37271726 1 Comprehensive multi-omics study of the molecular perturbations 1 induced by simulated diabetes on coronary artery endothelial cells 2 3 Aldo Moreno-Ulloa1,2*, Hilda Carolina Delgado-De la Herrán1,3, Carolina Álvarez-Delgado3, 4 Omar Mendoza-Porras4, Rommel A. Carballo-Castañeda1 and Francisco Villarreal5,6 5 6 1MS2 laboratory, Biomedical Innovation Department, Center for Scientific Research and Higher 7 Education of Ensenada (CICESE), Baja California, México 8 2Specialized Laboratory in Metabolomics and Proteomics (MetPro), CICESE, México 9 3Mitochondrial Biology Laboratory, Biomedical Innovation Department, Center for Scientific 10 Research and Higher Education of Ensenada (CICESE), Baja California, México 11 4CSIRO Livestock and Aquaculture, Queensland Bioscience Precinct, 306 Carmody Rd, St 12 Lucia, QLD, Australia 13 5School of Medicine, University of California, San Diego, CA, USA 14 6San Diego VA Healthcare System 15 16 17 18 * To whom correspondence should be addressed: Biomedical Innovation Department, CICESE 19 Carretera Ensenada-Tijuana No. 3918, Zona Playitas, CP. 22860, Ensenada, B.C. Mexico, 20 Phone: +52(646)175-05-00 ext. 2721, E-mail: amoreno@cicese.mx 21 22 23 24 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 2 Abstract 25 Coronary artery endothelial cells (CAEC) exert an important role in the development of 26 cardiovascular disease. Dysfunction of CAEC is associated with cardiovascular disease in 27 subjects with type 2 diabetes mellitus (T2DM). However, comprehensive studies of the effects 28 that a diabetic environment exerts on this cellular type scarce. The present study characterized 29 the molecular perturbations occurring on cultured bovine CAEC subjected to a prolonged diabetic 30 environment (high glucose [HG] and high insulin [HI]). Changes at the metabolite and peptide 31 level were assessed by untargeted metabolomics and chemoinformatics, and the results were 32 integrated with proteomics data using published SWATH-based proteomics on the same in vitro 33 model. Our findings were consistent with reports on other endothelial cell types, but also identified 34 novel signatures of DNA/RNA, aminoacid, peptide, and lipid metabolism in cells under a diabetic 35 environment. Manual data inspection revealed disturbances on tryptophan catabolism and 36 biosynthesis of phenylalanine-based, glutathione-based, and proline-based peptide metabolites. 37 Fluorescence microscopy detected an increase in binucleation in cells under treatment that also 38 occurred when human CAEC were used. This multi-omics study identified particular molecular 39 perturbations in an induced diabetic environment that could help unravel the mechanisms 40 underlying the development of cardiovascular disease in subjects with T2DM. 41 42 43 Keywords: SWATH-Proteomics; Metabolomics; Type 2 Diabetes Mellitus; Endothelial cells; 44 Feature-Based Molecular Networking 45 46 47 48 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 3 1. Introduction 49 Damage to coronary artery endothelial cells (CAEC) leads to coronary endothelial dysfunction, 50 which is associated with the development of cardiac pathologies in subjects with and without 51 coronary atherosclerosis (1). Subjects with type 2 diabetes mellitus (T2DM) are particularly at 52 increased risk of myocardial infarction (2) and coronary endothelial dysfunction has been 53 implicated in the prognosis (3). A high-glucose (HG) environment —hallmark of T2DM— leads to 54 nitric oxide signaling, cell cycle (4), apoptosis (5), angiogenesis (6), and DNA structure impairment 55 (7). However, given the intrinsic heterogeneity of the endothelium, the molecular perturbations 56 caused by HG vary accordingly with the type of studied endothelial cells (8, 9). For instance, 57 human microvascular endothelial cells showed increased gene expression of endothelial nitric 58 oxide synthase, superoxide dismutase 1, glutathione peroxidase 1, thioredoxin reductase 1 and 59 2 compared to the regulation observed in human umbilical vein endothelial cells (HUVEC) when 60 cultured in HG for 24 h. Furthermore, the response of endothelial cells to HG is influenced by the 61 duration of exposure (10, 11) as demonstrated in bovine aortic and human microvascular 62 endothelial cells where cell proliferation and apoptosis were higher at <48 h compared to 8 weeks 63 of exposure (10). In another example of time-dependent response, increased apoptosis (derived 64 from DNA fragmentation) and tumor necrosis factor alpha protein levels were reported in human 65 coronary artery endothelial cells (HCAEC) after only 24 h of incubation with HG (5). Hence, the 66 molecular response to HG cannot be generalized among endothelial cell types. Previously we 67 reported impaired mitochondrial function/structure and nitric oxide signaling in HG treated HCAEC 68 for 48 h (12). However, a 72 h study documented an increased in pro-inflammatory cytokines (13) 69 and oxidative stress in HCAEC (14). The long-term (>72 h) effect of HG in CAEC has not been 70 as extensively documented compared to other endothelial cell types. Characterizing the effect of 71 HG on CAEC may allow us to identify key signaling pathways (or specific biomolecules) 72 associated with the development of endothelial dysfunction and cardiac pathologies. 73 Here, liquid chromatography coupled to mass spectrometry (LC-MS2)-based untargeted 74 metabolomics and SWATH-based quantitative proteomics data, as well as bio- and chemo-75 informatics were used to characterize the molecular perturbations occurring in Bovine Coronary 76 Artery Endothelial Cells (BCAEC) under a prolonged diabetic environment. 77 78 2. Methods 79 2.1 Chemical and reagents 80 Recombinant human insulin was purchased from Sigma Aldrich (St. Louis, MO, USA). Antibiotic-81 antimitotic solution, trypsin-EDTA solution 0.25%, Hank’s Balanced Salt Solution (HBSS) without 82 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 4 phenol red, Dulbecco’s Modified Eagle’s Media (DMEM) with glutamine, Fetal Bovine Serum 83 (FBS), Hoechst 33258, Pentahydrate (bis-Benzimide)-FluoroPure™, and methanol-free 84 formaldehyde (16% solution) were obtained from Thermo Fisher Scientific (Waltham, MA, USA). 85 Methanol, Acetonitrile, and water were Optima™ LC-MS Grade and obtained from Fisher 86 Scientific (Hampton, NH, USA). Ethanol LiChrosolv® Grade was obtained from Merck KGaA 87 (Darmstadt, Germany). Rabbit anti-Von Willebrand factor (vWf) antibody and goat anti-rabbit IgG 88 conjugated to Alexa Fluor 488 were obtained from Abcam (Cambridge, MA, USA). 89 90 2.2 Cell culture 91 BCAEC were purchased from Cell applications, Inc. (San Diego, CA, USA) and grown as 92 previously described (15). In brief, cells were grown with DMEM (5.5 mmol/L glucose, 93 supplemented with 10% FBS and 1% antibiotic-antimitotic solution) at 37 oC in an incubator with 94 a humidified atmosphere of 5 % CO2. Before experiments, cells were switched to DMEM with 1% 95 FBS for 12 h to maintain the cells under a quiescent state. The model to simulate diabetes is 96 described in (15) (Figure 1). Endothelial cells were cultured for 12 days to determine the chronic 97 molecular perturbations caused by simulated diabetes and to avoid the early (within 48 h) cell 98 proliferation effects caused by HG (10, 16). In brief, cells were first treated with 100 nmol/L insulin 99 (high-insulin, HI) in normal glucose (NG, 5.5 mmol/L in DMEM) for 3 days (17) and then 100 maintained in high-glucose (HG, 20 mmol/L in DMEM) and constant HI for 9 days. This sequential 101 scheme tried to mimic the pathophysiological conditions that occur in T2DM patients, wherein 102 hyperinsulinemia precedes hyperglycemia (18). Cells were used at passages between 6 to 12. 103 The control group did not receive HI nor HG treatment. For selected experiments (binucleation 104 analysis), HCAEC (55 years old Caucasian male, history of T2DM for >5 years) were purchased 105 from Cell Applications, Inc. and subjected to the same conditions as BCAEC but using MesoEndo 106 Growth Medium (Cell Applications, Inc.) to induce proliferation. For simulated diabetes, HCAEC 107 were treated with HI and HG as with BCAEC but, MesoEndo Growht Medium was used instead. 108 For consistency, the group that underwent simulated diabetes (HG + HI) will be referred to as the 109 “experimental group”. All experiments were carried out in triplicate. 110 111 2.3 Immunofluorescence 112 As previously described (15), 100,000 cells per well were seeded onto 12-well plates (Corning® 113 CellBIND®) and exposed to simulated diabetes. Thereafter, BCAEC and HCAEC were washed 114 with PBS to remove dead cells and debris. Cells were fixed, permeabilized, and blocked as 115 described before (19). Cells were then incubated with a polyclonal antibody against the vWf 116 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 5 (1:400, 3% BSA in PBS) overnight at 4oC and thereafter washed 3x with PBS. Alexa Fluor 488-117 labeled anti-rabbit (1:400 in PBS) was then used as a secondary antibody for 1 h at RT and 118 washed 3x with PBS. As a negative control, cells were incubated only with secondary antibody to 119 assess for non-specific binding. Cell nuclei were stained with Hoechst 33258 (2 µg/ml in HBSS) 120 for 30 min and washed 3x with PBS. Fluorescent images were taken in at least three random 121 fields per condition using an EVOS® FLoid® Cell Imaging Station with a fixed 20x air objective. 122 Image analysis was performed through ImageJ software (version 2.0.0). 123 124 2.4 Metabolite extraction 125 Cells were seeded at 300,000 cells per well in 6-well plates (Corning® CellBIND®) and treated as 126 above. After HG and HI conditions, metabolites were extracted following a published protocol for 127 adherent cells with some modifications (20) (Figure 1). In brief, after washing the cells 3 x with 128 PBS, 500 µL of a cold mixture of methanol: ethanol (50:50, v:v) were added to each well, covered 129 with aluminum foil, and incubated at -800C for 4 h. Cells were then scrapped using a lifter (Fisher 130 Scientific, Hampton, NH, USA), and the supernatant was transferred to Eppendorf tubes before 131 centrifugation for 10 min at 14,000 rpm at 40C. The supernatant was transferred to another tube 132 and dried down by SpeedVac™ System (Thermo Fisher Scientific, Waltham, MA, USA). Samples 133 were reconstituted in water/acetonitrile 95:5 v/v with 0.1% formic, centrifuged at 14,000 rpm for 134 10 min at 4o C. The particle free supernatant was recovered for further LC-MS2 analysis. 135 136 2.5 LC-MS2 data acquisition for metabolomics 137 Metabolites were loaded into an Eksigent nanoLCâ 400 system (AB Sciex, Foster City, CA, USA) 138 with a HALO Phenyl-Hexyl column (0.5 x 50 mm, 2.7 µm, 90 Å pore size, Eksigent AB Sciex, 139 Foster City, CA, USA) for data acquisition using the LC-MS parameters previously described with 140 some modifications (21). In brief, the separation of metabolites was performed using gradient 141 elution with 0.1% formic acid in water (A) and 0.1% formic acid in ACN (B) as mobile phases at a 142 constant flow rate of 5 µL/min. The gradient started with 5% B for 1 min followed by a stepped 143 increase to 100%, B over 26 min and held constant for 4 min. Solvent composition was returned 144 to 5% B for 0.1 min. Column re-equilibration was carried out with 5% mobile phase B for 4 minutes. 