Ellagic Acid Triggers the Necrosis of Differentiated Human Enterocytes Exposed to 3-Nitro-Tyrosine: An MS-Based Proteomic Study

To study the molecular basis of the toxicological effect of a dietary nitrosated amino acid, namely, 3-nitrotyrosine (3-NT), differentiated human enterocytes were exposed to dietary concentrations of this species (200 μM) and analyzed for flow cytometry, protein oxidation markers and MS-based proteomics. The possible protective role of a dietary phytochemical, ellagic acid (EA) (200 μM), was also tested. The results revealed that cell viability was significantly affected by exposure to 3-NT, with a concomitant significant increase in necrosis (p < 0.05). 3-NT affected several biological processes, such as histocompatibility complex class II (MHC class II), and pathways related to type 3 metabotropic glutamate receptors binding. Addition of EA to 3-NT-treated cells stimulated the toxicological effects of the latter by reducing the abundance of proteins involved in mitochondrial conformation. These results emphasize the impact of dietary nitrosated amino acids in intestinal cell physiology and warn about the potential negative effects of ellagic acid when combined with noxious metabolites.


Introduction
Oxidative stress is related to assorted health disorders such as type 2 diabetes mellitus, hypertension, atherosclerosis, pulmonary diseases, systemic inflammatory response syndrome, aging, cancer and Alzheimer's disease, among others [1,2]. Reactive oxygen species (ROS) and reactive nitrogen species (RNS) are involved in oxidative stress through the imbalance between ROS/RNS and antioxidant defenses [2]. Specifically, 3-nitro-L-tyrosine (3-NT), a nitrosation product of tyrosine, is formed during processing of nitrite-added foods (cured foods) [3,4] and is recognized as an indicator of protein nitration in foods and other biological systems [5]. In addition, 3-NT has been used as a biomarker of acute and chronic inflammatory processes, which indicates the involvement of the nitrosation product in pathological conditions [6,7]. While the potential toxicity of 3-NT has been documented, the underlying molecular mechanisms are poorly understood. Blanchard-Fillion et al. [7] reported that 3-NT is transported into rat neuronal cells via the L-aromatic amino acid transporter and is mis-incorporated into proteins such as α-tubulin and decreases the synthesis of dopamine. In these cells, 3-NT is eventually able to induce caspase-3 mediated apoptosis which the authors associated with the onset of neurological disorders such as Parkinson's disease [7]. Consistently, Zhang and Wei [8] found that 3-NT induced apoptosis in cardiomyocytes in a type 2 diabetes rat model. Neutralization of 3-NT counteracted the apoptosis and diabetic myocardiopathy in rats, which reasonably incriminates the

Quantification of Protein Carbonyls
Both α-aminoadipic semialdehyde (α-AS) and γ-glutamic semialdehyde (γ-GS), two main protein carbonyls, were quantified in differentiated human enterocytes. The procedure reported by Utrera et al. [17] was replicated with slight changes described as follows. Briefly, 100 µL of the cell lysates were treated with 1 mL of cold 10% trichloroacetic acid (TCA) solution, and then proteins were precipitated with centrifugation at 2240× g for 5 min at 4 • C. The supernatant was removed, and the resulting pellet was treated with 1 mL of cold 5% TCA. Proteins were then precipitated by applying cold 5% TCA and centrifugation at 2240× g for 5 min. Protein-bound carbonyls were then derivatized with p-amino-benzoic acid (ABA) followed by a subsequent acid hydrolysis in 6M HCl and 110 • C for 18 h. The hydrolysates were evaporated at 40 • C in a vacuum concentrator, and the generated residues were reconstituted with 200 µL of Milli-Q water and then filtered through hydrophilic polypropylene GH Polypro (GHP) syringe filters (0.45 µm pore size, Pall Corporation, NJ, USA) for HPLC analysis. Details on the chromatography apparatus, as well as on the separation, elution and identification of the compounds of interest, were published in Utrera et al. [17]. Standard α-AS-ABA and γ-GS-ABA were synthesized according to the protocol described by Akagawa et al. [18]. Identification of α-AS and γ-GS in samples was carried out by comparing their retention times with those of the synthesized standard compounds. A standard curve of ABA was prepared by injecting increasing known concentrations of this fluorescent species (0.1 to 0.5 mM). The peaks corresponding to both α-AS-ABA and γ-GS-ABA were manually integrated and the resulting areas plotted against the ABA standard curve. Results are expressed as nmol of protein carbonyl per mg of protein.

Endogenous Antioxidant Enzymes
The activity of two endogenous antioxidant enzymes, namely, catalase (CAT) and superoxide dismutase (SOD), was determined by spectrophotometric methods. One unit (U) of catalase was defined as the amount of protein from cell lysate needed to decompose Antioxidants 2022, 11, 2485 4 of 26 1 mmol of H 2 O 2 per min. One unit of SOD was defined as the amount of protein from cell lysate required to inhibit pyrogallol autoxidation by 50%.

