nmol/l insulin over the time course. (d) Effect of insulin-resistant conditions on actin remodelling in WT-IR-Neph cells stimulated with 100 nmol/l insulin over the time course (black line, basal conditions; black dotted line, diabetic media). (e) Bar graph of data from (c, d) for 100 nmol/l insulin at 10 min. A significant increase in percentage of WT-IR-Neph podocytes positive for actin remodelling was observed; this response was lost in WT-IR-Neph podocytes exposed to insulin-resistant conditions. *p = 0.036 Mann–Whitney test, n = 3

Fig. 10

Proposed mechanism of podocyte IR degradation and its consequences. Normal insulin signalling to podocytes leads to phosphorylation of IRS proteins, predominantly IRS-2 [54], and activation of downstream signalling events, including the phosphorylation of Akt and mTOR. Nephrin is also required for insulin-stimulated glucose uptake (in part via interaction of nephrin with VAMP2 [39, 40] and insulin-stimulated actin remodelling). Hyperinsulinaemia, occurring in diabetes and insulin resistance, causes an increase in proteasomal and lysosomal degradation of the podocyte IR, attenuating downstream signal transduction. Ultimately, loss of podocyte IRs may exacerbate albuminuria and features of diabetic nephropathy, dysregulating the ER stress response [17], VEGF-A secretion [56], ROS production [57], autophagy [57], cell viability [20] and actin remodelling, and reducing glucose uptake. ER, endoplasmic reticulum; GSV, GLUT storage vesicle; UPR, unfolded protein response; VAMP2, vesicle-associated membrane protein 2