Glycated albumin stimulates fibronectin gene expression in glomerular mesangial cells: Involvement of the transforming growth factor-β system

Kidney International, Vol. 53 (1998), pp. 631– 638

Glycated albumin stimulates fibronectin gene expression in glomerular mesangial cells: Involvement of the transforming growth factor-b system FUAD N. ZIYADEH, DONG CHEOL HAN, JONATHAN A. COHEN, JIA GUO, and MARGO P. COHEN Departments of Medicine and Biochemistry, University of Pennsylvania, and Exocell, Inc. Philadelphia, Pennsylvania, USA

Glycated albumin stimulates fibronectin gene expression in glomerular mesangial cells: Involvement of the transforming growth factor-b system. Albumin modified by Amadori glucose adducts, formed in increased amounts in diabetes, stimulates collagen IV production and gene expression in renal glomerular mesangial cells, and induces mesangial matrix accumulation accompanied by increased mRNA encoding a1 (IV) collagen and fibronectin in diabetic animals. These effects contribute to the pathogenesis of diabetic nephropathy, and resemble biologic activities of the cytokine TGF-b1, which also has been causally implicated in diabetic renal disease. We postulated that Amadori-modified glycated albumin modulates TGF-b1 expression in mesangial cells, and that TGF-b1 participates in mediating the glycated albumin-induced increases in mesangial cell matrix production. To test this hypothesis, we measured mRNA encoding TGF-b1, the TGF-b Type II receptor and fibronectin, a key matrix component of the TGF-b1 tissue response, after incubation of mesangial cells with glycated albumin. Steady state levels of the mRNAs encoding for these proteins were stimulated when mesangial cells were cultured in the presence of albumin containing Amadori glucose adducts compared with levels in cells cultured with the nonglycated, glucose-free counterpart. The glycated protein-induced changes in mRNA expression were observed with concentrations of glycated albumin encompassing those found in clinical specimens and in media containing physiologic (5.5 mM) glucose concentrations, indicating that they were due to the glucosemodified protein and not to a hyperglycemic milieu. Further, they were accompanied by increased translated fibronectin protein, which was prevented with TGF-b neutralizing antibody, as was the glycated albumininduced increase in fibronectin mRNA. The findings indicate that Amadori-modified glycated albumin stimulates mesangial cell TGF-b1 gene expression by mechanisms that are operative under normoglycemic conditions. These data provide the first link between elevated glycated serum albumin concentrations and increased TGF-b1 bioactivity in the pathogenesis of mesangial matrix accumulation in diabetes.

The Diabetes Control and Complications Trial (DCCT) demonstrated that improved glycemic control, evidenced by decreased mean glycated hemoglobin levels, lowered the risk for development of renal and other complications of diabetes [1]. These results conclusively established hyperglycemia as an important

Key words: glycated albumin, fibronectin, TGF-b, type II receptor, mesangial matrix in diabetes, diabetes mellitus. Received for publication July 7, 1997 and in revised form October 10, 1997 Accepted for publication October 10, 1997

© 1998 by the International Society of Nephrology

risk factor in diabetic nephropathy, although they did not define mechanisms by which elevated glucose concentrations in diabetes promote renal dysfunction. The recognition that hyperglycemia accelerates nonenzymatic glycation, and that glucose-derived modifications of proteins alter their structural and functional properties, have implicated increased glycation as a mechanistic link between hyperglycemia and the pathogenesis of microvascular disease in diabetes. Nonenzymatic glycation is a condensation reaction between glucose and reactive protein amino groups, yielding Schiff base intermediates that undergo Amadori rearrangement to form stable protein-glucose adducts [2–5]. In vivo, circulating glycated proteins principally exist as Amadori products and their concentration, driven by the ambient glucose concentration, is increased in diabetes as a result of hyperglycemia [2, 3, 6 – 8]. Recently, we have provided evidence strongly supporting a causal role for elevated concentrations of glycated albumin in the development of diabetic nephropathy, a disorder characterized pathologically by an expansion of the renal glomerular mesangial matrix that leads to an encroachment of cellular elements and a loss of filtration area [9 –12]. Serum containing concentrations of glycated albumin that are found in diabetic subjects stimulates renal glomerular mesangial cell production and gene expression of Type IV collagen, the predominant constituent of the expanded glomerular mesangial matrix that is seen in diabetes [13, 14]. Increased elaboration of Type IV collagen and increased mRNA encoding for a1 (IV) collagen accompanied by decreased cell replicative capacity and transcriptional activation of the a1 (IV) collagen gene can be induced by incubation of mesangial cells with purified glycated albumin in concentrations representing those associated with diabetes [15, 16]. When diabetic db/db mutant mice are treated with monoclonal antibodies (A717) that specifically neutralize excess plasma concentrations of Amadori-modified glycated albumin, renal cortical overexpression of mRNAs encoding a1 (IV) collagen and fibronectin is prevented, mesangial matrix expansion is reduced, and changes reflecting compromised renal function are improved [17–20]. These salutary effects are independent of blood glucose concentration, and are not observed in db/db mice treated with immunoglobulin that is not reactive with glycated albumin. The effects of glycated albumin on mesangial cell biology resemble those of transforming growth factor-beta (TGF-b), a

