Characterization of the plasma proteomic profile of Fabry disease: Potential sex- and clinical phenotype-specific biomarkers

Translational Research 269 (2024) 47–63 Available online 21 February 2024 1931-5244/© 2024 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/bync/4.0/). Characterization of the plasma proteomic profile of Fabry disease: Potential sexand clinical phenotype-specific biomarkers Laura L´ opez-Valverde a , b , 1 , María E. V´ azquez-Mosquera a , b , 1 , Crist´ obal Col´ on-Mejeras a , b , Susana B. Bravo b , c , Sofía Barbosa-Gouveia a , b , J. Víctor ´ Alvarez a , b , Rosario S´ anchez-Martínez d , Manuel L´ opez-Mendoza e , M´ onica L´ opez-Rodríguez f , g , Eduardo Villacorta-Argüelles h , María A. Goicoechea-Diezhandino i , Francisco J. Guerrero-M´ arquez j , Saida Ortolano k , Elisa Leao-Teles l , ´ Alvaro Hermida-Ameijeiras a , b , 2 , * , María L. Couce a , b , 3 , * a Unit of Diagnosis and Treatment of Congenital Metabolic Diseases. RICORS-SAMID, CIBERER. University Clinical Hospital of Santiago de Compostela, Choupana s/n, Santiago de Compostela, A Coru˜ na 15706, Spain b Health Research Institute of Santiago de Compostela (IDIS), University Clinical Hospital of Santiago de Compostela, Choupana s/n, Santiago de Compostela, A Coru˜ na 15706, Spain c Proteomic Platform, University Clinical Hospital of Santiago de Compostela, Choupana s/n, Santiago de Compostela, A Coru˜ na 15706, Spain d Internal Medicine Department, Alicante General University Hospital-Alicante Institute of Health and Biomedical Research (ISABIAL), Pintor Baeza 12, Alicante 03010, Spain e Department of Nephrology, Hospital Universitario Virgen del Rocío, Manuel Siurot s/n, Sevilla 41013, Spain f Internal Medicine Department, Hospital Universitario Ram´ on y Cajal, IRYCIS, Colmenar Viejo, Madrid 28034, Spain g Faculty of Medicine and Health Sciences, Universidad de Alcal´ a (UAH), Av. de Madrid, Alcal´ a de Henares 28871, Spain h Department of Cardiology, Complejo Asistencial Universitario de Salamanca, P◦. de San Vicente 58, Salamanca 37007, Spain i Nephrology Department, Hospital Gregorio Mara˜ n´ on, Dr. Esquerdo 46, Madrid 28007, Spain j Department of Cardiology, Internal Medicine Service, Hospital de la Serranía, San Pedro, Ronda, M´ alaga 29400, Spain k Rare Diseases and Pediatric Medicine Research Group, Galicia Sur Health Research Institute-SERGAS-UVIGO, Clara Campoamor 341, Vigo 36213, Spain l Centro de Referˆ encia de Doenças Heredit´ arias do Metabolismo, Centro Hospitalar Universit´ ario de S˜ ao Jo˜ ao, Prof. Hernˆ ani Monteiro, Porto 4200-319, Portugal ARTICLE INFO Keywords: Fabry disease Biomarkers Proteomics Plasma Clinical phenotypes Sex ABSTRACT Fabry disease (FD) is a X-linked rare lysosomal storage disorder caused by deficient α -galactosidase A ( α -GalA) activity. Early diagnosis and the prediction of disease course are complicated by the clinical heterogeneity of FD, as well as by the frequently inconclusive biochemical and genetic test results that do not correlate with clinical course. We sought to identify potential biomarkers of FD to better understand the underlying pathophysiology and clinical phenotypes. We compared the plasma proteomes of 50 FD patients and 50 matched healthy controls using DDA and SWATH-MS. The >30 proteins that were differentially expressed between the 2 groups included proteins implicated in processes such as inflammation, heme and haemoglobin metabolism, oxidative stress, coagulation, complement cascade, glucose and lipid metabolism, and glycocalyx formation. Stratification by sex revealed that certain proteins were differentially expressed in a sex-dependent manner. Apolipoprotein A-IV was upregulated in FD patients with complications, especially those with chronic kidney disease, and apolipoprotein C-III and fetuin-A were identified as possible markers of FD with left ventricular hypertrophy. All these proteins Abbreviations: α -GalA, α -galactosidase A; AUC, area under the curve; CKD, chronic kidney disease; Clin FD, FD with clinical complication; Clin FemFD, female FD with some clinical complication; CTL, control; DDA, data-dependent acquisition; DIA, data-independent acquisition; ERT, enzyme replacement therapy; FD, Fabry disease; GFR, glomerular filtration rate; GB3, globotriaosylceramide; LC-MS/MS, liquid chromatography-tandem mass spectrometry; LSD, lysosomal storage disease; Lyso-GB2, lyso-lactosylceramide; Lyso-GB3, globotriaosylsphingosine; LV, left ventricle; LVH, left ventricular hypertrophy, MRI, magnetic resonance imaging; MLR, multiple linear regression; No Clin FD, FD without some clinical complication; No Clin FemFD, female FD without clinical complication; NT-proBNP, N-terminal probrain natriuretic peptide; SWATH, sequential window acquisition of all theorical fragment ion spectra; TpI, troponin I; VUS, variants of unknown significance. * Corresponding authors. E-mail addresses: [email protected] (´ A. Hermida-Ameijeiras), [email protected] (M.L. Couce). 1 These authors contributed equally 2 (+34 981951263) 3 (+34 981950162) Contents lists available at ScienceDirect Translational Research journal homepage: www.elsevier.com/locate/trsl https://doi.org/10.1016/j.trsl.2024.02.006 Received 23 November 2023; Received in revised form 25 January 2024; Accepted 20 February 2024 Translational Research 269 (2024) 47–63 48 had a greater capacity to identify the presence of complications in FD patients than lyso-GB3, with apolipoprotein A-IV standing out as being more sensitive and effective in differentiating the presence and absence of chronic kidney disease in FD patients than renal markers such as creatinine, glomerular filtration rate and microalbuminuria. Identification of these potential biomarkers can help further our understanding of the pathophysiological processes that underlie the heterogeneous clinical manifestations associated with FD. Introduction Fabry Disease (FD; OMIM #301500) is an inherited X-linked lysosomal storage disease (LSD) caused by pathogenic variants in the GLA gene (OMIM *300644) that induce a deficiency in the activity of the lysosomal enzyme α -galactosidase A ( α -GalA, EC 3.2.1.22). 1,2 This deficiency leads to the accumulation of globotriaosylceramide (GB3) and other naturally occurring glycosphingolipids such as globotriaosylsphingosine (lyso-GB3) in plasma, urine, and multiple cell types, primarily endothelial, renal, cardiac, and neuronal cells. 