Similar Ehlers–Danlos Syndrome Profiles Produced by Variants in Multiple Collagen Genes

Abstract

Background: Despite increased attention to double-jointedness or joint hypermobility as seen in connective tissue dysplasias like Ehlers–Danlos syndrome, improved clinical DNA correlations are needed to reduce decadal delays in diagnosis. Methods: To this end, patterns of history (among 80) and physical (among 40) findings are compared for 121 Ehlers–Danlos syndrome patients with recurring variants in collagen type I, II, III, V, VI, VII, IX, XI, and XII genes and novel ones in type XV, XVII, XVIII, and XXVII. Results: A recognizable tissue laxity–dysautonomia profile that transcended collagen biochemical class, triple helix component, mutation type, or presence of accessory DNA variants was defined with a few exceptions. Patients with novel variations experienced severe symptoms at younger ages (6–10 versus 14–18 years) and those with collagen type III variations had more than one significant difference in finding frequencies (spinal disk issues 75% versus 49%; bloating-reflux 100% versus 69%; migraines or menorrhagia 92% versus 53%). Conclusions: These results suggest that collagen DNA variants of diverse gene and molecular type can demonstrate EDS disposition and hasten its diagnosis when distress and disease become manifest.

1. Introduction

The most common and typical connective tissue dysplasia [1] is Ehlers–Danlos syndrome (EDS), a multisystem disorder highlighted by joint hypermobility and skin elasticity [2,3,4]. If this laxity extends to other connective tissues like muscle and the vessel linings, then a substantial proportion of the double-jointed (10–20% of people) will have less appreciated complications of the neuromuscular (migraines, daily headaches, neuropathy, myalgia, and muscle weakness [5]) and autonomic systems (lower body blood pooling and venous/pelvic congestion from distensible vessels [6,7]). Symptoms of “fight or flight” adrenergic stimulation to replete cerebral circulation include the tachycardia, anxiety, and chronic fatigue of postural orthostatic tachycardia syndrome or POTS [8,9] and the reactive skin/airways and food–medicine intolerances of mast cell activation syndrome (MCAS [9,10]) while those of parallel “rest and digest” cholinergic suppression produce the irregularity, reflux, and nausea of irritable bowel syndrome (IBS [11]). Failure to appreciate hypermobility as a threshold trait and take subjective findings of dysautonomia seriously has delayed or derailed EDS diagnoses (e.g., fibromyalgia, chronic fatigue syndrome) as reported previously [12,13] and reiterated here.

Timely diagnosis of a disease with powerful preventive measures (wise exercise, bracing, pain management [14]) and practical therapies (salt, hydration—[8]; allergy–dietary management including lower gluten-dairy—[10,11]) could be improved by genomic analyses that screen the multiple genes required to produce this ancient [15,16] and ubiquitous [17] connecting tissue. Several to all-gene (panel to whole exome) sequencing has related over 300 genes to the tissue laxity–dysautonomia of EDS [13,18,19,20,21], insights compromised by fallacies persisting from the era of one-gene or targeted DNA tests. First of these is the assumption that all pathogenic variants in the collagen type III alpha-1 chain gene (COL3A1, OMIM120180 [22]) diagnosed vascular EDS (OMIM130050 [23]) while those in collagen type V alpha-1 or alpha-2 genes (COL5A1, OMIM120215 [24], or COL5A2, OMIM120190 [25]) diagnosed classical EDS (OMIM130000 [26]; OMIM130010 [2,3,4,27]). Second is the evolutionarily improbable assumption that the hypermobile type of EDS (OMIM130020 [28]), with symptoms equally reflective of joint-skin and neuro-autonomic derangement as these other types, is associated with no gene changes [2,29].

Critical for suspecting and understanding EDS is a holistic documentation of findings that range from the joint-skin laxity highlighted by dermatologists Ehlers and Danlos (joint pain, unusual scarring), to bends of skeleton (deformations like scoliosis, flat feet), genitourinary issues (menorrhagia, pelvic congestion, urinary urgency/infection), and the neuro-autonomic symptoms mentioned above [2,3,4,5,6,7,8,9,10,11,12,13,14]. As collagen gene changes are among the first described and best accepted of DNA changes contributing to EDS, we employed a systematic evaluation to ask if known and novel collagen DNA variants would contribute to a general finding profile that would foster early diagnosis of EDS before its symptoms become disabling.

Systematically documented tissue laxity–neuro-autonomic finding profiles are compared in 19 of the 44 genes associated with 13 of the 28 types of human collagens that were found variant in 121 EDS patients having whole exome sequencing [13,20]. All but the network-forming category of human collagens (COL4 proteins) was represented—genes encoding proteins belonging to the fibrillar (COL1, 2, 3, 5, 11, 27), fibril-associated with interrupted triple helices (FACITS—COL 9, 12), collagen VI-like (COL6, 7), membranous (COL17), and collagens with multiple triple helix domains and interruptions [the multiplexins COL15, 18)—as having variations in EDS patients.

