Proteomic discoveries in hypermobile Ehlers–Danlos syndrome reveal insights into disease pathophysiology

Abstract

Hypermobile Ehlers–Danlos Syndrome (hEDS) is a poorly understood connective tissue disorder that lacks molecular diagnostic markers. This study aimed to identify proteomic signatures associated with hEDS to define underlying pathophysiology and to inform objective diagnostic strategies with therapeutic potential. An unbiased mass spectrometry–based proteomic analysis of serum from female hEDS patients (n = 29) and matched controls (n = 29) was conducted. Differentially abundant proteins were analyzed through pathway enrichment and gene ontology pipelines. Prioritized candidate biomarker proteins were verified in expanded patient and control cohorts via ELISA. Cytokine array profiling was conducted to assess immune signaling patterns. Proteomic analysis revealed 35 differentially expressed proteins in hEDS, with 43% involved in the complement cascade and 80% linked to immune, coagulation, or inflammatory pathways. Pathway analyses confirmed enrichment in complement activation, coagulation, and stress responses. ELISA validation showed significant reductions in C1QA, C3, C8A, C8B, and C9 in hEDS patients, consistent across age and sex. Cytokine profiling revealed alterations in nodal immune cell mediators in hEDS patients, supporting a model of dysregulated inflammatory response. Our findings indicate a systemic immune dysregulation, particularly involving the complement system and profibrotic cytokines, as a common feature in hEDS pathophysiology. These findings challenge the traditional view of hEDS as solely a connective tissue disorder and support a revised paradigm that includes innate immune dysfunction. This immune involvement may contribute to disease pathophysiology and inform the development of biologically based diagnostic tools, enabling earlier diagnosis and guiding future therapeutic strategies.

Introduction

Hypermobile Ehlers–Danlos syndrome (hEDS) is a heritable connective tissue disorder characterized by generalized joint hypermobility, chronic pain, and a broad range of multisystemic complications.^1–3 Individuals with hEDS often experience a range of comorbidities that extend beyond classical collagen-related disorders. These can include dysautonomia and mast cell activation disease, as well as gastrointestinal and neuropsychiatric symptoms.^1,2,4 Clinical presentation and symptom severity are variable, with manifestations ranging from mild to significantly impacting daily function.⁵ Among the 14 subtypes of Ehlers–Danlos syndrome (EDS), hEDS is the most prevalent, affecting approximately 1 in 500 to 1 in 3,100 individuals.⁶ The condition is diagnosed more frequently in women and may appear sporadically or follow an autosomal dominant inheritance pattern within families. Unlike other EDS subtypes, hEDS lacks a definitive genetic or molecular marker, factors that likely contribute to its diagnostic uncertainty and underreporting of its true prevalence. Diagnosis relies on subjective clinical criteria that capture only a narrow picture of the full disease spectrum and are inconsistently applied in clinical practice.^7,8 As a result, patients frequently endure diagnostic delays of over a decade, with their symptoms commonly dismissed or misattributed to other medical conditions or psychological factors.^9,10

The lack of objective biomarkers remains a major obstacle in the diagnosis and management of hEDS, further complicating timely diagnosis and appropriate care. While few studies have identified clinically useful biomarkers to date,^11,12 proteomic and genomic markers can offer insight into disease susceptibility, progression, and treatment response. Proteomic biomarkers often reflect dynamic biological processes and can serve both diagnostic and prognostic functions, providing critical insights into disease pathophysiology and therapeutic potential, independent of underlying polygenic contributions. In this study, we employed an unbiased proteomic platform alongside cytokine profiling arrays to identify serum protein biomarkers that distinguish individuals with hEDS from matched controls. Subsequent pathway analysis and targeted validation were conducted to refine candidate biomarkers and assess their associations with clinical phenotypes, yielding novel insights into the underlying pathophysiology of hEDS and identifying potential targets for therapeutic intervention.

