FGF1^ΔHBS delays the progression of diabetic nephropathy in late-stage type 2 diabetes mouse model by alleviating renal inflammation, fibrosis, and apoptosis

Abstract

Elderly adults are at higher risk for developing diabetic complications including diabetic nephropathy (DN), contributing to excess morbidity and mortality in elderly individuals. A non-mitogenic variant of fibroblast growth factor 1 (FGF1^ΔHBS) was demonstrated to prevent DN in an early-stage (2-month-old) type 2 diabetes (T2D) mouse model. The present study aimed to investigate the potential therapeutic effects of FGF1^ΔHBS against the progression of renal dysfunction in a late-stage T2D mouse model with established DN. Nine-month-old db/db mice were administered FGF1^ΔHBS every other day for 3 months. db/db mice at 12-month-old without FGF1^ΔHBS treatment exhibited high blood glucose level and elevated urine albumin-to-creatinine ratio. FGF1^ΔHBS treatment effectively reversed hyperglycemia, delayed the development of renal dysfunction, and reduced kidney size and weight. Furthermore, FGF1^ΔHBS treatment significantly prevented the progression of renal morphologic impairment. FGF1^ΔHBS treatment demonstrated anti-inflammatory and anti-fibrotic effects, with significantly decreased protein levels of key pro-inflammatory cytokines and pro-fibrotic factors in kidney. Moreover, FGF1^ΔHBS treatment greatly decreased apoptosis of renal tubular cells, accompanied by significant downregulation of the proapoptotic protein and upregulation of the antiapoptotic protein and peroxisome proliferator-activated receptor α (PPARα) expression in kidney. Mechanistically, FGF1^ΔHBS treatment directly protected mouse proximal tubule cells against palmitate-induced apoptosis, which was abolished by PPARα inhibition. In conclusion, this study demonstrated that FGF1^ΔHBS delays the progression of renal dysfunction likely through activating PPARα to prevent renal tubule cell death in late-stage T2D, exhibiting a promising translational potential in treating DN in elderly T2D individuals by ameliorating renal inflammation, fibrosis and apoptosis.

1. Introduction

Today more than half of all diabetic individuals in the United States are over 60 years old, and diabetes prevalence peaks between 65 and 74 years old [1]. The high burden of comorbidities, physical disability, cognitive impairment, malnutrition, and increased susceptibility to the complications of dysglycemia and polypharmacy contribute to excess morbidity and mortality in elderly diabetic individuals with complications, including diabetic nephropathy (DN) [2,3]. These characteristics challenge therapeutic management in elderly diabetic individuals with DN. The conventional therapies to prevent and/or treat DN include rigorous blood glucose and blood pressure control, renin angiotensin inhibitors application, lipids lowering, along with lifestyle and dietary interventions. There are no treatment strategies available that specifically target the pathogenesis of DN [4]. Furthermore, in clinical situations many factors interrupt the diagnosis and intervention of the onset diabetic complications [5]. Consequently, DN is usually diagnosed long after it begins in elderly patients. Therefore, development of therapies to treat or reverse pre-existing DN are urgently needed.

Fibroblast growth factor 1 (FGF1), a member of paracrine FGF sub-family, exhibits mitogenic activity in cells from a variety of tissues including liver, vasculature, and skin [6,7]. Recently, FGF1 was shown to exert an unexpected metabolic role in obesity and diabetes [8,9]. Our previous study demonstrated FGF1 protected against DN in type 1 diabetes and type 2 diabetes (T2D) through its anti-inflammatory effect [10]. However, the mitogenic activity of FGF1 increases the risk of stimulating tumorigenesis, thus limits the direct application of FGF1 for diabetes treatment [11–13]. Fortunately, our recent study showed the mitogenic and metabolic functions of FGF1 could be uncoupled by generating a non-mitogenic FGF1 variant with 3 substitutions of heparin-binding sites (i.e., FGF1^ΔHBS). This FGF1^ΔHBS dramatically reduced its proliferative potential yet maintained its full metabolic activity of native FGF1 in vitro and in vivo [14]. Our collaborators have demonstrated preventive effects of FGF1^ΔHBS against DN and adriamycin-induced chronic nephropathy in mouse models [15]. More importantly, our recent study demonstrated FGF1^ΔHBS could reverse nonalcoholic fatty liver disease in late-stage db/db T2D mice [16], indicating a promising potential for application of FGF1^ΔHBS in treatment and/or reversal of established DN.