3,4 FD is a rare disease with an estimated overall incidence of 1 per 40,000 live births. The most recent newborn screening studies estimate the prevalence of FD between 1 in 1500 and 1 in 8000, although many of those diagnosed do not develop clinical symptoms. 5–7 Clinically, the disease is characterized by broad phenotypic variability, ranging from a severe multiorgan involvement (the so-called classic phenotype) to atypical late-onset forms. 8 The classic phenotype, which features little to no α -GalA activity, emerges during childhood and is characterized by angiokeratomas, neuropathic pain, gastrointestinal disturbances, hypohidrosis, intolerance to heat and exercise, and cornea verticillata, progressing in adulthood to severe organ dysfunction, including kidney and heart disease and cerebrovascular events. The late-onset phenotype, in which onset typically occurs in adulthood, involves significant residual α -GalA activity and is associated with single organ damage, usually renal or cardiac involvement. 9,10 Moreover, affected women can have very mild or severe forms due to random X-chromosome inactivation. 11,12 Kidney and heart dysfunction are the main complications in FD. Kidney disorder in FD is characterized by proteinuria and a decrease in glomerular filtration rate (GFR) which, over the years, can lead to renal failure, 11,13 while FD heart disease can manifest as infiltrative hypertrophic cardiomyopathy, arrhythmias, coronary artery disease, valve disorders or heart failure. 13,14 The prognosis includes progressive damage to vital organs with age, which can lead to organ failure. End-stage renal disease and cardiovascular or cerebrovascular complications significantly worsen life expectancy. 15 Current treatments are based on enzyme replacement therapy (ERT) and/or the use of pharmacological chaperones. Treatment initiation depends on the degree of organ involvement and can improve quality of life as well as disease course. 16 For these reasons, early diagnosis is essential to improve management. However, this is difficult in FD patients, particularly in women, due to the highly variable clinical presentation, the frequent inconclusive results of biochemical and genetic tests, and the lack of correlation between genotype, α -GalA activity, and clinical severity of the disease. 1,13,17 Biomarker identification in FD research is a growing field that has significant potential to overcome current challenges associated with delayed diagnosis and to enhance the effectiveness of care. Proteomics technologies applied to biological fluids from FD patients can potentially help identify disease-related protein markers. Indeed, recent studies have sought to identify urinary 18–22 and plasma 23–28 biomarkers using proteomics approaches. Here, we describe the plasma proteome profile of a Fabry patient cohort, and identify proteins linked to factors such as sex and main associated complications, in order to further our knowledge of biomarkers for potential use in clinical practice. This study is the deepest untargeted plasma proteomics study of FD patients performed to date, and reveals previously unknown plasma proteome patterns linked to the most prevalent clinical manifestations, notably chronic kidney disease (CKD) and left ventricular hypertrophy (LVH). Material and methods Experimental design This study was approved by the Galician (Spain) Research Ethics Committee of Santiago-Lugo (Xunta de Galicia/Servizo Galego de Saúde, registration code: 2020/438 [Ana Estany Gestal; committee technical secretariat]) and was conducted in accordance with the legal and regulatory provisions for the use of human biological samples, in compliance with the Declaration of Helsinki. Informed consent was obtained from each study participant or guardian/legal representative if available. The overall purpose of the study was to identify plasma biomarkers in FD to improve early detection and diagnosis, define clinical phenotypes, and facilitate the monitoring of the disease. This clinical study had an international, multicentre, prospective, open design, with a control group protocol, and collected biological samples from 50 patients diagnosed with FD and 50 sexand age-matched healthy subjects. All participants were recruited over a 24-month period, from October 2020 until September 2022, and all procedures took place in a single visit. The overall study workflow is depicted in Fig. 1. Study population Biological samples from patients with a diagnosis of FD, both treated and naïve, who were recruited through clinicians from FD reference centres in Spain and Portugal. The ageand sex-matched group of healthy controls were recruited from volunteers and nonmedical staff at the University Clinical Hospital of Santiago. The following inclusion criteria were applied: (i) male or female, aged 6–75 years; (ii) informed consent provided by participants or their legal representatives; (iii) FD patients: diagnosis biochemically confirmed by a decrease in the enzymatic activity of α -GalA and/or by the presence of pathogenic variants in GLA; (iv) healthy controls: with no family history of lysosomal storage disorders and no clinical signs of FD, considered healthy with no previous history of diabetes mellitus, atherosclerotic vasculopathy, or other inflammatory diseases. The exclusion criteria were as follows: (i) inconclusive genetic diagnosis (i.e. carriers of variants of unknown significance [VUS] or variants with conflicting interpretations of pathogenicity); (ii) any medical or psychological disorder that could interfere with the patient’s ability to provide informed consent; (iii) inability or unwillingness to provide informed consent; (iv) participation in a study with an investigational drug within 3 months before providing informed consent. Sample collection Two blood samples (3.5 ml each) per study participant were collected directly into purple K2-EDTA tubes and stored at 4◦C for up to 24 h. Subsequently, these samples were centrifuged at 10,000 g for 10 min at 4◦C, and plasma and leukocytes were collected separately and frozen at -20◦C for further analysis. Clinical variables Information related to the following variables was collected from each FD patient: demographic variables (sex, age, age at diagnosis, and time of evolution); genotype/phenotype (GLA variant, type of L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 49 phenotype: [classic or late-onset]); associated CKD, defined as abnormalities of kidney structure or function, present