Materials and methods

Cohort recruitment and phenotyping

Participants were recruited through the Ehlers–Danlos Society’s HEDGE (Hypermobile Ehlers–Danlos Genetic Evaluation) study. Self-reported clinical data regarding hEDS diagnosis, phenotypic features, and comorbidities are presented in detail elsewhere.¹² Serum and plasma samples from hEDS patients (n = 41) and controls (n = 38) were obtained from the Ehlers–Danlos Society (n = 41) under the Medical University of South Carolina Institutional Review Board (protocol #00117828). All samples were processed through Reprocell, Inc and the Ehlers–Danlos Society and stored under standardized conditions to ensure consistency in downstream analyses. The majority of samples were female (n = 67), with 6 male samples in each cohort. All samples were age and sex-matched, with a mean age of 37 years across both groups. General clinical phenotypes as charted by the Ehlers–Danlos Society for hEDS patients included in the study can be found in Fig. S1.

Sample preparation and liquid chromatography–tandem mass spectrometry analysis

Age and sex-matched pairs of serum samples (n = 29 for hEDS and n = 29 control) were randomized prior to plating. Matching was used to reduce confounding biological variability related to age and sex. For this analysis, only female samples were used. For each analysis, 1.5 µL of undepleted serum was processed using the EasyPep 96-well Sample Preparation Kit (Thermo Scientific) following the manufacturer’s protocol. In brief, proteins were solubilized, reduced with dithiothreitol, alkylated with iodoacetamide, and enzymatically digested with Lys-C and trypsin. The digestion was quenched, and peptides were desalted and dried under vacuum. Peptides (1 µg) were reconstituted in 6 µL of 5% acetonitrile with 0.2% formic acid and transferred to autosampler vials. Each sample was spiked with 1 µL of indexed retention time standard peptides (Biognosys) to monitor instrument performance, retention time drift, and sensitivity. In addition, for quality control purposes, pooled reference samples, generated by combining aliquots from all individual patient samples, were run after every 15 individual samples throughout the acquisition to assess digestion efficiency, technical reproducibility, and instrument sensitivity.

Proteomics was performed by liquid chromatography–tandem mass spectrometry on an Easy1200 nanoLC coupled to an Orbitrap Fusion Lumos Mass Spectrometer (Thermo Scientific). Digested peptides (1 µg) were separated by C18 reverse-phase chromatography (Acclaim PepMap RSLC column, 75 µm × 25 cm, 2 µm particle size, 100 Å pore size) at 300 nL/min. Solvent A consisted of 5% acetonitrile with 0.2% formic acid; solvent B consisted of 80% acetonitrile with 0.2% formic acid. The liquid chromatography gradient was as follows: 0% to 35% solvent B for 90 min; 35% to 90% B for 5 min; 90% B for 5 min; 90% to 0% B for 5 min; and 0% B for 20 min. The electrospray ionization voltage was set at 2.2 kV, and the ion transfer tube was maintained at 300 °C. Survey scans were collected in the Orbitrap at 60,000 resolution over an m/z range of 380 to 980, with an automatic gain control (AGC) target of 100% and a maximum injection time of 100 ms. Tandem mass spectra were acquired in data independent acquisition (DIA) mode at 15,000 resolution using 60 variable windows of 10 m/z each, with an AGC target of 200%, a maximum injection time of 50 ms, and a collision energy of 35%.

Database searching and quantification

Raw data were searched against a target-decoy version of the UniProt human proteome database (downloaded 15 March 2023 with 20,422 entries) using Spectronaut 18.6 (Biognosys) with the library free, DirectDIA+ (deep) algorithm. Identifications were filtered at a 1% false discovery rate at the spectrum, peptide, and protein levels. To increase stringency of protein identification, a minimum of 2 unique, proteotypic peptides per protein was required. Fixed modification of cysteine with carbamidomethyl and variable modifications of protein N-terminal acetylation and oxidized methionine were included. A minimum peptide length of 7 and maximum peptide length of 52 amino acids were specified. Quantification was performed based on the extracted ion chromatograms of 3 to 6 MS2 fragment ions per peptide. Database search results were imported into Perseus v1.6.15 (Max Planck Institute) for processing and statistical analysis. Log₂-transformed protein intensities were filtered to retain proteins quantified in 100% of the samples. Statistical comparisons between age-matched groups were performed using paired Student t-tests with Perseus V1.6.15. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD062941.¹³

AdvaitaBio and gene ontology analyses

Proteomic datasets with associated P values were analyzed using AdvaitaBio software to perform pathway enrichment and generate Circos plots for biological processes, cell types, cellular localization, and disease pathways.^14,15 Intersecting protein nodes identified in our mass spectrometry dataset revealed functional overlap across multiple pathways, highlighting key points of convergence within these biological systems.

