Protein Susceptibility to Peroxidation by 4-Hydroxynonenal in Hereditary Hemochromatosis

Iron overload caused by hereditary hemochromatosis (HH) increases free reactive oxygen species that, in turn, induce lipid peroxidation. Its 4-hydroxynonenal (HNE) by-product is a well-established marker of lipid peroxidation since it reacts with accessible proteins with deleterious consequences. Indeed, elevated levels of HNE are often detected in a wide variety of human diseases related to oxidative stress. Here, we evaluated HNE-modified proteins in the membrane of erythrocytes from HH patients and in organs of Hfe−/− male and female mice, a mouse model of HH. For this purpose, we used one- and two-dimensional gel electrophoresis, immunoblotting and MALDI-TOF/TOF analysis. We identified cytoskeletal membrane proteins and membrane receptors of erythrocytes bound to HNE exclusively in HH patients. Furthermore, kidney and brain of Hfe−/− mice contained more HNE-adducted protein than healthy controls. Our results identified main HNE-modified proteins suggesting that HH favours preferred protein targets for oxidation by HNE.

Iron toxicity is triggered by the accelerated generation of highly reactive hydroxyl radical (-OH) in the Fenton and Haber-Weiss reactions [5], causing oxidative damage to cells. Several studies have shown that oxidative stress plays an important role in the development of HH pathology [6][7][8][9][10][11][12], including ferroptosis, described as cell death elicited by lipid peroxidation under the catalysis of iron ions [13]. Increased circulating iron results in lipid peroxidation production above normal with increased protein oxidation [14][15][16]. Lipid peroxidation as a free radical chain reaction is induced by reactive oxygen species and produces a wide variety of oxidation products, among which HNE is the most toxic [5]. HNE reacts with several cellular components, such as DNA, proteins and other accessible molecules. High HNE concentrations cause cell cycle and differentiation arrest, concluding in cell death [17,18]. Although protein modification by HNE take place mainly by Michael

Identification of HNE-Modified Proteins in Erythrocyte Membrane of HH Patients
Potential oxidative damage to erythrocyte membrane proteins in human HH was analysed by 2D immunoblotting to detect HNE oxidative adducts of isolated membrane proteins ( Figure 1A). Spots of interest were excised and digested with trypsin from the stained spots at the Coomassie blue-stained duplicate gels ( Figure 1B and Supplementary Figure S1) and analysed by MALDI-TOF/TOF to generate a peptide mass fingerprint.
Different HNE-bound protein spots were identified on erythrocytes samples from eight HH patients and healthy control people ( Figure 1A). While five protein spots were common in both patients and controls (c, e, h, g, l), six protein spots were observed exclusively in the HH patients (a, b, d, f, I, j) and only one was present uniquely in the healthy controls (m). Patients numbered 1 to 4 were adults aged 41 to 50 years (see Section 4), while those numbered 5 and 6 were children aged 13 and 10 years. common in both patients and controls (c, e, h, g, l), six protein spots were observed exclusively in the HH patients (a, b, d, f, I, j) and only one was present uniquely in the healthy controls (m). Patients numbered 1 to 4 were adults aged 41 to 50 years (see methods section), while those numbered 5 and 6 were children aged 13 and 10 years.  Table 1 summarises the sample distribution of the nine identified proteins. The seven patient-specific proteins were guanine nucleotide-binding protein g(i)/g(s)/g(t) subunit beta-1, actin 1, actin 2, spectrin alpha chain, ankyrin 1, CD55 and CD44. The only two proteins found in both groups were spectrin beta chain and band 3 anion transport protein. No unique protein was found in the healthy controls.  Table 1 summarises the sample distribution of the nine identified proteins. The seven patient-specific proteins were guanine nucleotide-binding protein g(i)/g(s)/g(t) subunit beta-1, actin 1, actin 2, spectrin alpha chain, ankyrin 1, CD55 and CD44. The only two proteins found in both groups were spectrin beta chain and band 3 anion transport protein.
No unique protein was found in the healthy controls.
Briefly, the proteins only identified in sample patients correspond to the following cell functions: guanine nucleotide-binding protein G(I)/G(S)/G(T) subunit beta-1 (P62873) is a modulator in transmembrane signalling systems required for the GTPase activity, for replacement of GDP by GTP and for G protein-effector interaction; cytoskeletal proteins actin cytoplasmic 1 and actin cytoplasmic 2 (beta actin, P60709 and gamma actin, P63261) contribute to cell motility and various biological processes such as sensing environmental forces, vesicular transport, moving over surfaces and cell division; ankyrin 1 (fragment H0YBS0; P16157) attaches integral membrane proteins to cytoskeletal elements; spectrin α chain (P02549) is the major constituent of the cytoskeletal network underlying the ery-throcyte plasma membrane; CD55 (H3BLV0) regulates the complement system preventing the formation of the membrane attack complex; CD44 antigen (H0YD13) is a transmembrane glycoprotein involved in cell-cell communication, cellular adhesion and migration. Spectrin beta chain (H0YJE6, P11277) was identified in both HH patients and healthy individuals, which could suggest a high susceptibility of this protein to modification by HNE. In addition, band 3 anion transporter protein (P02730) was also found to be modified by HNE in both control and patient samples and being an important integral erythrocyte membrane glycoprotein could also act cooperatively with the cytoskeleton. It should be noted that some proteins (spectrin beta chain, band 3 anion transport protein and ankyrin 1) were identified at more than one spot, suggesting post-translational modification or physiological proteolysis that results in molecular weight and/or charge of the corresponding protein.