145 Potential carryover was minimized with a blank run (1 µL buffer A) between sample experimental 146 samples. The eluate from the LC was delivered directly to the TurboV source of a TripleTOF 147 5600+ mass spectrometer (AB Sciex, Foster City, CA, USA) using electrospray ionization (ESI) 148 under positive mode. ESI source conditions were set as follows: IonSpray Voltage Floating, 5500 149 V; Source temperature, 350°C; Curtain gas, 20 psi; Ion source gases 1 and 2 were set to 40 and 150 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 6 45 psi; Declustering potential, 100 V. Data was acquired using information-dependent acquisition 151 (IDA) with high sensitivity mode selected, automatically switching between full-scan MS and 152 MS/MS. The accumulation time for TOF MS was 0.25 s/spectra over the m/z range 100-1500 Da 153 and for MS/MS scan was 0.05 s/spectra over the m/z 50-1500 Da. The IDA settings were as 154 follows charge state +1 to +2, intensity 125 cps, exclude isotopes within 6 Da, mass tolerance 50 155 mDa, and a maximum number of candidate ions 20. Under IDA settings, the ‘‘exclude former 156 target ions’’ was set as 15 s after two occurrences and ‘‘dynamic background subtract’’ was 157 selected. Manufacturer rolling collision energy (CE) option was used based on the size and 158 charge of the precursor ion using formula CE=m/z x 0.0575 + 9. The instrument was automatically 159 calibrated by the batch mode using appropriate positive TOF MS and MS/MS calibration solutions 160 before sample injection and after injection of two samples (<3.5 working hours) to ensure a mass 161 accuracy of <5 ppm for both MS and MS/MS data. Instrument performance was monitored during 162 data acquisition by including QC samples (pooled samples of equal volume) every 4 experimental 163 samples. Data acquisition of experimental samples was also randomized. 164 165 2.6 Metabolomics data processing 166 Mass detection, chromatogram building and deconvolution, isotopic assignment, feature 167 alignment, and gap-filling (to detect features missed during the initial alignment) from LC-MS2 168 datasets was performed using XCMS (https://xcmsonline.scripps.edu) (22) and MZmine (23) 169 software. The XCMS pipeline was used for normalization of feature area and statistical analysis. 170 To identify or annotate the metabolites at the chemical structure and class level, the MS2-171 containing features extracted with MZmine were further analyzed using the Global Natural 172 Products Social Molecular Networking (GNPS) (24), Network Annotation Propagation (NAP) (25) 173 and MS2LDA (26) in silico annotation tools, and Classyfire automated chemical classification (27), 174 as previously described (21) with some modifications. The confidences of such annotations are 175 level 2 (probable structure by library spectrum match) and level 3 (tentative candidates) in 176 agreement with the Metabolomics Standards Initiative (MSI) classification (28). Molecular 177 networking, NAP, and Classyfire outputs were integrated using the MolNetEnhancer workflow 178 (29). Molecular networks were visualized using Cytoscape version 3.8.2 (30). In addition, 179 chemical substructures (co-occurring fragments and neutral losses referred to as “mass2motifs” 180 [M2M]) were recognized using the MS2LDA web pipeline (http://www.ms2lda.org) to further 181 annotate metabolites (level 3, MSI). The detailed processing parameters for XCMS and MZmine 182 pipelines are found in the supporting information. 183 184 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 7 2.7 Peptidomics data processing 185 For peptide identification, raw .wiff and .wiff.scan files (same files used for MZmine and XCMS) 186 from the experimental and control groups were analyzed separately using ProteinPilot software 187 version 4.2 (Ab Sciex, Foster City, CA, USA) with the Paragon algorithm. MS1 and MS2 data were 188 searched against the Bos taurus SwissProt sequence database (6006 reviewed 189 proteins+common protein contaminants, February 2019 release). The parameters input was: 190 sample type, identification; digestion, none; Cys alkylation, none; instrument, TripleTOF 5600; 191 special factors, none; species, Bos taurus; ID focus, biological modifications, and amino acid 192 substitutions; search effort, thorough ID. False discovery rate analysis was also performed. All 193 peptides were exported and those with a >90% confidence were linked to the corresponding 194 feature extracted by the XCMS algorithm using their accurate mass and retention time 195 information. For peptide quantification, we employed the normalized feature abundances (MS1 196 level) generated by XCMS. A significance threshold of p<0.05 (Welch’s t test) was utilized. 197 198 2.8 Proteomics data reprocessing 199 The SWATH-based proteomics data (identifier PXD013643), hosted in ProteomeXchange 200 consortium via PRIDE (31), was reanalyzed with some modifications. The parameters used to 201 build the spectral library remained the same (15), while the parameter for peptides per protein 202 was set to 100 in the software SWATH® Acquisition MicroApp 2.0 in PeakView® version 1.2 (AB 203 Sciex, Foster City, CA, USA). The obtained protein peak areas were exported to Markerview™ 204 version 1.3 (AB Sciex, Foster City, CA, USA) for further data refinement, including assignment of 205 IDs to files and removal of reversed and common contaminants. Peak areas were exported in a 206 .tsv file, and normalized with NormalyzerDE online version 1.3.4 (32). The NormalyzerDE pipeline 207 comprises 8 different normalization methods (Log2, variance stabilizing normalization, total 208 intensity, median, mean, quantile, CycLoess, and robust linear regression). The results of 209 qualitative (MA plots, scatter plots, box plots, density plots) and quantitative (pooled intragroup 210 coefficient of variation [PCV], median absolute deviation [PMAD], estimate of variance [PEV]) 211 parameters were compared between the normalization methods to select the most appropriate. 212 213 2.9 Bioinformatic analysis of proteomics data 214 Proteins that passed the significance threshold were first converted to their corresponding Entrez 215 Gene (GeneID) using https://www.uniprot.org/uploadlists/ and then transformed to their human 216 equivalents using the ortholog conversion feature in https://biodbnet-217 abcc.ncifcrf.gov/db/dbOrtho.php. Bioinformatic analysis was done on OmicsNet website platform 218 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 8 (https://www.omicsnet.ca/) (33, 34). First, a protein-protein interaction (PPI) molecular network 219 (first-order network containing query or seeds molecules and their immediate interacting partners) 220 using STRING PPI database was built (35) and then pathway enrichment analysis was performed 221 using the built-in REACTOME and the Kyoto Encyclopedia of Genes and Genomes (KEGG) 222 databases. To visualize modules (functional units) contained in the molecular network the 223 WalkTrap algorithm (within OmicsNet platform) was employed. Hypergeometric test was used to 224 compute p-values. 225 226 2.10 Integrative analysis of proteomics and metabolomics data 227 The molecular interactions between the proteins and metabolites differentially abundant between 228 HG + HI and NG were determined in OmicsNet (32, 33). The lists of proteins (EntrezGene ID) 229 and metabolites (HMDB ID) were loaded to build a composite network using protein-protein 230 (STRING database selected) and metabolite-protein (KEGG database selected) interaction types. 231 The primary network relied on the metabolite input. Pathway enrichment analysis was performed 232 using the built-in REACTOME and KEGG databases. Hypergeometric test was used to compute 233 p-values. 234 235 2.11 Statistical analysis 236 All experiments were performed in triplicate. Based on the accuracy (determination of real fold-237 changes) of SWATH-based quantification (36), proteins with a fold change ≥ 1.2 or ≤ 1/1.2 and a 238 p-value <0.05 (Welch’s t-test) were considered as differentially abundant between NG and HG + 239 HI conditions. For the metabolomics data, features with a fold change ≥ 1.3 or ≤ 1/1.3 and a p-240 value <0.05 (Welch’s t-test) were considered as differentially abundant. We did not apply multiple-241 test corrections to calculate adjusted p-values, because this process could obscure proteins or 242 metabolites with real changes (true-positives) (37). Instead, the analysis was focused on top-243 enriched signaling pathways (adjusted p-value <0.01) that allowed us to determine a set of 244 interacting proteins and metabolites with relevant biological information and contributes in 245 reducing false positives. For multivariate statistical analysis and heatmap visualization, 246 Metaboanalyst 4.0 (https://www.metaboanalyst.ca) was utilized. Principal component analysis 247 (PCA) was used to assess for sample clustering behavior and inter-group variation. No scaling 248 was used for PCA and heatmap analysis. Software PRISM 6.0 (GraphPad Software, San Diego, 249 CA) was used for the creation of volcano plots and column graphs. 250 251 2.12 Data availability 252 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 9 The raw datasets supporting the metabolomics results are available in the GNPS/MassIVE public 253 repository (38) under the accession number MSV000084307. The specific parameters of the tools 254 employed for metabolite annotation are available on the following links: for classical molecular 255 networking, 256 https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=604b3d077e00430a9bc288eebf154b9b; for 257 FBMN 258 https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=5e2839037969442e868d9df21309d561; for 259 NAP, 260 https://proteomics2.ucsd.edu/ProteoSAFe/status.jsp?task=96cda48c0df64d3398a8f9088907afb261 5; for MS2LDA, http://ms2lda.org/basicviz/summary/1197/ (need to log-in as a registered or guest 262 user); for MolNetEnhancer, 263 https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=de80b9c765e042ffab7767a3101054fd. The 264 quantitative results generated using the XCMS platform can be accessed after logging into the 265 following link https://xcmsonline.scripps.edu and searching for the job number 1395724. SWATH 266 data is accessible on the ProteomeXchange with dataset identifier PXD013643. 267 268 3. Results 269 270 Untargeted metabolomics 271 Overall 5571 features or potential metabolites were detected using XCMS and MZmine, wherein 272 957 (~18%) features were commonly identified in both platforms (Figure 2A). Based on the 273 relative quantification using XCMS, 140 and 82 features were detected with reduced and 274 increased abundances respectively in the experimental group compared to the control group 275 (Figure 2B). The effects of HG and HI in the experimental group are observed by PCA analysis 276 wherein the experimental samples clustered away from the control group (Figure 2C). The 277 consistency of the LC-MS equipment is apparent by the clustering of the QC samples (Figure 278 2C). Further, the heatmap visualization of the top 100-modulated metabolites exhibited the 279 different distribution patterns among groups (Figure 2D). Using the GNPS platform for automatic 280 metabolite annotation, 106 compounds (excluding duplicates and contaminants) were putatively 281 annotated with a level 2 confidence annotation (MS2 spectral match) (Table S1) in agreeance 282 with the MSI classification (28). Some metabolites identified by the GNPS platform could not be 283 quantified because they were not detected by the XCMS algorithm during feature area 284 normalization and quantification. Moreover, GNPS Molecular Networking aligned the MS2-285 containing features (n=1,013) based on their structural similarity, creating 118 independent 286 networks or clusters with at least two connected nodes (Figure 3A). The use of MolNetEnhancer 287 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 10 workflow allowed to putatively identify chemical classes (level 3, MSI) for 56 of the 118 288 independent networks. The top-10 most abundant annotated chemical classes and associated 289 metabolites are shown in Figure 3A. Three-clusters from the network were further analyzed 290 because they contained annotated metabolites by spectral matching, which facilitates the 291 annotation of other cluster’s nodes. Cluster 1 revealed two metabolites linked to the 292 organonitrogen compounds class with reduced abundance in the experimental group (Figure 3B). 