Heparin affects cytosolic glucose responses of hyperglycemic dividing mesangial cells

Mesangial expansion underlies diabetic nephropathy, leading to sclerosis and renal failure. The glycosaminoglycan heparin inhibits mesangial cell growth, but the molecular mechanism is unclear. Here, rat mesangial cells (RMCs) were growth-arrested in the G0/G1 phase of cell division, stimulated to divide in normal glucose (5.6 mm) or high glucose (25.6 mm) with or without heparin, and analyzed for glucose uptake. We observed that RMCs entering the G1 phase in normal glucose with or without heparin rapidly cease glucose uptake. RMCs entering G1 in high glucose sustained glucose uptake for the first 3 h, and high-glucose exposure of RMCs only in the first 8 h of G1 induced the formation of an extracellular monocyte-adhesive hyaluronan matrix after cell division was completed. Moreover, a low heparin concentration under high-glucose conditions blocked glucose uptake by 1 h into G1. Of note, glucose transporter 4 (glut4) localized on the RMC surface at G0/G1 and was internalized into G1 cells under normal glucose conditions with or without heparin within 30 min. We also noted that, under high-glucose conditions, glut4 remained on the RMC surface for at least 2 h into G1 and was internalized by 4 h without heparin and within 1 h with heparin. These results provide evidence that the influx of glucose in hyperglycemic dividing RMCs initiates intermediate glucose metabolism, leading to increased cytosolic UDP sugars, and induces abnormal intracellular hyaluronan synthesis during the S phase of cell division.

Mesangial expansion underlies diabetic nephropathy, leading to sclerosis and renal failure. The glycosaminoglycan heparin inhibits mesangial cell growth, but the molecular mechanism is unclear. Here, rat mesangial cells (RMCs) were growth-arrested in the G 0 /G 1 phase of cell division, stimulated to divide in normal glucose (5.6 mM) or high glucose (25.6 mM) with or without heparin, and analyzed for glucose uptake. We observed that RMCs entering the G 1 phase in normal glucose with or without heparin rapidly cease glucose uptake. RMCs entering G 1 in high glucose sustained glucose uptake for the first 3 h, and high-glucose exposure of RMCs only in the first 8 h of G 1 induced the formation of an extracellular monocyte-adhesive hyaluronan matrix after cell division was completed. Moreover, a low heparin concentration under high-glucose conditions blocked glucose uptake by 1 h into G 1 . Of note, glucose transporter 4 (glut4) localized on the RMC surface at G 0 /G 1 and was internalized into G 1 cells under normal glucose conditions with or without heparin within 30 min. We also noted that, under high-glucose conditions, glut4 remained on the RMC surface for at least 2 h into G 1