Hfe −/− Mice: A Model to Explore HH Tissue Damage
The previously described Hfe −/− mouse model of HH [49][50][51] was followed at three time points (3, 5 and 7 months of age) and in both sexes to discriminate potential damage accumulated with age and in gender susceptibility. Haemoglobin concentration in peripheral blood was significantly higher in Hfe −/− mice than in wild-type mice in both sexes at all ages ( Figure 2A). Baseline haemoglobin values (mean ± standard deviation) for the control females were 12.98 ± 0.34 (g/dL) and for males 12.65 ± 0.31 (g/dL). Liver expression of hamp1 mRNA, encoding peptide hepcidin-1 that limits intestinal iron absorption and iron recycling by macrophages to maintain iron homeostasis [52], was significatively reduced in Hfe −/− mice at the 5-and 7-month age when considering both sexes ( Figure 2B). of hamp1 mRNA, encoding peptide hepcidin-1 that limits intestinal iron absorption and iron recycling by macrophages to maintain iron homeostasis [52], was significatively reduced in Hfe −/− mice at the 5-and 7-month age when considering both sexes ( Figure 2B). Hamp (hepcidin gene) mRNA expression is also shown without gender separation (bottom right). The data for each Hfe −/− mouse were normalised to controls of the same age and sex (indicated by dotted lines). Representative graphs show mean ± SEM from 2 independent experiments (each experiment n = 6 per bar/sex group). Statistical significance is indicated as **** p < 0.0001, ** p < 0.01 and * p < 0.05 in comparison to age-and sex-matched controls.

HNE Modifications in Kidney, Brain, Heart and Liver Proteins
The extent of protein carbonylation mediated by HNE in the kidney, brain, heart and liver of male and female Hfe −/− mice at 3, 5 and 7 months of age was assessed by Western blotting using anti-HNE antibodies. More differences were found in the protein carbonylation patterns of each gender than between the Hfe −/− and control mice. A pattern of five bands of modified kidney proteins at 94, 81, 63, 54 and 43 kDa was distinguished in males but only a 94 kDa band in females ( Figure 3A). The greatest HNE immunoreactivity with high number of protein bands and strong signal was observed in the brain. Both female and male mice showed six bands at 129, 104, 94, 76, 64 and 32 kDa. Furthermore, while four HNE-immunoreactivity bands were found in male heart at 53, 41, 32 and 23 kDa, such signals were barely visible in females. On the other hand, HNE signals in liver were weak and almost imperceptible in both sexes, hence, this organ was not considered in protein identification studies. Immunoblots of biological replicates are shown in Supplementary Figures S2-S7. To further identify HNE-modified proteins in the kidney, brain and heart, the corresponding bands were excised from the Coomassie brilliant bluestained parallel gel, digested, immuno-enriched and analysed by MALDI TOF/TOF. Hamp (hepcidin gene) mRNA expression is also shown without gender separation (bottom right). The data for each Hfe −/− mouse were normalised to controls of the same age and sex (indicated by dotted lines). Representative graphs show mean ± SEM from 2 independent experiments (each experiment n = 6 per bar/sex group). Statistical significance is indicated as **** p < 0.0001, ** p < 0.01 and * p < 0.05 in comparison to age-and sex-matched controls.

