Abstract
Chronic elevated glucose levels and activation of the renal renin-angiotensin system have been implicated in the pathogenesis of diabetic nephropathy. We tested the ability of lisofylline (LSF), a novel antiinflammatory compound, to prevent extracellular matrix (ECM) accumulation and growth factor production by human mesangial cells (HMCs) cultured in chronic elevated glucose (HG) or angiotensin II (AngII). HMCs were cultured in normal glucose (NG) (5.5 mm) and in HG (25 mm) for 7 d or with 10–7m AngII for 4 h with or without LSF. Levels of the ECM protein fibronectin and TGF-β in media were shown to increase in HG compared with NG. LSF decreased HG-induced fibronectin and TGF-β production to control levels. Increased expression of collagen type IV and laminin was observed in AngII-cultured HMCs. LSF protected HMCs from the AngII induction of these key matrix proteins. cAMP-responsive binding element phosphorylation was significantly higher in both HG and AngII-cultured HMCs. LSF reduced phosphorylation of both cAMP-responsive binding element and p38 MAPK compared with control. These data demonstrate that LSF protects HMCs from HG- and AngII-mediated ECM deposition by the reduction of matrix protein secretion possibly through regulation of TGF-β production and modulation of the p38 MAPK pathway. These results suggest that LSF may provide therapeutic benefit for prevention or treatment of diabetic nephropathy.
DIABETES MELLITUS IS the most common cause of end-stage renal failure, and the incidence of diabetic renal complications is increasing. Approximately 40% of patients with type 1 diabetes and 10% of patients with type 2 diabetes develop renal failure (1). Activation of the renin-angiotensin pathway and elevated glucose are important factors in the initiation and development of diabetes mellitus-related complications such as diabetic nephropathy (DN). The initial stages of DN are marked by an increased glomerular filtration rate and an accumulation of mesangial extracellular matrix (ECM) leading to glomerular hypertrophy. At more advanced stages of DN, overt proteinuria, tubulointerstitial fibrosis, hypertension, and reduction in creatinine clearance develop. Patients at this later stage rapidly progress to require dialysis or transplantation, as there are limited specific treatments for advanced renal disease.
Human mesangial cells (HMCs) provide a useful in vitro model for studying the early stages of DN as they produce and are receptive to a variety of growth factors and cytokines (2). Hyperglycemia and inflammatory cytokines have been shown to regulate mesangial cell function leading to increased glomerular injury (3). It is generally accepted that TGF-β is important in the development of kidney disease (4). Studies have shown that TGF-β is an important regulator of extracellular matrix production in mesangial cells (5, 6). TGF-β regulates ECM deposition in part through modulation of the Smad signaling pathway (7). R-Smads, receptor Smads 2 and 3, form a complex with TGF-β type II receptor, which then binds to a comediator Smad 4. This complex then enters the nucleus where it can directly promote target gene expression or combine with other transcription factors, i.e. SP-1 and AP-1, to target a wide variety of genes involved in fibrogenesis. Connective tissue growth factor (CTGF) and collagen types I and IV have been shown to be targets of the TGF-β/Smad signaling pathway (8, 9). CTGF is a profibrogenic cytokine that has been shown to promote fibronectin synthesis in cultured mesangial cells (10). As an autocrine growth factor, CTGF acts on the cells that produce it to proliferate and produce collagen. CTGF expression is promoted through both TGF-β and ras/MEK/ERK signaling pathways (8).