for >3 months, with implications for health and classified based on cause; GFR category (GFR is the volume of fluid filtered from the renal glomerular capillaries into the Bowman’s capsule per unit of time; decreased GFR corresponds to <60 mL/min/1.73m 2 ); albuminuria category (microalbuminuria corresponds to >30 mg/L); 29 associated LVH, defined as the thickening of the walls (>9 mm in women and >10 mm in men, determined by echocardiography) of the left ventricle and increase in its mass (linear method: >95 g/m 2 in women and >115 g/m 2 in men; 2D method: >88 g/m 2 in women and >102 g/m 2 in men), which can interfere with the heart’s ability to pump blood into the aorta; 30 stroke, defined as an episode of acute neurological dysfunction caused by ischemia or haemorrhage that persists >24 h or until death; 31 therapy (no therapy, ERT or chaperone therapy, age at therapy initiation, and therapy duration); biochemical variables (lyso-GB3 levels, troponin I, N-terminal pro-brain natriuretic peptide [NT-proBNP], creatinine, GFR, and microalbuminuria); maximal left ventricular (LV) wall-thickness and LV mass determined by echocardiography and/or cardiac magnetic resonance imaging (MRI); and α -GalA activity. Information related to the following variables was collected from each healthy control: age, sex, lyso-GB3 levels, and α -GalA activity. Biochemical analysis Lyso-GB3 determination was based on measuring the amount of accumulated substrate present in plasma. It needed a minimum of 100 μ L plasma volume (for determination in duplicate). As a standard solution, 500 μ g/ml Lyso-Lactosylceramide (Lyso-Gb2 – Internal standard) dissolved in DMSO/methanol was used. After incubation for 2 h at 37◦C and filtering of the supernatant, readings were taken using liquid chromatography-tandem mass spectrometry (LC-MS/MS) (HPLC 1200 Series, Agilent Technologies, and AB Sciex API 4000) in positive polarity. For measurement of α -GalA enzymatic activity in leukocytes, samples were sonicated and centrifuged for 10 min at 10,000 rpm at 4◦C, and the supernatant collected. The Bradford technique was used to quantify total protein. For measurement of α -GalA activity, a solution of 4-methylumbelliferyl-alpha-D-galactoside in citrate-phosphate buffer (substrate) was added to the solution, as well as N-acetyl-D-galactosamine, an inhibitor of alpha-galactosidase B but not alpha-galactosidase A. After 20 h incubation at 37◦C, the reaction was stopped with an ethylenediamine solution. The resulting fluorescence, caused by liberation of 4-methyl-umbelliferone from the substrate, is considered proportional to the enzymatic activity of the sample (excitation, 355 nm; emission, 460 nm; SPECTRAmax M2 Molecular Devices). Plasma proteomic analysis To ensure global and quantitative protein identification and quantification, an equal amount of protein was first depleted to eliminate high abundance proteins, as previously described. 32,33 Next, samples were concentrated on a 10 % SDS-PAGE gel. The protein band was detected by Sypro-Ruby fluorescent staining (Lonza, Basel, Switzerland), excised, and processed for in-gel and manual tryptic digestion, as previously described. 34 Briefly, the band was washed with 50 mM ammonium bicarbonate in 50 % methanol (HPLC grade) and dehydrated in acetonitrile (HPLC grade). The gel was reduced in 10 mM dithiothreitol (Sigma-Aldrich) in 50 mM ammonium bicarbonate (Sigma-Aldrich) at 56◦C for 30 min and alkylated with 55 mM iodoacetamide (Sigma-Aldrich) in 50 mM ammonium bicarbonate for 20 min at room temperature in darkness. Subsequently, the gel sections were rinsed with 50 mM ammonium bicarbonate in 50 % methanol and dehydrated with acetonitrile. Tryptic digestion was performed by adding modified porcine trypsin (Promega) at a final concentration of 20 ng/µL in 20 mM ammonium bicarbonate for 16 h at 37◦C. To extract the Fig. 1. Workflow of the overall clinical study. α -GalA, α -galactosidase A; CTL, control; DDA, data-dependent acquisition; FD, Fabry disease; LC-MS/MS, liquid chromatography-tandem mass spectrometry; lyso-GB3, globotriaosylsphingosine; SWATH, sequential window acquisition of all theorical fragment ion spectra. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 50 tryptic peptides generated, three 20-min incubations were performed in 40 µL 60 % acetonitrile and 0.5 % formic acid, and the samples pooled and vacuum dried in a SpeedVac for storage at -20◦C for further analyses. Digested peptides (>4 µg of each sample) were separated using reverse phase chromatography. The gradient was created using a micro liquid chromatography system (Eksigent Technologies nanoLC 400, Sciex) coupled to a high-speed Triple TOF 6600 mass spectrometer (Sciex, Foster City, C) with a microflow source, as described previously. 32 Qualitive analysis was performed using a data-dependent workflow (DDA) as described previously. 32 Further details can be found in the Supplementary Material and Methods. Quantitative analysis was performed by sequential window acquisition of all theoretical fragment ion spectra (SWATH-MS). To build the mass spectrometry (MS/MS) spectral libraries, pools of peptide solutions per condition were analysed using a shotgun data-dependent acquisition (DDA) approach by micro-LC-MS/MS (Sciex, USA), as described above and previously. 32,35,36 After creation of the library, protein quantification was performed for individual samples by SWATH-MS (a data-independent acquisition [DIA] method for sample acquisition) using a TripleTOF® 6600 LC-MS/MS system (Sciex, USA) (see Supplementary Material and Methods for details). Functional analysis Analysis of the qualitative plasma proteome was performed using the open-access software FunRich (Functional Enrichment Analysis Tool version 3.1.3; http://funrich.otg/index.html) for functional enrichment (Gene Ontology (GO)-Biological processes). Statistical analysis Epidemiological and statistical software was used (EPIDAT 4.0) to estimate the number of biological samples needed to achieve statistical power to identify potential significant differences between two groups. For identification of diagnostic biomarkers, candidate biomarkers were required to be present in 90 % of FD patients and ≤10 % of controls; assuming a 1:1 patient/control ratio, a confidence level of 95 %, and a statistical power of 100 %, 50 blood samples were required from patients and controls, respectively. Statistical analysis was performed using GraphPad Prism 8 (GraphPad