Neo-Epitopes Generated on Hydroxyl Radical Modified GlycatedIgG Have Role in Immunopathology of Diabetes Type 2

Glycoxidation plays a crucial role in diabetes and its associated complications. Among the glycoxidation agents, methylglyoxal (MG) is known to have very highglycationpotential witha concomitant generation of reactive oxygen species (ROS) during its synthesis and degradation. The presentstudy probes the MG and ROSinduced structural damage to immunoglobulin G (IgG) and alterations in its immunogenicity in diabetes type 2 patients (T2DM). Human IgG was first glycated with MG followed by hydroxyl radical (OH•) modification. Glycoxidation mediated effects on IgG were evaluated by various physicochemical techniques likeultraviolet (UV) and fluorescence spectroscopy, 8-anilinonaphthalene-1-sulfonic acid (ANS) binding studies, carbonyl andfree sulfhydryl groups assay, matrix assisted laser desorption ionization mass spectrometry-time of flight (MALDI-TOF), red blood cell (RBC) haemolysis assay, Congored (CR) staining analysis and scanning electron microscopy (SEM). The results revealed hyperchromicityin UV, advanced glycation end product (AGE)specific and ANS fluorescence, quenching in tyrosine and tryptophan fluorescence intensity,enhanced carbonyl content,reduction in free sulfhydryl groups,pronounced shift in m/z value of IgGand decrease in antioxidant activity in RBC induced haemolysis assayupon glycoxidation. SEM and CRstaining assay showed highly altered surface morphology in glycoxidised sample as compared to the native. Enzyme linked immunosorbent assay (ELISA) and band shift assay were performed to assess the changes in immunogenicity of IgG upon glyoxidation and its role in T2DM. The serum antibodies derived from T2DM patients demonstrated strong affinity towards OH• treated MG glycatedIgG (OH•-MG-IgG) when compared to native IgG (N-IgG) or IgGs treated with MG alone (MG-IgG) or OH• alone (OH•-IgG). This study shows the cumulating effect of OH• on the glycation potential of MG. The results point towards the modification of IgG in diabetes patients under the effect of glycoxidative stress, leading to the generation of neo-epitopes on theIgG molecule and rendering it immunogenic.


Introduction
There is an overwhelming literature supporting the indulgence of reactive oxygen species (ROS)and reactive carbonyl species (RCS) in severe pathogenesis of aging, cancer, diabetes and its associated complications [1,2]. The non-enzymatic synthesis of glycated adducts formed by the reaction of proteins withreducing sugar contribute in the pathogenesis of diabetic complications via free radical generation that promote carbonyl formation, fragmentation and cross linking of proteins [3][4][5]. Among the sugar derivatives,methylglyoxal (MG) is a reactive dicarbonyl compound having20,000 times more glycatingpotential than glucose [6].It is produced by degeneration of lipid peroxidation products (LPP), autoxidation of sugars, dephosphorylation of polyol pathways and glycolytic intermediates such as glyceraldehyde-3-phosphate (G3P) and dihydroxyacetone phosphate (DHAP) as well as oxidation of hydroxyacetone and aminoacetone [7,8]. MGreacts with a variety of biological macromolecules forming fluorescent and non-fluorescent crosslinks [8][9][10][11].Previous literature has reported that the concentration of MG in diabetes patients increases many folds in lens, blood and kidney [12][13][14][15].Adirect link between free radical generation and MG toxicityis well known [16]. ROS production by MG was first described in 1993 and since then, the mutual interdependency between free radicals and MG is widely reported [17].Diabetes patients have elevated plasma MG levels that inactivate antioxidant enzymes and thereby accumulate an oxidative stress [18][19][20][21]. MG is a key player in the modification of amino acids,nucleic acids [14,22] and specific binding of MG modified proteins leads to immunological complications in diabetes patients [10,15,23,24].This work aims to study the hydroxyl radical(OH • ) mediated structural perturbations in MG glycated immunoglobulin G (IgG) byvarious biophysical and biochemical techniques like ultraviolet (UV) and fluorescence spectroscopy, 8-anilinonaphthalene-1-sulfonic acid (ANS) binding studies, estimation of carbonyl content and free sulfhydryl groups, matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), red blood cell (RBC)haemolysis assay, congored(CR)staining analysis and scanning electron microscopy(SEM). Furthermore, this work demonstratesthe changes in immunogenicity of IgG upon OH • -MG mediatedglycoxidation and its role in the immunopathology of diabetes type 2 (T2DM).