This study aimed to investigate the therapeutic effects of FGF1^ΔHBS against the progression of renal dysfunction in a late-stage T2D mouse model with established DN. Our data demonstrate for the first time that FGF1^ΔHBS delays the progression of DN in a late-stage db/db T2D mouse model by reversing hyperglycemia, suppressing progression of renal dysfunction and promoting renal morphological changes. Simultaneously, FGF1^ΔHBS protected against the progression of DN is accompanied by a significant inhibition of renal inflammation, fibrosis and apoptosis.

2. Materials and methods

2.1. Late-stage db/db T2D mouse model

Male db/db (BKS.Cg-Dock7^m +/+ Lepr^db/J, Stock # 000642) mice were purchased from Jackson Laboratory (Bar Harbor, ME). Blood and kidney tissues of 9-month-old male db/db mice and age- and sex-matched littermate db/m control mice were collected to characterize the features of established DN. For the therapeutic study, 9-month-old db/db mice were administered with FGF1^ΔHBS (0.5 mg/kg body weight) or phosphate buffered solution (PBS) vehicle via intraperitoneal injection every other day for 3 months according to our previous study [16]. At the indicated time-points, the fasting blood glucose levels and body weight were recorded, and the spot urine was collected and stored for the final renal function evaluation. Blood glucose levels were determined by a FreeStyle complete blood glucose monitor (Abbott Diabetes Care Inc., Alameda, CA). All mice were euthanized at the age of 12 months. Kidney weights were normalized by tibial length, and kidney tissues were collected for the following studies. All mice were housed in a temperature-controlled room (25 °C) on a 12:12-h light/dark cycle and allowed standard rodent diet and water ad libitum. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the University of Louisville.

2.2. Bone mineral and fat composition test

Bone mineral and fat composition were tested by using Lunar PIXImus Dual-Energy X-Ray Absorptiometer (DEXA) scanning (Lunar PIXImus densitometer, WI, USA). The fat tissue content, bone mineral content (BMC) and bone mineral density (BMD) were analyzed with PIXImus software.

2.3. Kidney function analysis

Urine albumin-to-creatinine ratio (UACR) in spot urine was measured to evaluate the renal function [17]. Urinary albumin was measured using a Mouse Albumin ELISA Quantification Kit (E90–134, Bethyl Laboratories, Montgomery, TX). Creatinine was detected using a QuantiChrom Creatinine Assay Kit (DICT-500, Bioassay Systems, Hayward, CA) according to the manufacturer’s instructions.

2.4. Kidney histopathological analysis

Kidney paraffin-embedded sections were processed as previously described [10]. Kidney tissues were fixed in 10% formalin, embedded in paraffin, and sectioned at 5 μm thickness using a microtome. After deparaffinization and rehydration, the kidney sections were subjected to histological staining using standard protocols. Hematoxylin and eosin (H&E) stained sections were used to observe the changes of glomerular size, histopathological changes, renal tubular lesion, and renal epithelial cell histopathology. Sirius-red staining was performed to evaluate renal fibrosis [18]. Periodic Acid-Schiff (PAS) staining was performed to observe the mesangial matrix accumulation in glomerular and tubular lesions [19,20]. All images of stained sections were taken using an BX43 Olympus microscope with a DP74 digital camera (Olympus, Tokyo, Japan). Image-Pro Plus 6.0 image analysis software (Media Cybernetics, Inc. Rockville, MD) was used to quantify results for each staining as needed. The positive staining areas of each assay were analyzed double-blinded from random fields of each section.

2.5. Western blot analysis

Renal tissues were homogenized in RIPA buffer containing protease inhibitor cocktail to extract the protein. The protein concentration was determined by the bicinchoninic acid method (Bio-Rad, Hercules, CA). Thirty micrograms of protein obtained from each sample was loaded onto 8–12% sodium dodecyl sulphate polyacrylamide gels and electro-transferred onto a nitrocellulose membrane (Bio-Rad, Hercules, CA, USA). The membrane was blocked with 5% non-fat milk in tris-buffered saline with 0.05% Tween 20 (TBST) for 1 h and incubated with the following primary antibodies: Ki-67, transforming growth factor β (TGFβ), fibronectin, collagen I, intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion protein 1 (VCAM-1), tumor necrosis factor α (TNFα), cluster of differentiation 68 (CD68) (Abcam; 1:1000), proliferating cell nuclear antigen (PCNA) (Cell Signaling Technology, 1:1000), connective tissue growth factor (CTGF) (Santa Cruz Biotechnology; 1:1000), plasminogen activator inhibitor-1 (PAI-1) (BD Biosciences; 1:1000) and β-actin antibodies (Santa Cruz Biotechnology; 1:3000) in TBST 4 °C overnight. β-Actin was used as a loading control. The membranes were washed with TBST 3 times and incubated with appropriate secondary antibodies (Cell Signaling Technology, 1:5000) conjugated with horseradish peroxidase for 1 h at room temperature. After washing with TBST 3 times, immunoreactive bands were detected using enhanced chemiluminescence reagents (Bio-Rad, Hercules, CA). The abundance of the target proteins was analyzed using Image Lab analysis software (Bio-Rad, Hercules, CA) and normalized against respective loading controls.