293 Library spectral match (level 2, MSI) suggest PC(16:0/18:1(9Z)) and PC(18:0/18:2(9Z,12Z)) as 294 putative candidates, which was supported by MS2LDA phosphocholine-substructure recognition 295 (Figure 3C). In cluster 2, glutathione-based metabolites (MSI level 3) were detected through 296 fragments m/z 308.0925, 233.0575, 179.0475, and 162.0225 retrieved by the M2M_453 297 substructure and associated with glutathione structure using mzCloud in silico predictions (Figure 298 4A). The precursor ion at m/z 713.1472 and glutathione (annotated at level 2, MSI) were detected 299 with increased abundance in the experimental group. MS2LDA visualization, at the M2M level, 300 correlated with the GNPS molecular networking clustering (Figure 4B). In cluster 3, various 301 phenylalanine-based metabolites were putatively annotated aided by MS2LDA substructure 302 recognition (Figure 4C and 4D). Within this cluster, glutamyl-phenylalanine (annotated at level 2, 303 MSI) and the precursor ions at m/z 297.1802 and 487.1548 presented with increased abundance 304 in the experimental vs. control group. On the other hand, various aminoacids were annotated 305 (level 2, MSI) by GNPS spectral matching and manual inspection of data (Table S2). Threonine, 306 valine, proline, leucine, serine, glutamic acid, methionine, and tyrosine presented increased 307 abundance (fold change range 1.3-1.7, p<0.05) in the experimental vs. control group. Particularly, 308 metabolites linked to the catabolism of tryptophan via the serotonin and kynurenine pathway (39) 309 were annotated (level 2, MSI), including melatonin, acetyl serotonin, and kynurenine (Table S1). 310 However, only kynurenine was significantly elevated in the experimental group. The full list of 311 annotated metabolites, differential abundances and another relevant feature information is shown 312 in Table S2. 313 314 Peptidomics 315 Experimental and control datasets were analyzed separately to identify the peptides and their 316 biological modifications. The complete list of peptides identified by ProteinPilot between the 317 experimental and control groups are described in Table S3. Proline oxidation was the most 318 frequent biological modification detected in the experimental group datasets. We identified 8 and 319 12 peptides with a confidence of >90% in the control and experimental group, respectively. 320 Differential abundance of 2 proline-rich peptides was observed in the experimental group 321 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 11 compared to the control group. An additional tripeptide was manually annotated with a LPP 322 sequence (Table S4). 323 324 Proteomics 325 The re-analysis of the SWATH data (PXD013643 dataset) facilitated the identification of 952 326 quantifiable proteins (717 proteins with at least 2 unique peptides, 1% false discovery rate) and 327 no missing values among technical and biological replicates (Table S5). Sample datasets were 328 normalized using 8 different methods to select the most appropriate based on quantitative and 329 qualitative parameters on our dataset. Quantile normalization produced a better qualitative and 330 quantitative profile and was selected to further process our data (Figure S1). PCA analysis of 331 normalized data denoted a clear separation of the groups suggesting overall differences in their 332 proteomes (Figure 5A). Differential abundance analysis revealed 32 and 33 proteins with 333 increased and decreased abundance in the experimental group (Figure 5B). Further, the 334 heatmap visualization of the top 50-modulated proteins exhibited the different distribution patterns 335 among the experimental and control groups (Figure 5C). To obtain a molecular insight we 336 performed a functional enrichment analysis using a network-based approach. First, we created a 337 composite network comprising PPI between the modulated proteins by simulated diabetes (seed 338 proteins) and their immediate interacting partners (highest confidence >0.9) retrieved from 339 STRING Database (incorporated in OmicsNet platform). The principal network using the up-340 modulated proteins consisted of 461 proteins, 709 edges and 18 seed proteins (nodes with blue 341 shadow) and is illustrated in Figure 5D. Eight modules or clusters were generated, that may 342 represent relevant complexes or functional units (40). The 5 most significant (adjusted p-value 343 <0.05) REACTOME and KEGG pathways on the global network are shown in Table 1. Two 344 modules contained multiple seed proteins and were linked to DNA/RNA and protein metabolism 345 pathways using the WalkTrap algorithm (Figure 5D). On the other hand, the principal network 346 using the down-modulated proteins consisted of 488 proteins, 513 edges and 18 seed proteins 347 identified eleven modules wherein one module (with 2 seed proteins) indicated associations with 348 mitochondrial function pathways (Figure 5E). 349 350 Integration of Metabolomics and Proteomics 351 The signaling pathways perturbed by simulated diabetes were identified by a composite network 352 of interacting metabolites and proteins using OmicsNet built-in databases. Figure 6 illustrates the 353 composite bi-layered metabolite-PPI network using the up-modulated molecules (under simulated 354 diabetes) comprised of 9 metabolites (seed metabolites), 177 edges, and 166 proteins (5 seed 355 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 12 proteins). The 10 top-most enriched signaling pathways identified in the composite network are 356 shown in Table 2. The two principal modules highlighted by the WalkTrap algorithm were linked 357 to glutathione and amino acid metabolism. We noted a smaller interaction between Acyl-protein 358 thioesterase 1 (LYPLA1) and a phosphatidylcholine metabolite when simultaneously analyzing 359 up- and down-modulated proteins and metabolites. No significant composite network was 360 identified using the down-modulated proteins and metabolites. 361 362 Cellular morphology 363 To better understand the effects that simulated diabetes exerts on endothelial cells the changes 364 on cellular structure endpoints were evaluated. The endothelial nuclei morphology in the BCAEC 365 control and experimental groups were evaluated using fluorescent-staining and image analysis. 366 We also evaluated the presence of vWF (marker of endothelial cells) in BCAEC and HCAEC, to 367 reveal the cellular boundary and to demonstrate their endothelial phenotype (41). We noted an 368 increase in the percentage of binucleated BCAEC in the experimental group compared to the 369 control group (top panel Figure 7A and 7B). A similar result with larger nuclei, was observed 370 when using HCAEC as a human in vitro model (bottom panel Figure 7A and 7B). Finally, as 371 expected, we observed a typical intracellular localization of vWF and a 100% positivity in 372 endothelial cells. 373 374 4. Discussion 375 This study investigated the molecular perturbations occurring in coronary endothelium cells 376 subjected to prolonged simulated diabetes that facilitated the identification of signaling pathways 377 and specific molecules that could be associated with the development of cardiovascular disease. 378 To achieve this, we employed a MS-based multi-omics approach coupled to fluorescence 379 microscopy to detect structural changes. Endothelial cells cover the inner surface of blood vessels 380 and are distributed across the body. Their functions include: acting as a mechanical barrier 381 between the circulating blood and adjacent tissues as well as modulating multiple functions in 382 distinct organs (42). These regulatory functions vary according to localization and vascular bed-383 origin (43). HG blood levels are detrimental to endothelial cells function in T2DM leading to 384 coronary endothelial dysfunction and development of CVD (44, 45). The molecular effects of HG 385 on endothelial cells have been previously characterized (4, 6, 7, 10, 11); nevertheless, the 386 endothelial cell types used in these studies are not intrinsically involved in CVD. The present study 387 used an in vitro model involving endothelial cells that modulate the heart function, CAEC (46). 388 Our model not only used HG (20 mmol/L) to simulate diabetes (4, 6, 7, 10, 11) but first induced 389 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 13 insulin resistance to mimic the pathophysiological conditions that occur in T2DM wherein 390 hyperinsulinemia precedes hyperglycemia (18). Diabetes was simulated for up to 12 days to 391 mimic chronic HG exposure and to prevent measuring cell proliferation known to occur in early 392 HG (10, 16). Despite a lack of apparent increase in cell proliferation in the experimental group 393 compared to control group after twelve days, an increase in overall protein abundance was 394 detected by Bradford assay (data not shown) and inferred from total ion chromatogram (TIC) of 395 MS (Figure S1A). We suggest that protein synthesis is increased as a consequence of the higher 396 presence of bi-nucleated CAEC (with increased DNA/RNA metabolism) under HG + HI compared 397 to that in the control cohort (Figure 7A and 7B). Previous studies have shown reduced endothelial 398 cell proliferation (mostly in HUVEC) after long-term (7-14 days) HG exposure (4, 11, 47-53), 399 accompanied by an increase in protein synthesis (53). This MS-based methodological pipeline 400 that included appropriate controls during data acquisition (QC) and processing (e.g., 401 normalization, filtering, annotation, dereplication, etc.), allowed the identification of global 402 changes in the metabolome of CAEC under HG + HI. Specifically, increased abundance of valine, 403 leucine, tyrosine, serine, leucine, proline, methionine, and glutamic acid in cells under HG 404 conditions was observed; and this is consistent with reports on human aortic endothelial cells 405 (54). Notably, several clinical studies have established a direct relationship between 406 prevalence/incidence of T2DM and increased levels of valine, leucine and tyrosine in serum and 407 plasma (55-59). Our results support the role of CAEC in contributing to the elevated pool of amino 408 acids seen in circulation under a HG environment. We speculate that increased levels of these 409 amino acids could result from either increased production or reduced degradation as suggested 410 in endothelial cells (immortalized cell line, EA.hy 926) that transition from a glycolytic metabolism 411 towards lipid and amino acid oxidation when challenged by HG (60). Furthermore, evidence of 412 increased tryptophan catabolism was identified through the kynurenine pathway. In this regard, a 413 non-significant decrease of ~ 40% in the abundance of tryptophan was detected. However, a 414 significant increase of ~ 450% in kynurenine (tryptophan’s main metabolite) (61) between the HG 415 + HI group and NG group was also observed, which is a key finding as elevated plasma levels of 416 kynurenine are known to increase CVD risk (62, 63). This novel finding contributes to expanding 417 the understanding of amino acid metabolism in endothelial cells under simulated diabetes. Acetyl 418 serotonin and melatonin which are components of the serotonin pathway that degrades 419 tryptophan (64) were also detected with only minor abundancy increases (20-30%) in the HG + 420 HI group compared to control. Differences in glutathione (cysteine-glutamic acid-glycine, 421 tripeptide) metabolism in CAEC