Lisofylline (LSF), 1-(5-R-hydroxyhexyl)-3,7-dimethylxanthine, is a novel antiinflammatory compound that has been shown to protect cultured rat pancreatic islets and pancreatic β-cells from cytokine damage (11, 12). LSF in vivo can also protect against autoimmune destruction of islets and reduce diabetes development in the nonobese diabetic mouse model (13). LSF has been used to reduce inflammation in lung injury studies (14). Of particular relevance to DN, LSF has been shown to reduce TGF-β release from bone marrow cells when given in vivo (15). In addition, new information from our studies in isolated pancreatic β-cells shows that LSF can maintain normal mitochondrial membrane potential and ATP levels and block apoptosis induced by inflammatory cytokines (11). LSF blocks lipid peroxide formation and reactive oxygen species (ROS), in addition to having antioxidant and free radical scavenging properties (16). Given the emerging information for the role of the mitochondria and ROS in leading to diabetes complications (17), this suggests that antiinflammatory agents that protect mitochondria and reduce downstream inflammatory cascades may provide a new way to prevent or retard renal damage due to diabetes. In this study, we examined the ability of LSF to protect mesangial cells from glucose- and angiotensin II (AngII)-mediated ECM production. We also examined the regulation of the TGF-β and p38 MAPK signaling pathways as possible targets of action of LSF.
Materials and Methods
Culture of normal HMCs
HMCs from normal tissue were obtained from Clonetics (CC-2559, San Diego, CA). HMCs were cultured in DMEM containing 1 g/liter glucose (Life Technologies, Inc., Carlsbad, CA) supplemented with 20% ν-Serum (BD Biosciences, Franklin Lakes, NJ). The cells were maintained at 37 C in a humidified incubator supplied with 5% carbon dioxide. HMCs were cultured to 70–80% confluency in 100-mm tissue culture-treated dishes. For high glucose (HG) studies, on d 1, the culture medium was changed and glucose added (adjusted to 4.5 g/liter) for HG dishes. After 2 d in HG culture, the medium was changed and cells were cultured in HG in the presence of LSF. Dose-response curves for LSF were performed in multiple systems, and 20 μm was found to be the lowest significantly effective concentration (data not shown). This concentration was also chosen as it corresponds to the approximate concentration of LSF in blood when administered by injection in clinical trials (18). LSF concentrations greater than 40 μm may have nonspecific effects and was therefore used at 20 μm for all studies. LSF was added on d 3 and every 24 h for 5 d for a total of 7 d in HG conditions. Alternatively, HMCs were cultured in normal glucose medium with addition of 10–7m AngII with or without 20 μm LSF for 4 h. To be relevant to DN, AngII was used at 10–7m to correlate with the maximal levels of AngII seen in human kidney (19). LSF was kindly provided by Dr. Jack Singer (Cell Therapeutics Inc., Seattle, WA).
Collection of HMC media and cell lysates
After a 7-d (HG) or 4-h (AngII) culture, media were collected and aliquoted. Cell culture dishes were placed on ice and washed with PBS. Subsequently, 300 μl New England Biolabs (Beverly, MA) lysis buffer containing phenylmethylsulfonylfluoride and protease inhibitor was added to each plate and allowed to incubate on ice for 5 min. Cells were collected, sonicated on ice, and centrifuged at 4 C for 5 min at 8000 rpm. The supernatant was collected for protein assay and Western blot analysis.
Analysis of fibronectin and TGF-β in culture media
HMC culture media samples were analyzed by ELISA method for fibronectin and TGF-β levels. For fibronectin, ELISA kit ECM300 (Chemicon, Temecula, CA) was used according to the manufacturer’s instructions. For TGF-β, acid-activated media samples were analyzed by ELISA kit DB100 (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions. ELISA plates were read and values calculated using a SpectraMax 190 plate reader.