2.6. Immunohistochemistry

Kidney sections were processed via a standard immunohistochemical staining protocol as previously described [19]. After deparaffinization and hydration, sections were incubated in 0.3% H₂O₂ in methanol for 30 min and then were subjected to antigen retrieval using retrieval solution in a pressure cooker for 20 min. Sections were blocked with 5% bovine serum albumin in PBS for 30 min at room temperature. Subsequently, the sections were incubated with the following primary antibodies: TGFβ (Abcam; 1:300), fibronectin (Abcam; 1:300), collagen I (Abcam; 1:500), CTGF (Abcam; 1:500), collagen IV (Abcam; 1:500), ICAM (Abcam; 1:400), VCAM-1 (Abcam; 1:300), TNFα (Abcam; 1:300) and CD68 (Abcam; 1:1000) at 4 °C overnight. After washing with PBS three times, sections were incubated with horseradish peroxidase-conjugated secondary antibodies (Abcam; 1:500) for 1 h at room temperature. The reaction was colored with 3,3-diaminobenzidine (DAB, Vector Laboratories) and counterstained with hematoxylin. The results were imaged at a magnification of 400× using a BX43 Olympus microscope with a DP74 digital camera (Olympus, Tokyo, Japan). The positively stained area was quantified in randomly selected fields per section using Image-Pro Plus 6.0 image analysis software.

2.7. Apoptosis assay

A terminal deoxynucleotidyl transferase (TdT) dUTP nickend-labeling (TUNEL) assay was used to detect cell apoptosis in kidney. TUNEL assay was performed using the DeadEnd^™ Fluorometric TUNEL System (Cat No: G3250, Promega Corporation, US) according to the manufacturer’s instructions. In brief, sections were deparaffinized and rehydrated. Sections were treated with 20 μg/mL proteinase K for 10 min and then washed with PBS for 5 min. After immersing in 4% formaldehyde solution for 5 min and washing with PBS, the sections were covered with 100 μL of equilibration buffer for 10 min, and then were incubated with 50 μl of rTdT buffer at 37 °C for 60 min followed by terminating the reaction in 2× saline-sodium citrate buffer at room temperature for 15 min. After washing with PBS, sections were mounted with gold antifade reagent with 4^′,6-diamidino-2-phenylindole (DAPI) to stain nuclei. Images were taken immediately under a BX43 Olympus microscope with a DP74 digital camera (Tokyo, Japan) using a standard fluorescein filter set. The Image-Pro Plus 6.0 image analysis software was used to quantify the TUNEL-stained nuclei.

2.8. Cell culture and treatment

Mouse proximal tubule (BUMPT) cell line was obtained from Boston University School of Medicine, Boston, MA. BUMPT cells were cultured in DMEM containing 1 mmol/L d-glucose, 10% fetal bovine serum (FBS), 100 units/mL penicillin and 100 μg/mL streptomycin, and incubated in a humidified atmosphere of 5% CO₂ at 37 °C, and passaged every 2 days by trypsinization. BUMPT cells were seeded in 6-well culture plates at an initial density of 20,000 cells/well and grown for 24 h before treatment.

Palmitate (Pal, Cat No: P9767, Sigma-Aldrich) was dissolved in 50% ethanol and mixed with bovine serum albumin (BSA) to make Pal-BSA stock solution. BUMPT cells were exposed to the Pal-BSA solution to mimic the renal tubule cell damage in diabetic mice as described in our previous studies [16,21]. The peroxisome proliferator-activated receptor (PPARα) inhibitor (GW6471, Cat No: 11697, Cayman Chemical) was dissolved in dimethyl sulfoxide (DMSO). After serum starvation for 24 h, BUMPT cells were exposed to GW6471 (10 μmol/L) or DMSO in FBS free medium for 8 h. Then BUMPT cells were pretreated with FGF1^ΔHBS (100 nmol/L) for 1 h, followed by exposure to 90 μmol/L Pal-BSA or BSA in the presence and absence of FGF1^ΔHBS for additional 16 h. Then cells were collected to detect PPARα expression and apoptotic indicators.

2.9. Statistical analysis

Differences between two groups were determined by the unpaired two-tail Student t-test. All data were expressed as mean ± SEM. When there were two factors and multiple groups involved, F-tests for two-way Analysis of Variance (ANOVA) were used to determine whether there was a group difference. If there was a group difference, post-hoc t-tests were applied to examine which groups were significantly different. Differences were considered significant when P < 0.05.