Of particular interest was whether patients having changes in the collagen type III alpha-1 chain (COL3A1) and collagen type V alpha-1/alpha-2 (COL5A1/A2) genes would have profiles consistent with their associated vascular [30] or classical [4] EDS types and whether collagen gene changes in the 66% meeting criteria for hypermobile EDS [29] would correlate sufficiently to aid diagnosis of this EDS type (2).

2. Materials and Methods

Prior reports [13,20] describe how the 1979 EDS patients were evaluated and how their DNA testing was coordinated. Table S1 of the Supplementary Materials in reference [20] lists the categories and findings of the systematic history–physical evaluations; Tables S2–S4 [20] list the altered genes and DNA variants from the 568 of 973 EDS patients who had DNA sequencing (532 of 906 from whole exome sequencing). The collagen gene variations shown here are a subset of those DNA results; only methods relevant to the current data are repeated or expanded here.

2.1. Patient Evaluations

Evaluations of 1979 patients (1899 diagnosed with EDS, 80 not meeting diagnostic criteria—[29]) were conducted from July 2011 through October 2020 in the Dallas private practice of author GW. Frequent findings in the first 638 EDS patients were incorporated into systematic history (80-finding) and physical (40-finding) forms that were used to evaluate 1261 of the EDS and 64 of the Not EDS patients. All the Not EDS patients presented with increased hypermobility quantified by Beighton scores [31]. Most (95%) of the EDS patients shared this hypermobility but presented with complications of joint laxity (joint pain, injury, skeletal deformations like scoliosis—50%), autonomic imbalance (POTS—33%; IBS—6.8%; MCAS symptoms—1.8%), or the neuromuscular system (Chiari deformation—3.2%; headaches or neuropathy—2.3%).

2.2. DNA Testing

DNA testing of 967 EDS patients, 906 by whole exome sequencing and 61 by allele/gene panel testing [13,20], used standard methods for NextGen DNA sequencing [32], with all but earlier patients having conjoint microarray analysis [33] when it was available. A concurrent 461 patients evaluated for developmental disability also had DNA testing, many with independent [34] microarray analysis. The very different genes altered in the 82 of these latter patients were documented as strong support for contributing to rather than coincidental DNA change in these patient groups [20]. All but 5 EDS patients were tested through the GeneDx Company (Gaithersburg, MD, USA), their requisitions including consent for the anonymous sharing of DNA results as did clinic intake forms that included consent for medical genetic evaluation/treatment.

Qualifications of DNA variants by the testing company (variants of uncertain significance or likely pathogenetic/pathogenetic—[35,36]) were given additional qualifiers using a clinical protocol [13] that qualified individual DNA changes as variants of uncertain, moderate, strong, or evidenced diagnostic utility (VUDU, VMDU, VSDU, or VEDU). The clinical protocol begins with consensus qualifications (impact on product function, prevalence in affected or normal patients) and adds concordance of variant/EDS symptoms in relatives and how well action of the altered gene would fit with the tissue laxity/neuro-autonomic mechanisms of EDS [13]. Variants additional to the one considered most relevant to EDS, occurring in 37% of all 566 EDS patients with relevant DNA results [20] and in 48% of those with collagen gene variants, could upgrade the final medical DNA qualification (MDna 1–4+) scores if their altered gene was considered to have a synergistic contribution to EDS [13,37]. The 121 patients with collagen DNA variants in this study all had 3–4+ medical diagnostic utility scores; the qualifications of individual variants are listed in Results.

2.3. Patient and DNA Databases

The 1979 EDS and 735 developmental disability patients having outpatient evaluations were entered into a password-protected MS Excel© GW patient database as described [19,20], approved by the North Texas IRB (centered at Medical City Hospital, Dallas, TX, USA) in 2014 (exempt protocol number 2014-054, approved 12 November 2015). The 1261 EDS patients with systematic evaluations were transferred to a deidentified EDS1261GW1-23 database that included only sex and age range demographics, type of evaluation, source of referral, detailed history–physical findings, and positive/negative but not specific DNA results. These data are the source for this article and are included in the previously mentioned Supplementary Materials of reference [20]. Patient numbers are scrambled in the DNA variant list (Supplemental Excel Table S3) compared to the patient findings of Table S6 [20]; qualified researchers interested in matching DNA and clinical findings are able to contact author GNW at golder.wilson@ttuhsc.edu for the connecting code.

2.4. Statistics

Clinical findings were tallied from all or select patients (those with collagen gene changes over the age of 10 years) in the EDS1261GW1-23 database. Statistical averages and standard deviations were calculated using Excel formulae. Significant differences at the p < 0.05 level were determined using the latter formulae and online resources [38], the latter comparing means by two-tailed t and proportions by N-1 chi-squared tests.