and was internalized by 4 h without heparin and within 1 h with heparin. These results provide evidence that the influx of glucose in hyperglycemic dividing RMCs initiates intermediate glucose metabolism, leading to increased cytosolic UDP sugars, and induces abnormal intracellular hyaluronan synthesis during the S phase of cell division.
Mesangial expansion is the principal glomerular lesion in diabetic nephropathy (DN) 2 that reduces the area for filtration and eventually leads to sclerosis and renal failure (1,2). However, mesangial extracellular matrix expansion and sclerosis are preceded by phenotypic activation and transient proliferation of glomerular mesangial cells, followed by prominent glomerular infiltration of monocytes and macrophages (3,4). Glomerular monocytes and macrophages have been prominently identified in DN in both animal models (4) and diabetic humans (5) and appear to have a key role in the induction of mesangial matrix expansion, hypercellularity, and onset of proteinuria (6,7). A previous study (8) as well as our own (9,10) have shown that there is a significant increase in the hyaluronan matrix in glomeruli during the first week after induction of diabetes in rats by streptozotocin, which is coincident with glomerular monocyte/macrophage influx (4). There is compelling evidence for a causal link between increased glomerular hyaluronan matrix and monocyte/macrophage accumulation and that, in diabetic glomeruli, the structure of the hyaluronan matrix can mediate monocyte adhesion and activation, thereby contributing directly to the sclerotic process.
Hyaluronan is a linear glycosaminoglycan composed of repeating disaccharide units of GlcNAc and D-glucuronic acid with alternating ␤-1,4 and ␤-1,3 glycosidic bonds. It is a major ubiquitous component of extracellular matrices. The number of repeat disaccharides in a completed hyaluronan molecule can reach 20,000 or more, a molecular mass of more than 8 million Da, and a length of more than 20 m. Formation of the monocyte-adhesive hyaluronan matrix by rat renal mesangial cells (RMCs) is a multiphase process that requires a PKC signaling pathway activated in RMCs dividing in hyperglycemic glucose, synthesis and accumulation of hyaluronan in intracellular compartments (13-24 h from G 0 /G 1 ) that initiates an endoplasmic reticulum stress/autophagic response, and cyclin D3-mediated formation of the abnormal extracellular monocyte-adhesive hyaluronan matrix after completion of cell division (24 -48 h from G 0 /G 1 ) (10). Understanding these cellular and molecular events will provide significant insights into the mechanisms controlling cellular responses to hyperglycemia that initiate the progression of DN.
Heparin is a highly sulfated, highly polyanionic glycosaminoglycan with a repeating disaccharide that contains a hexuronic acid (either glucuronic acid or iduronic acid) and glucosamine (either N-acetylated or N-sulfated). It is synthesized as a proteoglycan (serglycin) that is found in mast cells. Heparin has been shown to inhibit mesangial cell growth in both experimental renal disease models (11)(12)(13)(14)(15) and in cell culture (16,17). Particularly pertinent to this study is that heparin has been shown to prevent albuminuria and mesangial expansion and sustain glomerular function in Streptozotocin-treated diabetic rats (11,12) and suppress mesangial cell proliferation and matrix expansion in experimental mesangioproliferative glomerulonephritis (18). However, the molecular and cellular mechanisms underlying the beneficial roles of heparin in DN are still unknown.
Our previous studies (19 -21) support our proposal for a receptor for heparin on the surface of quiescent, growth-arrested G 0 /G 1 RMCs based on the observations that binding of heparin to RMCs is specific, rapid (5-10 min), saturable (within 60 min), and reversible; that Scatchard analysis of heparin binding indicates a single class with ϳ6.6 ϫ 10 6 binding sites/cell (K d ϭ 1.6 ϫ 10 Ϫ8 M) in quiescent cells; that surface-bound heparin can be internalized and degraded; that the affinity and number of heparin binding sites are affected by the stage of RMC growth; and that heparin acts at the RMC surface to affect both PKC-dependent and -independent pathways.
RMCs can be growth-arrested by serum starvation for 48 h. At this time, the cells are quiescent because less than 10% of the cells express proliferating cell nuclear antigen, a marker for cell proliferation, and this number increases to ϳ60% 16 h after serum stimulation; serum-starved RMCs incorporate very little [ 3 H]thymidine into DNA during 1 h of labeling, whereas a transient burst of DNA synthesis is observed 16 h after serum stimulation; no c-fos mRNA is detected in starved RMCs by Northern blotting analysis, whereas transient expression of c-fos mRNA is observed between 0.5-1 h after serum stimulation; ϳ75% of starved RMCs are arrested in G 0 /G 1 phase, as determined using flow cytometry analysis, and progress to S phase by 18 h after serum stimulation; and the G 1 /S boundary is reached at ϳ12 h, and the cell cycle is completed ϳ24 h after cells reenter G 1 . Therefore, after 48 h of serum starvation, RMCs are quiescent at G 0 /G 1 and re-enter the cell cycle with good synchrony upon serum stimulation (19 -21). Therefore, this serum-starved RMC model is an excellent one to investigate the responses to high glucose during a synchronized cell cycle and to find out to what extent heparin impacts the high-glucoseinduced responses. This study provides evidence for the critical role of glucose uptake during the first 4 h of G 1 phase in the responses of RMCs to high glucose in the absence or presence of heparin.