Inc) and SPSS Statistics (version 26.; IBM Corp). All statistical tests were two sided, and a p-value <0.05 was considered statistically significant. The details of experiments can be found in the Fig. legends and tables. Clinical and biochemical characteristics were compared between groups using the Mann-Whitney U-test for quantitative variables and Pearson’s chi-squared and Fisher’s exact tests for qualitative variables. The hypergeometric test was used to select the enriched biological processes and Fisher’s exact test to compare between groups. To analyse the quantitative proteomic results, normality tests (Shapiro-Wilk test) were performed if possible, and the log2FC was calculated with the mean or median. Significant differences between groups were assessed using the Mann-Whitney U-test or Student’s t-test. Proteins with p-values <0.05 and log2FC ≥+0.6 and ≤-0.6 were selected. In comparisons in which groups had a n ≥15, the p-value of those proteins influenced by confounding factors (sex, age) was adjusted by binary logistic regression. For regression, the variables were rescaled using Min-Max scaling due to the magnitude of the values. Spearman and Pearson rank analyses were used to analyse the correlation between proteins and lyso-GB3, or between proteins and other biochemical variables. The areas under the curve (AUC) of the ROC curves were calculated using GraphPad Prism 8. Results Clinical, biochemical, and genetic characteristics of the FD cohort Fifty FD patients and 50 control (CTL) subjects, matched by sex and age, were enrolled in this study. FD patients were 44 % female, and the median age was 47.50 (10–75) years. The median follow-up time since diagnosis was 6.5 (0–34) years. Twenty patients had the classical phenotype and 30 had the late-onset phenotype. Thirty-one patients were receiving specific treatment for FD with ERT (n=21, 67.7 %) or chaperone therapy (n=10, 32.3 %), with a median age at therapy initiation of 47 (8–62) years. The treated FD patients in our cohort were predominantly male and older, with a later diagnosis than untreated patients. Moreover, these treated patients had more complications, mainly involving the kidney and heart, and showed greater alterations in lyso-GB3 levels and enzymatic activity. Interestingly, 31.6 % of untreated patients had organ damage and were therefore susceptible to specific treatment for FD. Three of the most common manifestations in FD were assessed: stroke (12 % of patients); CKD (44 %); and LVH (58 %). Biochemical variables related to these conditions were altered in FD patients, as detailed in Table 1. Plasma lyso-GB3 levels were increased in FD patients, with a median of 74.20 (3.2–1767) ng/mL (normal range, 0.3–8.25 ng/mL), while α -GalA activity in leucocytes was reduced, at 1.05 (0–23.9) nmol/h/mg (normal range: 4.8–23 nmol/h/mg). Male FD patients were more likely to be receiving treatment, had a higher incidence of CKD and LVH, and showed greater alterations in lyso-GB3 levels and reduced enzymatic activity, as well as greater alterations in other variables, in particular NT-proBNP, LV wall thickness, and microalbuminuria (Table 1). As detailed in Fig. 2, the two most frequent genetic variants of GLA in the study cohort were c.713G>A (n=14, 25 %) and c.902G>A (n=7, 12.5 %), both missense variants. Plasma proteome of FD patients vs healthy controls An initial qualitative analysis of the FD and CTL plasma proteome using MS/MS identified 335 and 313 proteins, respectively (Supplementary Table 1 and Supplementary Table 2) (Fig. 3a). Of these, 271 were common to both groups, 64 were exclusive to the FD group, and 42 were exclusive to the CTL group. Analysis of the processes associated with the proteins unique to each group showed a greater proportion of proteins associated with cell communication/signal transduction and metabolism in the FD group, and of those associated with protein metabolism and cell growth and/or maintenance in the CTL group, with statistical significance reached for the latter effect (Fig. 3b). SWATH-MS analysis detected 8 downregulated (FC ≤0.65; p<0.05) and 28 upregulated (FC ≥+1.5; p<0.05) proteins in the FD vs CTL groups (Fig. 3c) (Supplementary Table 3). By studying the correlation between these differentially expressed proteins and lyso-GB3, we detected significant correlations between lyso-GB3 and the upregulated proteins HBA (r =0.361, p =0.010), HBB (r =0.390, p =0.005), PRDX2 (r =0.323, p =0.022), CAH1 (r =0.390, p =0.005), DIAC (r =-0.358, p =0.011), IGHG3 (r =-0.302, p =0.033), FA9 (r =-0.320, p =0.023), and GPX3 (r =-0.299, p =0.035). Analysis stratified by sex (Supplementary Table 4 and Supplementary Table 5) revealed 14 and 11 proteins upregulated exclusively in female and male FD patients, respectively. In females, expression of ANT3, HRG, FINC, G3P, 6PGD, FA9, and SODC differed most clearly from the profile observed in males, while in males, 1433Z, S10A8, CD99, HBG2, A2GL, FHR2 differed most clearly from the profile observed in females. Proteins downregulated in the FD versus CTL groups included C1QB and PLMN in women and CO5 and vitronectin (VTNC) in men (Fig. 4a). Moreover, irrespective of sex, 12 proteins were upregulated (FC ≥1.5; p<0.05) and 5 downregulated (FC ≤0.65; p<0.05) in the FD versus CTL groups (Fig. 4b). Analysis of the ROC curves allowed us to filter out markers for which AUC values exceeded 0.75, and revealed upregulation of HBD, HBB, HBA, DIAC, DEF3, FCN2, APOH, and S10A9, L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 51 and downregulation of ITIH, CETP, and CO3 (Fig. 4c). The same approach was applied to proteins that were differentially expressed in the FD group for only one sex, revealing upregulation of ANT3, HRG, and FINC in women and 1433Z in men, and downregulation of C1QB and PLMN in women and CO5 and VTNC in men (Fig.s 4d-e). Clinical phenotypes Comparison of FD patients with clinical complications (CKD and/or LVH and/or Stroke; Clin FD) (n =35) versus those without (No Clin FD) (n =15) revealed no significant differences in lyso-GB3 or α -GalA levels (Fig. 5a). Analysis of plasma protein expression in FD patients with an associated complication showed upregulation of APOA4, APOC3, and QSOX1 and downregulation of IGHG3 and FETUA (Fig. 5b) (Supplementary Table 6). ROC curves were generated to determine which markers best distinguished between these 2 FD subgroups. Lyso-GB3 was unable to distinguish between the 2 groups (AUC=0.6610). AUC