Sample Preparation for LC-MS/MS Based Proteomics
The procedure carried out for the preparation of tryptic peptides from cultured cells has already been described by Díaz-Velasco et al. [15]. Samples were treated with 0.5 mL of lysis buffer pH 7.5 (100 mM Tris-HCl, 50 mM NaCl, 10% glycerol, 0.5 M EDTA pH 8.5) containing 100 mM PMSF (Phenylmethansulfonylfluorid) and 100 µg/mL Pepstatin in a 1:100 proportion. Samples were stirred and sonicated (Branson Ultrasonics, Danbury, CT, USA). Lysates were incubated on ice for 1 h and centrifuged at 9032× g for 10 min at 4 • C. Supernatants were analyzed for protein concentration. Aliquots containing 50 µg of proteins were partially run in SDS-PAGE (4% stacking and 12% separating). Runs were stopped as soon as samples reached the separating part of the gel to be digested in-gel. The gel was stained with Coomassie blue R250 and each lane was cut into 1 mm 3 pieces and subjected to in-gel digestion. Samples were incubated with 0.5 M DTT in 50 mM ammonium bicarbonate for 20 min at 56 • C for protein reduction. The resulting free thiol (-SH) groups were alkylated by incubating the samples with 0.55 M iodoacetamide in 50 mM ammonium bicarbonate for 15 min in the dark at room temperature (22 • C). Afterwards, samples were treated with 50 mM ammonium bicarbonate, ProteaseMAX (Promega, Madison, WI, USA) and trypsin (Promega, USA) and incubated for 2 h at 37 • C. The proteolysis was stopped by adding formic acid. Supernatants were dried in vacuo and subsequently reconstituted with loading buffer (98% milli-Q water, 2% acetonitrile and 0.05% trifluoroacetic acid).

Label-Free Quantitative Proteomic Analyses
A Q-Exactive Plus mass spectrometer coupled to a Dionex Ultimate 3000 RSLCnano (Thermo Scientific, Waltham, MA, USA) analyzed 5 µg from each digest. Data was collected using a Top15 method for MS/MS scans following the procedure described by Delgado et al. [19], with some modifications as described in [15]. Comparative proteome abundance and data analysis were conducted using MaxQuant software (version 1.6.0.15.0) and Perseus (v 1.6.14.0) to organize the data and perform statistical analysis. Carbamidomethylation of cysteines was set as a fixed modification; oxidation of methionines and acetylation of N-terminals were set as variable modifications. Database searching was carried out against Homo sapiens UNIPROT protein database. The maximum peptide/protein false discovery rates (FDR) were set to 1% based on comparison to a reverse database. The MaxLFQ algorithm was used to generate normalized spectral intensities and infer relative protein abundance. Proteins were only retained in final analysis if they were detected in at least two replicates from at least one treatment, and the proteins that matched to a contaminant database or the reverse database were removed. Quantitative analysis was performed using a T-test to compare treatments with the control. Fold change (FC) is expressed as Log 2 . This refers to the degree of quantity change for a particular protein between cells (control vs. treatment). If FC is <1, the change denotes a decrease in the concentration of protein in treated samples (vs. control) while fold change >1 indicates a significant increase in the concentration of that protein in the treated sample (vs. control). When a given protein is only present in one of the groups, fold change cannot be measured and such condition is denoted in tables as "C" (if protein is only present in C, in other words, exposure to whatever treatment leads to a complete depletion of the protein) or 3-NT (if protein is only present in 3-NT-exposed cells) or 3-NT+EA (if protein is only present in cells exposed to the combination of both species). A qualitative analysis was also performed to detect proteins that were found in at least three replicates of a given group of samples (i.e., treated) and were undetectable in the counterpart group (i.e., control). Proteins satisfying one of these two aforementioned criteria were identified as discriminating proteins, and their corresponding genes were grouped by biological processes and molecular functions using ClueGO (v. 2.5.6) [20]. To define term-term inter-relations and functional groups based on shared genes between the terms, the Kappa score was established at 0.4. Three GO terms Antioxidants 2022, 11, 2485 5 of 26 and 4% of genes covered were set as the minimum required to be retained in the final result. The p-value was corrected by Bonferroni step down and set as p < 0.05.

Statistical Analysis
All experiments were carried out five times and each individual sample was measured twice for flow cytometry. Data was analyzed for normality and homoscedasticity, and the effect of the exposure to 3-NT and 3-NT+EA was evaluated by Analysis of Variance (ANOVA). Tukey's test was used for multiple comparisons of the means. The effect of the incubation time on the same measurements was assessed by Student's t-test. SPSS (version 15.0) was used for statistical analysis of the data and the significance level was set at p < 0.05.

Flow Cytometry and Protein Oxidation Markers
In this study, the parameters analyzed by flow cytometry were cell viability, ROS occurrence, apoptosis and necrosis. The results showed a significant decrease in the percentage of live cells among differentiated human enterocytes exposed to 3-NT compared to control cells (p < 0.05) ( Figure 1A). The decreased viability of cells exposed to 3-NT was manifested in a significant increase of necrotic events (p < 0.05) ( Figure 1A). Cells exposed to 3-NT showed a significant decrease in ROS occurrence (p < 0.05) ( Figure 1B) and no significant effect was observed in the percentage of live cells for caspase-3-mediated apoptotic events ( Figure 1A) as compared to control counterparts. Cells exposed to the combination of 3-NT+EA showed a significant decrease in the percentage of live cells (p < 0.05) ( Figure 1C), a significant reduction in the percentage of live cells of apoptotic events (p < 0.01) and a significant and remarkable increase in necrotic events (p < 0.01), as compared to cells exposed only to 3-NT. No significant differences were found in ROS occurrence between non-treated cells and 3-NT+EA ( Figure 1D). To elucidate whether the effects observed in the 3-NT+EA group were a result of the occurrence of EA or the interaction between both species, an additional assay was performed to assess the impact of EA alone. This phytochemical showed a mild (non-significant) positive effect on human enterocytes as compared with a control group, with a clear trend to an increased viability and decreased apoptotic and necrotic events ( Figure 1E-G).
The analysis of protein oxidation markers α-AS and γ-GS was assessed by the detection of early (specific protein carbonyls, α-AS and γ-GS) and advanced oxidation protein products (AOPPs) (Figure 2). The results show that the exposure to 3-NT led to a significant accretion of protein carbonyls in enterocytes, while this increase was not significant for AOPPs. The addition of EA to 3-NT-treated cells caused another significant and highly remarkable increase in protein carbonyls and AOPPs as compared to both Control and 3-NT counterparts.  The analysis of protein oxidation markers α-AS and γ-GS was assessed by the detection of early (specific protein carbonyls, α-AS and γ-GS) and advanced oxidation protein products (AOPPs) (Figure 2). The results show that the exposure to 3-NT led to a significant accretion of protein carbonyls in enterocytes, while this increase was not significant for AOPPs. The addition of EA to 3-NT-treated cells caused another significant and highly remarkable increase in protein carbonyls and AOPPs as compared to both Control and 3-NT counterparts.