ELISA analyses

ELISA validation was conducted on an expanded cohort to confirm and extend the proteomic findings across a wider range of individuals. ELISAs for C1QA, C8a, C8b, C3, and C9 were quantified in hEDS patient (n = 41) and control (n = 38) serum samples according to manufacturer protocols (Abcam). The plates were read using a SpectraMax iD3 Multimode Microplate reader, first kinetically at 600 nm until the highest standard optical density reached 0.8 or plateaued for 2 min, at which point the stop solution was added and the absorbance was measured at 450 nm. The samples were run in triplicate, and the intra-assay coefficient of variation (%CV) was set at <10%. In cases of high CV, the Grubb test was used to identify the outlying replicate, which was then removed from analysis. Optical density values were converted to serum concentrations by interpolating against the standard curve. Data distribution was assessed using the Kolmogorov–Smirnov test in GraphPad Prism 10 software. Outliers were identified and removed using the robust regression and outlier removal method (Q = 10%). Group differences were analyzed using GraphPad Prism 10 unpaired t-test or one-way ANOVA, as appropriate.

Cytokine analyses

Cytokine arrays were conducted to explore broader immune system alterations in hEDS. Simultaneous profiling of 105 human cytokines was performed using matched patient–control (n = 13 each) serum sample pairs according to manufacturer instructions (Proteome Profiler Human XL Cytokine Array Kit, R&D Systems). Signal detection was carried out by chemiluminescence using SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific). Quantification was performed using FIJI (ImageJ) software. Standardized masks defining regions of interest were applied to the scanned membranes to identify cytokine-specific antibody spots. Duplicate spot intensities were quantified via automated image analysis, with integrated densities calculated averaged between duplicate spots, and background correction applied using the array’s internal negative control spots, as recommended by the kit protocol. Spots below the limit of detection resulting in negative values after background subtraction were set to zero. Statistical analysis was conducted using unpaired, 2-tailed nonparametric Mann–Whitney t-tests in GraphPad Prism 10. Differences were considered statistically significant at P < 0.05 and highly significant at P < 0.01. Log₂ fold change and -log₁₀  P values were calculated and plotted.

Correlation matrix

A correlation matrix was generated based on serum samples from individuals with hEDS included in both cytokine and ELISA experiments (n = 11) to identify relationships among complement components and circulating cytokines. All complement components were included, along with cytokines that had a P < 0.05 when compared to controls. Pearson correlation coefficient (r) was calculated in GraphPad Prism 10.

Western analyses of plasma samples

Fibronectin (FN1) and type I collagen (COL1) fragments were assayed using the same methodology and plasma samples described by Ritelli et al (2025)¹². All samples were obtained from the same individuals, via the Ehlers–Danlos Society, and the same blood draws as those used in the Ritelli et al study, ensuring direct comparability. Western blot procedures were first replicated exactly as described, then repeated using more standardized protocols, including optimized blocking times, antibody dilutions, and protein handling techniques consistent with current best practice. The same primary antibodies were used: goat anti-human COL1 antibody (#AB758, Merck-Millipore) and anti-human fibronectin antibody (#F3648, Sigma-Aldrich, Merck Life Science). Secondary antibodies included HRP-conjugated anti-rabbit IgG (#A8275, Sigma-Aldrich, Merck Life Science) and anti-goat IgG (#401515, EMD Millipore, Merck Life Science). Protein detection was performed using enhanced chemiluminescence (ECL) with Pierce ECL Substrate (Thermo Scientific #32209) and/or SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific #34094). Because serum albumin (∼66 kDa) can interfere with detection of proteins of similar molecular weight, including FN1 fragments (∼52 kDa) and COL1 fragments (∼45 kDa), we performed albumin depletion using Abcam’s Albumin Depletion Kit (ab241023). This step markedly improved signal clarity by reducing background interference, yielding cleaner and more distinct bands for the proteins of interest.