631

632

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells

multifunctional cytokine polypeptide with potent fibrogenic activity [21–23]. TGF-b1 bioactivation is associated with the modulation of cell growth and stimulation of collagen production by renal tubular and glomerular mesangial cells in culture, and these effects are prevented by TGF-b neutralizing antibodies [24 –26]. Several investigators have reported that TGF-b1 mRNA is elevated in kidneys of rodents with streptozotocin-induced insulinopenic diabetes [27–30]. Such findings, and reports that serum, urine and renal concentrations of the TGF-b protein are increased in rodents with insulin-deficient diabetes [27, 29 –31], have helped foster the hypothesis that TGF-b1 plays a causative role in diabetic renal disease. It has been postulated that matrix overproduction leading to renal fibrosis evolves from increased TGF-b1, initially induced as a reparative process in response to tissue injury, but later representing chronic failure to terminate tissue repair [32]. We and others have suggested that, in this scenario, the stimulus triggering increased TGF-b1 in diabetes may be elevated concentrations of glucose or angiotensin II, both of which have been reported to induce TGF-b1 mRNA expression in mesangial cells [24, 26, 33, 34]. However, the possibility that glycated albumin influences TGF-b1 expression, or that TGF-b1 participates in mediating the glycated protein-induced increases in mesangial cell matrix production, has not been examined. The goal of the present experiments was to determine whether glycated albumin modulates the TGF-b system in glomerular mesangial cells. To this end, we examined the effect of glycated albumin on steady state levels of mRNA encoding TGF-b1, the Type II TGF-b receptor, and fibronectin, a key matrix component of the TGF-b1 tissue response. We focused on the Type II receptor because it is the primary signaling receptor that binds the free ligand. The Type I receptor does not directly bind the free ligand, and it is best described as the transducer [21]. Our approach to these studies adhered to conditions in which the glycated protein was represented by the Amadori modified construct and not advanced glycation end products (AGE) [14, 35, 36]. Further, we examined effects in normal (5.5 mM) glucose concentration to ensure that observed responses were due to the glycated protein and not induced by elevated media glucose concentration. We report that exposure of mesangial cells in culture to glycated albumin in physiologic glucose concentration stimulates gene expression of TGF-b1, the TGF-b Type II receptor, and fibronectin, and the production of translated fibronectin protein. The glycated albumin-induced stimulation of fibronectin mRNA and protein were prevented in the presence of anti-TGF-b neutralizing antibody. METHODS Cell culture Murine mesangial cells in culture were derived from glomeruli isolated from SJL/J mice with a graded sieving technique and plated to explant [37, 38]. Mesangial cells were selected and subcultured according to previously described criteria [37], including morphology, ability to grow in medium containing D-valine instead of L-valine, and the presence of cytoplasmic filaments (desmin and vimentin), angiotensin II binding capacity, and contractile response to angiotensin II. Cells were stabilized by virus transformation with SV-40 and were carried in Dulbecco’s modified Eagle’s medium (DMEM) containing 5.5 mM glucose and 10% fetal calf serum (FCS). The suitability of transformed

mesangial cells and their parallelism with non-transformed cells with respect to growth and response to media manipulations have been documented [14, 25]. To initiate experiments under varying culture conditions, mesangial cells were plated into 25 cm2 plastic flasks, grown to confluence, rested for one day in serum-free medium containing 5.5 mM glucose and then grown for 48 hours in fresh media under experimental conditions described below. Experimental culture conditions The experimental conditions were initiated after the rest period upon the addition of fresh media containing 5.5 mM glucose, 1% FCS and purified glycated or nonglycated albumin (200 to 600 mg/ml), without or with TGF-b neutralizing antibody or irrelevant immunoglobulin (30 mg/ml). Results of preliminary experiments established that low concentrations of FCS maintained uniform cell growth, but did not obscure responses induced by glycated albumin or high glucose concentration. Glycated albumin was added in concentrations representing those found in clinical specimens, and encompassing those which previously have been found to stimulate mesangial cell a1 (IV) collagen production, mRNA expression and gene activation [8, 15, 16]. In nondiabetic individuals the glycated form comprises approximately 1.0% of serum albumin, and is equivalent to concentrations between 300 to 400 mg/ml of glycated albumin; the concentration of glycated albumin is in this range or is increased 11⁄2- to threefold in diabetic subjects, according to recent glycemic status [5, 8, 39]. Where indicated, the experimental conditions employed 5% nonglycated or glycated serum in place of the purified proteins, without or with A717 monoclonal antibody (see below). For study of elaboration of translated fibronectin, incubations under the various experimental conditions were conducted in serum free media to ensure that fibronectin measured in the post-incubation media represented that secreted by the cells. Glycated protein preparation Glycated albumin was prepared from albumin, purified from human plasma by chromatography on Affi gel Blue and DEAESepharose, obtaining a homogeneous band of approximately 66,000 molecular weight on SDS gel electrophoresis. The glycated protein was separated from the nonglycated protein by affinity chromatography on phenylboronate agarose (PBA), which binds Amadori adducts and not advanced glycation end products (AGE). This protocol has been shown to yield glycated albumin containing '1 mol glucose/mol albumin and reactive with the A717 monoclonal antibody that specifically recognizes albumin modified by Amadori glucose adducts and that does not react with AGE-albumin [15–17]. The epitope defined by the A717 monoclonal antibody encompasses an albumin domain containing a deoxyfructosyllysine residue and is structurally and immunochemically distinct from those in AGE. The PBA pass-through, used as source material for nonglycated albumin, contained , 0.05 mol glucose/mol albumin, migrated on SDS-PAGE as a single 66 kDa band, and did not react with the A717 monoclonal antibody, as has been described [17]. Glycated serum was prepared by incubating sterile-filtered normal human serum for four days at 25°C with 28 mM glucose in phosphate buffered saline (PBS), followed by dialysis against PBS to remove free glucose and again sterile-filtering. Serum incubated for the same period in PBS without glucose, dialyzed and sterile filtered, was used as control (normal serum; NS). The