HNE Modifications in Kidney, Brain, Heart and Liver Proteins
The extent of protein carbonylation mediated by HNE in the kidney, brain, heart and liver of male and female Hfe −/− mice at 3, 5 and 7 months of age was assessed by Western blotting using anti-HNE antibodies. More differences were found in the protein carbonylation patterns of each gender than between the Hfe −/− and control mice. A pattern of five bands of modified kidney proteins at 94, 81, 63, 54 and 43 kDa was distinguished in males but only a 94 kDa band in females ( Figure 3A). The greatest HNE immunoreactivity with high number of protein bands and strong signal was observed in the brain. Both female and male mice showed six bands at 129, 104, 94, 76, 64 and 32 kDa. Furthermore, while four HNE-immunoreactivity bands were found in male heart at 53, 41, 32 and 23 kDa, such signals were barely visible in females. On the other hand, HNE signals in liver were weak and almost imperceptible in both sexes, hence, this organ was not considered in protein identification studies. Immunoblots of biological replicates are shown in Supplementary  Figures S2-S7. To further identify HNE-modified proteins in the kidney, brain and heart, the corresponding bands were excised from the Coomassie brilliant blue-stained parallel gel, digested, immuno-enriched and analysed by MALDI TOF/TOF. Homogenates from each organ were separately resolved on 12% polyacrylamide gels. HNE (4-hydroxynonenal) modified proteins were detected by Western blot using HNE specific antibodies. (A) Western blot analysis of HNE adducts in kidney, brain, and heart proteins from female and male control (C) and Hfe −/− (H) mice at 3, 5 and 7 months of age. Images are representative of n = 3 per group. (B) Coomassie brilliant blue-stained gels where HNE-immunoreactive bands protein bands were excised. Arrows indicate the protein bands that were excised from the gel and the molecular weight are expressed in KDa on the left side of the arrow.
In the brain, the six HNE-immunoreactive bands were equally detected in male and female mice ( Figure 3A). Bands 1 and 4 were both identified as Na + /K + -exchanging ATPase subunit alpha-3 which, as a complete enzyme (EC 7.2.2.13), catalyses the hydrolysis of ATP coupled with the exchange of sodium and potassium ions across the plasma membrane (Table 2). Band 3, mitochondrial aconitase hydratase (EC:4.2.1.3), catalyses the conversion of citrate to isocitrate in the tricarboxylic acid (TCA) cycle. In band 5, two proteins were identified: heat shock cognate 71 kDa protein with a key role in protein quality control system and serum albumin, which was also found in kidney as band 3. In band 6 was identified mitochondrial malate dehydrogenase (EC:1.1.1.37) which, as the final step in the TCA cycle, reversibly catalyses the NAD/NADH-dependent oxidation of malate to oxaloacetate. In the heart, a higher number of bands detected by anti-HNE antibody was found in male than in female mice ( Figure 3A). Within band 1, the alpha and beta subunits of mitochondrial membrane ATP synthase (F1F0 ATP synthase or Complex V: EC 7.1.2.2) were identified in males and females ( Table 2). The rest of HNE-immunoreactive bands were only found in male tissue: band 2, mitochondrial creatine kinase-S-type (EC:2.7.3.2), catalyses the transfer of phosphate between ATP and various phosphogens (e.g., creatine phosphate) playing a central role in energy transduction; band 3, mitochondrial malate dehydrogenase (EC 1.1.1.37) which was also found in the brain as band 6; band 4, with two proteins identified: mitochondrial ATP synthase F(0) complex subunit B1 (F1F0 ATP synthase or Complex V: EC 7.1.2.2) and apolipoprotein A-I, this last participating in the reverse transport of cholesterol from tissues to the liver for excretion.
Additionally, semi-quantitative analysis of anti-HNE reactivity signals of Hfe −/− in comparison to control samples (ran in the same Western blot: Figure 3A and Supplementary  Figures S2-S7) was performed by densitometry of those bands listed in Table 2. Since band intensity remained unchanged within age groups, data are shown grouped according by HH status although, when statistically significant, they are also compared by sex (Figure 4).  Figure 3A and identified in Table 2 from (A) kidney, (B) brain and (C) heart of control (C) and Hfe −/− (H) mice. Analysis of male (m) and female (f) mice is shown separately only in cases of statistically significant results. When band analysis is shown for males only, it is indicated in brackets (males) at the top. Total Hfe −/− mice or control n = 18; male/female Hfe −/− or control mice n = 9. Values are expressed as relative band intensity (%) considering 100% the mean value of n = 18 control mice or n = 9 male/female control mice as indicated (mean ± SEM). Statistical comparisons were performed between groups using Student unpaired t-tests or Mann-Whitney U-tests. * p < 0.05, *** p ≤ 0.001.
In the kidney, levels of total HNE-modified protein signal from Hfe −/− mice were significantly higher (p = 0.001) than those found in healthy mice ( Figure 4A, kidney). Of the five HNE-modified protein bands from the kidney, only mitochondrial sarcosine dehydrogenase was also present in females ( Figure 3A). Its band intensity shown in Figure 4A (band 1) of all Hfe −/− mice irrespective of gender (p = 0.0156) and in Hfe −/− females (p = 0.04) was significantly higher than in the control mice. Similarly, HNE modification was greater in aconitase hydratase protein ( Figure 4A band 2) from male Hfe −/− mice (p = 0.022) than in the matched control mice. In the brain, the six bands observed in both sexes showed an overall increase in HNE-protein signal in Hfe −/− mice than in the controls ( Figure 4B brain). However, when treated individually, only mitochondrial aconitate hydratase of males showed a significant intensity increase than in the control mice ( Figure 4B band 3). Finally, the HNE signal of the heart proteins showed no differences between Hfe −/− and the control mice, either globally ( Figure 4C heart) or in any of the individual bands.

