Influence of Acute High Glucose on Protein Abundance Changes in Murine Glomerular Mesangial Cells

2.1. Cell Culture

The rat GMC line CRL-2573 (ATCC) maintained normal growth media (DMEM: 5 mM D-glucose, 15% FBS) under 5% CO₂ at 37°C. The cells (passages 10–15) were plated in Corning T25 flasks and cultured until 70–80% confluence was reached. Normal media were removed from cells and replaced with DMEM supplemented with 0.5% FBS/5 mM D-glucose. After 24 h, media were removed and replaced with isoosmotic-normal glucose (NG^∗) media (DMEM-5 mM D-glucose, 20 mM mannitol, and 0.5% FBS) or high glucose (HG) media (DMEM: 25 mM D-glucose, 0 mM mannitol, and 0.5% FBS), for 2 h. For 2DE analysis, after 2 h treatment, the total protein was collected as previously described [14] using IPG rehydration buffer supplemented with protease inhibitors.

2.2. Cell Viability

Cell viability was determined after 2 h HG and NG^∗ treatment using the MTT assay [15] as described by the manufacturer (Sigma, St. Louis, MO, USA).

2.3. Two-Dimensional Electrophoresis (2DE) and Image Acquisition

2DE experiments were conducted as reported previously [14]. Murine GMC protein (75 μg) was rehydrated overnight into IPG (pH 3–10; 7 cm; Invitrogen) strips. The strip was focused for a total of 1200–1300 Vh with a final 30 min focusing period at 2000 V constant. Proteins were separated in the second dimension on 4–12% Bis-Tris mini gels (8 cm × 8 cm). The gel slabs were fixed in 10% methanol and 7% acetic acid and then transferred to SYPRO-Ruby protein gel stain (Molecular Probes, Oregon, USA) for 18 hours. Gels were scanned using a PerkinElmer ProXpress CCD-based digital imager at 50 μm resolution. The gel/stain exposure and emission acquisition times were varied to maximize the detector response while avoiding detector saturation. The image files were matched, reference gels were created, and spot volumes were determined using Progenesis Discovery software (Nonlinear Dynamics, Newcastle upon Tyne, UK). A student’s t-test is used to evaluate all matched spot pairs. Protein spots that were found to have variable spot volumes between samples were statistically compared by spot mean and SEM.

2.4. Proteomic Analyses

Protein gel spots were digested as previously described [14]. MALDI-TOF and TOF/TOF MS data were acquired on the tryptic digests using an AB4700 Proteomics Analyzer (Applied Biosystems, Foster City, CA) and analyzed using Matrix Science Mascot (ver. 2.0) as described previously [16]. Data was analyzed assuming (a) monoisotopic peptides masses, (b) cysteine carbamidomethylation, (c) variable oxidation of methionine, (d) maximum of one missed trypsin cleavage, and (e) a mass accuracy of greater than 150 ppm for MS data and 0.3 Da for MS-MS data against the SwissProt (release 52.0, 20070307) protein database (261513 sequences; 95638062 residues) constrained to the mammalian (50870 sequences) taxa. Limitation of the original protein mass was not employed within the Mascot search. Protein identifications were accepted for protein identifications that include using MASCOT MS + MS/MS analysis with significant MOWSE scores (p < 0.05; for MS MOWSE score of 60 which equals significance and for MS/MS MOWSE peptide ion score alone of 40 which equals significance).