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells

633

Fig. 1. Transforming growth factor-beta1 (TGF-b1) and fibronectin gene expression in mesangial cells cultured for 48 hours in media containing 5.5 mM glucose and 5% normal (NS) or glycated (GS) serum. Representative Northern blots from two experiments are shown.

glycated serum (GS) contained diabetic concentrations of glycated albumin, which were approximately twofold those in normal serum when measured by enzyme-linked immunoassay using the murine monoclonal antibody that specifically recognizes deoxyfructosyllysine epitopes in albumin modified by Amadori glucose adducts [8, 17, 39]. Albumin, the principal serum protein, contributes 80 to 90% of the total glycated serum protein concentration. Antibody preparation A717 was harvested from ascitic fluid of mice injected with the A717 cell line, purified by affinity chromatography on Protein G, and digested with papain to produce Fab fragments. Specificity and reactivity of the Fab fragments with the glycated albumin epitope was confirmed as described [18, 19]. Affinity purified murine IgG, unreactive with glycated albumin or other serum proteins, was used as control. The TGF-b antibody was a neutralizing mouse monoclonal antibody kindly provided by Dr. B. Fendly (Genentech, S. San Francisco, CA, USA). This antibody (2G7; IgG2b k) is panselective against all mammalian isoforms of TGF-b (b1, b2, b3). The concentration employed (30 mg/ml) neutralizes more than 80% of the TGF-b bioactivity determined by the mink lung epithelial bioassay [40]. This antibody also has been shown to prevent early features of diabetic renal disease in streptozotocin-diabetic mice [30]. An irrelevant isotype-matched murine immunoglobulin (HP6001; ATCC, Rockland, MD, USA) was used as control. RNA hybridization analysis RNA was extracted from mesangial cells harvested after culture under the described conditions and hybridized with 32P-labeled cDNA probes as previously described [14, 41]. Cells were denatured in 4 M guanidinium thiocyanate, 25 mM sodium citrate pH 7.0, 0.5% sarcosyl and 0.1 M 2-mercaptoethanol. Twenty to 30 mg of total RNA were electrophoresed on a 1.2% agarose gel, transferred onto nylon membranes, and ultraviolet wave crosslinked. Integrity and equal loading of samples were assessed by methylene blue staining of the transferred RNA. The membranes were hybridized with labeled cDNA inserts (32P-deoxycytidine 59-triphosphate; Amersham) using random priming. After hybridization, blots were washed for 30 minutes in 2 3 SSC with 0.1% SDS at 62°C, followed by two washes of 15 minutes in 0.1 3 SSC with 0.1% SDS at 62°C. The membranes were autoradiographed with intensifying screens at 270°C for up to 10 days. Blots were

then stripped for two hours at 62°C with 5 mM Tris, 0.2 mM EDTA (pH 8.0), and 5% sodium pyrophosphate, and subsequently rehybridized with cDNA fragment encoding an RNA standard to account for small variations in RNA loading and transfer. The loading standards employed were GAPDH mRNA and mRNA encoding for mouse ribosomal protein L32 (mrpL32) [42]. Exposed films were scanned with a laser densitometer, and mRNA levels were calculated relative to those of the loading standard. The cDNA probes encoding murine TGF-b1 and the TGF-b Type II receptor were first synthesized by the polymerase chain reaction (PCR) using murine kidney cDNA as template and specific oligonucleotide primers based on the published cDNA sequences [43– 45]. The PCR products were cloned into the pCRII TA cloning system (Invitrogen, LaJolla, CA, USA). Nucleotide sequencing of the PCR products confirmed identity of the probes. The cDNA probe for fibronectin was a 649 nucleotide AvaI/HindIII fragment that codes for most of the human fibronectin 39 untranslated regions [46], specifically hybridizes with 8.8 kb fibronectin mRNA, and recognizes mouse and rat fibronectin. Analysis of translated proteins Fibronectin in the media collected at the end of the experimental period was measured by an ELISA (Exocell, Philadelphia, PA, USA), in which fibronectin in standard or sample was bound to gelatin that had been immobilized onto microtiter wells, and was detected with HRP-conjugated monoclonal antibody against fibronectin [47]. RESULTS Initial experiments examined the ability of glycated serum to stimulate mesangial cell gene expression of TGF-b1 and fibronectin, using cultures of mesangial cells made quiescent by serum deprivation for one day and then treated for 48 hours with media made 5% in NS or GS. The media glucose concentration in these studies was 5.5 mM. On Northern blot analysis, cells grown in media supplemented with GS expressed higher levels of TGF-b1 and fibronectin mRNA than did cells grown in media supplemented with NS (Fig. 1). Densitometric scanning of the exposed films with normalization in relation to the GAPDH loading standard confirmed that steady state levels of TGF-b1 and fibronectin mRNA were increased in cells exposed to GS compared with levels in cells cultured with NS; the TGF-b1:GAPDH mRNA relative ratio was 1.34 and the fibronectin:GAPDH relative ratio

634

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells

Fig. 2. Fibronectin gene expression in mesangial cells cultured for 48 hours in media containing 5.5 mM glucose and 5% normal (NS, lane 1) or glycated serum (GS, lanes 2 and 3). Cells were cultured in the presence of murine immunoglobulin unreactive with glycated albumin or other serum proteins (lanes 1 and 2) or with A717 anti-glycated albumin monoclonal antibody (lane 3). Representative Northern blots from two experiments are shown.