Hepatic Expression of Iron-Related mRNA Genes in Hfe −/− Mice
To understand liver dysfunction caused by HH in mice associated to the oxidative modifications observed, the expression of a group of four genes related to the disease was also analysed at 3, 5 and 7 months of age ( Figure 5). Hepatic gene expression of two cytokines, tumour necrosis factor-α (tnf ) acting as hepcidin mRNA inhibitor [53] and interleukin 6 (il6) acting as hepcidin mRNA inducer [54] showed no significant changes. Glutathione peroxidase-1 (gpx1) and superoxide dismutase 2 (sod2) encoding two oxidative stress response enzymes showed some changes. Thus, gpx1 at 3-month-old female Hfe −/− mice showed a significant enhanced gpx1 expression compared to the control and Hfe −/− males of the same age. Hfe −/− males showed a lower but significative gpx1 expression at 5 months of age, but notably increased at 7 months of age. In the case of sod2, no expression differences were observed between any group. In the kidney, levels of total HNE-modified protein signal from Hfe −/− mice were significantly higher (p = 0.001) than those found in healthy mice ( Figure 4A, kidney). Of the five HNE-modified protein bands from the kidney, only mitochondrial sarcosine dehydrogenase was also present in females ( Figure 3A). Its band intensity shown in Figure 4A (band 1) of all Hfe −/− mice irrespective of gender (p = 0.0156) and in Hfe −/− females (p = 0.04) was significantly higher than in the control mice. Similarly, HNE modification was greater in aconitase hydratase protein ( Figure 4A band 2) from male Hfe −/− mice (p = 0.022) than in the matched control mice. In the brain, the six bands observed in both sexes showed an overall increase in HNE-protein signal in Hfe −/− mice than in the controls (Figure 4B brain). However, when treated individually, only mitochondrial aconitate hydratase of males showed a significant intensity increase than in the control mice ( Figure 4B band 3). Finally, the HNE signal of the heart proteins showed no differences between Hfe −/− and the control mice, either globally ( Figure 4C heart) or in any of the individual bands.