were also found, suggesting an increased response to oxidative 422 stress (65). In line with this observation, previous research reported a glutathione-dependent 423 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 14 reaction to ambient HG in artery-derived endothelial cells (66, 67) but the same could not be 424 observed in vein-derived endothelial cells (68, 69). This emphasizes the different responses to 425 HG among endothelial phenotypes. Here, novel evidence is provided of the up-regulation of 426 glutathione-based metabolites. The composite protein network suggested an increase in 427 glutathione metabolism supported by elevated levels of oxidized glutathione and, one of its 428 synthetic precursors, glutamic acid. At the protein level, peroxiredoxin (PRDX2 and PRDX6) and 429 thioredoxin (TXN2, mitochondrial) showed increased abundances in the experimental group, 430 which are part of the cells natural enzymatic defense against oxidative stress (70). The 431 substructure analysis of metabolomics data facilitated identifying glutamic acid- and 432 phenylalanine-based metabolites, presumably di- or tri-peptides, including the annotated 433 metabolite glutamyl-phenylalanine. Furthermore, the CAEC peptidome analysis suggested an 434 increase in proline-containing peptides. This type of peptide is of particular interest because of 435 their resistance to non-specific proteolytic degradation, body distribution and remarkable 436 biological effects (71-74). Yet, the precise function of such phenylalanine-, glutamine-, and 437 proline-based peptides remains to be characterized in CAEC. We can only speculate that they 438 are the result of a compensatory mechanism to reduce glucose cellular damage. Also, increased 439 protein abundance of core and regulatory subunits from the proteasome complex (PSMA4 and 440 PSMD3) was found in cells under simulated diabetes. This suggests an increased protein 441 degradation and subsequent peptide formation in response to HG. Metabolomic profiling also 442 revealed changes in the lipidome of CAEC challenged with HG + HI, wherein a reduction in 443 phosphatidylcholine (PC) lipids and subsequent increase in phosphocholine were noted. 444 Changes in the phospholipidomic profile of bovine aortic endothelial cells treated with HG for 24 445 h has also been reported in a lipidome study (75). Here, proteomics and metabolomics data were 446 manually integrated and this allowed to determine critical roles for PAFAH1B2 and LYPLA1 in 447 mediating the degradation of PC lipids (Figure 8). PAFAH1B2 was found to be up-regulated in 448 this study and it is known to be associated with inflammation and higher levels of lysoPC (76). As 449 a result, PAFAH1B2 could increase the pool of lysoPC lipids, further exacerbating inflammation 450 in the cardiovascular system (77). On the other hand, LYPLA1 has a lysophospholipase activity 451 that can hydrolyze a range of lysophospholipids, including LysoPC, thereby generating a fatty 452 acid and glycerophosphocholine as products (78). Increased levels of phosphocholine (~ 460%) 453 were detected in HG treated cells compared to control, that could be associated with the 454 degradation of LysoPC lipids. It should be noted that the use of pathways databases such as 455 KEGG and REACTOME possess some limitations when dealing with lipid metabolites because 456 its chemical diversity is not well annotated/defined within the databases. For example, KEGG 457 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 15 provides a chemical class identifier instead of individual identity to lipids, constricting their 458 biological importance (79). Thus, based on our manual inspection of the metabolomics-459 proteomics data and in line with the evidence, we suggest that simulated diabetes evokes 460 inflammation on BCAEC and that PAFAH1B2 and LYPLA1 play a role in modulating such 461 process. 462 Previously, we reported the multinucleation of CAEC cultured under simulated diabetes (15). This 463 type of cell possesses ≥2 nuclei. Here, we replicated our previous findings of increased 464 binucleation in BCAEC. The same outcome was obtained when using HCAEC as a human in vitro 465 model (Figure 7A and 7B), validating the binucleation process in other CAEC. After refinement 466 of LC-MS2 data and bioinformatics re-processing of published SWATH-based datasets of BCAEC 467 under simulated diabetes (15), molecular signatures and pathways that could be linked to the 468 binucleation process were found (Figure 8). For instance, we noted an increased abundance of 469 proteins, under simulated diabetes, with reported nuclei localization and linked to DNA 470 metabolism, including ribosomal proteins RPS7, RPS13, and RPL9 (80). Further, we observed 471 an increased abundance of proteasome proteins, PSMA4 and PSMD3, which are linked to protein 472 metabolism (81). Hence, we infer that the CAEC binucleation occurs as a compensatory 473 mechanism to increase the cell capacity to metabolize the excess of ambient glucose by 474 increasing the cell metabolic machinery (transcription/translation processes). Although an 475 increase in cell proliferation could boost a coordinated increase of ribosomal and proteasome 476 proteins, we do not believe this is the case here, as mentioned before. After 4-5 days of simulated 477 diabetes, cells occupied 100% of the well's plate surface, thereby impeding to harbor more cells 478 because endothelial cells grow as a monolayer. This is consistent with findings stating that when 479 endothelial cells become highly confluent, they stop growing due to cell-cell contact, even in the 480 presence of growth factors (82). In support of this, up-stream (CTGF and CD62) (83, 84) (Table 481 S5) and down-stream proteins (FABP4) (85) (Table S5) involved in angiogenesis and proliferation 482 were down-regulated by simulated diabetes. Importantly, there is evidence (not in endothelial 483 cells) of cellular processes contributing to the stimulation of cellular binucleation without increases 484 in cell proliferation, including cellular enhancement of antimicrobial defenses (86), senescence 485 (87), and malignancy (88). Various mechanisms have been linked to the binucleation process, 486 such as cytokinesis failure, cellular fusion, mitotic slippage, and endoreduplication (89). The 487 elucidation of the exact molecular mechanisms leading to the binucleation process of CAEC is 488 beyond the scope of our study. 489 In conclusion, this study applied an integrated multi-omics and bioinformatics/chemoinformatics 490 approach to characterize the molecular perturbations that simulated diabetes exerts on CAEC. 491 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 16 We confirmed several independent studies that reported alterations at protein and metabolite 492 levels in endothelial cells of different sources than coronary vessels. Metabolomics, identified 493 alterations in amino acid, peptide, and phospholipid metabolism. Notably, the chemoinformatic 494 analysis identified unreported alterations of phenylalanine-, glutathione-, and proline-based 495 peptides on coronary endothelium under simulated diabetes. Proteomics provided evidence of 496 reduced mitochondrial mass and angiogenesis. The integration of proteomics and metabolomics 497 identified increased glutamic acid metabolism and suggested that the antioxidant enzymes are 498 involved in protecting the cells from oxidative stress. Fluorescence microscopy reported the 499 appearance of non-proliferative binucleated CAEC cells as a mean to metabolize the excess of 500 ambient glucose. Overall, our study improved the understanding of the molecular disturbances 501 caused by simulated diabetes that could mediate CAEC dysfunction and may be relevant in the 502 context of CVD in subjects with T2DM. 503 504 505 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 17 5. Acknowledgements 506 This work was derived in part from the Thesis Project of H.C.D.H. at the Posgrado en 507 Ciencias de la Vida, CICESE. We thank Alan G. Hernández-Melgar for his invaluable 508 technical assistance with the NormalyzerDE software. 509 510 511 512 513 514 515 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 18 6. Funding 516 Part of this work was supported by CICESE (Grant No. 685109 to AMU and Internal 517 Project No. 685-110 from CAD), NIH R01 DK98717 (to FV), and VA Merit-I01 BX3230 (to 518 FV). 519 520 521 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 19 7. Conflict of interest 522 Dr. Villarreal is a co-founder and stockholder of Cardero Therapeutics, Inc. 523 524 525 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 20 8. Author contributions 526 A.M.U. contributed to the study conception and design, data acquisition, formal analysis, 527 methodology, project administration, and funding acquisition. H.C.D.H., L.D.M, and R.A.C.C. 528 contributed to the data acquisition, formal analysis and interpretation of some experiments. 529 C.A.D., and F.V. contributed to funding acquisition and resources. O.M.P contributed to data 530 interpretation and critical revision of manuscript. All authors contributed to the drafting, revising, 531 and approval of the final version of the manuscript. 532 533 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 21 References 534 535 1. Halcox, J. P.; Schenke, W. H.; Zalos, G.; Mincemoyer, R.; Prasad, A.; Waclawiw, M. A.; 536 Nour, K. R.; Quyyumi, A. A., Prognostic value of coronary vascular endothelial dysfunction. 537 Circulation 2002, 106, (6), 653-8. 538 2. Lind, M.; Wedel, H.; Rosengren, A., Excess Mortality among Persons with Type 2 539 Diabetes. N Engl J Med 2016, 374, (8), 788-9. 540 3. Gutierrez, E.; Flammer, A. J.; Lerman, L. O.; Elizaga, J.; Lerman, A.; Fernandez-Aviles, 541 F., Endothelial dysfunction over the course of coronary artery disease. Eur Heart J 2013, 34, 542 (41), 3175-81. 543 4. Lorenzi, M.; Cagliero, E.; Toledo, S., Glucose toxicity for human endothelial cells in 544 culture. Delayed replication, disturbed cell cycle, and accelerated death. Diabetes 1985, 34, (7), 545 621-7. 546 5. Kageyama, S.; Yokoo, H.; Tomita, K.; Kageyama-Yahara, N.; Uchimido, R.; Matsuda, 547 N.; Yamamoto, S.; Hattori, Y., High glucose-induced apoptosis in human coronary artery 548 endothelial cells involves up-regulation of death receptors. Cardiovasc Diabetol 2011, 10, 73. 549 6. Dubois, S.; Madec, A. M.; Mesnier, A.; Armanet, M.; Chikh, K.; Berney, T.; Thivolet, C., 550 Glucose inhibits angiogenesis of isolated human pancreatic islets. J Mol Endocrinol 2010, 45, 551 (2), 99-105. 552 7. Lorenzi, M.; Montisano, D. F.; Toledo, S.; Barrieux, A., High glucose induces DNA 553 damage in cultured human endothelial cells. J Clin Invest 1986, 77, (1), 322-5. 554 8. Patel, H.; Chen, J.; Das, K. C.; Kavdia, M., Hyperglycemia induces differential change in 555 oxidative stress at gene expression and functional levels in HUVEC and HMVEC. Cardiovasc 556 Diabetol 2013, 12, 142. 557 9. Pala, L.; Pezzatini, A.; Dicembrini, I.; Ciani, S.; Gelmini, S.; Vannelli, B. G.; Cresci, B.; 558 Mannucci, E.; Rotella, C. M., Different modulation of dipeptidyl peptidase-4 activity between 559 microvascular and macrovascular human endothelial cells. Acta Diabetol 2012, 49 Suppl 1, 560 S59-63. 561 10. Esposito, C.; Fasoli, G.; Plati, A. R.; Bellotti, N.; Conte, M. M.; Cornacchia, F.; Foschi, A.; 562 Mazzullo, T.; Semeraro, L.; Dal Canton, A., Long-term exposure to high glucose up-regulates 563 VCAM-induced endothelial cell adhesiveness to PBMC. Kidney Int 2001, 59, (5), 1842-9. 564 11. Baumgartner-Parzer, S. M.; Wagner, L.; Pettermann, M.; Grillari, J.; Gessl, A.; 565 Waldhausl, W., High-glucose--triggered apoptosis in cultured endothelial cells. Diabetes 1995, 566 44, (11), 1323-7. 