PCR analysis of HMCs
Total RNA was isolated from HMCs using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Then 2 μg of total RNA was treated with DNAse I at room temperature for 10 min. The reaction was stopped by the addition of EDTA. DNAse-treated total RNA was reverse transcribed using iScript (Bio-Rad, Hercules, CA) according to the manufacturer’s protocol. For measurement of collagen type IV α-1, laminin β-1, or CTGF mRNA abundance, 2 μl of cDNA from each experimental group was used. Primer sequences used were as follows: collagen type IV α-1, sense 5′-gaagggtgatccaggtgaga-3′ antisense 5′-ctccctttttcccctttgtc-3′; laminin, sense 5′-atgtgcaggcataacaccaa-3′ antisense 5′-tccaggaattgttcccagag-3′; CTGF, sense 5′-caagggcctcttctgtgact-3′ antisense 5′-tggagattttgggagtacgg-3′; and β-actin, sense 5′-tgacggggtcacccacactgtgcccatcta-3′ antisense 5′-ctagaagcatttgcggtggacgatggaggg-3′. PCR conditions for collagen type IV, laminin, CTGF, and β-actin were as follows: 94 C for 2 min, followed by 30 cycles of 94 C for 30 sec, 60 C for 45 sec, and 72 C for 1 min, with a final extension time of 5 min at 72 C. Bands were analyzed on a 1.0% agarose gel in 1× Tris-acetate-EDTA. PCR for β-actin was performed as a control.
Western blot analysis of HMC cell lysates
For protein assay, total HMC cell lysates were analyzed using a Bio-Rad kit according to the manufacturer’s instructions. Protein values were read at 750 λ and compared with a standard curve on the SpectraMax 190 plate reader.
From each sample, 20 μg of protein was separated on a SDS NuPAGE 4–12% Bis-Tris gel (Invitrogen NP0321) according to manufacturer’s protocol. Gels were transferred onto nitrocellulose membrane (0.45 μm) (Invitrogen LC2001). After transfer, blots were blocked using Blocker Blotto (Pierce, Rockford, IL; 37530) and washed three times in Tris-buffered saline/Tween 20 (TBST). The primary antibody, phospho-cAMP response element binding protein (phospho-CREB) (Cell Signaling, Beverly, MA; 9191S) or phospho-p38 (R&D Systems AF869), was diluted 1:1000 in TBST with 5% BSA and added to the blot overnight at 4 C. Blots were washed three times in TBST and then incubated with the secondary antibody, antirabbit Ig-horseradish peroxidase (Amersham Biosciences, Piscataway, NJ; NA934V) or bovine antigoat IgG-horseradish peroxidase (Santa Cruz Biotechnology, Santa Cruz, CA; sc-2350) diluted 1:2000 in Blocker BLOTTO at room temperature. Blots were developed using enhanced chemiluminescence Western blotting detection reagents (Amersham Biosciences RPN2106).
Blots were stripped with Restore Western blot stripping buffer (Pierce, no. 21059) and reprobed for tubulin (Sigma, St. Louis, MO; T-9026) to ensure equal loading for all samples. Phospho-p38 values were normalized to total p38 (Cell Signaling 9212). Western blot values were obtained using an Eagle Eye imaging system (Stratagene, La Jolla, CA), and all values were normalized to tubulin.
Statistical analysis
Data for all experiments were analyzed by ANOVA and Fisher’s protected least-significant difference test using the Statview 6.0 software program. Data are represented as the mean ± se of three different experiments unless otherwise noted in the figure legends.
Results
Effect of LSF on glucose-mediated fibronectin and TGF-β secretion by HMCs
We examined the ability of LSF and analogs to reduce glucose-mediated fibronectin production by HMCs. Figure 1 shows that fibronectin levels were significantly higher in HG conditions [1.63 ± 0.6 μg/mg protein in normal glucose (NG) vs. 2.78 ± 0.7 in HG; P < 0.02]. LSF significantly reduced fibronectin levels in HG cultured cells (1.77 ± 0.4 μg/mg protein in HG + LSF; P < 0.03). LSF alone had no significant effect on fibronectin levels in NG (data not shown).
LSF reduces fibronectin secretion by HMCs cultured in HG. Medium was collected from HMCs grown in HG for 7 d with daily addition of 20 μm LSF as described in Materials and Methods. ECM protein fibronectin was measured by ELISA (ECM300, Chemicon). Samples were normalized to total cellular protein. Values are expressed as fibronectin/protein (μg/mg). #, Significantly higher than NG control by ANOVA (P < 0.02); *, significantly lower than HG culture by ANOVA (P < 0.03).