3. Results

3.1. Established DN in late-stage T2D mice

T2D phenotypes of 9-month-old db/db mice were characterized in our most recent report [16]. Compared with the age- and sex-matched littermate db/m control mice, 9-month-old db/db mice exhibited renal hypertrophy reflected by obviously increased ratio of kidney weight to tibia length (Fig. 1A) and renal dysfunction reflected by dramatically elevated UACR (Fig. 1B). In addition, 9-month-old db/db mice exhibited obvious renal morphological and histopathological abnormalities: glomerular hypertrophy indicated by H&E staining of the enlarged glomerular size (Fig. 1C, D), glomerular remodeling mirrored by PAS staining of the abundant mesangial matrix accumulation in glomeruli (Fig. 1E, F), and tubular interstitial fibrosis reflected by Sirius Red staining of collagen deposition (Fig. 1G, H).

Fig. 1.

Features of established diabetic nephropathy (DN) in 9-month-old db/db diabetic mice. (A) The ratio of kidney weight to tibia length. (B) Urine albumin-to-creatinine ratio (UACR) in spot urine. (C) Representative images of H&E-stained kidney sections and (D) glomerular size quantified from images of H&E staining. (E) Representative images of PAS-stained kidney sections and (F) mesangial matrix expansion (% glomerular area) from images of PAS staining. (G) Representative images of Sirius Red staining and (H) quantified Sirius Red positive-stained area of kidney sections. Quantitative data are normalized to those of db/m control mice and expressed as mean ± SEM, n = 3 for db/m mice, n = 4 for db/db mice. *P < 0.05, **P < 0.01, vs. db/m littermate controls.

3.2. FGF1^ΔHBS delays the progression of DN in late-stage T2D mice

To extend the previous observation [15] that FGF1^ΔHBS treatment initiated at 2 months of age ameliorates DN in db/db mice, here we explore the therapeutic effects of FGF1^ΔHBS on DN in a late-stage T2D mice. FGF1^ΔHBS treatment was initiated at 9 months and continued through 12 months of age in db/db mice. During this period, db/db mice without treatment maintained constantly high blood glucose levels (Fig. 2A) and exhibited gradually deteriorating renal function indicated by the initial high UACR at 9 and 10 months of age, and dramatically increased UACR at 11 and 12 months of age (Fig. 2B). Whereas 3 months of FGF1^ΔHBS treatment almost completely normalized blood glucose levels (Fig. 2A) and remarkably suppressed the gradual elevation of UACR that was observed in db/db mice without treatment (Fig. 2B). These findings indicate a significantly delayed progression of DN by FGF1^ΔHBS.

Fig. 2.

FGF1^ΔHBS delays the progression of DN in late-stage db/db mice. 9-month-old db/db mice were administered with FGF1^ΔHBS (0.5 mg/kg body weight) or PBS vehicle via intraperitoneal injection every other day for 3 months. (A) Blood glucose level. (B) UACR in spot urine. Quantitative data are expressed as mean ± SEM, n = 7 for PBS group, n = 5 for FGF1^ΔHBS group. ^#P < 0.01 and ^&P < 0.01 vs. db/db mice at 9-month-old and 10-month-old, respectively. (C) The kidney size and (D) the ratio of kidney weight to tibia length. (E) Representative images of H&E staining of glomeruli and renal tubules in kidney sections. Blue arrows indicate dilated glomerular blood capillaries; yellow arrows indicate exfoliation of epithelial cells in the renal tubular lumen; black arrows indicate renal tubular vacuolar lesions. (F) Glomerular size quantified from images of H&E staining. (G) Representative images of PAS staining of glomeruli and renal tubules in kidney sections and (H) mesangial matrix expansion (% glomerular area) quantified from images of PAS staining, normalized to PBS group. (I) Representative images of Sirius Red staining of glomeruli and renal tubules in kidney sections and (J) quantified Sirius Red positive-stained area of kidney sections, normalized to PBS group. Scale bars = 50 μm. Quantitative data are expressed as mean ± SEM, n = 7 for PBS group, n = 5 for FGF1^ΔHBS group. *P < 0.05, **P < 0.01 and ***P < 0.001 vs. PBS group,