567 12. Ramirez-Sanchez, I.; Rodriguez, A.; Moreno-Ulloa, A.; Ceballos, G.; Villarreal, F., (-)-568 Epicatechin-induced recovery of mitochondria from simulated diabetes: Potential role of 569 endothelial nitric oxide synthase. Diab Vasc Dis Res 2016, 13, (3), 201-10. 570 13. Liu, T.; Gong, J.; Chen, Y.; Jiang, S., Periodic vs constant high glucose in inducing pro-571 inflammatory cytokine expression in human coronary artery endothelial cells. Inflamm Res 2013, 572 62, (7), 697-701. 573 14. Liu, T. S.; Pei, Y. H.; Peng, Y. P.; Chen, J.; Jiang, S. S.; Gong, J. B., Oscillating high 574 glucose enhances oxidative stress and apoptosis in human coronary artery endothelial cells. J 575 Endocrinol Invest 2014, 37, (7), 645-51. 576 15. Hilda Carolina Delgado De la Herrán, L. D.-M., Carolina Álvarez-Delgado, Francisco 577 Villarreal, Aldo Moreno-Ulloa, Formation of multinucleated variant endothelial cells with altered 578 mitochondrial function in cultured coronary endothelium under simulated diabetes. bioRxiv 579 2019. 580 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 22 16. Li, X. X.; Liu, Y. M.; Li, Y. J.; Xie, N.; Yan, Y. F.; Chi, Y. L.; Zhou, L.; Xie, S. Y.; Wang, P. 581 Y., High glucose concentration induces endothelial cell proliferation by regulating cyclin-D2-582 related miR-98. J Cell Mol Med 2016, 20, (6), 1159-69. 583 17. Madonna, R.; De Caterina, R., Prolonged exposure to high insulin impairs the 584 endothelial PI3-kinase/Akt/nitric oxide signalling. Thromb Haemost 2009, 101, (2), 345-50. 585 18. Zaccardi, F.; Webb, D. R.; Yates, T.; Davies, M. J., Pathophysiology of type 1 and type 2 586 diabetes mellitus: a 90-year perspective. Postgrad Med J 2016, 92, (1084), 63-9. 587 19. Moreno-Ulloa, A.; Miranda-Cervantes, A.; Licea-Navarro, A.; Mansour, C.; Beltran-588 Partida, E.; Donis-Maturano, L.; Delgado De la Herran, H. C.; Villarreal, F.; Alvarez-Delgado, C., 589 (-)-Epicatechin stimulates mitochondrial biogenesis and cell growth in C2C12 myotubes via the 590 G-protein coupled estrogen receptor. Eur J Pharmacol 2018, 822, 95-107. 591 20. Kirkwood, J. S.; Maier, C.; Stevens, J. F., Simultaneous, untargeted metabolic profiling 592 of polar and nonpolar metabolites by LC-Q-TOF mass spectrometry. Curr Protoc Toxicol 2013, 593 Chapter 4, Unit4 39. 594 21. Moreno-Ulloa, A.; Sicairos Diaz, V.; Tejeda-Mora, J. A.; Macias Contreras, M. I.; Castillo, 595 F. D.; Guerrero, A.; Gonzalez Sanchez, R.; Mendoza-Porras, O.; Vazquez Duhalt, R.; Licea-596 Navarro, A., Chemical Profiling Provides Insights into the Metabolic Machinery of Hydrocarbon-597 Degrading Deep-Sea Microbes. mSystems 2020, 5, (6). 598 22. Gowda, H.; Ivanisevic, J.; Johnson, C. H.; Kurczy, M. E.; Benton, H. P.; Rinehart, D.; 599 Nguyen, T.; Ray, J.; Kuehl, J.; Arevalo, B.; Westenskow, P. D.; Wang, J.; Arkin, A. P.; 600 Deutschbauer, A. M.; Patti, G. J.; Siuzdak, G., Interactive XCMS Online: simplifying advanced 601 metabolomic data processing and subsequent statistical analyses. Anal Chem 2014, 86, (14), 602 6931-9. 603 23. Pluskal, T.; Castillo, S.; Villar-Briones, A.; Oresic, M., MZmine 2: modular framework for 604 processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC 605 Bioinformatics 2010, 11, 395. 606 24. Aron, A. T.; Gentry, E. C.; McPhail, K. L.; Nothias, L. F.; Nothias-Esposito, M.; 607 Bouslimani, A.; Petras, D.; Gauglitz, J. M.; Sikora, N.; Vargas, F.; van der Hooft, J. J. J.; Ernst, 608 M.; Kang, K. B.; Aceves, C. M.; Caraballo-Rodriguez, A. M.; Koester, I.; Weldon, K. C.; 609 Bertrand, S.; Roullier, C.; Sun, K.; Tehan, R. M.; Boya, P. C.; Christian, M. H.; Gutierrez, M.; 610 Ulloa, A. M.; Tejeda Mora, J. A.; Mojica-Flores, R.; Lakey-Beitia, J.; Vasquez-Chaves, V.; 611 Zhang, Y.; Calderon, A. I.; Tayler, N.; Keyzers, R. A.; Tugizimana, F.; Ndlovu, N.; Aksenov, A. 612 A.; Jarmusch, A. K.; Schmid, R.; Truman, A. W.; Bandeira, N.; Wang, M.; Dorrestein, P. C., 613 Reproducible molecular networking of untargeted mass spectrometry data using GNPS. Nat 614 Protoc 2020. 615 25. da Silva, R. R.; Wang, M.; Nothias, L. F.; van der Hooft, J. J. J.; Caraballo-Rodriguez, A. 616 M.; Fox, E.; Balunas, M. J.; Klassen, J. L.; Lopes, N. P.; Dorrestein, P. C., Propagating 617 annotations of molecular networks using in silico fragmentation. PLoS Comput Biol 2018, 14, 618 (4), e1006089. 619 26. van der Hooft, J. J.; Wandy, J.; Barrett, M. P.; Burgess, K. E.; Rogers, S., Topic 620 modeling for untargeted substructure exploration in metabolomics. Proc Natl Acad Sci U S A 621 2016, 113, (48), 13738-13743. 622 27. Djoumbou Feunang, Y.; Eisner, R.; Knox, C.; Chepelev, L.; Hastings, J.; Owen, G.; 623 Fahy, E.; Steinbeck, C.; Subramanian, S.; Bolton, E.; Greiner, R.; Wishart, D. S., ClassyFire: 624 automated chemical classification with a comprehensive, computable taxonomy. J Cheminform 625 2016, 8, 61. 626 28. Schymanski, E. L.; Jeon, J.; Gulde, R.; Fenner, K.; Ruff, M.; Singer, H. P.; Hollender, J., 627 Identifying small molecules via high resolution mass spectrometry: communicating confidence. 628 Environ Sci Technol 2014, 48, (4), 2097-8. 629 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 23 29. Ernst, M.; Kang, K. B.; Caraballo-Rodriguez, A. M.; Nothias, L. F.; Wandy, J.; Chen, C.; 630 Wang, M.; Rogers, S.; Medema, M. H.; Dorrestein, P. C.; van der Hooft, J. J. J., 631 MolNetEnhancer: Enhanced Molecular Networks by Integrating Metabolome Mining and 632 Annotation Tools. Metabolites 2019, 9, (7). 633 30. Shannon, P.; Markiel, A.; Ozier, O.; Baliga, N. S.; Wang, J. T.; Ramage, D.; Amin, N.; 634 Schwikowski, B.; Ideker, T., Cytoscape: a software environment for integrated models of 635 biomolecular interaction networks. Genome Res 2003, 13, (11), 2498-504. 636 31. Perez-Riverol, Y.; Csordas, A.; Bai, J.; Bernal-Llinares, M.; Hewapathirana, S.; Kundu, 637 D. J.; Inuganti, A.; Griss, J.; Mayer, G.; Eisenacher, M.; Perez, E.; Uszkoreit, J.; Pfeuffer, J.; 638 Sachsenberg, T.; Yilmaz, S.; Tiwary, S.; Cox, J.; Audain, E.; Walzer, M.; Jarnuczak, A. F.; 639 Ternent, T.; Brazma, A.; Vizcaino, J. A., The PRIDE database and related tools and resources 640 in 2019: improving support for quantification data. Nucleic Acids Res 2019, 47, (D1), D442-641 D450. 642 32. Willforss, J.; Chawade, A.; Levander, F., NormalyzerDE: Online Tool for Improved 643 Normalization of Omics Expression Data and High-Sensitivity Differential Expression Analysis. J 644 Proteome Res 2019, 18, (2), 732-740. 645 33. Zhou, G.; Xia, J., Using OmicsNet for Network Integration and 3D Visualization. Curr 646 Protoc Bioinformatics 2019, 65, (1), e69. 647 34. Zhou, G.; Xia, J., OmicsNet: a web-based tool for creation and visual analysis of 648 biological networks in 3D space. Nucleic Acids Res 2018, 46, (W1), W514-W522. 649 35. Szklarczyk, D.; Franceschini, A.; Wyder, S.; Forslund, K.; Heller, D.; Huerta-Cepas, J.; 650 Simonovic, M.; Roth, A.; Santos, A.; Tsafou, K. P.; Kuhn, M.; Bork, P.; Jensen, L. J.; von 651 Mering, C., STRING v10: protein-protein interaction networks, integrated over the tree of life. 652 Nucleic Acids Res 2015, 43, (Database issue), D447-52. 653 36. Muntel, J.; Kirkpatrick, J.; Bruderer, R.; Huang, T.; Vitek, O.; Ori, A.; Reiter, L., 654 Comparison of Protein Quantification in a Complex Background by DIA and TMT Workflows 655 with Fixed Instrument Time. J Proteome Res 2019, 18, (3), 1340-1351. 656 37. Pascovici, D.; Handler, D. C.; Wu, J. X.; Haynes, P. A., Multiple testing corrections in 657 quantitative proteomics: A useful but blunt tool. Proteomics 2016, 16, (18), 2448-53. 658 38. Wang, M.; Carver, J. J.; Phelan, V. V.; Sanchez, L. M.; Garg, N.; Peng, Y.; Nguyen, D. 659 D.; Watrous, J.; Kapono, C. A.; Luzzatto-Knaan, T.; Porto, C.; Bouslimani, A.; Melnik, A. V.; 660 Meehan, M. J.; Liu, W. T.; Crusemann, M.; Boudreau, P. D.; Esquenazi, E.; Sandoval-Calderon, 661 M.; Kersten, R. D.; Pace, L. A.; Quinn, R. A.; Duncan, K. R.; Hsu, C. C.; Floros, D. J.; Gavilan, 662 R. G.; Kleigrewe, K.; Northen, T.; Dutton, R. J.; Parrot, D.; Carlson, E. E.; Aigle, B.; Michelsen, 663 C. F.; Jelsbak, L.; Sohlenkamp, C.; Pevzner, P.; Edlund, A.; McLean, J.; Piel, J.; Murphy, B. T.; 664 Gerwick, L.; Liaw, C. C.; Yang, Y. L.; Humpf, H. U.; Maansson, M.; Keyzers, R. A.; Sims, A. C.; 665 Johnson, A. R.; Sidebottom, A. M.; Sedio, B. E.; Klitgaard, A.; Larson, C. B.; P, C. A. B.; Torres-666 Mendoza, D.; Gonzalez, D. J.; Silva, D. B.; Marques, L. M.; Demarque, D. P.; Pociute, E.; 667 O'Neill, E. C.; Briand, E.; Helfrich, E. J. N.; Granatosky, E. A.; Glukhov, E.; Ryffel, F.; Houson, 668 H.; Mohimani, H.; Kharbush, J. J.; Zeng, Y.; Vorholt, J. A.; Kurita, K. L.; Charusanti, P.; McPhail, 669 K. L.; Nielsen, K. F.; Vuong, L.; Elfeki, M.; Traxler, M. F.; Engene, N.; Koyama, N.; Vining, O. B.; 670 Baric, R.; Silva, R. R.; Mascuch, S. J.; Tomasi, S.; Jenkins, S.; Macherla, V.; Hoffman, T.; 671 Agarwal, V.; Williams, P. G.; Dai, J.; Neupane, R.; Gurr, J.; Rodriguez, A. M. C.; Lamsa, A.; 672 Zhang, C.; Dorrestein, K.; Duggan, B. M.; Almaliti, J.; Allard, P. M.; Phapale, P.; Nothias, L. F.; 673 Alexandrov, T.; Litaudon, M.; Wolfender, J. L.; Kyle, J. E.; Metz, T. O.; Peryea, T.; Nguyen, D. 674 T.; VanLeer, D.; Shinn, P.; Jadhav, A.; Muller, R.; Waters, K. M.; Shi, W.; Liu, X.; Zhang, L.; 675 Knight, R.; Jensen, P. R.; Palsson, B. O.; Pogliano, K.; Linington, R. G.; Gutierrez, M.; Lopes, N. 676 P.; Gerwick, W. H.; Moore, B. S.; Dorrestein, P. C.; Bandeira, N., Sharing and community 677 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 24 curation of mass spectrometry data with Global Natural Products Social Molecular Networking. 678 Nat Biotechnol 2016, 34, (8), 828-837. 679 39. Bender, D. A., Biochemistry of tryptophan in health and disease. Mol Aspects Med 1983, 680 6, (2), 101-97. 681 40. Poyatos, J. F.; Hurst, L. D., How biologically relevant are interaction-based modules in 682 protein networks? Genome Biol 2004, 5, (11), R93. 683 41. Muller, A. M.; Hermanns, M. I.; Skrzynski, C.; Nesslinger, M.; Muller, K. M.; Kirkpatrick, 684 C. J., Expression of the endothelial markers PECAM-1, vWf, and CD34 in vivo and in vitro. Exp 685 Mol Pathol 2002, 72, (3), 221-9. 686 42. Aird, W. C., Phenotypic heterogeneity of the endothelium: II. Representative vascular 687 beds. Circ Res 2007, 100, (2), 174-90. 688 43. Aird, W. C., Endothelial cell heterogeneity. Cold Spring Harb Perspect Med 2012, 2, (1), 689 a006429. 690 44. Widlansky, M. E.; Gokce, N.; Keaney, J. F., Jr.; Vita, J. A., The clinical implications of 691 endothelial dysfunction. J Am Coll Cardiol 2003, 42, (7), 1149-60. 692 45. Ganz, P.; Vita, J. A., Testing endothelial vasomotor function: nitric oxide, a multipotent 693 molecule. Circulation 2003, 108, (17), 2049-53. 694 46. Paulus, W. J.; Vantrimpont, P. J.; Shah, A. M., Paracrine coronary endothelial control of 695 left ventricular function in humans. Circulation 1995, 92, (8), 2119-26. 696 47. Abe, M.; Ono, J.; Sato, Y.; Okeda, T.; Takaki, R., Effects of glucose and insulin on 697 cultured human microvascular endothelial cells. Diabetes Res Clin Pract 1990, 9, (3), 287-95. 698 48. Du, X. L.; Sui, G. Z.; Stockklauser-Farber, K.; Weiss, J.; Zink, S.; Schwippert, B.; Wu, Q. 699 X.; Tschope, D.; Rosen, P., Introduction of apoptosis by high proinsulin and glucose in cultured 700 human umbilical vein endothelial cells is mediated by reactive oxygen species. Diabetologia 701 1998, 41, (3), 249-56. 702 49. Graier, W. F.; Grubenthal, I.; Dittrich, P.; Wascher, T. C.; Kostner, G. M., Intracellular 703 mechanism of high D-glucose-induced modulation of vascular cell proliferation. Eur J Pharmacol 704 1995, 294, (1), 221-9. 705 50. Kamal, K.; Du, W.; Mills, I.; Sumpio, B. E., Antiproliferative effect of elevated glucose in 706 human microvascular endothelial cells. J Cell Biochem 1998, 71, (4), 491-501. 