Figure 2.
Protein carbonyls (nmol/mg protein of α-AS and γ-GS) (black bars) and advanced oxidation protein products (Pentosidine; fluorescence intensity) (grey bars) on differentiated human enterocytes upon exposure to 200 μM 3-NT and 200 μM 3-NT+200 μM EA for 72 h. Asterisks on top of bars denote significant differences between group of samples in ANOVA (*** p < 0.001).(a-c) Different letters on top of bars denote significant differences between means in post-hoc Tukey tests (p < 0.05).

Proteomic Analyses
The MS-based proteomic platform enabled the identification of 2398 proteins in total. All these proteins were identified with at least two peptides and a FDR <1%. Quantitative (p < 0.05) and qualitative (only detected in one condition) changes in protein abundance were studied (Tables S1 and S2). To analyze the possible harmful effects of 3-NT, a first comparison was made between the proteomes from non-treated control cells versus 3-NTtreated cells. In this case, 150 proteins were significantly influenced by 3-NT, with 56 among those being found in lower abundance in cells treated with 3-NT, and 15 being only found in non-treated control samples. On the other hand, 63 proteins were detected in higher quantity in cells exposed to 3-NT and 16 were only found in 3-NT-treated cells.
To study the hypothetical protective role of EA against 3-NT, an additional comparison was made between the proteomes from 3-NT-treated cells versus 3-NT+EA-treated cells. In this case, 1214 proteins were significantly influenced by 3-NT+EA; among those, 481 were detected in lower abundance in cells exposed to 3-NT+EA, 365 were found in higher quantity in 3-NT+EA treated cells, 23 were only detected in the presence of 3-NT+EA and 345 were only found in 3-NT treated cells. For a systematized and comprehensible description and discussion of results, discriminating proteins were grouped by biological processes and molecular functions. The comparison between non-treated control cells versus 3-NT-treated cells (possible harmful effect of 3-NT) is shown in Graphs SG1A-SG1D

Proteomic Analyses
The MS-based proteomic platform enabled the identification of 2398 proteins in total. All these proteins were identified with at least two peptides and a FDR <1%. Quantitative (p < 0.05) and qualitative (only detected in one condition) changes in protein abundance were studied (Tables S1 and S2). To analyze the possible harmful effects of 3-NT, a first comparison was made between the proteomes from non-treated control cells versus 3-NT-treated cells. In this case, 150 proteins were significantly influenced by 3-NT, with 56 among those being found in lower abundance in cells treated with 3-NT, and 15 being only found in non-treated control samples. On the other hand, 63 proteins were detected in higher quantity in cells exposed to 3-NT and 16 were only found in 3-NT-treated cells.
To study the hypothetical protective role of EA against 3-NT, an additional comparison was made between the proteomes from 3-NT-treated cells versus 3-NT+EA-treated cells. In this case, 1214 proteins were significantly influenced by 3-NT+EA; among those, 481 were detected in lower abundance in cells exposed to 3-NT+EA, 365 were found in higher quantity in 3-NT+EA treated cells, 23 were only detected in the presence of 3-NT+EA and 345 were only found in 3-NT treated cells. For a systematized and comprehensible description and discussion of results, discriminating proteins were grouped by biological processes and molecular functions. The comparison between non-treated control cells versus 3-NT-treated cells (possible harmful effect of 3-NT) is shown in Graphs SG1A-SG1D in the Supplementary Material. The comparison between 3-NT-treated cells versus 3-NT+EAtreated cells (3-NT vs. hypothetical protective role of EA) is shown in Graphs SG2A-SG2D in the Supplementary Material. Specific terms for each of these processes and full details of discriminating proteins and associated genes are provided in Supplementary Material Tables S3-S10. Tables 1 and 2 show a selection of representative proteins from each relevant biological process and molecular function affected by the presence of 3-NT and 3-NT+EA. Only discriminating proteins having a defined biological significance are presented in the following sections. 1 Fold change (FC) indicates the degree of quantity change for a particular protein between cells (control vs. 3-NT); FC <1 denotes a decrease in the concentration of protein in the treated sample (vs. control); FC >1 indicates a significant increase in the concentration of the protein in treated sample (vs. control). When a given protein is only present in one of the groups, fold change cannot be measured and such condition is denoted as "C" (if protein is only present in C cells) or 3-NT (if protein is only present in 3-NT-exposed cells). 3-NT3-NT.