Meanwhile, FGF1^ΔHBS treatment improved renal morphology, which was evident as obviously reduced kidney size (Fig. 2C) and weight (Fig. 2D). Furthermore, FGF1^ΔHBS treatment remarkably improved the renal histopathological features: from the observation of H&E staining, db/db mice showed an abnormal kidney structure in the form of irregular Bowman’s capsules with dilated glomerular blood capillaries in glomeruli, severe tubular vacuolar lesions and exfoliation of epithelial cells in renal tubules, and enlargement of glomerular size; these histopathological changes were largely improved by FGF1^ΔHBS treatment (Fig. 2E, F). Moreover, FGF1^ΔHBS treated db/db mice showed significant decreased PAS positive staining areas in glomeruli and renal tubules, indicating a greatly reduced accumulation of mesangial matrix in glomeruli and expansion of basement membrane in tubules (Fig. 2G, H). In addition, FGF1^ΔHBS treatment ameliorated renal fibrotic remodeling, which was reflected by significantly reduced Sirius Red staining of collagen deposition in glomeruli and renal tubules (Fig. 2I, J). The common side effects associated with current insulin-sensitizing therapies are weight gain and bone loss [22,23]. Here we evaluated the potential side effects of FGF1^ΔHBS in db/db mice. The results demonstrated FGF1^ΔHBS treatment significantly reduced body weight (Supplemental Fig. 1A) and fat composition (Supplemental Fig. 1B, C); whereas the BMC, BMD and total bone area were not affected (Supplemental Fig. 1D–F). As expected, FGF1^ΔHBS treatment significantly decreased renal cell proliferative marker Ki67 and PCNA expression (Supplemental Fig. 1G–I).

3.3. FGF1^ΔHBS attenuates renal inflammation in late-stage T2D mice

Inflammation has emerged as a key pathophysiological mechanism in DN [10]. To evaluate the anti-inflammatory effect of FGF1^ΔHBS in the kidney of late-stage T2D mice, we measured levels of the pro-inflammatory cytokines in renal tissues. FGF1^ΔHBS treatment significantly downregulated the protein expression of ICAM-1 (Fig. 3A, B), VCAM-1 (Fig. 3A, C), TNFα (Fig. 3A, D), and CD68 (Fig. 3A, E), indicating FGF1^ΔHBS alleviated the inflammatory responses in DN.

Fig. 3.

The effect of FGF1^ΔHBS on the expression of renal inflammatory factors in late-stage diabetic mice. (A) The inflammatory protein expressions of ICAM-1, VCAM-1, TNFα and CD68 were detected by western blot and β-actin was used as a loading control. The expression abundance of (B) ICAM-1, (C) VCAM-1, (D) TNFα and (E) CD68 was quantified by densitometry analysis. Quantitative data are expressed as mean ± SEM, n = 5 for PBS group, n = 4 for FGF1^ΔHBS. *P < 0.05 and **P < 0.01 vs PBS group.

To further confirm the anti-inflammatory activity of FGF1^ΔHBS in DN, we performed immunohistochemical staining of these key inflammatory cytokines in renal sections. The expression of ICAM-1 (Fig. 4A, B) and VCAM-1 (Fig. 4A, C) were predominantly localized in renal tubules. In contrast, the expression of TNFα (Fig. 4A, D) and CD68 (Fig. 4A, E) were predominantly localized in glomeruli, which were all significantly attenuated by FGF1^ΔHBS treatment. These findings demonstrate that FGF1^ΔHBS plays an essential role in controlling excessive inflammation in DN.

Fig. 4.

Effect of FGF1^ΔHBS on protein expression and localization of renal inflammatory factors in late-stage diabetic mice. (A) Representative images of immunohistochemical staining of inflammatory markers (ICAM-1, VCAM-1, TNFα and CD68) in kidney sections. Relative expression abundance of (B) ICAM-1, (C) VCAM-1, (D) TNFα and (E) CD68 on images of immunohistochemical staining was quantified using Image-Pro Plus 6.0 and normalized to PBS group. Scale bars = 50 μm. Quantitative data are expressed as mean ± SEM, n = 7 for PBS group, n = 5 for FGF1^ΔHBS group. **P < 0.01 vs PBS group.

3.4. FGF1^ΔHBS ameliorates renal fibrosis in late-stage T2D mice

Fibrosis is a key hallmark of progressive kidney disease in diabetes. TGFβ, a profibrotic factor, has emerged as a critical factor in renal fibrosis [24]. Treatment with FGF1^ΔHBS markedly decreased protein expression of TGFβ (Fig. 5A, B), accompanied by significant downregulation of downstream fibrotic mediators including CTGF (Fig. 5A, C) and PAI-1 (Fig. 5A, D) in kidneys. In addition, FGF1^ΔHBS treatment markedly reduced the expression of extracellular matrix protein fibronectin (Fig. 5A, E) and collagen I (Fig. 5A, F) as compared to the control group.

Fig. 5.