2.5. Confocal Microscopy

Confocal microscopy images were obtained as previously described [17]. Briefly, multichambered cover glass wells (Nunc, Naperville, CT) were seeded with GMC cells. Cells were serum starved with 0.5% FBS-NG medium 24 h before 2 h glucose treatment. Cells were rinsed three times with PBS that contained calcium and magnesium and fixed in 3.7% paraformaldehyde in PBS for 10 min, followed by permeabilization with 0.025% NP-40 in PBS for 15 min. Cells were incubated with primary antibody (1 : 250 anti-PHB in PBS/0.025% NP-40) at 20°C, rinsed five times with PBS/0.025% NP-40, and incubated with the Alexa Fluor 488 conjugated secondary antibody (1 : 1000) at 20°C. The cells were rinsed five times with PBS/0.025% NP-40, incubated with 300 nM DAPI for 5 min, and rinsed three times with PBS. Images were acquired using a Zeiss confocal microscope and analyzed using LSM510 software. Z scan analysis was performed by scanning at 1 μm intervals and three-dimensional reconstruction of the fluorescence images. The images for PHB and for DAPI were merged in a single image to elucidate PHB cellular distribution. Fluorescence intensity measurements (mean fluorescence intensity per μm²) were computed per cell (n = 4-5 cells per treatment replicate per treatment condition) and used to estimate differences in PHB nuclear and cytoplasmic distribution.

2.6. Protein Immunoblotting (IB)

1DE and 2DE protein immunoblots (IB) were conducted as previously described [14]. Total cell lysate samples were separated by 2DE (n = 3 HG, n = 3  NG^∗). For 2DE IB analysis, following IEF of mesangial proteins, the plastic backing of the IPG strips was trimmed off. The acidic most point of the strip was aligned in the IPG well of the Bis-Tris mini gels adjacent to the MW standard lane of the minigel. This procedure insured uniform alignment of IPG strips to the MW standards, in order to compare PHB migration pattern between experimental conditions. Following 1DE or 2DE electrophoresis and transfer, membranes were immunoblotted for PHB (Santa Cruz Biotechnologies, Santa Cruz, CA) at a 1 : 1000 dilution in 5% albumin in Tris-Tween-20 buffered saline (TTBS). PHB spots were imaged on film with luminol images aligned and quantified by densitometry analysis comparing the means of the acidic third and basic third of the PHB charge trains to the total train densitometry. Additional antibodies used for 1DE immunoblots were anti PDI (Stressgen; San Diego, CA) at a concentration of 1 : 10,000 and PSMA2 (Cell Signaling; Danvers, MA) at a concentration of 1 : 1000.

2.7. Analysis of Protein Expressional Networks

Ingenuity Pathways Analysis bioinformatic tool (Ingenuity Systems, Mountain View, CA) uses a curated database (Ingenuity Systems Knowledge Base) of previously published findings on mammalian biology from the public literature to evaluate proteins lists inclusive of expression ratios for protein expressional patterns. The purpose of the evaluation is to establish within the lists of provided expressional data relational networks of protein interactions (e.g., direct protein-protein interaction and transcriptional control). Analysis of submitted protein lists with expressional ratios using the Ingenuity knowledge base was used to identify direct interactions between mammalian orthologs.

Murine GMC proteins demonstrating statistically significant expression between 2 h HG and 2 h NG^∗ as well were analyzed by the Ingenuity Knowledge Base and Pathways Analysis tool. The data output identifies nodes characterizing individual proteins and edges characterizing biological relationships. Putative protein networks are rank ordered according to p value ($$), where the p value is a measure of random association of the listed proteins.

2.8. Statistical Analysis

Statistical analysis of relative spot pixel intensity from 2D gels (n = 5, 2 h each group) and analysis of PHB, PDI, or PSMA2 for HG versus NG^∗ expression by IB was performed using two-tailed, unpaired t-test. p values < 0.05 were considered significant.