Hepatic Expression of Iron-Related mRNA Genes in Hfe −/− Mice
To understand liver dysfunction caused by HH in mice associated to the oxidative modifications observed, the expression of a group of four genes related to the disease was also analysed at 3, 5 and 7 months of age ( Figure 5). Hepatic gene expression of two cytokines, tumour necrosis factor-α (tnf) acting as hepcidin mRNA inhibitor [53] and interleukin 6 (il6) acting as hepcidin mRNA inducer [54] showed no significant changes. Glutathione peroxidase-1 (gpx1) and superoxide dismutase 2 (sod2) encoding two oxidative stress response enzymes showed some changes. Thus, gpx1 at 3-month-old female Hfe −/− mice showed a significant enhanced gpx1 expression compared to the control and Hfe −/− males of the same age. Hfe −/− males showed a lower but significative gpx1 expression at 5 months of age, but notably increased at 7 months of age. In the case of sod2, no expression differences were observed between any group. Comparative liver mRNA expression of tnf, il6, gpx1 and sod2. Relative expression of tumour necrosis factor-α (tnf), interleukin 6 (il6), glutathione peroxidase-1 (gpx1) and superoxide dismutase 2 (sod2) in female and male 3-, 5-and 7-month-old Hfe −/− mice (H) and healthy controls (C) Figure 5. Comparative liver mRNA expression of tnf, il6, gpx1 and sod2. Relative expression of tumour necrosis factor-α (tnf ), interleukin 6 (il6), glutathione peroxidase-1 (gpx1) and superoxide dismutase 2 (sod2) in female and male 3-, 5-and 7-month-old Hfe −/− mice (H) and healthy controls (C) was calculated by the 2 −∆∆CT method. The results are expressed as arbitrary units normalised to β-actin to correct for mRNA quantity and integrity. The dotted lines represent normalized values of healthy controls. Data are shown as mean ± SEM, each with n = 6 per group, where * p ≤ 0.05 comparing to same-age controls, and ## p < 0.01 compared to same age Hfe −/− mice but different gender.