Results

Tandem mass spectrometry discovery platform

To identify potential proteomic biomarkers associated with hEDS, mass spectrometric analysis was performed on serum samples from affected female donors (n = 29) with a clinical diagnosis of hEDS and female age-matched controls (n = 29) (Table S1). Out of 478 proteins identified, 35 proteins exhibited significantly altered expression between hEDS patients and controls (P < 0.05; Table S2, Fig. 1A). Of these, 9 proteins (26%) were associated with kallikrein activity or clotting cascades while the most prominent differentially represented pathway was the complement cascade, encompassing 15 of the 35 differentially expressed proteins (43%). A heat map shows individual level changes of complement proteins identified (Fig. 1B). Altogether, 28 of 35 proteins (80%) were linked to immune response, coagulation, blood pressure regulation, or inflammatory processes. In stark contrast, structural proteins within the extracellular matrix (ECM) were largely absent, with ECM1 being the only structural protein that was differentially expressed in serum. Notably, we observed no significant differences in fibronectin or collagen levels, or their proteolytic fragments, between hEDS and control samples (Figs. S2 and S3).

Figure 1.

Significantly altered serum proteins in hEDS. (A) Table of proteins differentially expressed in individuals with hEDS compared to controls. Kallikrein/clotting factors are highlighted in purple and complement proteins in blue. (B) Heatmap of normalized, log₂ protein abundances based on mass spectrometry results between control samples and samples from patients with hEDS.

Pathway and gene ontology analysis

Pathway analysis of the proteomic dataset revealed convergent enrichment across multiple orthogonal analytical frameworks (Fig. 2). Subcellular localization profiling demonstrated that the majority of differentially abundant proteins were enriched within the extracellular milieu, including extracellular space, vesicles, and membrane-bound organelles (Fig. 2A). Gene ontology and pathway annotations indicated a predominant involvement in innate immune responses, complement activation (via both classical and alternative pathways), coagulation cascades, and cellular responses to inflammatory or oxidative stress (Fig. 2B). Cell type–specific expression analysis revealed that these proteins are predominantly synthesized by hepatocytes and other liver-derived cell populations, consistent with their known role in the systemic regulation of immune and hemostatic pathways (Fig. 2C). Disease association mapping reinforced these findings, with significant enrichment for disorders of complement dysregulation and coagulopathies, including complement component deficiencies and thrombotic microangiopathies (Fig. 2D). Integration of these datasets within canonical pathway frameworks localized most identified proteins to upstream regulatory nodes within the complement and coagulation networks, suggesting that these factors act as key drivers of immunologic and thrombo-inflammatory processes (Fig. 2E).

Figure 2.

Association of significant proteins with complement and other pathways. (A) Circos plot of significant proteins associated with cellular component Gene Ontology (GO) terms. (B) Circos plot of significant proteins associated with biological process GO terms. (C) Circos plot of cell types associated with significant proteins. (D) Circos plot of diseases associated with significant proteins. (E) Pathway map of complement and coagulation cascades, with significant proteins highlighted based on log₂ fold change.