Other Proteins of Biological Significance Found in Lower Relative Quantity in 3-NT-Treated Cells
Other biologically relevant proteins found in lower quantities were delta (24) snRNA-associated Sm-like protein LSm2 (LSM2) and factor 1 of TRPM8 channel-associated (TCAF1) (Tables 1 and S1).

Proteins Found in Higher Relative Quantity in Cells Exposed to 3-NT Compared to Control Counterparts Positive Regulation of Cyclic-Nucleotide Phosphodiesterase Activity
The biological process most widely affected by the exposure of 3-NT (87.01%, p < 0.01) was positive regulation of cyclic nucleotide phosphodiesterase activity (SG. 1C). Among the most affected pathways, we identified the nitric oxide synthase regulator activity, adenylate cyclase activator activity and positive and negative regulation of ryanodinesensitive calcium-release channel activity (Supplementary Table S5). The proteins affected in all these processes were calmodulin-1, calmodulin-2 and calmodulin-3 (CALM1, CALM2 and CALM3; Fold change: 1.29) (Tables 1 and S1-S5).

Type 3 Metabotropic Glutamate Receptor Binding
The most affected molecular function in the presence of 3-NT was type 3 metabotropic glutamate receptor binding (86.36%, p < 0.05) (SG. 1D) in which the aforementioned calmodulin proteins, are implicated (CALM1, CALM2 and CALM3) (Supplementary Table  S6). Other pathways affected were nitric oxide synthase regulator activity, adenylate cyclase activator activity and N-terminal myristoylation domain binding (Supplementary Table S6).

Proteins Found in Lower Relative Quantity in Cells Exposed to 3-NT+EA Compared to 3-NT Counterparts Intracellular Transport
The biological process most influenced by the exposure of 3-NT+EA (18.72%, p<0.01) was intracellular transport (SG. 2A). The cytoplasmic translational initiation process was one of the most affected in proteins such as subunit alpha of translation initiation factor eIF-2B (EIF2B1, only in 3-NT), subunit 2 of eukaryotic translation initiation factor 2(EIF2S2, only in 3-NT), eukaryotic initiation factor 4A-II (EIF4A2, fold change: 0.43), eukaryotic translation initiation factor 4E (EIF4E, fold change: 0.46), RNA-binding protein 4 (RBM4, fold change: 0.73) and monocarboxylate transporter 1 (SLC16A1, fold change: 0.40), among others (Tables 2 and S2-S7). These proteins are also implicated in other processes such as translational initiation, protein targeting to ER (endoplasmic reticulum) translation regulator activity and nucleic acid binding (Supplementary Table S7), which were, for this reason, also affected by 3-NT+AE exposure. Ribosomal proteins were also found in lower abundance such as La

Purine Nucleotide Metabolic Process
Proteins involved in the metabolic process of purine nucleotides were found at lower quantity in 3-NT+EA treated cells compared to 3-NT samples (8.15%, p < 0.01) (SG. 2A). Specific routes such as the cristae formation, mitochondrial ATP synthesis coupled proton transport and energy coupled proton transport, down electrochemical gradient, were severely decreased. Some relevant proteins in the aforementioned processes are the subunits alpha (ATP5F1A), beta (ATP5F1B) and gamma (ATP5F1C) of the mitochondrial ATP synthase (fold changes: 0.17; 0.31 and 0.50, respectively). The subunit delta of the same mitochondrial ATP synthase, (ATP5F1D) was not found in enterocytes when EA was combined to 3-NT. Other relevant mitochondrial proteins were also found in lower abundance such as the subunit B1 of the mitochondrial ATP synthase F(0) complex (ATP5PB, fold change: 0.50), the subunit d of the mitochondrial ATP synthase, (ATP5PD, fold change: 0.46), the subunit O of the mitochondrial ATP synthase, (ATP5PO, fold change: 0.42), the subunit 5B of the mitochondrial cytochrome c oxidase, (COX5B, fold change: 0.61) and the mitochondrial stomatin-like protein 2 (STOML2, fold change: 0.38) (Tables 2 and S2-S7). Other remarkable proteins implicated in the ATP biosynthetic process, such as pyruvate kinase PKM (PKM, fold change: 0.70) and calcium-binding mitochondrial carrier protein Ar-alar2 (SLC25A13, fold change: 0.74), were in lower abundance in cells exposed to 3-NT+EA than in those exposed only to 3-NT. Interestingly, various relevant proteins involved in the mitochondrial membrane organization were also remarkably found in lower quantities when EA was combined with 3-NT. Among the latter, we may emphasize AFG3-like protein the mitochondrial MICOS complex, (fold changes: 0.07; 0.47, respectively, while the latter was only present in 3-NT-treated cells), DnaJ homolog protein (member 11, subfamily C) (DNAJC11, fold change: 0.33), the mitochondrial EF-hand domain-containing protein 1 (LETM1, fold change: 0.40), the mitochondrial inner membrane protein OXA1L (OXA1L, only in 3-NT) and sorting and assembly machinery component 50 homolog (SAMM50, fold change: 0.60) (Tables 2 and S2-S7).
Within the processes implicated in the regulating activity of nitric oxide synthase, the type 3 metabotropic glutamate receptor binding, the N-terminal myristoylation domain binding, the adenylate cyclase activator activity and the nitric oxide synthase regulator activity were some of the most affected in human enterocytes by exposure to 3-NT+EA. Proteins found in lower quantity were CALM1, CALM2, CALM3, subunit beta of calcium/calmodulin-dependent protein kinase type II (CAMK2B, fold change: 0.66), phosphatidylinositol-binding clathrin assembly protein (PICALM, only in 3-NT), subunit alpha of guanine nucleotide-binding protein G(s) isoforms XLas (GNAS, fold change: 0.67), ATPase 4 of plasma membrane calcium-transporting (ATP2B4, only in 3-NT), epidermal growth factor receptor (EGFR, only in 3-NT), heat shock protein HSP 90-alpha (HSP90AA1, fold change: 0.82) and heat shock protein HSP 90-beta (HSP90AB1, fold change: 0.68) (Tables 2 and S2-S8).