The minimal exposure time that yields RMC responses to high glucose
Our previous studies have shown that serum-starved, nearconfluent RMC cultures stimulated to divide in 25.6 mM glucose (high glucose) and 10% FBS for 72 h form an extruded monocyte-adhesive hyaluronan matrix after division (9,10). To determine the minimal exposure time that yields maximal monocyte binding in response to high glucose concentration, serum-starved, near-confluent RMC cultures were treated with 10% FBS to stimulate cell division in normal glucose (5.6 mM) or high glucose for 72 h or with high glucose for 8 h that was changed to normal glucose and continued to 72 h. The cultures were cooled to 4°C, and U937 cells, a myeloid cell line used to monitor monocyte adhesion (22), were added for 1 h. Fig. 1 shows examples of the numbers of bound U937 cells for the three culture conditions. As shown previously, cultures in high glucose for 72 h bound significantly more U937 cells than those in normal glucose (9,10). The cultures that were changed to normal glucose at 8 h also showed significantly more U937 cell binding than normal-glucose cultures but also significantly less than the 72-h high-glucose cultures (Fig. 1). Parallel cultures were analyzed for hyaluronan content by FACE (9, 10). Interestingly, the significant increases in hyaluronan content compared with normal-glucose cultures were the same for both the 72-h and the 8-h high-glucose-treated cultures (Fig. 2). These results provide evidence that RMCs exposed to high glucose for only 8 h into G 1 are already reprogrammed to initiate intracellular hyaluronan synthesis, which does not start until the cells enter S phase (19 -21), and that the structure of the extracellular matrix extruded after completing cell division is monocyteadhesive, albeit not as much as observed in continuous 72-h high-glucose cultures.

Dividing RMCs block glucose transport
To determine the uptake of glucose by RMCs at the initial stage of the cell cycle, RMC cultures were stimulated to divide from G 0 in normal and high glucose with or without heparin. The fluorescently labeled glucose analog 2-(N-(7-nitrobenz-2oxa-1,3-diazol-4-yl)amino)-2-deoxyglucose (2-NBDG), which enters through glucose transporters and is retained in cells, was added to the cultures at time 0, and cultures were analyzed for fluorescence at 1, 2, 3, and 4 h. The fluorescence for the 2-h cultures is shown in Fig. 3, and the measured fluorescence in cultures at each time point is shown in the bar graphs. Fig. 3A shows that 2-NBDG uptake is much higher in the high glucose compared with the high glucose plus heparin already at 2 h. Strikingly, there was nearly no 2-NBDG uptake in cultures in normal glucose regardless of the presence or absence of heparin, providing evidence that dividing cells activate a mechanism to prevent glucose uptake as they enter G 1 of the cell cycle. However, 2-NBDG uptake in high glucose alone increased con-

Hyperglycemia induces hyaluronan synthesis
tinuously during the first 3 h but did not increase during the following 3-4 h (Fig. 3B, High). In contrast, although uptake of glucose occurred in the first hour in high glucose plus heparin, it was subsequently blocked, with no additional uptake after 1 h (Fig. 3B, High ϩ Heparin). These results provide strong evidence that dividing RMCs, and likely most dividing cells, rapidly stop glucose uptake as they enter G 1 phase of cell division; that they cannot stop glucose uptake when the glucose concentration is hyperglycemic (more than ϳ2.5 ϫ normal (9)) for at least 3 h, which reprograms the cells to synthesize hyaluronan in intracellular compartments in S phase; and that the presence of heparin reprograms the cells to stop glucose uptake by the end of the first hour into G 1 , which subsequently blocks intracellular hyaluronan synthesis (23).

glut4 is likely involved
glut4 is a glucose transporter that is expressed by RMCs and is found on quiescent mesangial cells of glomeruli and on quiescent smooth muscle cells of microvasculature in normal rat kidneys (24). To determine the role of glut4 in mediating early glucose uptake by RMCs, nonpermeabilized RMC cultures were stained for glut4 at different times up to 4 h in G 1 in normal and high glucose with and without heparin (Fig. 4). The growth-arrested (G 0 /G 1 ) culture (time 0) shows glut4 distributed over the RMC surface, which is internalized within 30 min in RMC cultures in normal glucose and nearly so in cultures in normal glucose plus heparin. In cultures in high glucose, glut4 remains on the surface 1 and 2 h from G 0 and is no longer on the surface by 4 h. This indicates that it has been internalized within the next 2 h, likely around 3 h, given the results for the uptake of 2-NBDG (Fig. 3), which showed continuous uptake for only 3 h. The presence of heparin in the high-glucose culture showed nearly complete internalization at 1 h, which is also consistent with the 2-NBDG uptake that was blocked by 1 h from G 0 . These data indicate that the intracellular mechanism required to block the initial hyperglycemic influx of glucose by heparin likely involves interaction with a cell surface receptor that reprograms the RMCs to remove glut4 from the cell surface by the end of the first hour after the cells enter G 1 phase.