We found that HG culture increased TGF-β secretion by 2-fold compared with NG (257 ± 91 pg/mg protein in NG vs. 612 ± 101 in HG; P < 0.01) (Fig. 2). LSF partially reduced the increase of TGF-β in HG culture (454 ± 80 pg/mg protein in HG + LSF). These data suggest that TGF-β is one possible target of LSF action.
LSF reduces TGF-β levels in HG-cultured HMCs. Medium was collected from HMCs grown in HG for 7 d with daily addition of 20 μm LSF as described in Materials and Methods. TGF-β was measured by ELISA (DB100, R&D Systems). Samples were normalized to total cellular protein. Values are expressed as TGF-β/protein (picograms per milligram). #, Significantly higher than NG control by ANOVA (P < 0.01).
AngII-induced matrix protein gene expression is blocked by LSF
Increased ECM deposition by mesangial cells is a hallmark of early renal dysfunction. PCR analysis for the matrix proteins collagen type IV α-1 and laminin β-1 was performed on HMC cDNA. β-actin was also measured as a control. Gene expression was increased for both collagen type IV (1.5-fold) and laminin (2-fold) with AngII treatment compared with control (Fig. 3). LSF completely blocked the AngII-induced increase in the gene expression of these key matrix proteins. LSF had no effect on basal matrix protein expression.
LSF protects HMCs from AngII induction of matrix protein expression. RNA was harvested from HMCs cultured with 10–7m AngII with the addition of 20 μm LSF as described in Materials and Methods. PCR analysis was performed using primers for human collagen type IV (B), laminin (A), and β-actin. Values are expressed as matrix protein band normalized to β-actin. #, Significantly higher than control by ANOVA (P < 0.001); *, significantly lower than AngII culture by ANOVA (P < 0.001). Data are representative of three experiments (n = 3).
AngII-induced CTGF expression is blocked by LSF
We tested the ability of LSF to block the increased expression of CTGF in HMCs cultured with 10–7m AngII for 4 h. As shown in Fig. 4, PCR analysis of AngII-cultured HMCs showed significantly higher expression of CTGF. When cultured in the presence of both 10–7m AngII and 20 μm LSF, CTGF expression was lowered back to control levels (P < 0.0001).
LSF protects HMCs from AngII induction of CTGF expression. RNA was harvested from HMCs cultured with 10–7m AngII with the addition of 20 μm LSF as described in Materials and Methods. PCR analysis was performed using primers for human CTGF. Data are shown as CTGF band density/β-actin band density. *, Significantly lower than HG culture by ANOVA (P < 0.0001). Data are representative of three experiments (n = 3).
LSF modulates p38 and CREB phosphorylation
Phosphorylated (active) CREB is known to induce ECM production (20). Figure 5 shows an approximately 3-fold increase in CREB phosphorylation in HMCs cultured in glucose or 10–7m AngII. LSF significantly reduced glucose- and AngII-induced phosphorylation of CREB (8.9 ± 0.9 in LSF + HG vs. 28 ± 1.4 in HG; P < 0.0001). Phosphorylated p38 was significantly higher in HG conditions (8.69 ± 0.12 in NG vs. 12.41 ± 0.38 in HG; P < 0.0002) and was reduced to NG control levels by LSF (8.78 ± 0.46 in HG + LSF; P < 0.0002) as shown in Fig. 6. There were no statistical differences in p38 or CREB phosphorylation when LSF was added to NG control (data not shown).
LSF reduces CREB phosphorylation in HMCs cultured in HG or 10–7m AngII. Proteins were harvested from HMCs grown in HG for 7 d (A) or 10–7m AngII for 4 h with addition of 20 μm LSF (B) as described in Materials and Methods. Phosphorylation of CREB protein was measured by Western blot as described in Materials and Methods. Values are expressed as p-CREB density/tubulin density. #, Significantly higher than NG control by ANOVA (P < 0.0001); *, significantly lower than HG culture by ANOVA (P < 0.0001).