values for APOC3, QSOX1, IGHG3, FETUA and APOA4 were higher than those obtained for lyso-GB3, and APOA4 was the only marker that surpassed the AUC threshold of 0.75 (Fig. 5c). While APO4 expression differed according to sex, statistical significance was maintained in both groups after adjusting the p-value, with significant upregulation observed in the Clin FD subgroup relative to both the No Clin FD and the CTL group (Fig. 5d). Due to the heterogeneity of clinical manifestations in women, we ensured an appropriate distribution of samples into 2 groups according to the absence or presence of complications, and compared the plasma proteome in female patients with (Clin FemFD; n =11) versus without (No Clin FemFD; n =11) clinical complications. There were no significant differences between groups in either lyso-GB3 levels or α -GalA enzymatic activity (Fig. 6a). Three proteins were downregulated in the Clin FemFD group (IGHG3, VTNC, and FCN2; Fig. 6b; Supplementary Table 7): for all 3 proteins AUC values were significantly higher than those obtained for lyso-GB3 (Fig. 6c). IGHG3 was the protein that best distinguished between the Clin FemFD and No Clin FemFD subgroups (AUC=0.8347), but VTNC was the only of the three proteins that marked the differences with both No Clin FemFD and CTL groups with a similar trend in the latter (Fig. 6d). Table 1 Clinical and biochemical characteristics of the cohort. Characteristic FD n =50* CTL n =50* p-value a Female FD n =22* Male FD n =28* p-value a Sex Female Male 22/50 (44) 28/50 (56) 25/50 (50) 25/50 (50) 0.548 Age (Y) 47.50 (17) 45.50 (19) 0.692 44.5 (17) 48 (16) 0.604 Age at diagnosis (Y) 40 (30) 37.50 (26) 42.50 (32) 0.597 Time of evolution (Y) 6.50 (12) 5.50 (16) 7 (13) 0.883 Phenotype Classic Late-onset 20/50 (40) 30/50 (60) 8/22 (36.4) 14/22 (63.6) 12/28 (42.9) 16/28 (57.1) 0.642 Treatment Treated Untreated 31/50 (62) 19/50 (38) 6/22 (27.3) 16/22 (72.7) 25/28 (89.3) 3/28 (10.7) <0.0001 Type of therapy ERT Chaperone 21/31 (67.7) 10/31 (32.3) 4/6 (66.7) 2/6 (33.3) 17/25 (68) 8/25 (32) >0.9 Age at therapy initiation (Y) 47 (12) 44 (12) 48 (20) 0.841 Therapy duration (Y) 3 (5) 5.50 (6) 3 (5) 0.596 CKD b 22/50 (44) 5/22 (22.7) 17/28 (60.7) 0.007 LVH c 29/50 (58) 7/22 (31.8) 22/28 (78.6) 0.001 Stroke 6/50 (12) 4/22 (18.2) 2/28 (7.1) 0.385 Lyso-GB3 (ng/mL) 74.20 (105.6) 3.45 (1.1) <0.0001 29.2 (65.6) 97.25 (210.7) <0.0001 α -GalA (nmol/h/mg) 1.05 (6.9) 17.45 (7.4) <0.0001 7.1 (6.2) 0.0 (0.6) <0.0001 Elevated Troponin I (>40 pg/mL) 12/40 (30) 4/17 (23.5) 8/23 (34.8) 0.443 Elevated NT-proBNP (>125 pg/mL) 23/44 (52.3) 6/19 (31.6) 17/25 (68) 0.017 Elevated LV wall thickness d (mm) 35/50 (70) 12/22 (54.5) 23/28 (82.1) 0.035 Elevated LV mass e (g/m 2 ) 15/39 (38.5) 4/17 (23.5) 11/22 (50) 0.092 Decreased GFR f (<60 mL/min/1.73m 2 ) 8/50 (16) 1/22 (4.5) 7/28 (25) 0.064 GFR stages 0.560 G1 (normal or high: ≥90) 22/50 (44) 12/22 (54.5) 10/28 (35.7) G2 (mildly decreased: 60-89) 20/50 (40) 9/22 (40.9) 11/28 (39.3) G3a (mildly to moderately decreased: 45-59) 3/50 (6) 1/22 (4.5) 2/28 (7.1) G3b (moderately to severely decreased: 30-44) 2/50 (4) 0/22 (0) 2/28 (7.1) G4 (severely decreased: 15-29) 2/50 (4) 0/22 (0) 2/28 (7.1) G5 (kidney failure: <15) 1/50 (2) 0/22 (0) 1/28 (3.6) Microalbuminuria (>30 mg/L) 23/49 (46.9) 6/21 (28.6) 17/28 (60.7) 0.042 Microalbuminuria stages 0.053 A1 (normal to mildly increased: <30) 26/49 (53.1) 15/21 (71.4) 11/28 (39.3) A2 (moderately increased: 30-300) 15/49 (30.6) 5/21 (23.8) 10/28 (35.7) A3 (severely increased: >300) 8/49 (16.3) 1/21 (4.8) 7/28 (25) Abbreviations: CTL, Control; CKD, Chronic Kidney Disease; ERT, Enzyme Replacement Therapy; FD, Fabry Disease; GFR, Glomerular Filtration Rate; LV, Left Ventricle; LVH, Left Ventricular Hypertrophy; (Y), Years. Bold font denotes significant values (p<0.05). * n/N ( %); median (RQ). a Mann-Whitney U test (quantitative variables); Pearson’s chi-squared and Fisher’s exact tests (qualitative variables). b CKD is defined as abnormalities of kidney structure or function, present for >3 months, with implications for health, and CKD is classified based on cause, GFR category, and albuminuria category. c LVH is defined as thickening of the walls of the left ventricle and an increase in is mass, which can interfere with the heart’s ability to pump blood into the aorta. d >9 mm in women and >10 mm in men (echocardiography). e Linear method: >95 g/m 2 in women and >115 g/m 2 in men; 2D method: >88 g/m 2 in women and >102 g/m 2 in men. f GFR is the volume of fluid filtered from the renal glomerular capillaries into Bowman’s capsule per unit of time. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 52 Fig. 2. GLA variant distribution in the FD cohort. The frequency of the variant types (missense, nonsense, and frameshift variant) is represented. In the 50 FD patients, 16 variants were repeated 56 times: 11 missense (n=43, 76.8 %), 3 nonsense (n=5, 8.9 %), and 2 frameshift (n=8, 14.3 %). FD, Fabry Disease. Fig. 3. Comparison of the plasma proteome of FD patients vs healthy controls. a) Venn diagram showing the proteins identified in FD (pink) and CTL plasma (blue). b) Biological processes associated with proteins identified only in FD (black bars) and proteins identified only in CTL (grey bars) plasma. c) Volcano plot depicting significantly (p<0.05) downregulated (Log2FC ≤-0.6, purple) and upregulated (Log2FC ≥+0.6, orange) proteins in FD vs CTL. CTL, control; FD, Fabry disease. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 53 CKD-FD Significantly lower α -GalA activity and higher lyso-GB3 levels were observed in the CKD FD (n=22) versus No Clin FD (n=15) groups (Fig. 7a). Expression levels of 4 proteins were significantly altered in the CKD FD versus the No Clin FD group: APOA4 and HPTR (upregulated), and A2MG and IGHG3 (downregulated), although A2MG was discarded from further analyses as significance was lost after adjusting for confounding factors (Fig. 7b) (Supplementary Table 8). Analysis of the ROC curves for these proteins, together with those of lyso-GB3 and the diagnostic markers commonly used in CKD (GFR, creatinine, and microalbuminuria), revealed highest AUC values for APOA4 and HPTR (0.8833 and 0. 