Clinical sampling
The study was undertaken on T2DM patients (n = 80; age >20 years), excluding those with micro and macro-vascular complications, type 1 diabetes (T1DM) and gestational diabetes (GDM).Healthy subjects (n = 20) of the same age group were takenas control. Blood was taken in clot activator vials and serum was separated by centrifugation at 3000 rpm for 10 min followed by heating at 56˚C for 30 min to inactivate complement proteins and stored in aliquots at -20˚C with 0.1% sodium azide as preservative [25].

Ethical statement
The study was approved by institutional ethics committee (certificate approval no. 1297/FM) at J. N. Medical College,AMU, Aligarh.Blood was taken in clot activator vials only after written informed consent from both the patients and healthy individuals and a proper record of all the patients and healthy individuals has been maintained.

Isolation of IgG
Blood from healthy individuals was obtained and allowed to coagulate at 37˚C for 30min. It was centrifuged at 3000 rpm for 10 min to obtain the serum which was heated at 56˚C for 30 min to inactivate complement proteins. IgG was isolated by affinity chromatography using Protein-Aagaroseaffinity column and its concentration was determined considering 1.4 O. D. 278 = 1mg/ml IgG [25].Homogeneity of the IgGwas checked on 7.5% SDS-PAGE and it was stored at -20˚C with 0.1% sodium azide as preservative.

Preparation of OH • treated MG glycatedIgG (OH • -MG-IgG)
IgG (1μM) was incubated under sterile conditions with MG (7.5 mM) in phosphatebuffer saline (PBS) (10 mM, pH 7.4) at 37˚C for 24 hours. It was followed by OH • modification (30 min) of theglycated sample. The OH • radicalwas generated by fenton's reaction. Assay tubes containing glycatedIgGwere incubated with 100 μM hydrogen peroxide (H 2 O 2 ) and 30 μM FeCl 2 /EDTAfor half an hour at 37˚C. The reaction was then stopped by using 0.1% trifluoroacetic acid (TFA) [4]. The effect of OH • -MG on IgG was studied with IgG alone, MG modified IgG (MG-IgG)and OH • modified IgG (OH • -IgG)as control samples. The samples were extensively dialyzed against sterile PBS (10mM, pH 7.4)and stored at -20˚C until further analysis.

UV-Absorption spectroscopy
UV absorption spectra of native and modified IgG were recorded in 250-400 nm wavelengthrange onShimadzu UV-1700 spectrophotometer [26]. Hyperchromicity at 280 nmwascalculated by following equation: Fluorescence spectroscopy To detect structural changes in the modified IgG, fluorescence spectra were taken on Shimadzu (RF-5301-PC) spectrofluorometerat 25±0.2˚C using quartz cuvette of 1 cm path length. The fluorescence intensities of tyrosine and tryptophan were monitored by exciting the samples at 275 and 295 nm and the emission spectra were recorded in the range of 280-400 and 300-400 nm respectively [27]. Decrease in fluorescence intensities in caseof modified samples were calculated using the following equation AGE-specific fluorescence was also monitored by exciting the samples at 370 nm and recording emission spectra in the 380-550 nm range [28]. Increase in fluorescenceintensity was calculated using the following equation

ANS binding studies
The conformational changes in the modified IgG were further evaluated byANS binding and recorded in terms of fluorescence.TheIgG to ANS molar ratio was 1:50 and the emission spectra were recorded in the wavelength range of 400-600 nm after exciting the samples at 380 nm [29].Increase in fluorescence intensity was calculated as given below: Carbonyl content Carbonyl content of native and modified IgG wasquantitated as per the published protocol [30].

Free sulfhydryl group content
Free sulfhydryl group contents of native and modified IgG were determined as per the established method [31].

MALDI-TOF studies
MALDI-TOF MSwas performed on 4800 plus MALDI-TOF mass spectrometer operating in a positive ion mode. 0.95 μl of 1 mg/ml samplewas spotted on a 384 well insert opti-TOF-stainless steel MALDI plate. Sinapinic acid dissolved in 10 mg/ml acetonitrile and 0.1% TFA was used as a matrix. Analysis was carried out using protein chip software.