707 51. Lorenzi, M.; Nordberg, J. A.; Toledo, S., High glucose prolongs cell-cycle traversal of 708 cultured human endothelial cells. Diabetes 1987, 36, (11), 1261-7. 709 52. Quagliaro, L.; Piconi, L.; Assaloni, R.; Martinelli, L.; Motz, E.; Ceriello, A., Intermittent 710 high glucose enhances apoptosis related to oxidative stress in human umbilical vein endothelial 711 cells: the role of protein kinase C and NAD(P)H-oxidase activation. Diabetes 2003, 52, (11), 712 2795-804. 713 53. McGinn, S.; Poronnik, P.; King, M.; Gallery, E. D.; Pollock, C. A., High glucose and 714 endothelial cell growth: novel effects independent of autocrine TGF-beta 1 and hyperosmolarity. 715 Am J Physiol Cell Physiol 2003, 284, (6), C1374-86. 716 54. Yuan, W.; Zhang, J.; Li, S.; Edwards, J. L., Amine metabolomics of hyperglycemic 717 endothelial cells using capillary LC-MS with isobaric tagging. J Proteome Res 2011, 10, (11), 718 5242-50. 719 55. Chen, S.; Akter, S.; Kuwahara, K.; Matsushita, Y.; Nakagawa, T.; Konishi, M.; Honda, T.; 720 Yamamoto, S.; Hayashi, T.; Noda, M.; Mizoue, T., Serum amino acid profiles and risk of type 2 721 diabetes among Japanese adults in the Hitachi Health Study. Sci Rep 2019, 9, (1), 7010. 722 56. Lai, M.; Liu, Y.; Ronnett, G. V.; Wu, A.; Cox, B. J.; Dai, F. F.; Rost, H. L.; Gunderson, E. 723 P.; Wheeler, M. B., Amino acid and lipid metabolism in post-gestational diabetes and 724 progression to type 2 diabetes: A metabolic profiling study. PLoS Med 2020, 17, (5), e1003112. 725 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 25 57. Lu, Y.; Wang, Y.; Liang, X.; Zou, L.; Ong, C. N.; Yuan, J. M.; Koh, W. P.; Pan, A., Serum 726 Amino Acids in Association with Prevalent and Incident Type 2 Diabetes in A Chinese 727 Population. Metabolites 2019, 9, (1). 728 58. Menni, C.; Fauman, E.; Erte, I.; Perry, J. R.; Kastenmuller, G.; Shin, S. Y.; Petersen, A. 729 K.; Hyde, C.; Psatha, M.; Ward, K. J.; Yuan, W.; Milburn, M.; Palmer, C. N.; Frayling, T. M.; 730 Trimmer, J.; Bell, J. T.; Gieger, C.; Mohney, R. P.; Brosnan, M. J.; Suhre, K.; Soranzo, N.; 731 Spector, T. D., Biomarkers for type 2 diabetes and impaired fasting glucose using a nontargeted 732 metabolomics approach. Diabetes 2013, 62, (12), 4270-6. 733 59. Wang, T. J.; Larson, M. G.; Vasan, R. S.; Cheng, S.; Rhee, E. P.; McCabe, E.; Lewis, G. 734 D.; Fox, C. S.; Jacques, P. F.; Fernandez, C.; O'Donnell, C. J.; Carr, S. A.; Mootha, V. K.; 735 Florez, J. C.; Souza, A.; Melander, O.; Clish, C. B.; Gerszten, R. E., Metabolite profiles and the 736 risk of developing diabetes. Nat Med 2011, 17, (4), 448-53. 737 60. Koziel, A.; Woyda-Ploszczyca, A.; Kicinska, A.; Jarmuszkiewicz, W., The influence of 738 high glucose on the aerobic metabolism of endothelial EA.hy926 cells. Pflugers Arch 2012, 464, 739 (6), 657-69. 740 61. Badawy, A. A., Kynurenine Pathway of Tryptophan Metabolism: Regulatory and 741 Functional Aspects. Int J Tryptophan Res 2017, 10, 1178646917691938. 742 62. Pedersen, E. R.; Tuseth, N.; Eussen, S. J.; Ueland, P. M.; Strand, E.; Svingen, G. F.; 743 Midttun, O.; Meyer, K.; Mellgren, G.; Ulvik, A.; Nordrehaug, J. E.; Nilsen, D. W.; Nygard, O., 744 Associations of plasma kynurenines with risk of acute myocardial infarction in patients with 745 stable angina pectoris. Arterioscler Thromb Vasc Biol 2015, 35, (2), 455-62. 746 63. Sulo, G.; Vollset, S. E.; Nygard, O.; Midttun, O.; Ueland, P. M.; Eussen, S. J.; Pedersen, 747 E. R.; Tell, G. S., Neopterin and kynurenine-tryptophan ratio as predictors of coronary events in 748 older adults, the Hordaland Health Study. Int J Cardiol 2013, 168, (2), 1435-40. 749 64. Polyzos, K. A.; Ketelhuth, D. F., The role of the kynurenine pathway of tryptophan 750 metabolism in cardiovascular disease. An emerging field. Hamostaseologie 2015, 35, (2), 128-751 36. 752 65. Aquilano, K.; Baldelli, S.; Ciriolo, M. R., Glutathione: new roles in redox signaling for an 753 old antioxidant. Front Pharmacol 2014, 5, 196. 754 66. Yuan, W.; Edwards, J. L., Thiol metabolomics of endothelial cells using capillary liquid 755 chromatography mass spectrometry with isotope coded affinity tags. J Chromatogr A 2011, 756 1218, (18), 2561-8. 757 67. Weidig, P.; McMaster, D.; Bayraktutan, U., High glucose mediates pro-oxidant and 758 antioxidant enzyme activities in coronary endothelial cells. Diabetes Obes Metab 2004, 6, (6), 759 432-41. 760 68. Felice, F.; Lucchesi, D.; di Stefano, R.; Barsotti, M. C.; Storti, E.; Penno, G.; Balbarini, 761 A.; Del Prato, S.; Pucci, L., Oxidative stress in response to high glucose levels in endothelial 762 cells and in endothelial progenitor cells: evidence for differential glutathione peroxidase-1 763 expression. Microvasc Res 2010, 80, (3), 332-8. 764 69. Kashiwagi, A.; Asahina, T.; Ikebuchi, M.; Tanaka, Y.; Takagi, Y.; Nishio, Y.; Kikkawa, R.; 765 Shigeta, Y., Abnormal glutathione metabolism and increased cytotoxicity caused by H2O2 in 766 human umbilical vein endothelial cells cultured in high glucose medium. Diabetologia 1994, 37, 767 (3), 264-9. 768 70. Hanschmann, E. M.; Godoy, J. R.; Berndt, C.; Hudemann, C.; Lillig, C. H., Thioredoxins, 769 glutaredoxins, and peroxiredoxins--molecular mechanisms and health significance: from 770 cofactors to antioxidants to redox signaling. Antioxid Redox Signal 2013, 19, (13), 1539-605. 771 71. Scocchi, M.; Tossi, A.; Gennaro, R., Proline-rich antimicrobial peptides: converging to a 772 non-lytic mechanism of action. Cell Mol Life Sci 2011, 68, (13), 2317-30. 773 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 26 72. Migliaccio, A.; Castoria, G.; de Falco, A.; Bilancio, A.; Giovannelli, P.; Di Donato, M.; 774 Marino, I.; Yamaguchi, H.; Appella, E.; Auricchio, F., Polyproline and Tat transduction peptides 775 in the study of the rapid actions of steroid receptors. Steroids 2012, 77, (10), 974-8. 776 73. Radicioni, G.; Stringaro, A.; Molinari, A.; Nocca, G.; Longhi, R.; Pirolli, D.; Scarano, E.; 777 Iavarone, F.; Manconi, B.; Cabras, T.; Messana, I.; Castagnola, M.; Vitali, A., Characterization 778 of the cell penetrating properties of a human salivary proline-rich peptide. Biochim Biophys Acta 779 2015, 1848, (11 Pt A), 2868-77. 780 74. Vanhoof, G.; Goossens, F.; De Meester, I.; Hendriks, D.; Scharpe, S., Proline motifs in 781 peptides and their biological processing. FASEB J 1995, 9, (9), 736-44. 782 75. Colombo, S.; Melo, T.; Martinez-Lopez, M.; Carrasco, M. J.; Domingues, M. R.; Perez-783 Sala, D.; Domingues, P., Phospholipidome of endothelial cells shows a different adaptation 784 response upon oxidative, glycative and lipoxidative stress. Sci Rep 2018, 8, (1), 12365. 785 76. De Keyzer, D.; Karabina, S. A.; Wei, W.; Geeraert, B.; Stengel, D.; Marsillach, J.; 786 Camps, J.; Holvoet, P.; Ninio, E., Increased PAFAH and oxidized lipids are associated with 787 inflammation and atherosclerosis in hypercholesterolemic pigs. Arterioscler Thromb Vasc Biol 788 2009, 29, (12), 2041-6. 789 77. Tselepis, A. D.; John Chapman, M., Inflammation, bioactive lipids and atherosclerosis: 790 potential roles of a lipoprotein-associated phospholipase A2, platelet activating factor-791 acetylhydrolase. Atheroscler Suppl 2002, 3, (4), 57-68. 792 78. Wang, A.; Dennis, E. A., Mammalian lysophospholipases. Biochim Biophys Acta 1999, 793 1439, (1), 1-16. 794 79. Marco-Ramell, A.; Palau-Rodriguez, M.; Alay, A.; Tulipani, S.; Urpi-Sarda, M.; Sanchez-795 Pla, A.; Andres-Lacueva, C., Evaluation and comparison of bioinformatic tools for the 796 enrichment analysis of metabolomics data. BMC Bioinformatics 2018, 19, (1), 1. 797 80. Zhou, X.; Liao, W. J.; Liao, J. M.; Liao, P.; Lu, H., Ribosomal proteins: functions beyond 798 the ribosome. J Mol Cell Biol 2015, 7, (2), 92-104. 799 81. Goldberg, A. L., Protein degradation and protection against misfolded or damaged 800 proteins. Nature 2003, 426, (6968), 895-9. 801 82. Vinals, F.; Pouyssegur, J., Confluence of vascular endothelial cells induces cell cycle 802 exit by inhibiting p42/p44 mitogen-activated protein kinase activity. Mol Cell Biol 1999, 19, (4), 803 2763-72. 804 83. Yu, Y.; Moulton, K. S.; Khan, M. K.; Vineberg, S.; Boye, E.; Davis, V. M.; O'Donnell, P. 805 E.; Bischoff, J.; Milstone, D. S., E-selectin is required for the antiangiogenic activity of 806 endostatin. Proc Natl Acad Sci U S A 2004, 101, (21), 8005-10. 807 84. Brigstock, D. R., Regulation of angiogenesis and endothelial cell function by connective 808 tissue growth factor (CTGF) and cysteine-rich 61 (CYR61). Angiogenesis 2002, 5, (3), 153-65. 809 85. Elmasri, H.; Ghelfi, E.; Yu, C. W.; Traphagen, S.; Cernadas, M.; Cao, H.; Shi, G. P.; 810 Plutzky, J.; Sahin, M.; Hotamisligil, G.; Cataltepe, S., Endothelial cell-fatty acid binding protein 4 811 promotes angiogenesis: role of stem cell factor/c-kit pathway. Angiogenesis 2012, 15, (3), 457-812 68. 813 86. Quinn, M. T.; Schepetkin, I. A., Role of NADPH oxidase in formation and function of 814 multinucleated giant cells. J Innate Immun 2009, 1, (6), 509-26. 815 87. Holt, D. J.; Grainger, D. W., Multinucleated giant cells from fibroblast cultures. 816 Biomaterials 2011, 32, (16), 3977-87. 817 88. Tse, G. M.; Law, B. K.; Chan, K. F.; Mas, T. K., Multinucleated stromal giant cells in 818 mammary phyllodes tumours. Pathology 2001, 33, (2), 153-6. 819 89. Celton-Morizur, S.; Merlen, G.; Couton, D.; Desdouets, C., Polyploidy and liver 820 proliferation: central role of insulin signaling. Cell Cycle 2010, 9, (3), 460-6. 821 822 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 27 Figure legends 823 824 Figure 1. Illustration of the methodology followed in this study. 825 826 Figure 2. Simulated diabetes induced changes in the metabolome of bovine coronary 827 artery endothelial cells (BCAEC). (A) Venn diagram of features identified among MZmine and 828 XCMS software (0.01 Da and 1 min retention time, thresholds) on LC-MS2 datasets. (B) Volcano 829 plot of all quantified metabolites displaying differences in relative abundance (> +/-30% change, 830 <0.05 p-value cut-offs) between BCAEC cultured in control (NG) media and simulated diabetes 831 (HG+ HI) for twelve days. Values (dots) represent the HG+HI/NG ratio for all metabolites. Red 832 and blue dots denote downregulated and upregulated metabolites in the HG + HI group vs. NG 833 group, respectively. (C) Principal Component Analysis (PCA) of LC-MS2 datasets. Data was log 834 transformed without scaling. Shade areas depict the 95% confidence intervals. (C) HeatMap of 835 the top 100 metabolites ranked by t-test. Abbreviations: NG, normal glucose; HG, high glucose; 836 HI, high insulin; QC, quality control. 837 838 Figure 3. Bovine coronary artery endothelial cells (BCAEC) metabolite molecular network. 839 (A) Molecular classes (according to Classyfire) of the metabolome identified by the 840 MolNetEnhancer workflow and visualized by Cytoscape version 3.8.2. Each node represents a 841 unique feature and the color of the node denotes the associated chemical class. The thickness of 842 the edge (connectivity) indicates the MS2 similarity (Cosine score) among features. The m/z value 843 of the feature is shown inside the node and is proportional to the size of the node. Three selected 844 clusters or connected features as relevant are shown. (B) Inset of cluster 1 denoting the presence 845 of phosphocholine (PC)-containing lipids. Significant differential abundant features among 846 simulated diabetes (HG+HI) and control (NG) groups are indicated with an asterisk (p-value 847 <0.05). (C) Characterization of features in (B) aided by substructure recognition by MSLDA 848 software using MS1 visualization in www.ms2lda.org. Fragment at m/z 184.0725 linked to a PC 849 head group by mzCloud in silico prediction (www.mzCloud.org). Abbreviations: M2M, mass2motif; 850 FC, fold change; NG, normal glucose; HG, high glucose; HI, high insulin. Chemical structures 851 were drawn by ChemDraw Professional version 16.0.1.4. 