Discussion
Our study shows a total of seven membrane proteins modified by HNE exclusively in HH patients, whereas no HNE target protein was identified singularly in healthy controls, suggesting a more oxidative circulating environment in HH. Furthermore, since HNE easily crosses membranes and protein-HNE adducts are detectable in all cellular compartments [55], the increased carbonylation observed in HH patients may probably be extensible to the interior of the red cell.
Two of the proteins modified in HH patients in this study were guanine nucleotidebinding protein g(i)/g(s)/g(t) subunit beta-1, and actin, both described as carriers of HNE binding sites [56,57]. Actin-HNE adducts in red cell membrane have also been detected in autism, suggesting a link between erythrocyte shape abnormalities, membrane oxidative stress damage and actin alteration [29]. Ankyrin that was observed as HNE modified in HH patients has been found carbonylated in renal disease, G6PD deficiency, sickle cell trait or malaria infection [58][59][60], suggesting a highly susceptible cytoskeleton protein in oxidative environments. Spectrins, also components of the membrane skeleton bound to ankyrin, band 4.1 and actin, and responsible for erythrocyte shape and membrane lipid asymmetry [61,62], were found to be modified in the HH patients and the β-chain was also found in the healthy patients, consistent with being a main target of HNE adduction in intact exposed erythrocytes [63]. Our results suggest a link between oxidative damage of erythrocyte membrane proteins and erythrocyte shape abnormalities [46,47], described in HH as caused by structural defects in the membrane skeleton [64] and in HNE-treated erythrocytes [65]. On the other hand, the biological significance of the HNE modifications found in CD55 and CD44 antigens that were unique to erythrocytes from HH patients may be related to the reported increased intravascular haemolysis associated with an increased oxidative state that reduces CD55 expression [66]. A limitation of our study is the small number of human samples available. It would be desirable to increase the biological samples per group to associate the potential presence and amount of HNE-protein adducts with the pathology of the disease.
Band 3 anion transporter protein detected at different spots on 2D gels in both HH patients and healthy controls can be explained by the fragmentation of both of its domains [67] as described during erythrocyte senescence [68]. In fact, both fragment sizes are distinguished in fresh extracts from erythrocyte membrane proteins [68]. Alternatively, as a glycoprotein, the multiple spots of band 3 anion transporter could be due to a heterogeneous carbohydrate glycosylation shown by different mobilities by isoelectric focusing [69]. Moreover, since spectrin was found in three spots (f, g and h), with two of them (g and h) containing only its beta chain while the highest molecular weight spot (f) contained its alpha chain together with an ankyrin 1 fragment, it can be hypothesised that these different associated cytoskeleton spots are due to the retention of part of their original bonds even after electrophoresis [62].
In HH patients, the symptoms usually become apparent after 40-60 years of age [70] and biochemical markers are progressively altered with age [71]. To study the increased toxicity induced by the lipid peroxidative product HNE in organs, Hfe −/− mice was chosen as a model of progressive increase in hepatic and plasma iron concentrations, ferritin alteration and transferrin saturation over the course of months [72][73][74] as HH is a disease in which iron accumulates in different tissues, increasing oxidative stress. However increased oxidation was not observed with an age difference of 4 months in C57BL/6J mice, which as an aging model [75][76][77][78][79] may not be sufficient to show a cumulative effect as the few studies analysing the age-related formation of 4-HNE adducts [80] or ROS [81] in wild-type C57BL/6J mouse tissues show some difference only at 12 months of age. Therefore, our results in the first half of the mouse life span seem to correspond to the late onset of iron accumulation and oxidative products in humans and, therefore, this mouse model may allow us to observe those early changes that can later lead to accumulated tissue damage. Thus, in our study, increased oxidative stress damage in Hfe −/− mice was shown early by the high presence of HNE-modified proteins in the kidney and brain in comparison to the controls. This is consistent with the observed cause of kidney injury by circulating iron [82,83] and the increased oxidative stress environment observed in the brain of mice carrying an HFE mutation [84]. In this regard, human renal impairment has been described in some HH patients [85,86], while patients carrying the HFE C282Y mutation in homozygosis develop marked iron deposition in dementia-relevant brain areas [87].
HNE-immunoblot analysis of kidney, brain, liver and heart proteins showed that, among all proteins, mitochondrial sarcosine dehydrogenase in Hfe −/− female kidney (band 1) and mitochondrial aconitase hydratase in kidney (band 2) and brain (band 3) of Hfe −/− male had the highest and most distinct signals relative to those corresponding to wild-type mice. Sarcosine dehydrogenase produces free formaldehyde in the conversion of sarcosine to glycine and its HNE adducts have been already described [88]. It has been described that aconitase, a key enzyme of the Krebs cycle, undergoes reversible inactivation by ROS [89,90], being a known biomarker of mitochondrial oxidative stress [91], since in addition to being an essential organelle for ATP generation, it is also a major producer of ROS [92]. Iron-driven oxidation requires direct interaction with cellular reducing and oxidizing equivalents such as the enriched mitochondrial with superoxide and hydrogen peroxide suppliers of electrons. Consequently, increased oxidation is likely to occur in iron and free electron overloaded microenvironments such as mitochondria, where its dysfunction would be mediated [93,94].
Beyond our objectives, we observed sex-specific differences in the number of HNEadducted proteins in the kidney and heart. A higher level of oxidative stress injury induced by iron overload has been previously described in male mice compared to females in HJV knockout mice with high-iron diets [95], and in mice inoculated with iron-disodium salt [96]. The kidney proteins aconitase hydratase, albumin, methanethiol oxidase, mitochondrial methylmalonate-semialdehyde dehydrogenase and alcohol dehydrogenase were observed to be HNE modified only in male mice. Therefore, although it is not known whether it is more frequent in men, the observation of an acceleration of renal disease progression by elevated lipid peroxidation products such as HNE and malondialdehyde [83,97] is relevant. In the heart, mitochondrial enzymes creatine kinase-S-type and malate dehydrogenase, as well as apolipoprotein A-I were uniquely HNE modified in males. These early increased oxidative stress markers identified in male Hfe −/− mice may differentially contribute to the mitochondrial dysfunction recognised as key player in cardiac diseases [98].
Our study is the first to describe the presence of the lipid peroxidation product HNE in erythrocytes from HH patients. In the present study, we have discovered HNE-modified proteins in the membrane of erythrocytes from HH patients associated with structural functions of the cytoskeleton and membrane receptors that have previously been linked to the maintenance of erythrocyte shape and appearance. In addition, we studied the presence of HNE in different tissues of Hfe −/− mice and it was the kidney and brain that had more modified proteins than the healthy controls. Our study provides evidence for an increased presence of HNE-protein adducts in HH disease. Thus, the determination of certain HNEbound erythrocyte proteins in the clinical follow-up of HH patients could be useful as a marker of oxidative toxicity and for monitoring disease progression. Moreover, even further investigation of the potential loss of activity of HNE-modified proteins that have been identified in mouse organs could contribute to the understanding of the pathogenesis of HH sequelae. Thus, it would be worthwhile to further study the presence of erythrocyte and tissue markers related to oxidative stress to assist with treatment and to more precisely monitor the pathophysiological status of patients.