RBC haemolysis assay
The antioxidant characteristics of native and modified IgG were evaluated by RBC haemolysis assay [32]. Healthy human blood samples from the J.N. Medical College were obtained in EDTA vials. Erythrocytes from the plasma were isolated and washed three times with isotonic saline (NaCl, 0.15 M), and centrifuged at 4000 rpm for 10 minutes at 4˚C. 100 μl (approx 1X 10 8 erythrocytes, 400,000 cells/μl final concentrations) of diluted RBC (1/10 in 0.15 M NaCl) were added into each well of flat bottom 96 well-plate. 10 μM each of nativeIgG(N-IgG), OH • -IgG,MG-IgG andOH • -MG-IgGwere then added to the well in duplicate. This was followed by addition of 20 μl of 200 mMof 2,2'-Azobis(2-amidinopropane) dihydrochloride (AAPH), a peroxyl radical initiator. Diluted RBC alone and RBC with AAPH were taken as positive and negative controls respectively. The samples were incubated at 37˚C and absorbance was recorded on microplate reader at 540 nm every 20 min until constant readings were achieved in each well. Results are expressed as time (in minutes) required for 50% of maximal inhibition in haemolysis of RBCs (H T 50 ) using the following equation [33].
A AAPH = Absorbance of AAPH at 540 nm A sample is the absorbance of IgG (native and modified) at 540 nm.

Spectacle crosslink's in modified IgG
CR staining was done to study aggregation as a result of protein (IgG) modification. Staining was done as per previously published procedure [34].Briefly, 200-400 μl of freshly prepared CR solution was put on slides having 10 μl of air dried IgG (5 μM). The slides were then examined under polarised light microscope. The detection of yellow or green colour birefringenceindicates the presence of the protein aggregates in the samples. Images of the stained slides were recorded with the camera attached to the microscope.

Scanning electron microscopy of modified IgG
SEM imaging was used to analyze changes in micro-architecture of the protein upon modification. 20 μl of 5 μMsampleswere air-dried and then adsorbed on cellulose ultra filtrationmembrane.The samples were then coated with gold and mounted overstainless steel grids covered with collodion film of carbon operating under low vacuum at 15kV. The samples were imaged under JSM-6510LV (JEOL JAPAN) scanning electron microscope at a 1000X resolution [35,36].

Direct binding ELISA
ELISA was performed on flat bottom polystyrene plates as described earlier [37]. Briefly, 96 wells polystyrene polysorpimmunoplates were coated with 100μl of N-IgG, MG-IgG, OH • -IgGand OH • -MG-IgG(10 μg/ml) in antigen coating buffer. The plates were incubated for 2 hr at 37˚C followed by overnight incubation at 4˚C. Each sample was coated in duplicate and half of the plate, devoid of antigen coating, served as control. The plates were washed 3 times with tris buffered saline containing tween 20 (TBS-T) and unoccupied sites were blocked with 150 μl of 2.5% 'fat free skimmed milk in tris buffered saline (TBS) for 4-5 hr at 37˚C. After incubation, the plates were washed thrice with TBS-T and coated for 2 hrwith sera or affinity purified IgG from diabetes patients. Evaluation of bound antibodies was done using anti-human IgG alkaline phosphatase conjugate using PNPP as substrate. Absorbance (A) of each well was monitored at 410 nm on an automatic microplate reader and the mean of duplicate readings for each sample was recorded. Results have been expressed as a mean ofA test -A control .

Inhibition ELISA
The antigenic specificity of antibodies was determined by competitive inhibition ELISA [38].
Varying amounts of inhibitors (0-20 μg/ml) were mixed with a constant amount of antiserum or affinity purified IgG and the mixture was incubated for 2 hrat 37˚C and overnight at 4˚C. The immune complex thus formed was loadedin the antigen coated wells instead of the serum. The remaining steps were the same as in direct binding ELISA. Percent inhibition was calculated using the formula: Band shift assay For the visual detection of antigen-antibody binding, gel retardation assay was performed [39]. Immune complexes were prepared by incubating constant amount of native/modified antigen with varying amounts of affinity purified IgG in PBS for 2hr at 37˚C and overnight at 4˚C. One-fourth volume of sample dye was added to the mixture and electrophoresed on 6% SDS-polyacrylamide gel for 3 hr at 80 V. The gels were visualized by staining with silver nitrate.

Statistical analysis
Data are presented as mean±SD. Statistical significance of the data was determined by Student's ttest, and a value of p < 0.05 was considered as statistically significant.