ELISA validation of complement pathway components

To validate findings from our discovery proteomics dataset, a separate cohort of serum samples was analyzed using ELISA-based quantification of specific complement proteins for which commercial reagents were readily available. Differentially expressed proteins from the mass spectrometry analysis, C1QA, C8A, C8B, and C9, were prioritized based on their collective amplitude of change and pathway commonality. Although C3 was not detected in the initial proteomics screen, it was assayed based on its biological relevance and the high prevalence of mast cell activation disease reported in this hEDS cohort (71%). In the full cohort analysis (Fig. 3A), those with hEDS exhibited significantly lower serum levels of C1QA (P < 0.01), C8a (P < 0.001), C8b (P < 0.0001), and C9 (P < 0.05) compared to controls. C3 was also significantly reduced in hEDS patients (P < 0.001), supporting the hypothesis of upstream complement involvement. To account for potential confounders, females with previous diagnoses of autoimmune diseases and all males were excluded (Fig. 3B). Reductions in C1QA (P < 0.01), C8a (P < 0.05), C8b (P < 0.05), and C3 (P < 0.01) remained statistically significant, while differences in C9 were no longer significant. Age-stratified analysis (Fig. 3C) revealed that reductions in C1QA were most prominent in hEDS subjects aged <37 years (P < 0.01). C8a and C8b levels were significantly decreased in both younger and older hEDS subgroups (P < 0.01 and P < 0.05, respectively), indicating age-independent dysregulation of these terminal complement components. Similarly, C3 was significantly lower in both age groups (P < 0.05 for <37 years; P < 0.01 for ≥37 years), while C9 levels showed no age-specific differences.

Figure 3.

ELISA analysis of complement proteins in serum. (A) Serum concentration of complement proteins in individuals with hEDS (blue) compared to controls (purple). Statistical significance was assessed by unpaired t-test. (B) Analysis excluding male participants and individuals with autoimmune diseases. Significance determined by unpaired t-test. (C) Stratification by age group. Statistical significance assessed by one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Together, these findings demonstrate consistent and widespread reductions in serum complement proteins in subjects with hEDS, implicating dysregulation of both classical and terminal complement pathways. While the patterns of reduction were generally robust across sex and age, our data suggest that these factors may modestly influence the magnitude of complement dysregulation.

Circulating cytokine analyses

To assess systemic inflammatory signaling in hEDS, profiling of 105 cytokines was performed using a targeted protein array in serum samples from subjects with hEDS and age- and sex-matched controls (n = 13 hEDS, n = 13 controls) (Fig. 4). Several cytokines had significant differential expression in the hEDS cohort. IGFBP-2 was the only elevated cytokine compared to controls. Myeloperoxidase (MPO), Serpin E1 (PAI-1), TGF-α, Pentraxin 3 (PTX3), transferrin receptor (TfR), hepatocyte growth factor (HGF), and IL-19 were all significantly downregulated, with TGF-α having the largest fold change (Fig. 4A, Table S4). To explore the relationship between complement alterations and circulating cytokines, a correlation matrix was generated for individuals with hEDS included in both experimental groups (n = 11) (Fig. 4B). Moderate to strong correlations (r > 0.4 or r < −0.4) were observed between several of the circulating cytokines, and between specific complement components. C3 and C4b (r = 0.89), C8a and c9 (r = 0.85), and MPO and Serpin E1 (r = 0.85) displayed strong positive correlations. Strong correlations were also noted between inflammatory cytokines such as IL-19 and Pentraxin 3 (r = 0.89), IL-19 and MPO (r = 0.77), and Pentraxin 3 and MPO (r = 0.81). IL-19 was also moderately correlated with C4b (r = 0.48) and C3 (r = 0.56). An inverse relationship was seen between C9 and C1QA (r = −0.52), TGF-α and C9 (r = −0.43), and TGF-α and C8a (r = −0.21).

Figure 4.

Circulating cytokines in hEDS. (A) Volcano plot of log₂ fold change vs -log₁₀  P value of serum cytokine quantification from array analysis on hEDS (n = 13) vs matched controls (n = 13). Blue indicates decreased cytokine expression in hEDS relative to controls, while red indicates increased expression. Statistical significance was assessed using a Mann–Whitney t-test. (B) Correlation matrix of complement proteins and cytokines in hEDS (n = 11). Pearson correlation coefficients (r) were calculated between significant complement components (validated by ELISA) and cytokines measured in the same individuals with hEDS. Dark red indicates a strong positive correlation, while dark blue indicates a strong negative correlation.