Human Subjects and Animal Model
Blood samples were obtained from six HH patients (Table 3). Patients 4, 5 and 6 were family related, being father, daughter and son, respectively. All HH patients were diagnosed by genetic testing. Two healthy volunteers served as a healthy control group. This study was approved by the Ethical Review Board at Research Institute Hospital 12 de Octubre, Madrid (Spain) (ethical approval no. 14/400 03 06 2015) and was conducted according to the guidelines laid down by the Helsinki Declaration, with written informed consent obtained for each adult participant or, in the case of children, a parent or a guardian of the child participant provided written informed consent on their behalf or the child's assent. A total male (n = 18) and female (n = 18) Hfe −/− mice on a C57BL/6J background (strain B6 129P2-Hfetm1gfn ⁄J from The Jackson Laboratory, Bar Harbor, ME, USA) [50] and male (n = 18) and female (n = 18) Hfe +/+ mice of the same genetic background were used as the healthy control (n = 6 per group of sex, age and HH condition). Age groups were 3, 5 and 7 months old. Animals were bred under standard conditions and supplied with food and water ad libitum. A standard rodent chow was chosen with an iron content of

Hemoglobin Concentration
To determine the hemoglobin concentration, a drop of blood from each subject or animal was placed into the HemoCue Hb 301 analyzer (HemoCueAB, Angelholm, Sweden).

Isolation of Human Erythrocyte Membrane Proteins
Erythrocyte membrane protein isolation followed previously published procedures [59,99]. Briefly, peripheral whole blood was stored for 5 days at 4 • C to allow maturation of reticulocytes to erythrocytes. Serum was discarded and RPMI (Catalog: #31800022, Thermo Fisher Scientific, Waltham, MA, USA), HEPES 25 mM (Catalog: #10204932, GE Healthcare Life Sciences, Chicago, IL, USA) (v/v) at pH 7.4 was added in a 1:1 ratio. Diluted blood cells were suspended in Lymphoprep (Catalog: #07811, Stemcell, Vancouver, BC, Canada) in a 1:1 ratio and centrifuged at 800× g for 20 min. White cell fraction were removed and erythrocytes were washed twice with RPMI with 2.5% HEPES (v/v) and 100 µM of 3,5-di-tert-4-butylhydroxytoluene (BHT; Catalog: #47168, Sigma, St Louis, MO, USA) to avoid oxidative modifications. Erythrocytes were washed in phosphate buffer and used to obtain membrane proteins. Hemolysate supernatants obtained at 9000× g during 20 min at 4 • C were discarded and washed and centrifugation was repeated (at least 5 times) until supernatants appeared colourless. 15 mL of 100 mM sodium carbonate (Na 2 CO 3 ) pH 11.0 was added to the precipitate and passed 5 times through a 25G needle after which it was kept in gentle agitation for 30 min at 4 • C and centrifuged at 20,000× g at 4 • C for 20 min. Membrane ghosts were resuspended in modified RIPA buffer with 50 mM Tris 50 mM NaCl, 3% 3-[(3-Colamidopropyl)-dimethylammonium]-propanesulfonate (CHAPS) (w/v), 0.5% decanoyl-N-methylglucamide (MEGA 10) (w/v), 100 µM BHT and a protease inhibitor cocktail (Catalog: #4693159001, Roche, Basel, Switzerland) and shaken vigorously every 5 min for 1 h keeping it at 4 • C. Finally, it was centrifuged for 1 h at 7000× g at 4 • C, and the supernatant was collected. Protein concentration was determined using a modified Bradford method (Catalog: #5000201, Bio-Rad, Hercules, CA, USA).

Protein Extracts from Mouse Tissues
Liver, kidneys, brain and heart were immediately dissected upon animal sacrifice and stored in modified RIPA buffer at −80 • C. For extract preparation organs were thawed on ice and disintegrated. Then were shaken vigorously every 5 min for 1 h at 4 • C and centrifuged for 1 h at 7000× g and the soluble supernatant was collected for analyses.

Two-Dimensional Electrophoresis and Immunoblot of Human Erythrocyte Membrane Proteins
For identification of post-translational modifications of erythrocyte membrane proteins, 55 µg of membrane protein fractions obtained from each patient were separated by two-dimensional electrophoresis on 12% polyacrylamide (w/v) gels under denaturing conditions. Subsequently, the two-dimensionally separated proteins were transferred to nitrocellulose membranes that were incubated in PBS containing 5% skimmed milk powder (w/v) and 0.05% polysorbate 20 (Tween-20) (v/v), for 30 min at room temperature on an orbital shaker. Later, nitrocellulose membranes were incubated with anti-HNE polyclonal antibodies (gently provided by Dr. Dario Méndez [31]) 1:500 16 h at 4 • C, washed and incubated with the peroxidase-conjugated secondary antibodies 1:4000 (Catalog: #GENA9310-1ML, Amersham ECL Mouse IgG, HRPlinked F(ab') 2 fragment; GE Healthcare Life Sciences, Chicago, IL, USA) for 1 h. After washing, antigen-antibody interaction was detected by chemiluminescence using the Western Lightning ECL Pro substrate (Catalog: #NEL120E001EA, Perkin Elmer, Waltham, MA, USA) for 5 min. Light emission from the reaction was detected in the AGFA CP 1000 equipment (AGFA). PDQuest software (Bio-Rad, Hercules, CA, USA) was used to analyse protein spots. Parallel run 2D gels were stained with Colloidal Coomassie Blue Staining Kit (Catalog: #LC6025, Invitrogen, Waltham, MA, USA) to compare with the membranes and then to excise spots of interest.