852 853 Figure 4. Peptide metabolites modulated by simulated diabetes in bovine coronary artery 854 endothelial cells (BCAEC). (A) Cluster 2 retrieved from the main molecular network linked to 855 glutathione and derivatives. The fragments of mass-2-motif (M2M)_453 colored in red are 856 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 28 characteristic of a glutathione core and the fragments are shown in red. (B) Features associated 857 with M2M_453 using MS1 visualization in www.ms2lda.org. (C) Cluster 3 retrieved from the main 858 molecular network linked to phenylalanine-based metabolites. A singular node at m/z 487.1548 859 is also shown. The fragments of M2M_59 colored in red are characteristic of a phenylalanine core 860 (Heuristic and Quantum Chemical predictions by www.mzCloud.org). (D) Features associated 861 with M2M_59 using MS1 visualization in www.ms2lda.org. In GNPS’s clusters (A and C), the 862 node’s color denotes the chemical class assigned to the cluster. The thickness of the edge 863 (connectivity) indicates the cosine score (MS2 similarity). The m/z value of the feature is shown 864 inside the node and is proportional to the size of the node. Significant differential abundant 865 features among simulated diabetes (HG+HI) and control (NG) groups are indicated with an 866 asterisk (p-value <0.05). In MS2LDA’s nodes (B and D), the green node represents the M2M and 867 squares indicate individual features. Edges represent connections to M2M. Significant differential 868 abundant features among groups are indicated with an asterisk (p-value <0.05). Abbreviations: 869 M2M, mass2motif; FC, fold change; NG, normal glucose; HG, high glucose; HI, high insulin. 870 Chemical structures were drawn by ChemDraw Professional version 16.0.1.4. 871 872 Figure 5. Simulated diabetes induced changes in the proteome of bovine coronary artery 873 endothelial cells (BCAEC). (A) Principal Component Analysis (PCA) of LC-SWATH-MS2 874 datasets. Data was log transformed without scaling. Shade areas depict the 95% confidence 875 intervals. No scaling was used. (B) Volcano plot of all quantified proteins (Quantile normalization) 876 displaying differences in relative abundance (> +/-20% change, <0.05 p-value cut-offs) between 877 BCAEC cultured in control (NG) media and simulated diabetes (HG+ HI) for twelve days. Values 878 (dots) represent the HG+HI/NG ratio for all proteins. Red and blue dots denote downregulated 879 and upregulated proteins in the HG + HI group vs. NG group, respectively. (C) HeatMap of the 880 top 50 metabolites ranked by t-test. Protein-Protein interactome (>0.9 confidence) using the list 881 of proteins with increased abundance (D) and reduced abundance (E) in the HG + HI group. 882 Colored circles denote modules or clusters which may represent relevant complexes or functional 883 units. The input proteins are illustrated with a blue shade and the gene ID is also shown. The 884 most representative pathway (containing more input proteins) for all modules is indicated in blue 885 letters. Abbreviations: NG, normal glucose; HG, high glucose; HI, high insulin. 886 887 Figure 6. 3D Integrative network of the proteomic and metabolomic perturbations caused 888 by simulated diabetes in bovine coronary artery endothelial cells (BCAEC). Composite 889 protein-metabolite network created by OmicsNet using the up-regulated proteins (red nodes) and 890 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 29 metabolites (magenta nodes) in the HG + HI group (simulated diabetes). Interacting proteins (<0.9 891 confidence) were retrieved from STRING Database and are shown as gray nodes. Abbreviations: 892 NG, normal glucose; HG, high glucose; HI, high insulin. 893 894 Figure 7. Increased cellular binucleation by simulated diabetes in bovine coronary artery 895 endothelial cells (BCAEC) and human coronary artery endothelial cells (HCAEC). (A) 896 Representative immunofluorescence micrographs showing the localization of the von-Willebrand 897 factor (vWf, 1:400, 3% BSA in PBS) in fixed and permeabilized cells. The nuclei were stained 898 using the dye Hoechst 33258 (2 µg/ml in HBSS). White arrows indicate binucleated cells. (B) 899 Quantification of binucleated cells in HCAEC and BCAEC under simulated diabetes (HG+HI) vs. 900 control (NG) group. Fluorescence images were taken in at least three random fields per condition 901 using an EVOS® FLoid® Cell Imaging Station with a fixed 20x air objective. Image analysis was 902 performed by ImageJ software (version 2.0.0). Abbreviations: NG, normal glucose; HG, high 903 glucose; HI, high insulin. 904 905 Figure 8. Summary illustration of study findings. Cellular structures were created using 906 Servier Medical Art templates, which are licensed under a Creative Commons Attribution 3.0 907 Unported License; https://smart.servier.com. Chemical structures were drawn by ChemDraw 908 Professional version 16.0.1.4. 909 910 911 912 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ 30 Supporting information 913 914 Table S1. List of all the putatively annotated metabolites by MS2 spectral matching against GNPS 915 public spectral libraries. 916 917 Table S2. List of putatively annotated (MS2 spectral matching) metabolites modulated by 918 simulated diabetes. 919 920 Table S3. List of all detected peptides by ProteinPilot Software using the metabolomics datasets. 921 922 Table S4 Putative annotated proline-peptides altered by simulated diabetes in Bovine Coronary 923 Artery Endothelial Cells by ProteinPilot Software and manual inspection. 924 925 Table S5. List of the detected peptides and proteins in all conditions for SWATH-based 926 quantification. 927 928 Figure S1. Proteomics data normalization results using NormalyzerDE. (A) Total intensity of raw 929 data before normalization. (B) Quantitative parameters of normalization algorithms (pooled 930 intragroup coefficient of variation [PCV], median absolute deviation [PMAD], estimate of variance 931 [PEV]). Qualitative parameters of normalization algorithms; (C) Box plots (D) MA plots, and (E) 932 Density plots. 933 934 Figure S2. Cellular confluence in control and experimental group. Representative 935 micrographs of Bovine Coronary Artery Endothelial Cells (BCAEC) cultured for 9 days with 5.5 936 mmol/L glucose (control group) and 20 mmol/L glucose+100 nmol/L insulin (simulated diabetes 937 or experimental group). Images were taken using an EVOS® FLoid® Cell Imaging Station with a 938 fixed 20x air objective. Abbreviations: NG, normal glucose; HG, high glucose; HI, high insulin. 939 940 941 942 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 8, 2021. ; https://doi.org/10.1101/2021.01.06.425584doi: bioRxiv preprint https://doi.org/10.1101/2021.01.06.425584 http://creativecommons.org/licenses/by-nc-nd/4.0/ Metabolites DDA Omics Integration GO analysis ProteomeXchange PXD013643 `` SWATH-based Proteomics `` Untargeted Metabolomics MeOH/EtOH (50:50, v:v) 0 1 2 3 4 5 6 7 8 9 10 11 12 Insulin 100 nmol/L Glucose 20 mmol/L Days Glucose 5.5 mmol/L Control Substructure annotation Ry Metabolite annotation COOH `` m/z m/z RT In te ns it y In te ns it y In te ns it y MS1 MS2 A B SC IEX Trip le TO F 5 6 0 0 LC-MS/MS Triple®TOF 5600+ A 957 (17.18%) Total features detected: 5571 2194 (39.96%) 2420 (43.44%) n=82n=140 B C D N orm alized m etabolite abundance PC1 P C 2 NG HG+HI QC HG+HI NG QC Glycerophospholipids Organooxygen compounds Fatty acyls Steroids and derivatives Glycerolipids Chemical class Indoles and derivatives Organonitrogen compounds Coumarins and derivatives Connectivity (Cosine Score) Precursor ion m/z value Carboxylic acids and derivatives Benzene and substituted derivatives Unknown 522.3415 194.1019 210.1343 286.0574 703.5731 376.259 376.2327 546.2845 575.2753 467.1657 786.5972 264.2317 393.2856 522.355 184.1104 238.1764 530.2946 413.6366 227.0673425.1357 307.0978 701.5589 760.5815 327.0837 367.3351 507.1058 390.0094 448.2113 418.3511 291.0476 538.2512 472.2715 505.105 161.096 246.0794 376.2308 283.0944 575.2728 264.0784 675.5421 283.1171 440.2142 699.612 * * PC(18:0/18:2(9Z,12Z)) O P O OO N+ O O O O PC(16:0/18:1(9Z)) O P O OO N+ O O O O Cluster 1 588.3571 403.2539 522.3415 289.0601 443.112 884.5907 413.1411 295.1284 367.2373 429.1976 194.1019 399.349 277.1856 514.128 414.6985 600.2445 181.6122 118.086 212.1178 386.2386 545.1027 271.0785 540.1645 664.4615 713.4423 404.2208 413.2847 455.1888 212.1424 345.1096 481.2106 401.3451386.3258 623.364 460.311 615.1719 462.2474 658.375 423.2506 301.169 475.3242 436.3157 133.0648 507.1667 369.3511 360.2117 550.1585 210.1343 391.1352 309.1828 462.2255 516.3869 567.2984 562.1643 299.1096 382.2582 238.1065 382.527 332.2428 349.1834 642.613 299.1441 490.2817 253.2143 500.2991 465.2575 449.1492 322.1153 299.0619 337.0757 436.1964 371.1907 367.1969 757.4697 316.2243 409.2795 371.0673 585.2858 244.1903 344.1112 360.3446 217.1478 390.2615 381.1105 587.2967 345.2034 301.0599 319.2257 273.0908 343.2142 490.1759 520.3321 699.5139 629.159 441.7374 292.1173 456.2792 396.0879 398.7747 309.1296 401.1592 313.2729 528.3076 419.1089 594.2896 655.3817 531.2728 477.1781 286.0574 114.0912 181.1218 536.1638 566.3353 789.4651 229.1418 185.0804 316.2842 349.2379 331.2598 258.1688 422.3033 495.2392 509.2865 550.2611 306.2272 378.2235 358.135 230.1749 387.083 442.8012 751.5126 305.6453 768.5299 455.247 444.3311 298.0777 703.5731 318.2987 358.1104 330.0751 632.3741 376.259 475.1924 414.1404 273.2528 273.1396 385.1528 213.1103 398.2412 446.2587 540.2848 801.4936 353.1448 283.1749 304.8924 504.2723 565.3329 355.1119 279.0797 593.1553 497.2902 202.0859 216.1608 376.2327 359.3151 400.3045 474.1614 711.1412 319.1052 486.0965 696.6349 407.1877 331.9947 230.1386 439.2532 506.2452 285.1309 504.1222 414.2137 498.4031 546.2845 509.1202 867.5631 475.2779 434.2584 817.4667 348.1855 575.2753 200.1057 407.0865 612.4165 375.0856 467.1657 564.358 377.2349 174.9548 386.2077 338.1949 314.2003 786.5972 415.2237 419.17 264.2317 313.2128 225.0906 233.1838 393.2856 257.0585 276.1806 527.156 377.0441 232.6427 297.1442 933.571 411.2079 430.2003 437.1154 197.116 239.9349 928.6203 261.2528 522.355 319.1367 328.1327 420.7869 129.069 403.1745 506.2405 271.0781 232.1639 244.2086 354.2356 889.5489 390.2321 544.3316 107.0795 288.2526 399.222 338.0442 223.0626 439.1149 298.2223 184.1104 369.0139 768.5354 294.1536 448.1644 301.1202 404.1451 242.116 304.2111 299.199 578.3353 172.1334 277.0957 315.264 232.1536 497.2898 678.5877 632.4042 567.3284 633.737 304.0625 284.1591 203.1064 272.2211 443.2961 238.1764 359.0625 385.2079 245.1861 713.1472 358.2426 423.136 229.1379 327.1659 302.1146 477.2296 297.154 548.157 502.0925 399.3618 394.0924 307.1737 301.1301 301.2823 796.5661 309.1275 693.2997 536.1732 316.2115 388.1274 483.2052 331.1683 551.3183 353.1478 473.2587 654.3295 558.7383 530.2946 508.3023 659.359 473.234 383.2783 443.1717 664.1098 231.1589 267.0061 500.2756 679.4094 433.2468 630.8588 409.1622 319.2253 250.1779 247.013 329.0054 691.4994 315.195 474.2317 358.2043 438.2635 239.1529 299.0617 180.1749 433.2028 291.1168 296.2214 300.0763 387.1685 320.1042 716.5212 587.3253 236.0713 197.0779 151.075 413.6366 531.3313 851.3932 227.0673 429.2316 246.1521 453.21 329.202 652.4089 270.0416 552.1533 371.0477 338.1339 613.1577 334.9126 374.1591 309.1524 522.3083 541.2215 415.2797 376.1159 734.6477 257.0572 283.6001 331.0834 374.0627 592.3889 246.1016 425.1357 331.2049 575.209 404.7346 348.1212 283.1205 242.0992 478.292 377.1814 345.1106 285.007 640.268 307.0978 240.1799 402.0959 701.5589 404.2348 661.5624 331.1573 353.2221 509.2995 265.1674 283.1263 416.0969 509.2353 487.1958 620.4355 332.0995 190.1434 263.0813 760.5815 453.2655 382.2214 271.1622 210.0494 309.1573 493.1866 488.2337 295.1274 273.1677 