Hyperglycemia induces hyaluronan synthesis Cytosolic UDP-glucose increases in hyperglycemic dividing RMCs
The results in Figs. 3 and 4 provide evidence that continuous early uptake of glucose for 3 h in G 1 from G 0 likely increases cytosolic UDP sugars in hyperglycemic dividing cells, which are necessary for providing the substrates for the subsequent intracellular HA synthesis initiated early in S phase. Therefore, we determined to what extent early robust entry of glucose impacted intracellular UDP-glucose, UDP-GlcNAc, and UDP-GalNAc, the metabolites of intermediary glucose metabolism. Rat G 0 /G 1 RMCs were stimulated to divide in normal and hyperglycemic medium for 6 h in the presence or absence of heparin. UDP sugars were extracted from the cell cultures and purified by ion exchange column chromatography. The sugars were then released by acid hydrolysis for analysis on FACE gels (25,26). The results show that UDP-glucose is greatly elevated within 6 h by hyperglycemic medium alone (Fig. 5). These data are consistent with the continuous 3-h increase of 2-NBDG uptake by RMCs in response to high glucose (Fig. 3) and show that high glucose did stimulate glucose uptake and the subsequent cytosolic intermediary glucose metabolism in RMCs at an early G 1 cell cycle stage. Importantly, the heparin treatment prevented high-glucose induced UDP-glucose production, consistent with the finding that heparin blocked early 2-NBDG entry into RMCs induced by high glucose. Interestingly, intracellular UDP-GlcNAc and UDP-GalNAc were maintained at basal levels regardless of the presence or absence of heparin at the 6-h time point. Further, there were no differences in terms of these intracellular UDP sugars when comparing the 6-h cultures with G 0 /G 1 cells. The ratios of UDP-glucose content in high-glucose culture to the ones in low-glucose and heparintreated high-glucose cultures are ϳ8 and ϳ4, respectively. In addition, high-glucose-induced intracellular UDP-glucose production was observed as early as 3 h after the cells entered G 1 phase (data not shown). The ratios of UDP-glucose content in high-glucose culture to the ones in low-glucose and heparintreated high-glucose cultures are ϳ3 and ϳ2, respectively. Thus, these data indicate that RMCs at the early stage of the cell cycle in normal glucose or high glucose plus heparin maintain levels of cytosolic glucose concentrations well into G 1 phase and that the 3-h influx of glucose in high glucose (Fig. 3) increases the cytosolic UDP-glucose concentration greatly during the 6 h into G 1 .
Our previous studies indicate that hyperglycemic cells initiate intracellular hyaluronan synthesis (ϳ13 h from G 0 ) shortly after entering S phase (9, 10, 21, 23). Thus, increased cytosolic glucose and UDP-glucose at times later than 6 h need to initiate the hexosamine biosynthesis pathway to synthesize UDP-GlcNAc and convert UDP-glucose to UDP-glucuronic acid (UDP-GlcUA), the cytosolic substrates for synthesis of hyaluronan into intracellular compartments (10,23).