was 2.77, where the ratios in cells cultured with NS were assigned arbitrary values of 1.0. Blocking experiments with the A717 monoclonal antibody indicated that the increased fibronectin mRNA induced by glycated serum largely derived from glycated albumin, which represents '80% of glycated serum proteins. The steady state level of mRNA encoding fibronectin in cells cultured with glycated serum in the presence of A717 was significantly lower than that in cells cultured with GS in the absence of A717 (Fig. 2), giving a relative ratio compared with NS of 1.2. A717 did not affect fibronectin mRNA in cells incubated with normal serum (not shown). Corroboration that glycated albumin in the glycated serum stimulated the TGF-b system was obtained by supplementing cell cultures with purified glycated versus nonglycated albumin (200 to 600 mg/ml) in media containing 1% FCS and physiologic (5.5 mM) glucose concentration, and examining mesangial cell levels of mRNA encoding for fibronectin, TGF-b1, and the TGF-b Type II receptor. Northern blot analysis revealed that the steady state levels of mRNA encoding for these three proteins were increased after incubation with glycated albumin compared with levels in control incubations supplemented with nonglycated albumin (Fig. 3). With 200 mg/ml glycated albumin, the respective relative ratios of mRNA to that of the loading standard were 2.13 for fibronectin, 2.23 for TGF-b1, and 1.83 for TGF-b Type II receptor. Further increases in the relative ratios were observed with higher concentrations (600 mg/ml) of glycated albumin (Table 1). The glycated albumin-induced increases in mRNAs encoding for these proteins were comparable to those observed when mesangial cells were cultured for 48 hours in the presence of a hyperglycemic milieu. Figure 4 depicts representative mRNA hybridization assays for fibronectin, TGF-b1 and TGF-b Type II receptor in cells incubated in media containing 1% FCS, 600 mg/ml nonglycated or glycated albumin, and 5.5 mM versus 25 mM glucose. In cells supplemented with nonglycated albumin, relative ratios (mRNA:loading standard) in cells cultured in 25 mM glucose compared with those in cells cultured in 5.5 mM glucose were 3.38 for fibronectin, 2.77 for TGF-b1 and 2.38 for TGF-b Type II receptor. Twenty-five mM glucose also augmented the

Fig. 3. RNA hybridization analysis of mRNA encoding fibronectin, transforming growth factor beta1 (TGF-b1), and TGF-b Type II receptor (TbIIR) in mesangial cells cultured for 48 hours with nonglycated (NA) or glycated (GA) albumin in 5.5 mM glucose and 1% FCS. Representative Northern blots from three experiments are depicted. Table 1. Glycated albumin effects on mesangial cell mRNA Supplement

Fibronectin

TGF-b1

TGF-b Type II receptor

Nonglycated albumin 600 mg/ml Glycated albumin 200 mg/ml Glycated albumin 600 mg/ml

1.0

1.0

1.0

2.13

2.23

1.83

2.28

2.60

2.34

Mesangial cells were cultured for 48 hours in DMEM containing 5.5 mM glucose, 1% FCS, and the indicated concentrations of nonglycated or glycated albumin. Data were obtained by densitometric scanning of autoradiographs of the hybridization assays performed with the corresponding cDNAs and with the mrpL32 cDNA probe. Values represent relative ratios compared with those in cells incubated with nonglycated albumin, assigned an arbitrary value of 1.0. Abbreviation TGF-b1 is transforming growth factor-beta1.

glycated albumin-induced increases in TGF-b1 and fibronectin steady-state mRNA levels relative to those observed in 5.5 mM glucose concentration, giving an additive effect on gene expression of TGF-b1 but not fibronectin (Table 2). The stimulated fibronectin gene expression induced by glycated albumin under normoglycemic conditions was accompanied by increases in the translated protein, measured by ELISA (Table 3).

635

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells

Table 3. Effect of TGF-b antibody on fibronectin production by glomerular mesangial cells Supplement

Antibody

Fibronectin ng/ml

None Nonglycated albumin Glycated albumin Glycated albumin Glycated albumin

— — — MIg Anti-TGF-b

514 6 106 512 6 112 2330 6 121a 2024 6 125a 1382 6 94ab

Results represent mean 6 SEM of four independent cultures in each of which media were collected after incubation for 48 hours with 600 mg/ml of nonglycated versus glycated albumin and 30 mg/ml of isotype-matched irrevelant murine immunoglobulin (MIg) or neutralizing TGF-b murine monoclonal antibody. a P , 0.05 compared with nonglycated albumin. b P , 0.05 compared with no antibody or with MIg.

Fig. 4. Effect of high glucose concentration on mesangial cell gene expression of fibronectin, transforming growth factor-b1 (TGF-b1), and TGF-b Type II receptor (TbIIR). Mesangial cells were cultured for 48 hours in media containing 1% FCS, 600 mg/ml nonglycated (NA) or glycated (GA) albumin, and 5.5 versus 25 mM glucose. Representative Northern blots from three experiments are depicted. Table 2. Glucose effects on mesangial cell mRNA Supplement

Glucose

Fibronectin

TGF-b1

TGF-b Type II receptor

NA NA GA GA

5.5 mM 25 mM 5.5 mM 25 mM

1.0 3.38 2.28 3.41

1.0 2.77 2.60 3.63

1.0 2.38 2.34 2.32

Mesangial cells were cultured for 48 hours in DMEM containing 5.5 or 25 mM glucose, 1% FCS, and 600 mg/ml of nonglycated (NA) or glycated (GA) albumin. Data are expressed as described in the legend to Table 1.