Discussion

This study reveals a distinct pattern of immune dysregulation in individuals with hEDS, challenging its traditional classification as a primary connective tissue disorder. These immune findings broaden the perspective on hEDS, and it is important to consider if they may be due to inherent disease processes, secondary inflammation, or associated or contributing comorbidities. Using an integrative proteomic approach, we identified 35 differentially expressed serum proteins, with a striking enrichment of components involved in complement activation, innate immunity, coagulation, and inflammatory signaling. Nearly half of these proteins belonged to the complement cascade, underscoring the potential role of immune-mediated mechanisms in hEDS pathobiology. In addition, abnormalities in vascular and coagulation-related proteins may potentially explain features of vascular fragility in hEDS patients.

Complement dysfunction emerged as one of the most compelling molecular signatures in hEDS. ELISA validation confirmed consistent reductions in classical (C1QA), terminal (C8A, C8B, C9), and central (C3) complement components, key regulators of both classical and alternative pathways, independent of age, sex, or autoimmune status (Fig. 3). Notably, decreased C8 and C9 levels may impair membrane attack complex formation, potentially compromising pathogen clearance and inflammatory regulation. This deficiency aligns with recent reports of increased infection rates in hEDS.¹⁶ Beyond antimicrobial defense, C1Q and C3 also modulate mast cell activation.¹⁷ Given that 71% of this hEDS cohort reported mast cell–related symptoms, a prevalence consistent with prior studies,^1,18,19 these findings suggest meaningful crosstalk between complement dysregulation and mast cell–mediated immune responses. Such interactions may contribute to ECM degradation, microvascular fragility, and enhanced nociceptive signaling, ultimately impacting connective tissue integrity, vascular permeability, immune cell trafficking, and chronic pain.^20–23 The consistency of complement protein reductions suggests a model of hyperactive complement consumption rather than impaired synthesis. This paradox is commonly observed in other inflammatory and autoimmune diseases and may underlie several of the multisystemic manifestations of hEDS.²⁴

Cytokine profiling also revealed a similar pattern of immune dysregulation and chronic inflammation in hEDS, evidenced by decreased levels of MPO, PAI-1, TGF-α, PTX3, TfR, HGF, and IL-19, along with upregulation of IGFBP-2. These cytokines are involved in processes such as innate immunity, ECM remodeling, oxidative stress, and tissue inflammation and repair, all processes with relevance to manifestations of hEDS. MPO and PTX3 are produced by activated neutrophils, and changes in these proteins suggest alterations in neutrophil-driven responses and tissue remodeling.^25–27 Notably, PTX3 can bind directly to complement proteins, such as C1q and mannose-binding lectin (MBL), thereby regulating both the classical and lectin pathways of complement activation.^28,29 Decreased Serpin E1 and TGF-α may indicate impaired tissue repair and disrupted ECM homeostasis.^30,31 Reductions in IL-19, a cytokine associated with Th2-type immune responses and mucosal inflammation, may align with the high prevalence of allergic, dermatologic, and gastrointestinal symptoms observed in individuals with hEDS.^3,32

Correlation analyses revealed several relationships among these proteins, which offer further insight into immune activity in hEDS. A proinflammatory cluster emerged, with correlations among IL-19 and Pentraxin 3, IL-19 and MPO, and PTX3 and MPO. This may support a combination of neutrophil activation (MPO), innate immune signaling (PTX3), and Th2 cytokine responses (IL-19). MPO and Serpin E1 also showed a strong correlation, indicating potential changes in neutrophil function in the setting of hEDS. Additionally, MPO, PTX3, and Serpin E1 showed strong correlations with complement components C3 and C4b, suggesting that these inflammatory cytokines are linked to complement activation in hEDS. Complement-related proteins also had relationships among each other, including a very strong correlation between C3 and C4b, which is unsurprising given their shared pathways. Components of the membrane attack complex, such as C8a and C9, and C8a and C8b, also showed strong correlations. Interestingly, an inverse correlation was observed between C1QA and C9, underscoring heterogeneity within the hEDS population. While complement components overall were lower in hEDS compared to controls, within-group comparisons revealed variability in complement activation among individuals with hEDS (Fig. 4B).