LSF reduces p38 phosphorylation in HMCs cultured in HG. Proteins were harvested from HMCs grown in HG for 7 d with addition of 20 μm LSF as described in Materials and Methods. Phosphorylated (active) p38 was measured by Western blot as described in Materials and Methods. Values are expressed as phospho-p38 density/tubulin density. #, Significantly higher than NG control by ANOVA (P < 0.0002); *, significantly lower than HG culture by ANOVA (P < 0.0002).
Discussion
DN remains a major complication of type I and type 2 diabetes. There are several cell types and multiple signaling pathways involved in DN including protein kinase C, activation of MAPK, advanced glycation end products-receptor for advanced glycation end products (AGE-RAGE), and increased oxidative stress and inflammation (21–23). Most investigators agree that an important aspect of the disease is an increase in TGF-β and fibronectin production and ECM deposition by HMCs (4, 24). In this study, we show for the first time that LSF reduces glucose-induced fibronectin levels completely to levels seen in normal glucose culture. Because LSF reduced TGF-β levels in HG culture, it suggests that reduction of this key growth factor is one mechanism of LSF action. Studies have shown that directly blocking TGF-β dramatically reduces glomerular fibrosis (8, 25, 26). TGF-β is a pluripotent growth factor having numerous functions in various cell types in the body. In addition to its role in wound healing and fibrosis, TGF-β is also involved in other biological activities such as suppression of autoimmune diseases and cancer suppression and metastasis (27–29). Because of its complex role, chronic systemic blockade of TGF-β may have adverse effects in vivo. LSF or related compounds may provide advantages as mediators of fibrogenesis as LSF reduces TGF-β levels in HG culture but does not block the release below control levels. CTGF has also been shown to be an important factor involved in TGF-β-related fibronectin production (8, 30–33). LSF was shown to block the AngII induction of CTGF expression in HMCs.
There is a significant amount of cross-talk between TGF-β and p38 MAPK signaling pathways (8, 34). p38 MAPK activation of CREB is involved in increased matrix expression by mesangial cells (35). Additionally, CREB plays a role in fibronectin transcription (36–38). We showed in this study that CREB phosphorylation was significantly increased in HG and AngII culture. p38 phosphorylation was also significantly increased in HG culture. LSF reduced phosphorylated p38 and CREB completely to control levels, possibly by acting as a phosphatase. Future studies are planned to address this potential action of LSF. These data suggest that LSF protects HMCs from glucose- and AngII-mediated ECM deposition at least in part by mediating p38 MAPK-associated signaling. Additional studies will be needed to further explore the upstream and downstream signaling mechanisms of LSF on p38 MAPK and TGF-β signaling. Furthermore, it is possible that LSF could be working by additional mechanisms such as reducing lipid peroxidation or preventing ROS release by mitochondria. The mitochondria have been implicated as the major source of how glucose induces complications (17). In other studies, we have shown LSF works to maintain normal mitochondrial function and membrane potential (11).
In summary, these results suggest that antiinflammatory agents such as LSF and analogs could provide therapeutic benefit to reduce glucose- and AngII-induced matrix production in DN by targeting TGF-β expression and downstream signaling.
Acknowledgments
This work was supported by a Center Grant from the Juvenile Diabetes Research Foundation and National Heart, Lung, and Blood Institute Grant 55798.
Abbreviations
- AngII
Angiotensin II
- CREB
cAMP response element binding protein
- CTGF
connective tissue growth factor
- DN
diabetic nephropathy
- HG
high glucose
- HMC
human mesangial cell
- LSF
lisofylline
- NG
normal glucose
- ROS
reactive oxygen species
- TBST
Tris-buffered saline with Tween 20