Organic Substance Catabolic Process
The catabolism of organic substances was also upregulated (10.78%, p < 0.01) by the influence of 3-NT+EA (SG. 2C). Proteins found in higher quantity in the catabolic process of L-serine were the iron-sulfur subunit of the mitochondrial succinate dehydrogenase

Other Proteins of Biological Significance Found in Higher Relative Quantity
Proteins of biological significance only found in 3-NT+EA treated cells that have not been included in the previous biological processes/molecular functions are alcohol dehydrogenase 4 (ADH4), glutathione S-transferase Mu 3 (GSTM3), ferroptosis suppressor protein 1 (AIFM2), heme oxygenase 1 (HMOX1) and subunit RPB1 of DNA-directed RNA polymerase II (POLR2A) (Tables 2 and S2).

Endogenous Antioxidant Enzyme Activity
Further to the full proteome of the cells under study, the activity of two major antioxidant enzymes, namely, catalase (CAT) and superoxide dismutase (SOD) was analyzed to gain further insight into the physiological response of the cell to the exposure to the tested dietary species (Figure 3). Consistent with the proteome results, the catalase activity in cells exposed to 3-NT-EA was significantly higher than in those treated with 3-NT alone, or in control. While superoxide dismutase was not identified as discriminating between treatments, the assessment of its activity revealed that enterocytes challenged with 3-NT+EA had significantly higher SOD activity than the other two counterparts.

Other Proteins of Biological Significance Found in Higher Relative Quantity
Proteins of biological significance only found in 3-NT+EA treated cells that have not been included in the previous biological processes/molecular functions are alcohol dehydrogenase 4 (ADH4), glutathione S-transferase Mu 3 (GSTM3), ferroptosis suppressor protein 1 (AIFM2), heme oxygenase 1 (HMOX1) and subunit RPB1 of DNAdirected RNA polymerase II (POLR2A) (Tables 2 and S2).

Endogenous Antioxidant Enzyme Activity
Further to the full proteome of the cells under study, the activity of two major antioxidant enzymes, namely, catalase (CAT) and superoxide dismutase (SOD) was analyzed to gain further insight into the physiological response of the cell to the exposure to the tested dietary species (Figure 3). Consistent with the proteome results, the catalase activity in cells exposed to 3-NT-EA was significantly higher than in those treated with 3-NT alone, or in control. While superoxide dismutase was not identified as discriminating between treatments, the assessment of its activity revealed that enterocytes challenged with 3-NT+EA had significantly higher SOD activity than the other two counterparts.

Discussion
Presently, the onset of oxidative and nitrosative stress is associated with multiple pathologies [1,2]. Dietary oxidized amino acids have been documented to induce redox imbalance and impairment of organic functions in human cells [15,21]. According to our hypotheses, 3-NT could be implicated in the assorted health disorders attributed to the intake of processed cured meat products, while the application of recognized antioxidants such as EA could contribute to counteract the redox-related biological impairments of 3-NT. The results from the present study revealed that 3-NT is certainly implicated in the impairment of specific biological functions in human enterocytes; however, the exposure to EA was not found to counteract such impairments. Contrary to our initial hypothesis, the combination of 3-NT and EA led to a severe impairment of the proteome of the enterocytes with remarkable increase in necrosis. Such cell injure can be visualized in the Picture S1 (Available as Supplementary Material). The most relevant biological impairments caused by 3-NT and 3-NT+EA in differentiated human enterocytes are discussed as follows. Figure 4 illustrates the metabolic pathways affected in these cells by the exposure to 3-NT and 3-NT+EA.
hypotheses, 3-NT could be implicated in the assorted health disorders attributed to the intake of processed cured meat products, while the application of recognized antioxidants such as EA could contribute to counteract the redox-related biological impairments of 3-NT. The results from the present study revealed that 3-NT is certainly implicated in the impairment of specific biological functions in human enterocytes; however, the exposure to EA was not found to counteract such impairments. Contrary to our initial hypothesis, the combination of 3-NT and EA led to a severe impairment of the proteome of the enterocytes with remarkable increase in necrosis. Such cell injure can be visualized in the Picture S1 (Available as Supplementary Material). The most relevant biological impairments caused by 3-NT and 3-NT+EA in differentiated human enterocytes are discussed as follows. Figure 4 illustrates the metabolic pathways affected in these cells by the exposure to 3-NT and 3-NT+EA.