161.1165 349.2384 187.1435 398.0898 428.1321 439.2902 321.1474 535.21 204.6112 446.2585 407.1732 654.19 455.3344 500.3049 409.2372 295.1285 547.331 625.2796 375.0536 242.1172 376.2281 209.0914 597.3382 370.1355 270.0645 541.2606 671.3161 276.1258 468.9809 287.0473 320.2047 277.1414 177.0627 504.3367 735.4868 461.2096 617.3879 589.3121 202.1068 148.0604 520.295 327.0837 347.1596 556.1906 297.1436 334.0427 206.1392 431.0663 911.594 336.1909 329.1818 608.3826 289.2475 367.3351 156.0483 478.1842 619.2416 404.847 465.1608 207.0614 605.4416 579.337 242.2114 310.0135 396.1853 341.206 397.2156 344.2247 433.1485 393.2099 346.1958 434.0664 398.7746 484.137 371.2276 216.1954 318.1811 399.1987 708.4868 435.1311 474.2324 383.0228 291.1928 316.1778 751.5081 467.6187 272.1853 739.6026 434.1648 269.0628 293.6761 744.5533 302.1702 637.1514 331.109 507.1058 416.158 369.1843 489.1045 233.0772 507.1304 407.1706 298.1856 404.1482 427.305 210.1843 857.3778 729.4168 348.2378 477.2335 409.1764 735.4996 779.5132 388.2532 478.2635 431.2514 208.1812 449.2862 576.4091 474.2906 594.1326 336.3255 637.8139 627.095 403.2751 353.2653 683.345 325.0998 331.2087 242.2836 450.2319 342.1657 188.066 339.2526 335.158 300.2013 659.2866 591.358 214.2521 686.4674 545.2051 348.2012 314.1799 231.1164 209.122 503.3049 297.1802 360.1322 403.2229 455.107 322.0628 229.0386 311.1988 316.061 300.0013 520.3407 250.1462 446.2248 387.1969 436.1368 480.1787 390.0094 334.2211 458.2322 399.3568 506.2726 235.1782 389.1216 452.2747 329.0849 502.2879 654.3297 385.2046 421.2828 433.2463 846.4398 402.2327 355.1755 325.0272 360.2022 263.1492 447.1673 402.1759 316.3205 610.3903 396.0943 157.0855 465.1713 256.0605 518.3147 773.4406 653.332 331.1648 448.2113 511.2145 606.138 294.2061 409.1837 198.1845 403.22 486.2007 740.5427 560.4098 345.2984 669.4164 480.8023 407.1805 611.3249 395.2195 333.1659 840.5653 491.1995 211.0939 377.1463 299.1804 431.1815 459.1706 329.1689 401.2022 302.1736 248.108 482.3244 823.5398 358.1599 228.0157 441.1515 247.1286 255.0077 385.1619 534.3515 310.0137 601.3545 307.5846 257.1651 682.583 205.0968233.128 311.0851 708.4518 208.9943 367.317 502.1256 221.117 404.2074 293.0018 418.3511 522.2898 383.0299 232.0109 187.0257 625.39 315.0742 365.157 387.3458 528.1033 363.1622 415.2539 399.268 291.0476 313.0673 288.3147 415.2449 667.3875 480.3075 602.3388 334.0404 261.1445 433.2063 331.9939 331.1366 243.1336 261.1307 313.068 543.1052 331.2228 329.1491 224.185 284.0988 497.7717 332.604 295.1349 538.2512 284.2213 257.2212 285.1892 260.1854 561.3954 343.1966 385.0872 544.3215 318.8867 301.1419 214.0972 377.2579 306.094 356.077 669.2021 717.6223 487.3739 402.1677 204.042 683.5415 592.1747430.2644 421.3172 484.1461 243.0875 597.8556 425.1561 779.5197 874.4691 610.2815 575.2665 362.0542 472.2715 504.3064 592.3891 245.137 414.2359 342.2119 305.2105 651.257 387.2096 376.3416 376.762 505.105 653.3624 613.3397 308.8983 247.132 368.1613 289.0533 237.0678 311.1281 790.3776 569.3138 328.2316 298.1252 521.1332 379.2408 307.5897 120.0809 459.1915 685.3904 443.2189 544.3519 348.3105 264.0825 648.3785 274.1444 541.1202 550.1643 306.2638 472.3602 219.1125 642.6193 290.1934 664.1265 415.2426 504.1223 528.1789 195.1005 290.1952 190.0339 471.1366 487.1548 331.1695 412.2519 416.2854 321.0536 533.4225 386.0815 204.1049 433.2634 353.0485 309.6428 354.0344 315.215 518.0863 409.1898 635.3845 796.5389 537.3384 227.1646 353.2061 480.164 752.5133 264.072 488.3571 308.0629 130.1588 603.2672 654.1961 861.4975 452.0319 293.1149 222.0541 266.173 546.1593 481.2613 198.0459 773.4417 297.2129 312.1869 161.096 357.0869 302.148 493.313 532.3832 373.3665 448.1767 287.1911 181.6125 845.5211 335.0784 289.1088 369.154 288.1805 509.3542 773.4915 641.3634 289.0603 277.1275 202.1436 320.2765 604.3512 355.0651 445.1738 561.3101 304.2108 439.2092 582.272 380.1061 639.3729 521.1369 334.0761 593.2762 324.1011 453.3427 335.1059 515.2178 540.1005 294.2273 380.1122 246.0794 365.2676 462.1453 376.2308 204.045 357.2989 268.0628 215.9822 725.3575 530.1589 184.1328 338.0815 385.1971 268.1542 343.2955 645.3443 283.0944 559.3043 462.248 313.1461 477.3414 529.3357 609.3057 373.1514 405.2599 545.1545 394.1956 385.1877 423.1984 592.1777 409.2473 325.2274 276.0625 436.062 320.1688 228.1955 867.3655 317.21 344.2278 509.2845 517.2837 584.2047 328.9155 341.1805 434.0686 287.2105 242.1159 221.1114 575.2728 597.2364 198.1485 362.3622 312.3253 543.2987 345.3354 315.2268 713.3509 478.0851 357.2366 610.367 264.0784 476.3052 340.0767 286.1397 636.4161 475.2773 652.1914 166.0861 675.5421 546.1562 271.181 387.1935 192.1223 331.1899 351.1808 507.2935 637.3037 225.1095 560.3254 631.3504 359.1477 269.0883 619.2691 164.1065 218.1383 283.1171 259.0201 434.2585 188.1999 385.2071 635.1389 493.2815 479.0776 526.2914 223.0963 429.3016 639.2251 534.3094 298.095 486.226 697.3627 513.3387 373.0739 277.0958 440.2142 711.1232 297.0668 285.0076 420.1996 246.0252 486.1428 516.2999 325.2119 455.2416 581.3643 300.2164 625.2127 421.2334 241.5671 239.1269 301.0709 384.1148 381.2962 299.1163 829.4161 366.1317 711.1342 362.2528 305.1569 548.3632 263.125 526.2574 610.1959 353.1168 337.0808 580.1723 282.1469 383.2029 699.612 408.7552 166.0742 359.7141 218.1194 283.1332 696.4348 671.3179 525.2874 553.3113 443.1129 332.2177 262.1188 539.267 437.2364 184.9848 265.5857 343.1425 188.07 292.096 432.2794 368.1451 261.1094 456.2802 Cluster 1 Cluster 3 Cluster 2 A C MS2LDA 760.5815 m/z * B M2M_526_Phosphocholine-based substructure 0 100 200 300 400 0 25 50 75 100 m/z R el at iv e In te ns ity 184.0725 OH P O OHO N+ Lo g2 F C 3.6 -2.6 1 Peptide metabolites C A 162.0225 m/z M2M_453_Glutathione-based substructure H + 308.0911 m/z OH H N OH O SH H2N O H N OH O SH N H NH2 HO O O O H2 N OH O SH O H N OH O SH N H O 233.059 m/z 162.0219 m/z 179.0484 m/z M2M_59_Phenylalanine-based substructure 166.0875 m/z 166.0875 m/z 166.0875 m/z 166.0875 m/z NH2 OH OH NH2 120.0825 m/z 120.0825 m/z 120.0825 m/z 120.0825 m/z [M+H]+ [M+H]+ MS2LDA M2M_453 B * 615.1719 m/z D MS2LDA M2M_59 *297.1802 m/z 487.1548 m/z * * 295.1285 m/z 277.1856 382.2582 382.527 371.1907 371.0673 306.2272 230.1749 442.8012 509.1202 403.1745 244.2086 385.2079 358.2426 477.2296 371.0477 331.1573 190.1434 295.1285 461.2096 347.1596 329.1818 371.2276 216.1954 318.1811 348.2378 477.2335 208.1812 297.1802 355.1755 331.1648 261.1445 331.1366 295.1349 257.2212 331.1695 264.072 302.148 509.3542 202.1436 334.0761 313.1461 320.1688 188.1999 610.1959 261.1094 N H O OH O HO O NH2 Glutamyl-phenylalanine [M+H]+ Unknown 118.086 481.2106 386.3258 615.1719 409.2795 217.1478 490.1759 629.159 358.1104 398.2412 313.2128 197.116 242.116 299.199 713.1472 502.0925 508.3023 679.4094 409.1622 716.5212 197.0779 338.1339 246.1016 210.0494 242.1172 291.1928 637.1514 311.1988 256.0605 294.2061 307.5846 285.1892 421.3172 305.2105 308.8983 307.5897 321.0536 308.0629 222.0541 266.173 277.1275 268.1542 343.2955 242.1159 513.3387 277.0958 301.0709 526.2574 580.1723 359.7141 283.1332 265.5857 * 308.0925 m/z 179.0475 m/z 233.0575 m/z 162.0225 m/z 179.0475 m/z 308.0925 m/z 308.0925 m/z 179.0475 m/z Precursor ion 629.159 m /z Precursor ion 713.1472 m /z Precursor ion 615.1719 m /z [M+H]+ *[M+H]+ 535.21233.0772 487.1548 Unknown * * [2M+H]+ Lo g2 F C -0.8 3.1 0 Lo g2 F C -1.1 1.3 0 Interacting protein Seed/input protein Connectivity E B n=32n=33 A Protein-protein interaction network symbology Mitochondrial function PSMD3 PSMA4 RPS13 RPL9 RPS7 MCM3 PPP2R2B YWHAQ UBE2N PRMT5 PRDX2 PRDX6 COPG1 DNA/RNA metabolism APRT DDX1 FIS1 MX1 H2AFV DHX9 CAV1 APEX1 DYNLL1 MYH10 RDX YWHAB GABARAPL2 RPL18A COX4I1 UQCRC1 CTGF LAMP1 CPSF6 B2M PDIA4 D C N orm alized protein abundance Protein metabolism-20 -10 0 10 20 -4 0 -2 0 0 20 40 Scores Plot PC 4 ( 14.8 %) P C 1 ( 36 .6 % ) HG+HI NG PC4 P C 1 NG HG+HI HG+HI NG PRDX6 TXN2 APRT Metabolites Proteins PRDX2 Glutamate Glutathione Proline Leucine Tyrosine 2-AminoadipateKynurenine Serine Methionine Threonine OAT HG+HING Nuclei vWF Nuclei vWF HG+HING B A HCAEC BCAEC ≈ 30% ≈ 58% Binucleation HG+HI BINUCLEATION Translation Nuclei Up-regulated Down-regulated Angiogenesis or cell proliferation CTGF AFABP CD62 CAV-1 RPS7 RPS13 RPL9 PSMA4 PSMD3 Integrated Analysis COX4I1 UQCRC1 NDUFB3 NDUFA7 Mitochondrial inner mas CAVIN3 Caveolae DNA and RNA metabolism NH2 H N O OH NH2O NH2 O OH Tryptophan SerotoninKynurenine Catabolism NH2 H N HO N H O OH NH2 O HO O H2N O OH Glutamyl-phenylalanine Phenylalanine-based metabolites Phenylalanine O P O OO N+ O O H R1 PAFAH1B2 Deacylation LysoPC lipids PC lipids OH P O OHO N+ Phosphocholine O P O OO N+ O O R2 R1 LYPLA1 Degradation Inflammation Catabolism Oxidative stress Peptides NH2 O HO O OHOrnitine OAT Glutamic acid NH O OH O H N OH OSH N HNH2 HO O O Ox-Glutathione Glutathione-based metabolites PRDX2 PRDX6 TXN2 ROS Glutathione Proline Mitochondria Table 1. Pathway enrichment analysis of up-regulated and down-regulated proteins in HG+HI group REACTOME Database Total Hits FDR Total Hits FDR Up-regulated Down-regulated Metabolism of RNA 339 142 1.26E- 100 Peptide chain elongation 178 77 5.81E- 48 Metabolism of mRNA 317 136 1.04E-97 Influenza Infection 185 78 6.52E- 48 Synthesis of DNA 95 75 2.37E-79 Nonsense Mediated Decay Independent of the Exon Junction Complex 184 77 4.04E- 47 DNA Replication 102 77 5.67E-79 Influenza Life Cycle 180 76 4.63E- 47 DNA Replication Pre-Initiation 80 68 3.49E-76 Eukaryotic Translation Elongation 186 77 4.63E- 47 M/G1 Transition 80 68 3.49E-76 Nonsense Mediated Decay Enhanced by the Exon Junction Complex 203 80 4.63E- 47 S Phase 122 78 4.62E-71 Nonsense-Mediated Decay 203 80 4.63E- 47 G1/S Transition 113 75 4.37E-70 Influenza Viral RNA Transcription and Replication 176 75 5.77E- 47 Assembly of the pre- replicative complex 63 57 5.61E-67 Viral mRNA Translation 176 75 5.77E- 47 Metabolism of RNA 339 142 1.26E- 100 Eukaryotic Translation Termination 178 75 1.42E- 46 KEGG Database Up-regulated Down-regulated Basal transcription factors 153 111 2.17E- 115 Basal transcription factors 153 88 8.15E- 73 Mismatch repair 45 43 1.68E-53 Nucleotide excision repair 135 46 3.62E- 24 SNARE interactions in vesicular transport 124 41 2.28E-22 Renal cell carcinoma 201 46 2.26E- 16 Base excision repair 36 18 3.46E-13 Endometrial cancer 204 45 1.88E- 15 Human papillomavirus infection 155 34 1.62E-12 Peroxisome 137 35 5.32E- 14 Chemical carcinogenesis 201 36 1.50E-10 Nicotine addiction 193 41 1.48E- 13 Hepatocellular carcinoma 76 18 4.74E-07 Ribosome biogenesis in eukaryotes 79 26 2.92E- 13 Human T-cell leukemia virus 1 infection 162 26 1.49E-06 Gap junction 199 41 3.40E- 13 Chronic myeloid leukemia 97 19 3.79E-06 Herpes simplex virus 1 infection 225 42 5.11E- 12 Notch signaling pathway 160 25 3.79E-06 Glutamatergic synapse 231 36 1.02E- 14 Table 2. Integrative pathway enrichment analysis of up-regulated proteins and metabolites in HG+HI group REACTOME Database Total Hits FDR KEGG Database Total Hits FDR Metabolism of amino acids and derivatives 190 44 8.74E-39 EGFR tyrosine kinase inhibitor resistance 1490 129 1.01E- 72 Metabolism 1490 85 1.33E-35 Glutathione metabolism 56 39 1.63E- 54 Glutathione conjugation 25 21 7.33E-33 Alanine, aspartate and glutamate metabolism 36 24 4.57E- 32 Phase II conjugation 74 25 2.04E-25 ABC transporters 75 27 1.09E- 26 Amino acid synthesis and interconversion (transamination) 18 15 5.97E-23 Cysteine and methionine metabolism 49 22 3.48E- 24 Biological oxidations 142 25 7.73E-18 Pancreatic cancer 82 23 7.55E- 20 tRNA Aminoacylation 42 13 5.19E-12 Drug metabolism - cytochrome P450 72 21 1.67E- 18 Glutathione synthesis and recycling 10 7 3.67E-09 Metabolism of xenobiotics by cytochrome P450 76 21 5.18E- 18 Sulfur amino acid metabolism 25 9 9.84E-09 Drug metabolism - other enzymes 79 20 2.45E- 16 Tryptophan catabolism 11 6 7.51E-07 mRNA surveillance pathway 73 19 8.78E- 16