UV absorbance spectroscopy
The

Intrinsic fluorescence assay
Fluorophoric properties of tyrosine and tryptophan amino acid residues were utilised to record changes incurred upon modification.

AGE specific fluorescence
Fluorescence spectroscopy was also used to detect formation of AGEs by exciting samples at 370 nm with emission spectra recorded in the range of 380-550 nm. N-IgG and OH • -IgGexhibited negligible fluorescence at 370 nm excitation indicating no fluorogenic AGEs. MG-IgG and OH • -MG-IgG showed 69.1% and 71.0% hyperchromicity compared to native protein underidentical experimental conditions (Fig 3). 84.4% increase in ANS fluorescence intensity for OH • -IgG, MG-IgG and OH • -MG-IgGsamples compared to that of N-IgG respectively.FurtherOH • -IgG, MG-IgG and OH • -MG-IgG showed a blue shift of 9 nm, 21 nm and 29 nm respectivelyin comparison to that ofN-IgG (Fig 4).

Aggregation and cross linking studies
Analysis of samples incubated with CR demonstrated significant aggregate formation in OH • -MG-IgGsamples compared to OH • -IgG, MG-IgGand N-IgG (Fig 8). These results have been further confirmed by SEM wherein modification with OH • , MG and OH • -MG resulted in the formation of elongated fibril like aggregates. The OH • -MG-IgG showed maximum aggregate formation. A remarkable difference was observed in the micrographs of N-IgGcompared to its modified counterparts (Fig 9).

AAPH-mediated RBC haemolysis assay
Antioxidant property of proteins is of paramount importance with potential beneficial effects in relieving oxidative stress [40].Owing to typical heart shaped structure and availability of reduced cysteine and methionine residues, the antioxidant property of human serum albumin (HSA) is well established [41].Intrinsic antioxidant capacity of HSA to protect erythrocytes from haemolysis has been evaluated earlier [33]. Using the protocol, we replicated the experiment forIgG (Fig 10). HT 50 (time required for 50% inhibition in haemolysis) for N-IgG, OH • -IgG, MG-IgGand OH • -MG-IgG was found to be 23±3 min, 60±5 min, 64±4 min and 96±6 min respectively. Each value is mean ±SDof 3 independent determinations.   Table 1.

IgG isolation from the sera of diabetes patients
IgG was isolated from 10 serum samples of T2DM patients showing high specificity forOH • -MG-IgG [33]. Elute showed a single symmetrical peak at 280 nm and a single band on 7.5% SDS-PAGE confirming the homogeneity of the IgG (data not shown).

Direct binding ELISA of isolated IgG
Thesaturation in the binding of the patient's IgG to OH • -MG-IgG was evaluated through direct binding ELISA on the microtitre plate coated with OH • -MG-IgG. The saturation was achieved at 40 μg/ml of IgG(data not shown).

Competitive inhibition ELISA of IgG isolated from diabetes patients
Inhibition ELISA of the IgG isolated from the sera of diabetes patients showed OH • -MG-IgGas the most powerful inhibitor followed by MG-IgG, OH • -IgGand N-IgG. The average percent

Gel retardation assay
Gel retardation studies showed a slight mobility shift when N-IgG (used as inhibitor) was incubated with increasing concentration of IgGfrom diabetes patient (serum sample #12). However, incubation of OH • -MG-IgG with IgGfrom diabetes patient resulted in a retarded mobility and proportional increase in the formation of high molecular weight immune complexes. Also, increase in molecular mass prevented the complete penetration of sample into the gel and its retention was observed at the origin of well (Fig 12). The microtitre plates were coated with 10 μg/ml of OH • -MG-IgG doi:10.1371/journal.pone.0169099.t002