One-Dimensional PAGE Electrophoresis and Immunoblot
Protein extracts from the different mouse tissues that were obtained were loaded in one-dimensional PAGE electrophoresis using precast denaturing 12% polyacrylamide (w/v) Mini-PROTEAN TGX™ Precast gels (Catalog: #4568043, Bio-Rad, Hercules, CA, USA). Samples (25 µg) were prepared in loading buffer containing 50 mM Tris (pH 6.8), 250 mM DTT, 2% SDS (w/v) and 0.25% bromophenol blue (Catalog: #1610404, BioRad, Hercules, CA, USA). Once proteins were separated, gels were transferred onto nitrocellulose membranes. Identification of HNE-modified proteins followed identical methodology using anti-HNE antibodies than that described in the previous section.

MALDI-TOF/TOF Analysis
Selected 1D or 2D gel spots either from human or mouse samples were manually excised and automatically digested using the Proteineer (Bruker Daltonics, Billerica, MA, USA). For MALDI-TOF/TOF analysis, samples were automatically acquired in an ABI 4800 MALDI-TOF/TOF mass spectrophotometer (Applied Biosystems, Waltham, MA, USA) as previously described [100]. The data obtained were subjected to search with the MASCOT version 2.3 algorithm with the constraints of decarboxyamidomethylation and species.

Analysis of mRNA Expression
Mice hepatocyte mRNA (n = 6 per group) was extracted to determine the expression of hamp, tnf, il6, sod2 and gpx. The tissue was previously stored at −80 • C in sections of approximately 2-5 mm in 300 µL of RNAlater solution (Catalog: #AM7020, Thermo Fisher Scientific, Waltham, MA, USA). mRNA was extracted in duplicate using the GeneJET RNA purification kit (Catalog: #K0702, Thermo Fisher Scientific, Waltham, MA, USA) including DNAse I (AM1906 Thermo Fisher Scientific, Waltham, MA, USA) digestion according to the manufacturer's instructions. Subsequently, the cDNA was synthesized using the High-Capacity cDNA Reverse Transcription Kit (Catalog: #4368814, Applied Biosystems, Waltham, MA, USA) for Real-Time Quantitative Reverse Transcription PCR (qRT-PCR) in an ABI 7000 Sequence Detection System (Applied Biosystems, Waltham, MA, USA). PCR reactions were performed for commercial mixtures of primers and specific probes corresponding to the sequences of the hamp, tnf, il6, sod2, gpx and β-actin gene, which was used as constitutive expression (Catalog: #4331182, Applied Biosystems assays Table 4). All PCR reactions were set with Maxima Probe/ROX qPCR Master Mix (2×) (#K0231, Thermo Fisher Scientific, Waltham, MA, USA). PCR reactions included an uracil DNA glycosylase pre-treatment of 2 min at 50 • C, an initial incubation of 10 min at 95 • C for polymerase activation, followed by 40 cycles (melting 15 s at 95 • C, annealing and extension 1 min at 60 • C). Relative changes in gene expression were calculated using the comparative 2 −∆∆CT method. Water was used as a negative control. Table 4. qRT-PCR primer and probe data.

Assay
Gene Gene Accession Number Amplicon Size

Statistical Analysis
Statistical analysis of quantitative variables was performed using the t Student parametric test or Mann-Whitney nonparametric test to find significant differences between groups. The Shapiro-Wilk test was used to calculate the normal distribution. Bartlett's test was used for homogeneity of the variance. All statistical analyses were performed using GraphPad Prism 9 (GraphPad Software Inc., San Diego, CA, USA). A p < 0.05 was considered significant.

Conflicts of Interest:
The authors declare no conflict of interest.