Potential immune and inflammatory findings are reinforced by gene ontology and pathway analyses (Fig. 2), which provide a systems-level framework for interpreting the proteomic alterations in hEDS. A key observation is the enrichment of pathways related to innate immunity, complement activation, coagulation, and responses to inflammatory and oxidative stress. Stress-related processes were notably enriched, underscoring the possibility that physiological or psychological stress may exacerbate immune dysfunction in hEDS, as has been previously reported.³³ In genetically or epigenetically predisposed individuals, immune activation from viral or environmental triggers may tip a regulatory threshold, initiating or exacerbating connective tissue dysfunction through sustained inflammatory and proteolytic activity. Similar mechanisms have been proposed in other heritable connective tissue disorders, where inflammation has been shown to influence ECM integrity and disease severity.^34,35 Additionally, many of the dysregulated proteins identified serve as functional upstream nodes in the complement and coagulation cascades, suggesting that they may serve as key modulators of disease initiation and progression (Fig. 2E). These findings align with and extend previous reports of immune and proteomic alterations in joint hypermobility syndromes. A proteomics study by Watanabe et al¹¹ and transcriptomic analyses by Chiarelli et al³⁶ similarly noted changes in immune-related gene and protein expression.

In addition to components of the complement and inflammatory pathways, our differential mass spectrometry analysis identified secreted proteins such as ADAMTS13, and others involved in kallikrein-related pathways. Although not the primary focus of this manuscript, these proteins are known to regulate ECM remodeling, blood pressure, coagulation, and immune cell function, highlighting their potential relevance to hEDS pathophysiology.

Paradoxically, classical structural ECM were not differentially represented between hEDS patients and controls in our analysis, an unexpected observation for a condition historically thought to arise from structural connective tissue defects. This finding also contrasts with previous reports describing elevated levels of collagen and fibronectin fragments in individuals with hEDS or HSD. Our mass spectrometry data confirmed the presence of these peptides across all samples; however, no significant differences in abundance were observed. To further explore this discrepancy, we conducted Western blotting as an additional approach to assess these potential fragments using an alternative methodology. Using the same patient samples and same protocol used by Ritelli et al (2025)¹², we detected bands consistent with their reported findings; however, these bands were present in all samples, including controls and at the same intensities (Fig. S2). Additional Western analyses employing a more conventional protocol, both with and without albumin depletion, and using samples collected in heparin or EDTA tubes, yielded similar results, albeit with markedly cleaner membranes (Fig. S3). Thus, although fibronectin and collagen-derived peptides were readily detectable by both mass spectrometry and Western blotting, we found no evidence of disease-specific enrichment. This highlights the need for further independent validation and replication cohorts and underscores the importance of proteomic reproducibility in hEDS biomarker discovery.

While complement and cytokine dysregulation may not be unique to hEDS and can occur in other autoimmune or autoinflammatory conditions, the consistent patterns observed in this cohort suggest that innate immune dysregulation may contribute to the multisystem manifestations of hEDS. Clinically, these immune signatures may augment diagnostic frameworks, enable biologically based patient stratification, and guide therapeutic decision-making. While these candidate biomarkers should not serve as standalone diagnostic tools, they may help uncover disease-relevant mechanisms and identify targets for future investigation. At this time, these findings should not replace comprehensive clinical evaluation or be used to deny patients a diagnosis, as more work is needed to validate these findings and determine their clinical utility.

The heterogeneous nature of hEDS and its comorbidities likely introduces variability in biomarker expression. Longitudinal studies will be essential to determine whether immune signatures track with disease progression or therapeutic response. Studies comparing across disease groups may help determine whether these alterations reflect a shared inflammatory profile among other inflammatory/autoimmune conditions, or a distinct pattern associated only with hEDS. While proteomics enables the identification of candidate biomarkers, mechanistic studies will be needed to establish causality and validate targets for intervention. This analysis is limited to serum-based profiling and does not capture immune or molecular activity within connective tissues, where additional mechanisms may be at play. Nevertheless, serum biomarkers offer a practical and accessible tool for enhancing diagnosis and guiding clinical management. Targeting dysregulated immune pathways through complement inhibition, mast cell stabilization, or cytokine-targeted therapies may hold potential for repurposing in hEDS, particularly in patients with severe multisystem involvement or immune-related comorbidities.