Impact of 3-NT and 3-NT+EA on Calmodulin-Dependent Intracellular Signaling Pathway
The calmodulin protein family is a group of biologically relevant proteins, which play a central role in Ca 2+ cell sensing, ion channel regulation and molecular signaling pathways with multiple physiological implications [22]. The higher relative quantity of 3 calmodulin proteins suggests that 3-NT exposure would have, as a primary molecular mechanism, the impairment of downstream biological processes depending on such proteins. As identified by CLUEGO, 3-NT upregulated the cyclic nucleotide phosphodiesterase activity and type 3 metabotropic glutamate receptor binding activity, which directly affects the nitric oxide synthase regulator activity, the adenylate cyclase activator activity, the regulation of ryanodine-sensitive calcium-release channel activity and the N-terminal myristoylation domain binding. These processes and functions accounted for around 87% of all proteomic upregulations caused by 3-NT exposure, which highlights the relevance of these molecular mechanisms in the potential toxicological effects of this nitrosated species.
Type 3 metabotropic glutamate receptors are constituted by seven transmembrane domains coupled to the G-protein signaling system [23]. The adenylate cyclase activator activity enables the formation of cAMP from ATP [24]. On the other hand, cyclic nucleotide phosphodiesterase activity is responsible for eliminating the excess of cAMP in the cell by converting it into 5 -AMP [24]. cAMP activates the catalytic subunits of protein kinase A (PKA), which goes to the nucleus, activating transcription factors [25]. PKA is also involved in the activation of ryanodine receptors (RYRs) [26], which can also be activated by IP3 through the phospholipase C pathway [27]. RYRs release calcium from rough endoplasmic reticulum (RER) to the cytosol, which, in turn, activates calmodulin proteins [26]. All these processes seemed to be increased by the effect of 3-NT exposure, and the consequences of such up-regulation have been described as harmful for enterocytes. In a recent study, Cunningham et al. [28] reported that activation of the calmodulin downstream molecular pathway in intestinal cells is initiated by an epithelial injury and leads to cell cycle arrest and inflammation processes. The inactivation of such molecular pathway protected against murine inflammatory bowel disease and colitis. It is, therefore, reasonable to consider that the higher abundance of calmodulin proteins found in enterocytes when treated with 3-NT may have harmful biological consequences. Consistently, proteins corresponding to DNA repair and RNA splicing were in higher quantities as compared to control counterparts. 3-NT-treated enterocytes showed impaired processes responsible for carrying out a normal synthesis of DNA and proteins (Table 1). These results are compatible with impaired cell cycle regulation. Consistently with our current study, Santulli et al. [27] observed the activation of RyRs by oxidative and nitrosative species and emphasized that these molecular mechanisms are implicated in the onset of various pathological conditions such as diabetes mellitus, hypertension and skeletal muscle disorders, among others.
3-NT is also involved in increasing nitric oxide synthase regulatory activity. Nitric oxide synthase (NOS) is an enzyme that catalyzes the conversion of L-arginine to L-citrulline producing nitric oxide (NO) [29], and has a binding domain to calmodulin. The cyclicnucleotide phosphodiesterase activity, also increased in 3-NT-treated cells, is involved in the activation of inducible nitric oxide synthase (iNOS) via adenylate cyclase/protein kinase A dependent of stimulatory G-protein [29]. At high levels of NO, or in the presence of ROS, the formation of peroxynitrite and its potential cytotoxic effects are enhanced [30]. Peroxynitrite can damage a variety of molecules in cells, including DNA and proteins, leading to apoptosis and necrosis [30], a situation which, in fact, occurred in the present study. Peroxynitrite plays a crucial role in chronic inflammatory diseases, diabetes, cancer and neurodegenerative disorders, among others [30], affect T cells and negatively influence the immune response [31], corresponding with the aforementioned obtained results, where MHC class II-restricted antigen presentation for CD4 + T cell-dependent was affected. Currently, 3-NT is recognized as a biomarker of up-regulated inducible NO synthase in inflammation [32]. Interestingly, 3-NT, a product of the nitrosative stress to proteins, would contribute to inducing further nitrosative stress by the activation of specific metabolic routes.
The molecular function of N-terminal myristoylation domain binding is affected by 3-NT exposure. N-myristylation of proteins is a co-translational lipidic modification of many eukaryotic proteins [33]. Myristoylation directly regulates the biological activity of endothelial nitric oxide synthase (eNOS), increasing calcium flows, and activating the calcium/calmodulin complex (calcium/CaM). This process is involved in the regulation of cell-signaling pathways in biological processes such as carcinogenesis and immune function [33]. Thus, for all of the above mentioned, 3-NT can be considered as an agonist of type 3 metabotropic glutamate receptors in cells, which stimulates different relevant pathways.
Acting as such, 3-NT promotes the deregulation of cAMP and calcium metabolism, which affects, in turn, the normal operation of the immune system through calmodulin.
The results obtained in EA-treated cells were coherent with those found in the literature, in which the ability of EA to enhance calcium metabolism, apoptosis and necrosis in tumor cells, is reported (revised by Mohammadinejad et al., [34]). However, enterocytes treated with EA led to an attempt to reverse some of the above-mentioned impaired biological functions (Supplementary Table S8). It was ignored whether a higher dose of EA would be necessary to compensate the effects of 3-NT on these pathways. Yet, the injurious effects of the combination of both species at these concentrations on enterocytes in terms of oxidative stress and necrosis, discussed in due course, suggest that, unlike our initial hypothesis, EA does not exert a protective effect against 3-NT-induced toxicity, but the opposite.