7697, respectively), even surpassing that of lyso-GB3 (Fig. 7c), and showed significantly greater differences in the CKD FD group relative to both the No Clin FD and CTL groups (Fig. 7d). In addition to showing a significant correlation with GFR (r =-0.399, p = 0.004), creatinine (r =0.499, p =0.00022), and microalbuminuria (r = 0.347, p =0.0135), APOA4 had a higher AUC than these markers (0.8833, p<0.0001). Comparison of FD patients with CKD as the sole complication (Only CKD FD, n =4) against the No Clin FD subgroup revealed that while differences in α -GalA activity persisted, differences in lyso-GB3 levels were attenuated (Fig. 8a). APOA4 remained significantly upregulated, APOH, CBPB2, and DIAC were upregulated, and ENOA was downregulated (Fig. 8b) (Supplementary Table 9). Analysis of ROC curves for these proteins showed higher AUC values than markers such as lyso-GB3, creatinine, and microalbuminuria whose AUC values failed to reach significance (p>0.05) (Fig. 8c). In comparisons of the Only CKD FD group with both the No Clin FD and CTL groups, Fig. 4. Comparison of the plasma proteome of FD patients vs healthy controls, stratified by sex. a) Venn diagram showing plasma proteins that were differentially expressed in female (purple) between male (blue) FD patients relative to their respective controls. b) Heatmap depicting differences in protein expression (z-score) between FD and CTL groups, regardless of sex. ROC curves of proteins for which AUC values were ≥0.75 in FD vs corresponding controls: male and female (c); female only (d); and male only (e). AUC, area under the curve; CTL, control; FD, Fabry disease; lyso-GB3, globotriaosylsphingosine. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 54 expression levels of the 5 aforementioned proteins were also significantly altered (Fig. 8d). LVH-FD Comparison of patients with LVH (LVH FD, n =29) versus the No Clin FD (n =15) group revealed significantly decreased α -GalA activity and increased lyso-GB3 levels (Fig. 9a), upregulation of APOC3, QSOX1, and APOC2, and downregulation of FETUA, IGHG3, SODC, and TSP4 with Log2FC limits set (Fig. 9b) (Supplementary Table 10). In subsequent analyses of the corresponding ROC curves, only AUC values for FETUA and APOC3 exceeded those of lyso-GB3, although only the APOC3 AUC value exceeded the 0.75 threshold (Fig. 9c). None of the AUC values for potential markers surpassed that of NT-proBNP (0.9477) or troponin I (0.8583). FETUA was the only protein for which expression levels differed significantly in the LVH FD group versus the No Clin FD and CTL groups (Fig. 9d). Comparison of FD patients with LVH as the sole complication (Only LVH FD, n =9) with the No Clin FD group showed no significant differences between groups in either lyso-GB3 levels or α -GalA activity (Fig. 10a). In terms of differentially expressed proteins, 6 were upregulated (TTHY, QSOX1, APOC2, APOC3, HEMO, and CO3) and 8 downregulated (IGHG3, FCN2, CATA, CRACD, IGD, FETUA, FHR1 and HRG) (Fig. 10b) (Supplementary Table 11): in all cases AUC values exceeded those of lyso-GB3, which failed to reach significance (0.7333). Only for TTHY (0.8395) and IGHG3 (0.8296) did AUC values exceed those of the biochemical markers NT-proBNP (0.8295) and troponin I (0.8295) (Fig. 10c), and with FETUA, CATA, and FCN2 are the ones with the best ROC curve (≥0.8) (Fig. 10d). CKDþLVH-FD Comparison of FD patients with both LVH and CKD (CKD+LVH FD, n =16) against those without clinical complications (No Clin FD, n =15) revealed that the former group had higher levels of lyso-GB3 (Fig. 11a) and markedly reduced α -GalA activity, which approached negligible levels. Of the proteins that were differentially expressed between groups (Fig. 11b) (Supplementary Table 12), APOA4, HPTR and APOC3 had AUC values that exceeded both the 0.75 threshold and the AUC of lysoGB3. APOA4 even outperformed troponin I, a marker of cardiac disease, Fig. 5. Comparison of the plasma proteome of FD patients with clinical complication vs those without. a) Plasma levels of lyso-GB3 and enzymatic activity of α -GalA in Clin FD and No Clin FD groups. b) Volcano plot depicting significantly (p<0.05) downregulated (Log2FC ≤-0.6, purple) and upregulated (Log2FC ≥+0.6, orange) proteins. c) ROC curves of proteins that were differentially expressed in Clin FD vs No Clin FD patients. d) Normalized MLR area of APOA4. α -GalA, α -galactosidase A; AUC, area under the curve; Clin FD, FD with clinical complication; CTL, control; FD, Fabry disease; lyso-GB3, globotriaosylsphingosine; MLR, multiple linear regression; No Clin FD, FD without clinical complication. *p<0.05. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 55 and markers of renal disease as shown in the previous sections (creatinine and glomerular filtration rate) (Fig. 11c). APOA4 and HPTR were the only proteins capable of differentiating the CKD+LVH from both the No Clin FD and CTL groups (Fig. 11d). Discussion We performed a plasma proteomic analysis to identify differences in protein expression in FD patients versus healthy controls, stratified or not by sex, in order to better understand the pathophysiology of the disease. Furthermore, we characterized the plasma proteome of subgroups of FD patients with the most common associated clinical complications to identify possible biomarkers of clinical evolution. We identified 17 proteins that were differentially expressed in FD patients versus healthy subjects, regardless of sex. These proteins are implicated in functions related to endothelial dysfunction and vasculopathy characteristic of FD, including inflammation, heme and haemoglobin metabolism, oxidative stress, coagulation, fibrinolysis, complement system activity, glycolysis, lipid and glycolipid metabolism and glycocalyx formation. 37–39 Analysis of ROC curves revealed the highest AUC values for HBA, HBB, HBD, DIAC, S10A9, DEF3, FCN2, CO3, APOH, ITIH1 and CETP, and significant correlations between HBA, HBB, and DIAC expression and lyso-GB3 levels. Increased expression of haemoglobin chains (HBA, HBB, and HBD) has been reported in various inflammatory conditions, 40–44 and can exert pro-oxidant effects and lead to endothelial dysfunction. 41,45 A previous study reported a similar trend in FD females and the opposite trend in FD males, 26 a discrepancy that may be due to differences in the recruitment of study participants and the sensitivity of the technique used. 46,47 Altered expression in FD patients of S10A9, Fig. 6. Comparison of the plasma proteome of female FD patients with clinical complication vs those without. A) Plasma levels of lyso-GB3 and enzymatic activity of α -GalA in Clin FemFD and No Clin FemFD groups. b) Volcano plot depicting significantly (p<0.05) downregulated (Log2FC ≤-0.6, purple) and upregulated (Log2FC ≥+0.6, orange) proteins. c) ROC curves of proteins that were differentially expressed in Clin FemFD vs No Clin FemFD groups. d) Normalized MLR area of IGHG3, VTNC, and FCN2. α -GalA, α -galactosidase A; AUC, area under the curve; Clin FemFD, Female FD with some clinical complication; CTL, control; FD, Fabry disease; FemFD, female FD; lyso-GB3, globotriaosylsphingosine; MLR, multiple linear regression; No Clin FemFD, Female FD without clinical complication. *p<0.05, **p<0.0001. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 62 (Au, Ag, Pt)-nanoparticles in human serum. Colloids Surfaces B Biointerfaces. 2019; 177:141–148. 33. ´ Alvarez J V, Bravo SB, Chantada-V´ azquez MP, et al. Plasma proteomic analysis in morquio a disease. Int J Mol Sci. 2021;22:6165. 34. Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing of proteins from silver-stained polyacrylamide gels. Anal Chem. 1996;68:850–858. 35. Ortea I, Rodríguez-ariza A, Chicano-g´ alvez E, Vacas MSA, G´ amez BJ. Discovery of potential protein biomarkers of lung adenocarcinoma in bronchoalveolar lavage fluid by SWATH MS data-independent acquisition and targeted data extraction. J Proteomics. 2016;138:106–114. 36. Ortea I, Ruiz-s´ anchez I, Ca˜ nete R, Caballero-villarraso J, Dolores M. Identification of candidate serum biomarkers of childhood-onset growth hormone deficiency using SWATH-MS and feature selection. J Proteomics. 2018;175:105–113. 37. Shen JS, Meng XL, Moore DF, et al. Globotriaosylceramide induces oxidative stress and up-regulates cell adhesion molecule expression in Fabry disease endothelial cells. Mol Genet Metab. 2008;95:163–168. 38. Rozenfeld P, Feriozzi S. Contribution of inflammatory pathways to Fabry disease pathogenesis. Mol Genet Metab. 2017;122:19–27. 39. Pollmann S, Scharnetzki D, Manikowski D, Lenders M, Brand E. Endothelial dysfunction in fabry disease is related to glycocalyx degradation. Front Immunol. 2021;12, 789142. 40. S¨ arkij¨ arvi S, Kuusisto H, Paalavuo R, et al. Gene expression profiles in Finnish twins with multiple sclerosis. BMC Med Genet. 2006;7:11. 41. Brunyanszki A, Erdelyi K, Szczesny B, et al. Upregulation and mitochondrial sequestration of hemoglobin occur in circulating leukocytes during critical illness, conferring a cytoprotective phenotype. Mol Med. 2015;21:666–675. 42. Kobayashi S, Ito A, Okuzaki D, et al. Expression profiling of PBMC-based diagnostic gene markers isolated from vasculitis patients. DNA Res Int J Rapid Publ Reports Genes genomes. 2008;15:253–265. 43. Leite GGF, Scicluna BP, van der Poll T, Salom˜ ao R. Genetic signature related to heme-hemoglobin metabolism pathway in sepsis secondary to pneumonia. NPJ Syst Biol Appl. 2019;5:26. 44. Derakhshani A, Safarpour H, Abdoli Shadbad M, et al. The role of hemoglobin subunit delta in the immunopathy of multiple sclerosis: mitochondria matters. Front Immunol. 2021;12, 709173. 45. Rother RP, Bell L, Hillmen P, Gladwin MT. The clinical sequelae of intravascular hemolysis and extracellular plasma hemoglobin: a novel mechanism of human disease. JAMA. 2005;293:1653–1662. 46. Bourassa S, Fournier F, Nehm´ e B, et al. Evaluation of iTRAQ and SWATH-MS for the quantification of proteins associated with insulin resistance in human duodenal biopsy samples. PLoS One. 2015;10, e0125934. 47. Jylh¨ a A, N¨ attinen J, Aapola U, et al. Comparison of iTRAQ and SWATH in a clinical study with multiple time points. Clin Proteomics. 2018;15:24. 48. Heizmann CW, Fritz G, Sch¨ afer BW. S100 proteins: structure, functions and pathology. Front Biosci. 2002;7:d1356–d1368. 49. Viemann D, Strey A, Janning A, et al. Myeloid-related proteins 8 and 14 induce a specific inflammatory response in human microvascular endothelial cells. Blood. 2005;105:2955–2962. 50. Inciarte-Mundo J, Frade-Sosa B, Sanmartí R. From bench to bedside: Calprotectin (S100A8/S100A9) as a biomarker in rheumatoid arthritis. Front Immunol. 2022;13, 1001025. 51. Joseph G, Tarnow L, Astrup AS, et al. Plasma alpha-defensin is associated with cardiovascular morbidity and mortality in type 1 diabetic patients. J Clin Endocrinol Metab. 2008;93:1470–1475. 52. Carbone F, Valente A, Perego C, et al. Ficolin-2 serum levels predict the occurrence of acute coronary syndrome in patients with severe carotid artery stenosis. Pharmacol Res. 2021;166, 105462. 53. Miyakis S, Giannakopoulos B, Krilis SA. Beta 2 glycoprotein I–function in health and disease. Thromb Res. 2004;114:335–346. 54. Rodríguez-Carrio J, Carrillo-L´ opez N, Ulloa C, et al. A subset of low density granulocytes is associated with vascular calcification in chronic kidney disease patients. Sci Rep. 2019;9:13230. 55. Lord MS, Melrose J, Day AJ, Whitelock JM. The interα -trypsin inhibitor family: versatile molecules in biology and pathology. J Histochem Cytochem Off J Histochem Soc. 2020;68:907–927. 56. Park ES, Choi JO, Park JW, Lee MH, Park HY, Jung SC. Expression of genes and their responses to enzyme replacement therapy in a Fabry disease mouse model. Int J Mol Med. 2009;24:401–407. 57. Beltrame MH, Catarino SJ, Goeldner I, Boldt ABW, de Messias-Reason IJ. The lectin pathway of complement and rheumatic heart disease. Front Pediatr. 2014;2:148. 58. Miller JJ, Aoki K, Moehring F, et al. Neuropathic pain in a Fabry disease rat model. JCI insight. 2018;3, e99171. 59. Winchester B. Lysosomal metabolism of glycoproteins. Glycobiology. 2005;15: 1R–15R. 60. Lepedda AJ, Fancellu L, Zinellu E, et al. Urine bikunin as a marker of renal impairment in Fabry’s disease. Biomed Res Int. 2013;2013, 205948. 61. Echevarria L, Benistan K, Toussaint A, et al. X-chromosome inactivation in female patients with Fabry disease. Clin Genet. 2016;89:44–54. 62. Rezaie AR, Giri H. Anticoagulant and signaling functions of antithrombin. J Thromb Haemost. 2020;18:3142–3153. 63. Lu Z, Cheng D, Yin J, et al. Antithrombin III protects against contrast-induced nephropathy. EBioMedicine. 2017;17:101–107. 64. Gao S, Nishibori M. Multiple roles of histidine-rich glycoprotein in vascular homeostasis and angiogenesis. Acta Med Okayama. 2021;75:671–675. 65. Gao S, Wake H, Gao Y, et al. Histidine-rich glycoprotein ameliorates endothelial barrier dysfunction through regulation of NF-κB and MAPK signal pathway. Br J Pharmacol. 2019;176:2808–2824. 66. Ekmekci H, Ekmekci OB, Sonmez H, Ozturk Z, Domanic N, Kokoglu E. Evaluation of fibronectin, vitronectin, and leptin levels in coronary artery disease: impacts on thrombosis and thrombolysis. Clin Appl Thromb Off J Int Acad Clin Appl Thromb. 