Our study provides strong evidence that immune dysregulation plays a role in hEDS. The observed complement and cytokine patterns, combined with emerging genetic insights, point to a complex interplay between genetic predisposition, immune system activation, and connective tissue dysfunction. Rather than being solely a primary connective tissue disease, hEDS may reflect a broader condition that includes rheumatologic and autoinflammatory components. These findings underscore the need for continued investigation into immune dysregulation in hEDS, which may yield important insights into its underlying pathophysiology and guide the development of targeted clinical interventions.

Author contributions

Molly Griggs (Data Curation [equal], Formal Analysis [equal], Investigation [equal], Methodology [equal], Project Administration [equal], Supervision [equal], Validation [equal], Writing–Original Draft [equal], Writing—Review & Editing [equal]), Victoria Daylor (Data Curation [equal], Formal Analysis [equal], Investigation [equal], Methodology [equal], Supervision [equal], Validation [equal], Writing—Review & Editing [equal]), Taylor Petrucci (Methodology [equal], Software [equal], Visualization [equal], Writing—Original Draft [equal], Writing—Review & Editing [equal]), Amy Weintraub (Data Curation [equal], Formal Analysis [equal], Investigation [equal], Methodology [equal], Software [equal], Validation [equal], Writing—Original Draft [equal], Writing—Review & Editing [equal]), Matthew Huff (Data Curation [equal], Methodology [equal], Software [equal], Visualization [equal]), Sofia Willey (Data Curation [equal]), Kathryn Byerly (Methodology [equal], Writing—Review & Editing [equal]), Brian Loizzi (Methodology [equal], Writing—Review & Editing [equal]), Jordan Morningstar (Conceptualization [equal]), Lauren Elizabeth Ball (Data Curation [equal], Formal Analysis [equal], Methodology [equal], Resources [equal], Software [equal]), Jennifer R. Bethard (Data Curation [equal], Formal Analysis [equal]), Richard Drake (Data Curation [equal], Formal Analysis [equal], Resources [equal]), Amol Sharma (Writing—Review & Editing [equal]), Josef K. Eichinger (Writing—Review & Editing [equal]), Michelle Nichols (Writing—Review & Editing [equal]), Steven Kautz (Writing—Review & Editing [equal]), Steven Shapiro (Writing—Review & Editing [equal]), Anne Maitland (Investigation [equal], Writing—Review & Editing [equal]), Sunil Patel (Writing—Review & Editing [equal]), Russell A. Norris (Conceptualization [equal], Data Curation [equal], Formal Analysis [equal], Funding Acquisition [equal], Investigation [equal], Methodology [equal], Project Administration [equal], Writing—Original Draft [equal], Writing—Review & Editing [equal]), and Cortney Gensemer (Conceptualization [equal], Data Curation [equal], Formal Analysis [equal], Funding Acquisition [equal], Investigation [equal], Methodology [equal], Project Administration [equal], Visualization [equal], Writing—Original Draft [equal], Writing—Review & Editing [equal])

Supplementary material

Supplementary material is available at ImmunoHorizons online.

Funding

This work was supported by the Fullerton Foundation, the Maltz Foundation, and the Ehlers–Danlos Society. This work was supported by the National Institutes of Health (NIH; grant number F32AI181339) and NIH Shared Instrumentation Grants S10 OD028692 and S10 OD025126. The work at the Medical University of South Carolina was performed in a facility constructed with support from the National Institutes of Health (grant number C06 RR018823) from the Extramural Research Facilities Program of the National Center for Research Resources.

Conflicts of interest

None declared.

Data availability

The mass spectrometry proteomics data generated and analyzed during this study have been deposited in the ProteomeXchange Consortium via the PRIDE partner repository under the dataset identifier PXD062941. Supplemental tables and additional supporting data are publicly available on Figshare.