Discussion
T2DM is a very complex and multifactorial metabolic disease characterized by insulin resistance and β cell failure that consequently leads tohyperglycemia [42]. The role of hyperglycemia has been the central focus in the understanding of diabetic complications that have potentially affected the life quality and expectancy and has assumed a state of huge financial burden on the global healthcare [43].It has become imperative to understand the different perspective of the causes of diabetes type 2 so as to understand its pathogenesis. Furthermore, the role of ROS,generated during various metabolic events in vivo, is believed to play a key role in pathological processes observed in T2DM [44]. This study has aimed towards understanding the cumulative effect of OH • and MG on the structural integrity of IgG and its effect on the generation of autoimmune response in T2DM.IgG was modified separately by OH • and MG, and the modification of MG-IgG by OH • was also studied.  [45,46].Fluorescence spectroscopy is a valuable parameter to monitor microenvironment around aromatic amino acid residues [47]. Our results indicate quenching in both tyrosine and tryptophan fluorescence.The observed quenching may be explained due to modification/destruction in tyrosine and tryptophan microenviroment [48], in addition to the formation of aggregates. It is the burial of fluorophoric amino acids during aggregation that causes quenching [49][50][51][52].Glycoxidised samples showing highest quenching followed by MG and OH • treated IgG in comparison to N-IgG clearly indicate maximum microenvironment change and aggregate formation in OH • -MG-IgG followed by MG-IgG and OH • -IgG in comparison to the native protein. Another fluorescence spectralcharacterstic of AGE formation was observed in MG-IgG and OH • -MG-IgG when the samples were excited at 370 nm with emission at 435 nm. Increase in ANS fluorescence intensity indicates reorganisation in protein conformation leading to exposure of hydrophobic patches and thus making the sample more accessible to ANS fluorophore [53,54]. Maximum unmasking of hydrophobic residues was observed in glycoxidised samples (OH • -MG-IgG). Glycated (MG-IgG) and hydroxylated (OH • -IgG) controls were comparatively less altered as can be observed from ANS fluorescence spectra. ANS fluorescence was lacking in N-IgGsample. Changes in carbonyl content, a prominent oxidation marker,was also studied. Accumulation of protein carbonyl owing to glycoxidation reactions is associated with various pathologies like diabetes and its associated complications, adult respiratory syndrome, pulmonary fibrosis, rheumatoid arthritis, parkinson's disease, alzheimer's disease, cystic fibrosis etc. The increase is due to oxidation of arginine, lysine, proline, histidine and other amino acid residues [55,56] The two in combination are highly potent to initiate a series of processes that ultimately lead to aggregation of IgG. Accumulation of these adducts/crosslinks/aggregates form the basis of a variety of pathologies such as renal damage, cataract, alzheimer's disease, hypertension and diabetes [63][64][65][66][67]. Several proteins like fibrinogen, hemoglobin, LDL, collagen, serum albumin, β-2 microglobulin have been reported to form aggregates post modification [68][69][70]. Antioxidant activity of non-enzyme proteins is widely known. The activity is imparted by amino acids such as cysteine, tryptophan, histidine, methionine and tyrosine. It has been reported that blood proteins contribute 10-50% of peroxyl radical scavenging activity of plasma [40,[71][72][73]. Effect of modifications on the antioxidant property of IgG was checked by free radical induced haemolysis assay. OH • -MG-IgG showed significantly reduced antioxidant activity followed by glycated and hydroxylated samples.N-IgG protected AAPH induced RBC haemolysis to the maximum extent and thus exhibited highest antioxidant activity. The decreased antioxidant activity in the modified IgG samples is due to ligand induced modification of amino acid side chains. Similar results have been reported earlier also [33]. In order to evaluate the role of structural modifications of IgG on its immunogenicity, we studied the generation of neo-epitopes upon modification of IgG,that could potentially lead to an aggressive autoimmune response in T2DM. The results indicate highly immunogenic nature of OH • -MG-IgG compared to that of MG-IgG, OH • -IgG andN-IgG. The competitive inhibition data showing substantially higher inhibition of diabetes auto-antibodies activity by OH • -MG-IgGexplains high specificity of these antibodies towards the modified epitopes on IgG, compared to that of glycated, hydroxylated andN-IgGsamples. The higher recognition of OH • -MG-IgG by the autoantibodiesfrom diabetes patients point towards epitope sharing between in vitro modified IgG and the the immunogenic epitopes that appear in diabetes patients. The solid phase immunoassay results were further confirmed by gel shift assay wherein incubation of OH • -MG-IgG, with increasing concentrations of the antibodies from diabetes patients, resulted in a retarded mobility and proportional increase in the formation of high molecular weight immune complexes. The results indicate the generation of auto-antibodies in diabetes patients against the modified epitopes of IgG.

Conclusion
From the above results we conclude that the damaging potential of glycoxidation is quite high as compared to glycation or oxidation alone. Glycoxidation is the result of glycation and generation of ROS (i.e. oxidative stress); both factors are dominant in diabetes. Appreciable recognition of the glycoxidatively modified epitopes by the antibodies of diabetes patients may be used as a parameter of disease progression.