Discussion
Daily intraperitoneal injection of heparin (55 mg/kg) in the diabetic rat model prevented proteinuria and nephropathy and sustained kidney function for 8 -10 weeks (11,12). However, the cellular and molecular mechanism(s) underlying this therapeutic role of heparin are still unclear. Our previous studies have shown that glomerular mesangial cells that divide in medium with glucose concentrations three times higher than normal or greater activate hyaluronan synthesis in intracellular compartments (10,23). This initiates autophagy and subsequent extrusion of a monocyte-adhesive hyaluronan matrix after completion of cell division.
This mechanism also occurs in vivo in the streptozotocin type 1 diabetic rat model, with extensive accumulation of hyaluronan in glomeruli and influx of macrophages, resulting in nephropathy, proteinuria, and kidney failure by 6 weeks. Our studies also demonstrated that heparin (2.0 g/ml) prevents intracellular hyaluronan synthesis and autophagy in vitro. However, after completing division, the cells synthesized an even larger extracellular monocyte-adhesive hyaluronan matrix. In vivo, glomerular hyaluronan in heparin-treated diabetic rats increased greatly in weeks 1-2 and declined to near control levels, i.e. a near-absence of hyaluronan, by 6 weeks, at which time the glomeruli had extensive numbers of macrophages. Therefore, these previous results provide evidence that dividing cells have a receptor that binds heparin and initiates the intracellular responses that inhibit activation of hyaluronan synthesis in intracellular compartments during division and that reprogram the cells to address the sustained glucose stress after division by initiating synthesis of a large monocyte-adhesive extracellular hyaluronan matrix. This study provides evidence that the mechanisms involved are initiated during the very early stage of RMC division, when entering G 1 phase.
Exposure of RMCs to high glucose for 8 h at early G 1 phase of the cell cycle was sufficient to induce formation of a monocyteadhesive hyaluronan matrix after cell division was completed. This result indicated that high glucose induced the intracellular mechanism that governs the RMC responses during subsequent cell cycle progression and after cell division is completed. Thus, this short treatment of RMCs entering G 1 with high glucose was enough to induce the subsequent RMC phenotypic activation responses, including intracellular hyaluronan synthesis early in S phase and the formation of a monocyte-adhesive hyaluronan matrix after division (9,10,21,23). This is consistent with previous studies in which high glucose induces activation of PKC in mesangial cells within 2-3 h after serum
Serum-starved G 0 /G 1 RMCs in normal glucose maintained glut4 on the cell surface and rapidly removed it when addition of serum stimulated re-entry into the G 1 cell cycle with subsequent blockage of 2-NBDG uptake. This provides evidence that glucose utilization is not necessary during the cell cycle and must be blocked. In contrast, re-entry of G 1 in high glucose sustained cell surface glut4 and 2-NBDG uptake for 3 h. This indicates that there was an early high-glucose-induced glucose influx into the cells in response to extracellular high glucose, leading to the activation of intracellular mechanisms that mediate the subsequent pathological responses. Therefore, preventing glucose entry into cells at the early stage of the cell cycle is required for normal cell division. This supports our hypothesis that major pathological responses are generated from hyperglycemic dividing cells.
Importantly, this high 2-NBDG-induced glucose influx into cells at the entry of G 1 stage of the cell cycle was blocked within 1 h by heparin by removal of glut4 from the cell surface. This indicates that direct interaction between heparin and cells at this early G 1 stage initiated intracellular pathways to prevent sustained glucose uptake into the cells, thereby preventing intracellular glucose activation of the mechanism(s) that interfere with normal cell cycle progression and compromise cell functions after division. This interpretation is supported by our previous finding that heparin interacts with cell surface receptors to regulate the mitogen-activated protein kinase-dependent and -independent pathways (19 -21).
Serum-starved G 0 /G 1 cells maintained glut4 on the cell surface, consistent with findings in other serum-starved cells, which have high activities that use glucose metabolic pathways to support their survival (24). Upon serum stimulation to re-enter the cell cycle in normal glucose, glut4 was rapidly removed from the cell surface to avoid excessive glucose influx interfering with subsequent normal cell cycle progression.
Because excessive glucose was inside of cells in high-glucose cultures, the question is whether this intracellular glucose activates the intermediate glucose metabolism during the early stage of the cell cycle. Our data demonstrated that UDP-glucose was greatly increased in cells cultured in high glucose for 6 h. The contents of this UDP-glucose in normal glucose with or without heparin were at basal levels equivalent to serumstarved cells, indicating that abnormal production of UDP-glucose by intermediate glucose metabolism is a pathological response of RMCs to sustained glucose uptake for 3 h in highglucose medium. As expected, heparin treatment prevented this activation of intermediate glucose metabolism for formation UDP-glucose by activating a mechanism to remove glut4 from the surface within 1 h and block further 2-NBDG uptake. This also demonstrated that one of the beneficial roles of heparin in diabetic nephropathy is to prevent the activation of high-glucose-induced intermediate metabolism by blocking excessive glucose uptake at the early stage of the RMC cell cycle.
In summary, this study provides strong evidence that highglucose-induced early responses in cell cycle progression are crucial for later changes in cell functions; high glucose induces excessive influx of 2-NBDG into cells through cell surface glut4, which activates abnormal intermediate cytosolic glucose metabolism; and interaction between heparin and a putative cell surface receptor on RMCs initiates a mechanism to remove the glut4 from the cell surface shortly after entering G 1 , preventing excessive glucose uptake and subsequent activation of intermediate glucose metabolism. One of our ongoing studies is to determine the cell surface receptor and what the signaling pathways are that remove glut4 and prevent excessive glucose influx to maintain normal RMC functions.