Impact of 3-NT and 3-NT+EA on Immune System and Steroid Hormones
The immune system is a network of cells and biological processes that protects an organism from internal and external hazardous substances. The major histocompatibility complex (MHC) is involved in the codification of proteins found on the surface of the cell recognizing potential antigens. Antigenic peptide-loaded MHC class II molecules (peptide-MHC class II) are expressed on the surface of professional antigen-presenting cells (APCs) such as macrophages, B cells, dendritic cells and thymic epithelial cells [35]. As the primary barrier between human organism and its environment, intestinal epithelial cells, as observed in the present study, express MHC class II in their surface, although the functional consequences of this expression are not fully understood [36]. All these cells present the antigen to antigen-specific CD4 + T cells, a mechanism that is essential for a specific and effective immune response [35]. In this study, processes involved in the antigen processing and presentation of exogenous peptide antigen via MHC class II were affected by the exposure of 3-NT. Proteins implicated in the motility of vesicles through microtubules (MYH10) [37], blebs formation (ROCK2) [38], vesicle trafficking (SNX9) [39], reorganization of actin cytoskeleton, exocytosis and early steps of protein synthesis (EIF2S3, EIF2S3B and EIF3D) [40] (Table 1) were significantly reduced in abundance in the presence of 3-NT compared to control counterparts. The lower relative abundance of tau protein binding function in 3-NT-treated cells is consistent with the impairment of the MHC class II mediated process (Supplementary Table S4). The tau protein binding was decreased in DCTN1 protein, which is an essential cofactor involved in most cellular functions of the microtubule motor cytoplasmic dynein [41] and in GSK3B protein, an essential key in protein tau phosphorylation. It is, hence, reasonable to hypothesize that the decrease of microtubule polymerization prevents vesicles of antigenic peptide-loaded MHC class II molecules from reaching the cell surface. Other proteins related to immunity, such as IRF3 [42] or LRBA [43], were also found in lower quantities in the presence of dietary concentrations of 3-NT. These results suggest that 3-NT impairs the ability of enterocytes to act as APCs and effectively contribute to protection against biological hazards. This could be explained by i) an impaired synthesis of proteins related to immunity, and ii) an impaired transportation of such proteins to the cell surface. These results are consistent with those published by Birnboim et al. [32] and Ahsan [44], where 3-NT is linked to systemic autoimmunogenic conditions and is related to diseases associated with immunological reactions where formation of Tyr-nitrated proteins has a major role [45]. Thus, 3-NT has a toxicological effect in the immune system through its implication in the decrease of MHC class II-restricted antigen presentation for CD4+ T cell-dependent, which is, as aforementioned, necessary for a suitable immune response. In addition, the results obtained revealed that the incorporation of EA to enterocytes did not seem to have an impact on this biological impairment caused by 3-NT.
In relation to steroid hormones, the proteins HSD3B1 and HSD3B2, involved in oxidation and isomerization in the biosynthesis of hormonal steroids [46], were not found in 3-NT treated cells. Steroid hormones are involved in a number of processes such as immune functions, inflammation, the control of metabolism and sexual development through sex steroids and corticosteroids [47]. Interestingly, protein HSD17B10 was found in terms dihydrotestosterone 17-beta-dehydrogenase activity and lipid oxidation (Supplementary  Table S9) in higher abundance in 3-NT+EA-treated cells. This protein, with hydroxysteroid dehydrogenase activity in steroid hormones [48], may be positive to cells as it may counteract the drop in proteins HSD3B1 and HSD3B2 caused by 3-NT. Therefore, EA may enable physiological biosynthesis of hormonal steroids in enterocytes and counteract, at some extent, the impairment caused by 3-NT.

Impact of 3-NT and 3-NT+EA on Antioxidant Defenses and Oxidative Stress
The analysis of specific protein oxidation markers in enterocytes revealed that the combination of 3-NT and EA remarkably promoted the onset of oxidative stress in enterocytes. These indicators of the oxidative damage to proteins are formed as a result of the oxidative deamination of alkaline amino acids (i.e., lysine) in the presence of ROS [49]. It is known that decreased and inefficient mitochondrial activity is linked to ROS generation, and, consequently, to oxidative stress [50]. Enterocytes exposed to 3-NT+EA showed a significant decrease in mitochondrial cristae formation and a lower abundance in proteins related to mitochondrial ATP synthase. The proteins found in lower abundance in 3-NT+EA-treated cells were related to the MICOS complex (mitochondrial contact site and cristae organizing system) ( Table 2). MICOS is a multi-subunit complex found in the inner mitochondrial membrane. Mitochondrial function and its architecture are closely related because an anomalous mitochondrial architecture leads to mitochondrial dysfunction [51]. MICOS is involved in human diseases such as amyotrophic lateral sclerosis, Alzheimer's and Parkinson's disease [51]. In agreement with the present results, several previous studies reported that EA inhibits mitochondrial respiration and had a negative impact on the restoration of both GSH and ATP, and on mitochondrial complex II function [52,53]. The remarkable impact of EA in promoting oxidative stress in 3-NT-treated enterocytes (as measured by protein oxidation) could be a plausible consequence of the disturbance effect of the phytochemical on mitochondrial activity. This could be one likely mechanism behind the profusely reported pro-oxidant activity of ellagic acid [54]. It is known that such pro-oxidant action depends on a number of factors, including the dose and occurrence of oxygen molecules and transition metals [54]. The present study reveals that 3-NT activates this pro-oxidant mechanism as exposure to EA alone, had, in general, protective effects on enterocytes against oxidative stress, apoptosis and necrosis ( Figure 1E-G). The molecular mechanism by which the combination of both species leads to such severe oxidative stress and necrosis is yet to be elucidated.
The oxidative damage to proteins measured in 3-NT+EA cells apparently contrasts with the lack of differences in ROS as assessed by flow cytometry. The lack of correspondence between both measurements is consistent with a previous study performed in human enterocytes exposed to 3-NT [9]. It is reasonable to hypothesize that, at the point of cell harvesting, ROS were already depleted while the effects of their pro-oxidative effects (oxidative damage and necrosis) were noticeable. Furthermore, the probe employed for cytometric ROS measurement detects, specifically, superoxide radicals, concentration of which commonly reflects mitochondrial activity. An impaired mitochondrial respiration is, in fact, compatible with the present results, which would actually explain the creation of a cellular pro-oxidative environment with other ROS that would, in turn, lead to the remarkable accretion of oxidized proteins.
The effect of 3-NT+EA on promoting the occurrence of detoxifying enzymes such as GSS and GSR (both involved in glutathione metabolism), GSTM3 and CAT in enterocytes, supports the hypothesis of enterocytes reinforcing the endogenous antioxidant defenses in response to the pro-oxidative environment caused by the combination of both chemical species. These enzymes are known to increase in cells in response to oxidative stress [55,56]. Furthermore, ADH4, which is well known for playing a relevant role in detoxifying ROS in enterocytes as a primary barrier to dietary pro-oxidative species [57], was only found in cells exposed to both 3-NT and EA. The proteomic results are consistent with additional analysis of the endogenous antioxidant activity of CAT and SOD. The activity of both enzymes was significantly higher in cells treated with 3-NT'+EA than in the other two groups of cells (3-NT, control). However, this attempt to strengthen the endogenous antioxidant defenses failed, as observed in the extent of oxidative damage to proteins and the severe necrosis found in enterocytes exposed to 3-NT+EA.