Fibronectin concentrations in cultures supplemented with nonglycated albumin did not differ significantly from those in media from cells cultured in unsupplemented media. However, fibronectin concentrations in media from cells cultured with glycated albumin were significantly increased compared with unsupplemented and nonglycated albumin-supplemented incubations. Inclusion of TGF-b antibody in culture media largely prevented the glycated albumin-induced increase in fibronectin, indicating that TGF-b participates in mediating this response. Confirmation that TGF-b participated in the glycated albumininduced stimulation of fibronectin gene expression under normoglycemic conditions was sought by Northern blot analysis after incubation of cells with glycated albumin in the presence of TGF-b neutralizing antibody. Figure 5 shows that anti-TGF-b prevented increased fibronectin mRNA in cells cultured with glycated albumin. On densitometric scanning, the relative fi-

Fig. 5. Effect of neutralizing transforming growth factor-b (TGF-b) antibody on mesangial cell gene expression of fibronectin. Cells cultured for 48 hours in media containing 1% FCS, 5.5 mM glucose, 400 mg/ml nonglycated (NA) or glycated (GA) albumin with anti-TGF-b (aT). Northern blot analysis of mRNA from cells incubated with high (25 mM) glucose media is shown for comparative purposes.

bronectin:GAPDH ratio compared with that for nonglycated albumin was 1.60 for glycated albumin and 0.91 for glycated albumin plus anti-TGF-b antibody. The relative ratio in cells incubated with 25 mM glucose, shown in Figure 5 and evaluated for comparative purposes, was 2.27. An isotype-matched immunoglobulin did not reduce fibronectin mRNA in cells cultured with glycated albumin (not shown). DISCUSSION The results presented in these experiments demonstrate that glycated albumin induces coordinate increases in mRNA encoding for TGF-b1, the TGF-b Type II receptor and fibronectin in renal glomerular mesangial cells. These findings provide evidence linking increased TGF-b1 gene expression to the glycated albumin-induced increases in extracellular matrix protein and mRNA synthesis that have been observed in glomerular mesangial cells in

636

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells

vitro and in renal cortex of diabetic animals in vivo [9, 14 –16, 18, 19]. High media glucose concentration has also been shown to stimulate gene expression of a1 (IV) collagen, fibronectin, and TGF-b1 in cultured mesangial cells [26, 48 –50]. This stimulation is accompanied by activation of protein kinase C, transient elevation of mRNA levels of c-fos and c-jun oncogenes and corresponding increases in c-fos and c-jun proteins, and delayed growth inhibition in association with bioactivation of TGF-b1 [25, 51–53]. Additionally, high glucose-induced increases in TGF-b are believed to involve protein kinase C signaling [54, 55]. Increased TGF-b1 therefore appears to be an important mediator of hyperglycemia with respect to the influence of elevated glucose concentrations on the elaboration and expression of mRNAs encoding for matrix macromolecules. However, in the present experiments increased levels of mRNA encoding for TGF-b1 and fibronectin in response to glycated albumin were observed in low physiologic glucose concentration (5.5 mM), and cannot be ascribed to an effect of high glucose in the culture media. The findings reported herein are the first to demonstrate that purified albumin modified by Amadori glucose adducts modulates TGF-b1 expression in mesangial cells, and that TGF-b1 participates in mediating glycated albumin-induced changes in mesangial matrix production. These observations afford new insight into the increased TGF-b1 gene expression and TGF-b1 protein content that have been observed in kidneys of diabetic animals, and suggest a dual-pronged mechanism linking glucose, glycated albumin, and TGF-b1 to the development of glomerular matrix accumulation in diabetes [27, 28, 32]. Increased glomerular TGFb1, which can be induced by either elevated glucose or glycated albumin, stimulates the production and gene expression of the extracellular matrix macromolecules fibronectin and Type IV collagen. Hyperglycemia is the driving force for increased albumin glycation, but the effects of glycated albumin do not require a hyperglycemic milieu to be operative. Extrapolated to the in vivo situation, this means that the influence of glycated albumin, which has a circulating half life of about two weeks, on the TGF-b system can continue after restoration of normoglycemia and can act independently of hyperglycemia. The observation that glycated albumin increased gene expression of the primary signal-transducing TGF-b Type II receptor deserves comment. Although up-regulation of the receptor could be a consequence of a decrease in the concentration of the TGF-b1 ligand [56], TGF-b1 expression actually was stimulated by glycated albumin. Since TGF-b1 participates in a positive feedback loop that auto-stimulates TGF-b1 expression [57], a decrease in translated TGF-b1 protein would be unlikely. Indeed, the decreased fibronectin production in response to the neutralizing anti-TGF-b antibody indicates that TGF-b protein bioactivity was increased in conditioned media when cells were incubated with glycated albumin. Moreover, experiments with cultured mesangial cells and in streptozotocin-diabetic rodents suggest that hyperglycemia can stimulate TGF-b Type II receptor expression independently of ligand concentration [30, 58, 59]. Elevated media glucose concentration increases Type II receptor mRNA and protein, as assessed by Western blot and by 125I-TGF-b1 binding studies [58, 59]. The results of our studies indicate that glycated albumin may also exert a direct influence on TGF-b Type II receptor expression. The demonstration of a relationship between glycated albumin and the TGF-b system is in keeping with the previously reported