Effect of FGF1^ΔHBS on the expression of renal fibrotic factors in late-stage diabetic mice. (A) The fibrotic protein expressions of TGFβ, CTGF, PAI-1, fibronectin and collagen I were detected by western blot and β-actin was used as a loading control. The expression abundance of (B) TGFβ, (C) CTGF, (D) PAI-1, (E) fibronectin, and (F) collagen I were quantified by densitometry analysis, respectively. Quantitative data are expressed as mean ± SEM, n = 5 for PBS group, n = 4 for FGF1^ΔHBS group. *P < 0.05 and **P < 0.01 vs PBS group.

We confirmed the effects of FGF1^ΔHBS on expression of renal fibrotic factors by immunohistochemical staining. In db/db mice, the expression of TGFβ (Fig. 6A, B) and CTGF (Fig. 6A, C) were primarily localized in renal tubules; whereas the expression of fibronectin (Fig. 6A, D), collagen I (Fig. 6A, E) and collagen IV (Fig. 6A, F) were localized in both glomeruli and renal tubules. All fibrotic factors were significantly decreased by FGF1^ΔHBS treatment. These findings demonstrate a potent anti-fibrotic activity of FGF1^ΔHBS in DN.

Fig. 6.

The effect of FGF1^ΔHBS on the protein expression and localization of renal fibrotic factors in late-stage diabetic mice. (A) The representative images of immunohistochemical staining of fibrotic markers (TGFβ, CTGF, fibronectin, collagen I and collagen IV) in kidney sections. The relative expression abundance of (B) TGFβ, (C) CTGF, (D) fibronectin, (E) collagen I, and (F) collagen IV on image of immunohistochemical staining was quantified using Image-Pro Plus 6.0 and normalized to PBS group. Scale bars = 50 μm. Quantitative data are expressed as mean ± SEM, n = 7 for PBS group, n = 5 for FGF1^ΔHBS group. *P < 0.05 vs PBS group.

3.5. FGF1^ΔHBS protects against renal tubular apoptosis in late-stage T2D mice

Apoptotic cell death is a pivotal event for the initiation and progression of renal dysfunction [25]. The balance between proapoptotic factors and antiapoptotic factors in apoptotic pathway determines the development of renal cell death [26]. In the present study, we observed significantly increased expression of the predominant antiapoptotic protein, Bcl-2 (Fig. 7A, B), and markedly decreased expression of the critical proapoptotic protein, Bax (Fig. 7A, C), after FGF1^ΔHBS treatment in db/db mice. Accordingly, there was a significant reduction of the cleaved-caspase 3 protein expression (Fig. 7A, D) in the kidneys of FGF1^ΔHBS treated db/db mice as compared to the untreated db/db mice. Furthermore, TUNEL staining confirmed renal cell death was prevented by FGF1^ΔHBS. Consistent with a previous report [27], db/db mice exhibited primarily cell apoptosis in renal tubules, which was markedly attenuated by FGF1^ΔHBS treatment (Fig. 7E, F).

Fig. 7.

The effect of FGF1^ΔHBS on renal apoptosis in late-stage diabetic mice. (A) The expressions of Bcl-2, Bax and cleaved-caspase 3 were detected by western blot and β-actin was used as a loading control. The expression abundance of (B) Bcl-2, (C) Bax, and (D) cleaved-caspase 3 was quantified by densitometry analysis. Quantitative data are expressed as mean ± SEM, n = 5 for PBS group, n = 4 for FGF1^ΔHBS group. *P < 0.05 and ***P < 0.001 vs PBS group. (E) Apoptotic cell death was detected by TUNEL staining of kidney sections. The sections were counterstained with DAPI and images were merged with the TUNEL staining (Merged). Glomerulus was circled in red. (F) Quantitative analysis of TUNEL positive nuclear number per field. Scale bars = 50 μm. Quantitative data are expressed as mean ± SEM, n = 7 for PBS group, n = 5 for FGF1^ΔHBS group. *P < 0.05 and ***P < 0.001 vs PBS group.

3.6. PPARα inhibition attenuates FGF1^ΔHBS protection against palmitate-induced apoptosis in mouse renal tubule cells

PPARα is abundantly expressed in the proximal tubules, and is implicated in beneficially regulating renal metabolism, inflammation, oxidative stress and cell survival [28]. In the present study, db/db mice treated with FGF1^ΔHBS showed a significant up-regulation of PPARα expression in kidney (Fig. 8A, B). To further explore the specificity of FGF1^ΔHBS effects in renal tubules, BUMPT cells were exposed to palmitate (Pal) in vitro to mimic diabetes-induced renal tubule cell damage. Paralleling the renal response to FGF1^ΔHBS treatment in vivo (Fig. 8A, B), Pal exposure of BUMPT cells downregulated PPARα expression (Fig. 8C–D) and induced apoptosis indicated by caspase 3 cleavage (Fig. 8C–E), which were markedly reversed by FGF1^ΔHBS treatment. Compared with DMSO vehicle control, inhibiting PPARα abolished the protective effects of FGF1^ΔHBS (Fig. 8C, E). These data indicate that PPARα activation may be important to FGF1^ΔHBS protection from apoptosis in renal tubules of DN.