Establishment of RMC cultures and assay for glucose uptake
RMC cultures were established from isolated glomeruli and characterized as described previously (31,32). RMCs were used between passages 5 and 15, when they still contract in response to angiotensin II and endothelin and exhibit growth suppression in the presence of heparin (1 g/ml), which are additional characteristics of mesangial cells (19,33,34). RMCs were cultured in RPMI 1640 medium containing 10% FBS and passaged at confluence by trypsin digestion for 5 min with a solution of 0.025% trypsin and 0.5 mM EDTA. To render cells quiescent (19), cultures at 40% confluence (2 ϫ 10 4 cells/cm 2 ) were washed with RPMI 1640 medium and placed in fresh medium containing 0.4% FBS for 48 h (yielding 70 -80% confluent cultures).
Glucose uptake by RMCs was assayed with 2-NBDG as described previously (35). Briefly, serum-starved RMCs on chamber slides were stimulated by 10% FBS in RPMI 1640 medium with concentrations of 5.6 and 25.6 mM D-glucose containing 3.9 mol 2-NBDG/1 mmol glucose in the presence or absence of 1 g/ml heparin. 1, 2, 3, and 4 h after stimulation, RMC images were collected on a fluorescence microscope, and the fluorescence intensities of the collected images were quantified with ImageJ software.

Immunohistochemistry
Paraformaldehyde-fixed RMC cultures on chamber slides were stained for glut4 with an antibody and for nuclei with 4Ј,6-diamidino-2-phenylindole, as described previously (9,10) or according to the manufacturer's instruction. Samples were treated with the antibody at 1:75 dilution, washed, and treated with anti-mouse IgG Cy3 antibody at 1:200 dilution. Stained samples were mounted in VectaShield containing 4Ј,6-diamidino-2-phenylindole (Vector Laboratories) for staining the nuclei of cells. Confocal images of the samples were obtained with a Leica TCS-NT laser-scanning confocal microscope equipped with four lasers for excitation at 351 and 561 wavelengths. The same settings of the confocal microscope and laser scanning were used for both control and treated samples.

Hyperglycemia induces hyaluronan synthesis Assay for monocyte adhesion
Serum-starved RMCs in 6-well plates were treated up to 72 h with 10% FBS and with concentrations of 5.6 and 25.6 mM D-glucose. U937 cells were cultured in suspension in RPMI 1640 medium containing 5% FBS and passaged at a 1:5 ratio (2 ϫ 10 5 cells/ml) every 48 h (22). Assays for monocyte adhesion were done as described previously (9,22). After washing, the cell cultures were imaged by microscopy with a Polaroid digital camera (9), and the numbers of monocytes per culture area were counted using Image-Pro software. Each culture was equally divided into four regions, and a culture area for imaging was picked randomly in each region. Streptomyces hyaluronidase treatment (1 turbidity reducing unit/ml at 37°C for 15 min) of RMCs before monocyte incubation was used to determine the extent of hyaluronan-mediated adhesion. Analysis of hyaluronan content in mesangial cell cultures was performed as described previously (9).

FACE analysis of reducing saccharides
Serum-starved RMCs in 6-well plates were treated for 6 h with 10% FBS and concentrations of 5.6 and 25.6 mM D-glucose with or without 1 g/ml heparin. The cell cultures were extracted with 1 ml of 75% cold ethanol for 10 min. The ethanol lysates were dried by centrifugal evaporation. The dried samples were dissolved in 200 l of Milli-Q water, and 200 l of chloroform was used to remove proteins from the samples. Then UDP sugar extracts were mixed with 50 l of Q-Sepharose beads, and, after washing with 100 l of Milli-Q water five times, the saccharides of UDP sugars bound to the beads were released in 50 l of 20% acetic acid at 100°C for 10 min. The acid supernatants were dried by centrifugal evaporation and redissolved in 20 l of 0.1 M ammonium acetate (pH 7.0). The saccharides were dried by centrifugal evaporation in microtubes and then subjected to reductive amination with 2-aminoacridone as described previously (9,25,26). At the end of incubation, the samples were each mixed with glycerol to 20%, and 5-l aliquots were then subjected to electrophoresis on Glyko Mono Composition gels with Mono running buffer from ProZyme Inc. (San Leandro, CA). Running conditions were 500 V at 4°C in a cold room for 1 h. Gels were imaged on an Ultra Lum transilluminator (365 nm). Images were captured with a Quantix cooled charge-coupled device camera from Roper Scientific/Photometrics and analyzed with the Gel-Pro Analyzer program, version 3.0 (Media Cybernetics).