importance of these factors in the pathogenesis of diabetic renal disease [18, 19, 30]. The present observations extend the recent in vivo studies in which we demonstrated that treatment directed against glycated albumin or TGF-b has a salutary influence on renal pathobiology in diabetes. Chronic treatment of db/db mice with anti-glycated albumin monoclonal antibodies ameliorates the cell biology, histomorphometric and functional changes of diabetic nephropathy in this model [18 –20]. Short-term treatment of streptozotocin-diabetic mice with neutralizing antibodies directed against TGF-b reduces kidney weight and glomerular hypertrophy, and attenuates the increase in extracellular matrix mRNA levels [30]. It will be interesting to examine whether glucose and glycated albumin activate TGF-b1 expression in mesangial cells through a common mechanism, such as protein kinase C signaling [54, 55, 60]. Although the mechanism by which glycated albumin stimulates TGF-b1 gene expression is speculative, it is possible that increased electronegativity of albumin from glycation of lysine amino groups enhances TGF-b1 retention in the pericellular milieu and/or confers proteoglycan-like properties that mimic Type III betaglycan receptor activity, which serves to anchor TGF-b1 in the pericellular environment and present it to other receptors [21, 22, 61– 64]. It is also possible that occupancy or modulation of low affinity TGF-b receptors by glycated albumin impedes termination of TGF-b1 receptor-mediated events. Glycated albumin may additionally stimulate matrix production through mechanisms unrelated to TGF-b1, as has been suggested by the identification of a cell associated ligand-receptor system for the Amadori-modified protein [65– 67]. This possibility is supported by the finding that anti-TGFb antibody significantly reduced but did not completely correct the glycated albumininduced increase in fibronectin production (Table 3). In summary, we have shown that albumin modified by Amadori glucose adducts stimulates the gene expression of TGF-b1, the TGF-b Type II receptor and fibronectin by murine mesangial cells in culture. These effects are observed in physiologic glucose concentration. The findings link increased glycated albumin in diabetes to increased TGF-b1 bioactivity in the pathogenesis of glomerular mesangial matrix accumulation. ACKNOWLEDGMENTS This research was supported in part by grants DK-44513, DK-45191, DK-38308 and DK-49455 from the National Institutes of Health. Dr. Han is a visiting scholar at the University of Pennsylvania and is supported by the Hyonam Kidney Laboratory, Soon Chun Hyang University, Seoul, Korea. Reprint requests to Dr. Margo P. Cohen, 3508 Market Street, Suite 420, Philadelphia, Pennsylvania 19104, USA. E-mail: ziyadeh @ mail.med.upenn.edu

REFERENCES 1. THE DCCT RESEARCH GROUP: The effect of intensive treatment of diabetes on the development and progression of long-term complications of insulin dependent diabetes mellitus. N Engl J Med 329:977– 986, 1993 2. BUNN HF, HANEY DM, KAMIN S, GABBAY KH, GALLOP PM: The biosynthesis of human hemoglobin A1c: Slow glycosylation in vivo. J Clin Invest 57:1652–1659, 1976 3. BUNN HF, GABBAY KH, GALLOP PM: The glycosylation of hemoglobin: Relevance to diabetes mellitus. Science 200:21–27, 1978 4. HIGGINS PJ, BUNN HF: Kinetic analysis of the nonenzymatic glycosylation of hemoglobin. J Biol Chem 256:5204 –5208, 1981

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells 5. COHEN MP: Diabetes and Protein Glycation. Clinical and Pathophysiologic Relevance. Philadelphia, JC Press, 1996 6. DAY JF, INGELBRETSEN CG, INGELBRETSEN WR, BAYNES JW, THORPE SR: Nonenzymatic glycosylation of serum proteins and hemoglobin: Response to changes in blood glucose levels in diabetic rats. Diabetes 29:524 –527, 1980 7. GUTHROW CE, MORRIS MA, DAY JF, THORPE SR, BAYNES J: Enhanced nonenzymatic glucosylation of serum albumin in diabetes mellitus. Proc Natl Acad Sci USA 76:4258 – 4261, 1979 8. COHEN MP, HUD E: Measurement of plasma glycoalbumin levels with a monoclonal antibody based ELISA. J Immunol Meth 122:279 –283, 1989 9. COHEN MP, ZIYADEH FN: Role of Amadori-modified nonenzymatically glycated serum proteins in the pathogenesis of diabetic nephropathy. J Am Soc Nephrol 7:183–190, 1996 10. STEFFES MW, BILOUS RW, SUTHERLAND DER, MAUER SM: Mesangial expansion as a central mechanism for loss of kidney function in diabetic patients. Diabetes 38:1077–1081, 1989 11. SAITO Y, KIDA H, TAKEDA SI, YOSHIMURA M, YOKOYAMA H, KOSHINO Y, HATTORI N: Mesangiolysis in diabetic glomeruli: Its role in the formation of nodular lesions. Kidney Int 34:389 –396, 1988 12. STEFFES MW, BILOUS W, SUTHERLAND DET, MAUER SM: Cell and matrix components of the glomerular mesangium in type I diabetes. Diabetes 41:679 – 684, 1992 13. ABRASS CK, PETERSON CV, RAUGI GS: Phenotypic expression of collagen types in mesangial matrix of diabetic and nondiabetic rats. Diabetes 37:1695–1702, 1988 14. COHEN MP, ZIYADEH FN: Amadori glucose adducts modulate mesangial cell growth and collagen gene expression. Kidney Int 45:475– 484, 1994 15. ZIYADEH FN, COHEN MP: Effects of glycated albumin on mesangial cells: Evidence for a role in diabetic nephropathy. Molec Cell Biochem 125:19 –25, 1993 16. COHEN MP, HUD E, WU VY, ZIYADEH FN: Albumin modified by Amadori glucose adducts activates mesangial cell Type IV collagen gene transcription. Molec Cell Biochem 151:61– 67, 1995 17. COHEN MP, HUD E: Production and characterization of monoclonal antibodies against human glycoalbumin. J Immunol Meth 117:121–129, 1989 18. COHEN MP, HUD E, WU VY: Amelioration of diabetic nephropathy with monoclonal antibodies against glycated albumin. Kidney Int 45:1673–1679, 1994 19. COHEN MP, SHARMA K, JIN Y, HUD E, WU VY, TOMASZEWSKI J, ZIYADEH FN: Prevention of diabetic nephropathy in db/db mice with glycated albumin antagonists: A novel treatment strategy. J Clin Invest 95:2338 –2345, 1995 20. COHEN MP, CLEMENTS RS, COHEN JA, SHEARMAN CW: Prevention of decline in renal function in the diabetic db/db mouse. Diabetologia 39:270 –274, 1996 21. MASSAGUE J: Receptors for the TGF-b family. Cell 69:1067–1070, 1992 22. SHARMA K, ZIYADEH FN: Biochemical events and cytokine interactions linking glucose metablism to the develpment of diabetic nephropathy. Semin Nephrol 17:80 –92, 1997 23. SHARMA K, ZIYADEH FN: Hyperglycemia and diabetic kidney disease: The case for Transforming Growth Factor-b as a key mediator. Diabetes 44:1139 –1146, 1995 24. ROCCO MC, CHEN Y, GOLDFARB FN: Elevated glucose stimulates TGF-b gene expression and bioactivity in proximal tubule. Kidney Int 41:107–114, 1992 25. WOLF G, SHARMA K, CHEN Y, ERICKSEN M, ZIYADEH FN: High glucose induced proliferation in mesangial cells is reversed by autocrine TGF-b. Kidney Int 42:647– 656, 1992 26. ZIYADEH FN, SHARMA K, ERICKSEN M, WOLF G: Stimulation of collagen gene expression and protein synthesis in murine mesangial cells by high glucose is mediated by autocrine activity of transforming growth factor-b. J Clin Invest 93:536 –542, 1994 27. YAMAMOTO T, NAKAMURA T, NOBLE N, ROUSLAHTI E, BORDER WA: Expression of transforming growth factor b is elevated in human and experimental diabetic nephropathy. Proc Natl Acad Sci USA 90:1814 – 1818, 1993 28. NAKAMURA TG, FUKUI M, EBIHARA I, OSADA S, NAGAOKA I, TOMINO