Fig. 8.

FGF1^ΔHBS upregulates PPARα in kidney and prevents palmitate (Pal)-induced apoptosis in BUMPT cells mediated by PPARα. (A) The protein expression of PPARα was detected by western blot and β-actin was used as a loading control. The expression abundance of (B) PPARα was quantified by densitometry analysis. Quantitative data are expressed as mean ± SEM, n = 5 for PBS group, n = 4 for FGF1^ΔHBS. **P < 0.01 vs PBS group. The BUMPT cells were exposed to PPARα inhibitor GW6471 (10 μmol/L) or DMSO control for 8 h, then pretreated with FGF1^ΔHBS (100 nmol/L) for 1 h and followed by exposure to Pal (90 μmol/L) for additional 16 h in the presence or absence of FGF1^ΔHBS. (C) The protein expression of PPARα and caspase 3 cleavage was detected by western blot and β-actin was used as a loading control. (D) The expression abundance of PPARα and (E) caspase 3 cleavage (cleaved/pro caspase 3 ratio) was quantified by densitometry analysis. Quantitative data were normalized to “DMSO + BSA + PBS” control group and expressed as mean ± SEM, and three independent experiments were performed. *P < 0.05, **P < 0.01, vs. indicated groups.

4. Discussion

Progressive DN increases the risk of death of diabetic patients and is particularly intense in an age-dependent response [29], leading to a challenge of clinical DN management in elderly diabetic individuals. We recently demonstrated FGF1^ΔHBS can reverse established nonalcoholic fatty liver disease in a late-stage T2D model [14]. Here, we further show FGF1^ΔHBS delays the progression of established DN in late-stage db/db mice by inhibiting renal inflammation, fibrosis and apoptosis.

The most clinically relevant finding of this study is FGF1^ΔHBS markedly delays the progression of DN in late-stage of T2D. Protection from kidney damage by FGF1^ΔHBS was previously shown only in an early-stage db/db mice (2–5 months old) [15]. To highlight the clinical significance of FGF1^ΔHBS in DN, late-stage db/db mice (9 months old) with serious renal damage and dysfunction (Fig. 1) were used to evaluate the potential therapeutic effects of FGF1^ΔHBS on the progression of DN in elderly diabetic individuals. Our findings demonstrate for the first time that FGF1^ΔHBS can alleviate the development of DN in elderly diabetic individuals, as indicated by slowed elevation of UACR, reduced kidney size and weight, and promoted renal morphology and histopathology at 12 months of age (Fig. 2). More importantly, as a recently discovered non-mitogenic insulin sensitizer, FGF1^ΔHBS chronic treatment did not show obvious side effects of body weight gain or bone loss (Supplemental Fig. 1) commonly associated with the conventional insulin sensitizing therapies [22,23]. In clinical situations, delayed diagnosis of DN is a common challenge for management in diabetic patients. Therefore, the ability of FGF1^ΔHBS to delay the progression of renal dysfunction and reverse the development of renal morphological and histopathological damage without obvious side effects demonstrates a promising clinical translational value in therapy of DN, especially in elderly individuals with diabetes.

Conventionally, strict control of blood glucose is a key strategy for DN therapy and/or prevention [30]. Here, FGF1^ΔHBS normalized elevated blood glucose levels in db/db mice (Fig. 2), which can be explained as a direct mechanism for how FGF1^ΔHBS protects against the progression of DN. However, recent studies suggest the beneficial effects of both native FGF1 and FGF1^ΔHBS variant on DN are independent of their capacity to lower blood glucose. Our previous study demonstrated native FGF1 ameliorated DN in type 1 diabetic mice despite no measurable effect on the blood glucose levels [10]. Our collaborators reported that FGF1^ΔHBS effectively ameliorated chronic kidney diseases in an adriamycin-induced nephropathy mouse model without hyperglycemia [15]. More importantly, we and our collaborators demonstrated native FGF1 and/or FGF1^ΔHBS could directly protect renal mesangial cells and podocytes from glucotoxicity and hepatocytes from lipotoxicity, respectively, independent of changes in exposure to glucose [15,16]. Thus, it is evident that FGF1^ΔHBS has actions independent of lowering blood glucose in protecting from DN.