3.1. Alteration of Protein Expression by Acute High Glucose

Based on the MTT assay results (data not shown), GMC viability did not statistically vary between 2 h HG and NG^∗ treatments. To determine proteins regulated by 2 h HG treatment, protein spot volume lists were curated by first estimating intergel variability in matched protein spot volumes (averaged CV for 20 matched spots = 0.17). Next, all intraglucose treatment matched gel spot volumes having a CV greater than 0.35 (2 × CV) were discarded. Fifty-one (51) protein spots had a spot volume CV of less than 0.35 and uncorrected t-test values of ≤0.05. Thirty-five protein spots had increased expression and 16 protein spots had decreased expression with 2 h HG treatment and all were analyzed using proteomic methods based on MASCOT MOWSE scoring including MALDI TOF/TOF peptide fragmentation (sequence tagging) data with a significance p value ≤ 0.05 for all the reported protein identities. A representative 2DE gel image (with annotations) and tabulated information for 51 regulated protein spots are provided (Figure 1; Table 1). In general, all proteins identified were observed migrating in the gels at the correct molecular weight plus or minus 10% except for gel spot 5. Cofilin-1 was identified migrating at a molecular weight of approximately 9000 Da and a pI of 5.3. Cofilin-1 nominally has a translated molecular weight of 18,749 Da and a pI of 8.2. Two additional cofilin-1 containing gel spots as well as one HSP10 containing gel spot were observed to focus on isoelectric points less than 0.5 pH units, more acidic than expected. Twenty-three proteins were observed to focus on isoelectric points greater than 0.5 pH units, more basic than expected. The remaining gel spots identified proteins within 0.5 pH units of the expected pI.

3.2. Analysis of Protein Expressional Networks

Bioinformatic analysis of protein expression in 2 h NG^∗ versus 2 h HG was achieved using the Ingenuity Knowledge Base and Pathways Analysis tools. The top three canonical pathways determined to be activated from 2 h acute high glucose exposure were actin-based motility by Rho, RhoA signaling, and the protein ubiquitination pathway. Analysis of protein expressional networks from murine GMC 2 h NG^∗ and 2 h HG protein expressional data suggested three primary expression networks. Network 1 (score 49) addressed cancer, reproductive system disease, and hematological disease and included 25 identified proteins out of 35 total network components (Figure 2(a)). Network 2 (score 19) addressed cell death and survival, drug metabolism, and lipid metabolism and included 9 identified proteins out of 35 protein nodes (Figure 2(b)). Network 3 (score 14) addressed cellular movement, cellular compromise, cellular function, and maintenance and was composed of 7 identified proteins out of 29 possible network proteins. Prominent nodes within Network 1 were centered on signaling proteins including proteins involved with ubiquitination, cyclin D, ERK1/ERK2 MAPKinase, HSP90, ROCK, and histones h3 and h4. Prominent nodes in network 2 were centered on the VEGF, TNF, TGFβ1, tumor protein 53 (TP53), and ubiquitination.

3.3. Immunochemical Analysis for the Effect of High Glucose on the Expression of Proteins

Immunoblot (IB) analyses of the selected proteins were used to confirm the 2DE findings. Prohibitin (PHB) was selected for confirmation as it was one of the most strongly regulated protein spots and was also a component of IPA Network 2 with direct interaction with a prominent network node of TNF. The expression by 1DE (Figures 3(a) and 3(b)) supported a trend in increased total PHB abundance, but 2DE IB analysis of 2 h GMC cells cultured in HG and NG^∗ showed a HG responsive and statistically significant (p < 0.02) increase in the acidic end of the PHB charge train (Figure 4). Confocal microscopy (Figures 5(a) and 5(b)) suggested that high glucose resulted in a statistically significant (p value < 0.0001) increased fractional abundance of PHB in the nucleus of the GMC.

Based on bioinformatics analysis defining regulation of protein ubiquitination pathways in one of the top three canonical pathways regulated, as well as a prominent node in protein expression Networks 1 and 2, we next analyzed the expression of proteins involved in protein homeostasis and found them to be regulated by 2DE analysis. Proteasome subunit alpha-type 2 (PSMA2) was confirmed to be increased in mesangial cells exposed to high glucose concentrations for 2 h (Figure 6). Comparative 2DE analysis also defined decreased expression of ER chaperone proteins such as PDI and GRP58, which may lead to increased unfolded protein load in mesangial cells and induction of proteasomal degradation processes. Immunoblot analysis of mesangial proteins for PDI confirmed 2DE findings of decreased expression of PDI (Figure 6).