29. 30.

31.

32. 33.

34.

35. 36.

37. 38. 39. 40.

41.

42. 43. 44. 45. 46. 47. 48. 49.

50.

637

Y, KOIDE H: mRNA expression of growth factors in glomeruli from diabetic rats. Diabetes 42:450 – 456, 1993 BOLLENINI JS, REDDI AS: Transforming growth factor-b1 enhances glomerular collagen synthesis in diabetic rats. Diabetes 42:1673–1677, 1993 SHARMA K, JIN Y, GUO J, ZIYADEH FN: Neutralization of TGF-b by anti-TGF-b antibody attenuates kidney hypertrophy and the enhanced extracellular matrix gene expression in STZ-induced diabetic mice. Diabetes 45:522–530, 1996 YANG C-W, HATTORI M, VLASSARA H, HE C-J, CAROME MA, YAMAMOTO E, ELLIOT S, STRIKER G, STRIKER LJ: Overexpression of transforming growth factor-b1 mRNA is associated with up-regulation of glomerular tenascin and laminin gene expression in non-obese diabetic mice. J Am Soc Nephrol 5:1610 –1617, 1995 BORDER WA, NOBLE NA: Transforming growth factor b in tissue fibrosis. N Engl J Med 331:1286 –1292, 1994 KAGAMI S, BORDER WA, MILLER DE, NOBLE NA: Angiotensin II stimulates extracellular matrix protein synthesis through induction of transforming growth factor b expression in rat glomerular mesangial cells. J Clin Invest 93:2431–2437, 1994 ANDERSON PW, ZHANG XY, TIAN J, CORREALE JD, XI XP, YANG D, GRAF K, LAW RE, HSEUH WA: Insulin and angiotensin II are additive in stimulating TGF-b1 and matrix mRNAs in mesangial cells. Kidney Int 50:745–573, 1996 SABBATINI M, SANSONE G, UCCELLO F, GILIBERTI A, CONTE G, ANDREUCCI VE: Early glycosilation products induce glomerular hyperfiltration in rats. Kidney Int 42:875– 881, 1992 ESPOSITO C, GERLACH H, BRENT J, STERN D, VLASSARA H: Endothelial receptor-mediated binding of glucose-modified albumin is associated with increased monolayer permeability and modulation of cellsurface coagulant properties. J Exp Med 170:1387–1407, 1989 KREISBERG JI, KARNOVSKY MJ: Glomerular cells in culture. Kidney Int 23:439 – 447, 1983 WOLF G, HABERSTROH U, NEILSON EG: Angiotensin II stimulates proliferation and biosynthesis of Type I collagen in cultured murine mesangial cells. Am J Pathol 140:95–107, 1992 HUD E, COHEN MP: Evaluation and performance characteristics of a novel ELISA using a monoclonal antibody to glycated albumin. Clin Chim Acta 185:157–164, 1989 ABE M, HARPEL JG, METZ DN, NUNES I, LOSKUTOFF DJ, RIFKIN DB: An assay for transforming growth factor-beta using cells transfected with a plasminogen activator- inhibitor-1 luciferase construct. Anal Biochem 216:276 –284, 1994 ZIYADEH FN, SNIPES E, WATANABE M, ALVAREZ RS, GOLDFARB S, HAVERTY TP: High glucose induces cell hypertrophy and stimulates collagen gene expression in proximal tubule. Am J Physiol 259:F704 – F714, 1990 MEYUHAS O, KLEIN A: The mouse ribosomal protein L7 gene. Its primary structure and functional analysis of the promoter region. J Biol Chem 265:11465–11473, 1990 LIN HY, WANG XF, NG-EATON E, WEINBERG RA, LODISH HF: Expression and cloning of the TGF-b Type II receptor, a functional transmembrane serine/threonine kinase. Cell 68:775–785, 1991 SHARMA K, ZIYADEH FN: Renal hypertrophy is associated with upregulation of TGF-b1 gene expression in diabetic BB rat and NOD mouse. Am J Physiol 265:F1094 –F1101, 1994 ROCCO M, NEILSON EG, HOYER JR, ZIYADEH FN: Attenuated expression of epithelial cell adhesion molecules in murine polycystic kidney disease. Am J Physiol 262:F679 –F686, 1991 MEYERS JC: Differential expression of Type I collagen and cellular fibronectin. Kidney Int 43:45–52, 1993 COHEN MP, CLEMENTS RS, COHEN JA, SHEARMAN CW: Glycated albumin promotes a generalized vasculopathy in the db/db mouse. Biochem Biophys Res Comm 218:72–75, 1996 AYO SH, RADNIK GA, GARONI JA, GLASS WF, KREISBERG JI: High glucose causes an increase in extracellular matrix proteins in cultured mesangial cells. Am J Pathol 136:1339 –1348, 1990 AYO SH, RADNIK GA, GLASS WF, GARONI JA, RAMPT E-R, APPLING DR, KREISBERG JI: Increased extracellular matrix and mRNA in mesangial cells grown in high glucose. Am J Physiol 260:F185–F191, 1991 HANEDA M, KIKKAWA R, HORIDE N, TOGAWA M, KOYA D, KAJIWARA N, OOSHIMA A, SHIGETA Y: Glucose enhances Type IV collagen