Impact of 3-NT and 3-NT+EA in Cell Viability, Apoptosis, Necrosis and Protein Repair
The flow cytometry assessment revealed that 3-NT caused a significant decrease in cell viability (p < 0.05) and a significant increase in necrosis (p < 0.05) ( Figure 1A). As discussed above, the cytotoxicity exerted by 3-NT on enterocytes may be derived from the upregulation of calcium-dependent biological functions that may have caused physiological impairments, including the onset of oxidative stress. While the increase of ROS was not detected by flow cytometry in 3-NT-treated cells, the accretion of oxidized proteins was significant as compared to CONTROL cells. The up-regulation of type 3 metabotropic glutamate receptors in the presence of 3-NT led to a higher abundance in a number of enzymes involved in important pathways of cellular metabolism, including nitric oxide synthase. This enzyme increases the amount of NO in cells, which makes the formation of RNS species possible, along with 3-NT, such as peroxynitrite. This species can damage DNA and proteins, increasing apoptosis and necrosis [30,31], with necrosis being the most characteristic mechanism induced by peroxinitrite [30]. In a previous study [9], a significant decrease of cell viability was accompanied by a significant increase in necrosis. Other studies have reported the same situation in cell cultures and animal models [8,58].
The significant decrease in cell viability (p < 0.05) in enterocytes exposed to 3-NT and EA, is reflected in a remarkable significant increase in necrosis compared to 3-NT treated cells (p < 0.01) ( Figure 1C). Cells exposed to 3-NT and EA seemed to enter into necrosis, which shows the severe chemical insult caused by the combination of these two chemical species. On the other hand, the significant decrease in apoptosis (p < 0.01) may be caused by of the ability of EA to counteract the upregulation of type 3 metabotropic glutamate receptors caused by 3-NT. Yet, the most plausible means of cytotoxicity caused by 3-NT+EA, which would explain the remarkable increase in oxidized proteins and cell necrosis, would be the severe mitochondrial disturbance. Such physiological impairment was manifested in the proteome by downregulation of the MICOS complex and other proteins implicated in the mitochondrial respiratory chain ( Table 2). As aforementioned, an abnormal mitochondrial conformation leads to mitochondrial dysfunction [51] and the onset of oxidative stress, which would explain, in turn, the significant and remarkable accretion of oxidized proteins and increased necrosis. Indeed, the protein CHCHD3, which plays an important role in maintaining the stability of the MICOS complex and the morphology of mitochondrial cristae [59], is remarkably diminished in 3-NT+EA-treated enterocytes along with other proteins related to this complex (Table 2).
Finally, the process of protein repair was also affected by the exposure to 3-NT+EA. Proteins related to the proteasome complex involved in the proteolytic degradation of most intracellular proteins [60] were found in higher abundance in the presence of 3-NT+EAtreated cells. Other proteins involved in post-replication repair and DNA damage checkpoint activation, such as RPA3 [61], were also increased in quantity as compared to 3-NTtreated cells. These results are coherent with the aforementioned attempt of the enterocytes to counteract the severe oxidative insult induced by the combination of 3-NT and EA. These efforts were, however, insufficient to avoid massive cell necrosis caused by these two chemical species.

Conclusions
We describe, for the first time, the underlying molecular mechanisms of the toxicological effects of 3-NT in human enterocytes. While this proteomics-based approach enables the study of these noxious effects from a comprehensive perspective, the confirmation of the impairment of all these downstream paths and interconnections of complex cellular mechanisms requires further specific studies. According to the generated data, the nitrosated amino acid 3-NT impairs several calcium-dependent physiological processes and affects the ability of enterocytes to act as antigen-presenting cells, compromising the immune response against potential biological threats. While it could be thought that EA would revert some of these mechanisms, the combination of 3-NT and the phytochemical led to compromised cell viability through anomalous mitochondria conformation and functionality, leading to severe oxidative stress and massive cell necrosis. These results highlight the importance of the study of dietary nitrosated amino acids in intestinal cell physiology, as they commonly occur in processed meat/dairy products where nitrite is used. Furthermore, the toxicological effects of 3-NT+EA should be studied for potential therapeutic studies in which cell death is deliberately induced.