2005;11:63–70. 67. Pecheniuk NM, Elias DJ, Deguchi H, Averell PM, Griffin JH. Elevated plasma fibronectin levels associated with venous thromboembolism. Thromb Haemost. 2008;100:224–228. 68. Plow EF, Hoover-Plow J. The functions of plasminogen in cardiovascular disease. Trends Cardiovasc Med. 2004;14:180–186. 69. Baker SK, Strickland S. A critical role for plasminogen in inflammation. J Exp Med. 2020;217, e20191865. 70. Cigna D, D’Anna C, Zizzo C, et al. Alteration of proteomic profiles in PBMC isolated from patients with Fabry disease: preliminary findings. Mol Biosyst. 2013;9: 1162–1168. 71. Hebert LA, Cosio FG, Neff JC. Diagnostic significance of hypocomplementemia. Kidney Int. 1991;39:811–821. 72. Hassan S, Sidransky E, Tayebi N. The role of epigenetics in lysosomal storage disorders: uncharted territory. Mol Genet Metab. 2017;122:10–18. 73. Chakravarty D, Ray AG, Chander V, et al. Systemic deficiency of vitronectin is associated with aortic inflammation and plaque progression in ApoE-Knockout mice. FASEB bioAdvances. 2022;4:121–137. 74. Wang F, Kohan AB, Kindel TL, et al. Apolipoprotein A-IV improves glucose homeostasis by enhancing insulin secretion. Proc Natl Acad Sci USA. 2012;109: 9641–9646. 75. Kronenberg F. Apolipoprotein L1 and apolipoprotein A-IV and their association with kidney function. Curr Opin Lipidol. 2017;28:39–45. 76. Kollerits B, Krane V, Drechsler C, et al. Apolipoprotein A-IV concentrations and clinical outcomes in haemodialysis patients with type 2 diabetes mellitus–a post hoc analysis of the 4D Study. J Intern Med. 2012;272:592–600. 77. Kronenberg F, K¨ onig P, Neyer U, et al. Multicenter study of lipoprotein(a) and apolipoprotein(a) phenotypes in patients with end-stage renal disease treated by hemodialysis or continuous ambulatory peritoneal dialysis. J Am Soc Nephrol. 1995;6:110–120. 78. Stangl S, Kollerits B, Lamina C, et al. Association between apolipoprotein A-IV concentrations and chronic kidney disease in two large population-based cohorts: results from the KORA studies. J Intern Med. 2015;278:410–423. 79. Boes E, Fliser D, Ritz E, et al. Apolipoprotein A-IV predicts progression of chronic kidney disease: the mild to moderate kidney disease study. J Am Soc Nephrol. 2006; 17:528–536. 80. Tøndel C, Bostad L, Hirth A, Svarstad E. Renal biopsy findings in children and adolescents with Fabry disease and minimal albuminuria. Am J kidney Dis Off J Natl Kidney Found. 2008;51:767–776. 81. van der Tol L, Svarstad E, Ortiz A, et al. Chronic kidney disease and an uncertain diagnosis of Fabry disease: approach to a correct diagnosis. Mol Genet Metab. 2015; 114:242–247. 82. Neto JTM, Kirsztajn GM, Pereira EM, et al. Proteomic profiling of engineered human immortalized podocyte cell model of Fabry disease. Mol Genet Metab. 2019; 126:S106–S107. 83. Shao B, de Boer I, Tang C, et al. A cluster of proteins implicated in kidney disease is increased in high-density lipoprotein isolated from hemodialysis subjects. J Proteome Res. 2015;14:2792–2806. 84. Wang K, Zelnick LR, Hoofnagle AN, et al. Alteration of HDL protein composition with hemodialysis initiation. Clin J Am Soc Nephrol. 2018;13:1225–1233. 85. Pieroni M, Moon JC, Arbustini E, et al. Cardiac involvement in fabry disease: JACC review topic of the week. J Am Coll Cardiol. 2021;77:922–936. 86. Tao Y, Xiong Y, Wang H, et al. APOC3 induces endothelial dysfunction through TNFα and JAM-1. Lipids Health Dis. 2016;15:153. 87. Ginsberg HN, Brown WV. Apolipoprotein CIII: 42 years old and even more interesting. Arterioscler Thromb Vasc Biol. 2011;31:471–473. 88. Ooi EMM, Barrett PHR, Chan DC, Watts GF. Apolipoprotein C-III: understanding an emerging cardiovascular risk factor. Clin Sci (Lond). 2008;114:611–624. 89. Jahnen-Dechent W, Heiss A, Sch¨ afer C, Ketteler M. Fetuin-A regulation of calcified matrix metabolism. Circ Res. 2011;108:1494–1509. 90. Chekol Abebe E, Tilahun Muche Z, Behaile T, Mariam A, et al. The structure, biosynthesis, and biological roles of fetuin-A: a review. Front cell Dev Biol. 2022;10, 945287. 91. Jung YS, Wang W, Jun S, et al. Deregulation of CRAD-controlled cytoskeleton initiates mucinous colorectal cancer via β-catenin. Nat Cell Biol. 2018;20: 1303–1314. 92. Wieczorek E, O˙ zyhar A. Transthyretin: from structural stability to osteoarticular and cardiovascular diseases. Cells. 2021;10:1768. 93. Domoto H, Iwaya K, Ikomi F, et al. Up-Regulation of antioxidant proteins in the plasma proteome during saturation diving: unique coincidence under hypobaric hypoxia. PLoS One. 2016;11, e0163804. 94. Sun J, Zhou C, Zhao Y, et al. Quiescin sulfhydryl oxidase 1 promotes sorafenibinduced ferroptosis in hepatocellular carcinoma by driving EGFR endosomal trafficking and inhibiting NRF2 activation. Redox Biol. 2021;41, 101942. 95. Ashouri R, Fangman M, Burris A, Ezenwa MO, Wilkie DJ, Dor´ e S. Critical role of hemopexin mediated cytoprotection in the pathophysiology of sickle cell disease. Int J Mol Sci. 2021;22:6408. 96. Jin JK, Blackwood EA, Azizi K, et al. ATF6 decreases myocardial ischemia/ reperfusion damage and links ER stress and oxidative stress signaling pathways in the heart. Circ Res. 2017;120:862–875. L. L´ opez-Valverde et al. Translational Research 269 (2024) 47–63 63 97. Jarlhelt I, Nielsen SK, Jahn CXH, et al. SARS-CoV-2 antibodies mediate complement and cellular driven inflammation. Front Immunol. 2021;12, 767981. 98. Irmscher S, Zipfel SLH, Halder LD, et al. Factor H-related protein 1 (FHR-1) is associated with atherosclerotic cardiovascular disease. Sci Rep. 2021;11:22511. 99. Li X, Zong J, Si S. Complement Factor H related protein 1 and immune inflammatory disorders. Mol Immunol. 2022;145:43–49. 100. Doerr A. DIA mass spectrometry. Nat Methods. 2015;12:35. 101. Ludwig C, Gillet L, Rosenberger G, Amon S, Collins BC, Aebersold R. Dataindependent acquisition-based SWATH-MS for quantitative proteomics: a tutorial. Mol Syst Biol. 2018;14:e8126. 102. Collins BC, Hunter CL, Liu Y, et al. Multi-laboratory assessment of reproducibility, qualitative and quantitative performance of SWATH-mass spectrometry. Nat Commun. 2017;8:1–11. 103. Anjo SI, Santa C, Manadas B. SWATH-MS as a tool for biomarker discovery: From basic research to clinical applications. Proteomics. 2017;17:1–55. 104. Krasny L, Bland P, Kogata N, et al. SWATH mass spectrometry as a tool for quantitative profiling of the matrisome. J Proteomics. 2018;189:11–22. L. L´ opez-Valverde et al.