638

51. 52. 53. 54. 55. 56. 57.

58.

Ziyadeh et al: Glycated albumin and TGF-b in mesangial cells

production in cultured rat glomerular mesangial cells. Diabetologia 34:198 –200, 1991 CRAVEN PA, DERUBERTIS FR: Protein kinase C is activated in glomeruli from streptozotocin rats. Possible mediation by glucose. J Clin Invest 83:1667–1675, 1989 AYO SH, RADNIK RA, GARONI J, TROYER DA, KREISBERG JI: High glucose increases diacylglycerol mass and activates protein kinase C in mesangial cell cultures. Am J Physiol 261:F571–F577, 1991 KREISBERG JI, RADNIK RA, AYO SH, GARONI J, SAIKUMAR P: High glucosed elevates c-fos and c-jun transcripts and proteins in mesangial cell cultures. Kidney Int 46:105–112, 1994 FUMO R, KUNCIO GS, ZIYADEH FN: PKC and high glucose stimulate collagen a1 (IV) transcriptional activity in a reporter mesangial cell line. Am J Physiol 36:F632–F638, 1994 CRAVEN PA, STUDER RK, FELDER J, PHILLIPS S, DERUBERTIS FR: Nitric oxide inhibition of transforming growth factor-b and collagen in mesangial cells. Diabetes 46:671– 681, 1997 CHOI ME, KIM E-G, HUANG O, BALLERMAN BJ: Rat mesangial cell hypertrophy in response to transforming growth factor-b1. Kidney Int 44:948 –958, 1993 KIM S-J, DENHEZ F, KIM K-Y, HOLT JT, SPORN MB, ROBERTS AB: Activation of the second promoter of the transforming growth factor-b gene by transforming growth factor b1 and phorbol esters occurs through the same target sequences. J Biol Chem 264:19373–19378, 1989 MOGYOROSI A, HOFFMAN BB, GUO J, JIN Y, ERICKSEN M, SHARMA K, ZIYADEH FN: Elevated glucose concentration stimulates expression of the Type II TGF-b receptor in glomerular mesangial cells. (abstract) J Am Soc Nephrol 7:1875A, 1996

59. LADSON-WOFFORD S, RISER BL, CORTES P: High extracellular glucose concentrations increase receptors for transforming growth factor-b in rat mesangial cells in culture. (abstract) J Am Soc Nephrol 5:696A, 1994 60. DERUBERTIS FR, CRAVEN PA: Activation of protein kinase C in glomerular cells in diabetes. Mechanisms and potential links to the pathogenesis of diabetic glomerulopathy. Diabetes 43:1– 8, 1994 61. CHEN RH, EBNER R, DERNYCK R: Inactivation of the Type II receptor reveals two receptor pathways for the diverse TGF-b activities. Science 260:1335–1338, 1993 62. LOPEZ-CASILLAS F, CHEIFITZ J, DOODY J, ANDRES L, LANE WAS, MASSAGUE J: Structure and expression of the membrane proteoglycan betaglycan, a component of the TGF-b receptor system. Cell 67:785– 795, 1991 63. LOPEZ-CASILLAS F, WRANA JL, MASSAGUE J: Betaglycan presents ligand to the TGF-b signalling receptor. Cell 73:1435–1444, 1993 64. LOPEZ-CASILLAS F, PAYNE HM, ANDRES JL, MASSAGUE J: Betaglycan can act as a dual modulator of TGF-b access to signalling receptors: Mapping of ligand binding and GAG attachment sites. J Cell Biol 124:557–568, 1994 65. WU VY, COHEN MP: Identification of aortic endothelial cell binding proteins for Amadori adducts in glycated albumin. Biochem Biophys Res Comm 193:1131–1136, 1993 66. WU VY, COHEN MP: Receptors specific for Amadori-modified glycated albumin on murine endothelial cells. Biochem Biophys Res Comm 198:734 –739, 1994 67. COHEN MP, WU VY: Evidence for a ligand receptor system mediating the biologic effects of glycated albumin in glomerular mesangial cells. Biochem Biophys Res Comm 207:521–528, 1995