Inflammation is the critical mechanism for development of DN [31]. Many studies suggest individuals who developed diabetes present characteristics of inflammation several years before diagnosis [32]. Consistent with the anti-inflammatory effects of FGF1^ΔHBS on DN in early-stage db/db mice [15], we observed FGF1^ΔHBS administration effectively inhibited expression of molecular markers involved in inflammatory cell recruitment, migration, infiltration and adhesion in glomeruli and renal tubules of late-stage db/db mice, as indicated by decreased ICAM-1, VCAM-1, TNFα and CD68 (Figs. 3 and 4).

Growing evidence indicates gradual glomerular and tubular fibrosis towards renal failure in DN [33]. Moreover, controlling excessive inflammation has great therapeutic potential of inhibiting progressive renal fibrosis [34]. Corresponding to the anti-inflammatory effects of FGF1^ΔHBS, we demonstrated FGF1^ΔHBS treatment suppressed renal fibrosis by inhibiting several essential profibrotic molecules, indicated by decreased TGFβ, CTGF, PAI-1, fibronectin, collagen I and collagen IV (Figs. 5 and 6). In addition, chronic fibrosis is strongly associated with renal apoptosis, which is particularly intense in DN development [25,35]. The commonly observed tubular fibrosis in end-stage renal disease [36] is primarily caused by the consequent replacement of the extracellular matrix during the progressive loss of renal tubular cells [37] and the increased secretion of fibrotic cytokines by inducers of apoptosis [38]. Evidence indicates inhibition of tubular cell apoptosis protects against renal development of fibrosis [39,40]. As expected, FGF1^ΔHBS effectively suppressed the intrinsic pathway of renal apoptosis through upregulating Bcl-2 and downregulating Bax, along with inhibiting caspase 3 cleavage (activation), eventually attenuating renal tubular cell apoptosis (Fig. 7). These findings imply FGF1^ΔHBS suppresses tubular apoptosis, which might be its primary mechanism protecting against DN in late-stage T2D mice.

Our findings reveal PPARα is a key downstream target for FGF1^ΔHBS protection against tubule cell apoptosis in DN. As a primary sensor of lipid metabolism, PPARα is implicated in metabolic regulation of the kidney [28]. Recently, PPARα was recognized to have beneficial actions in preventing renal fibrosis, inflammation, and apoptosis that are over its effects on renal metabolic regulation [41–43]. In our study, FGF1^ΔHBS treatment prevented renal tubular apoptosis, which coincided with an upregulation of PPARα expression in db/db mice (Fig. 8A, B) and in Pal-exposed BUMPT cells (Fig. 8C–E), substantiating a direct protective effect of FGF1^ΔHBS on renal tubule cells. Intriguingly, the direct protective effect of FGF1^ΔHBS on BUMPT cells could be abolished by a selective PPARα inhibitor (Fig. 8C–E), suggesting the PPARα pathway might be a key mechanism by which FGF1^ΔHBS protects against apoptosis in renal tubule cells. Although the abundance of PPARα in the kidney is well established, its role in DN is emerging. The mechanisms by which FGF1^ΔHBS activates PPARα pathway to prevent against DN are worthy of future investigation.

In summary, our findings illustrate the non-mitogenic FGF1^ΔHBS delays the progression of DN in a late-stage T2D mouse model via effectively inhibiting renal inflammation, fibrosis, and apoptosis. In addition to the previous study of FGF1^ΔHBS in protecting against DN [15], our present study provides a promising therapeutic approach for clinical translation to manage the progressive DN in elderly individuals with diabetes.

Abbreviations:

DN

diabetic nephropathy

T2D

type 2 diabetes

FGF1

fibroblast growth factor 1

UACR

urine albumin-to-creatinine ratio

PBS

phosphate-buffered saline

H&E

hematoxylin and eosin

PAS

periodic acid-Schiff

BSA

bovine serum albumin

TBST

tris-buffered salinewith Tween 20

ICAM-1

intercellular adhesion molecule 1

VCAM-1

vascular cell adhesion molecule 1

TNFα

tumour necrosis factor alpha

CD68

cluster of differentiation 68

PCNA

proliferating cell nuclear antigen

IL1β

interleukin 1 beta

TGFβ

transforming growth factor beta

CTGF

connective tissue growth factor

PAI-1

plasminogen activator inhibitor-1

Bcl-2

B-cell lymphoma 2

Bax

Bcl-2-associated X

GAPDH

glyceraldehyde 3-phosphate dehydrogenase

TUNEL

terminal deoxynucleotidyl transferase (TdT) dUTP nick end labeling

DAPI

4^′,6-diamidino-2-phenylindole

PPARα

peroxisome proliferator-activated receptor α