3. Results

3.1. Clinical Presentations

Patient 1: A 14-year-old male presented with symptoms of connective tissue weakness including those of joint-popping with movement, frequent sprains, and subluxations affecting his knees and shoulders. He developed early joint–muscle pain in his neck, temporomandibular joint, and shins and had a fracture of the growth plate of his wrist. Other complications included myopic vision correction by contacts, easy bruising, and flat feet. Neuromuscular symptoms included headaches, gastrointestinal ones of constipation, and reflux, orthostatic dizziness, and chronic fatigue among the 16 of 80 findings on the standard EDS history form. Family history included two siblings of 15 and 11 years and two half-sibs through his mother who did not have remarkable hypermobility or joint pain-injury. His mother has hypermobility, joint pain, and osteopenia, and his maternal grandmother has osteoporosis. Limited information on his father included a history of popping joints, muscle cramps, and flat feet. A physical examination documented a normal build (height 167 cm, 25th percentile, weight 59 kg, 50th percentile, his head circumference larger at 59 cm, 97th percentile). His normal facies and skeletal proportions were accompanied by a right deviation of the shoulder, increased joint laxity with the ability to perform 7 of 9 Beighton maneuvers, and skin elasticity shown by stretching of more than one inch around his jaw and on his mid-forearm. Diagnoses included a mild form of hypermobile EDS, whole exome sequencing showing heterozygous mutations in the collagen type 1 alpha-2 chain gene (COL1A2 c.1295G>A p.Arg432Gln, rs139446305, a substantial amino acid change), and in an intron of the collagen type 15 alpha-1 chain gene (COL15A1 c.12-6T>G, rs2118725299, that disrupted a splice site in the first intron and was also present in his symptomatic mother). The gene changes were qualified as variants of uncertain significance (VUS) by the testing company [35] and as respective variants of evidenced (VEDU) or uncertain (VUDU) diagnostic utility by the clinical protocol [13]. The COL15A1 variant was interpreted as having synergistic contribution to EDS, resulting in a medical diagnostic utility score of 4+ for the combined DNA result [13].

Patient 2: A 57-year-old female was evaluated because of her concerns about muscle weakness and a possible connective tissue dysplasia. She recognized early hypermobility with joint pain that along with injuries limited her activities by age 10. She required 15 skeletal surgeries (neck fusion, shoulder, knee, carpal tunnel, thumb) and later knee replacements. Her 29 of 80 history findings included popping joints, TMJ issues, subluxations (elbows, hips, knees, ankles), scaphoid fractures, diffuse spinal disk issues that led to the fusion surgery, flat feet with high arches, unusual skin scarring, menorrhagia, migraines, chronic regional pain syndrome, and carpal tunnel neuropathy combined with muscle aches. Dysautonomia complaints included the bowel irregularity of irritable bowel syndrome (IBS), the sleeping issues, heat–cold sensitivity, abnormal sweating of postural orthostatic tachycardia syndrome (POTS), the asthma/shortness of breath, and food–medication intolerances of mast cell activation syndrome (MCAS). Laboratory results included low thyroid levels and antibody deficiency like that of chronic variable immune deficiency. Family history showed asymptomatic parents and no biological children. Physical examination included a height at the 50th percentile (64 inches), weight at the 80th percentile (183 lb), and 14 of 40 typical physical findings, among then an arm span equal to height, forward neck curve, hypermobility evidenced by a Beighton score of 7/9, and the ability to perform the prayer sign behind her back (reverse Namaskar maneuver). She was diagnosed with a moderate form of hypermobile EDS, whole exome sequencing showing a heterozygous mutation in the collagen type XVII alpha-1 chain gene (COL17A1 c.3821G>A p.Gly1274Glu, rs749093771, replacing a glycine pivotal for triple helix formation, a new mutation that was not present in either parent). The variant was qualified as likely pathogenic [35] for a dominantly inherited corneal erosion disease (122400) [39]—a local manifestation of epidermolysis bullosae—by the testing company. The clinical qualification was as a variant of strong diagnostic utility (VSDU) that conferred a 3+ medical diagnostic utility for EDS [20].

Patient 3: A 52-year-old woman presented with problems of cervical hypermobility, vertigo, dizziness, headaches, and vision problems. She was noted to be hypermobile and clumsy during childhood, beginning dental care and orthodontics at the age of 8 years due to crowding that required multiple tooth extractions and braces. She was able to participate in sports as a teen but developed musculoskeletal issues with subluxations of the elbows, ligament tears, and multiple fractures. Other findings among 42 of 80 typical of EDS by history were chronic popping of her temporomandibular joint with a lopsided jaw, arthritis in multiple joints with neck and back pain that progressed with disk herniations from C2 to T3 and a need for C6-C7 vertebral fusion surgery, dry skin, endometriosis, ovarian cysts, bladder issues, headaches, poor balance, neuropathy, and muscle aches. Dysautonomia findings included the reflux, swallowing, and nausea of IBS, the sleep issues, brain fog, tachycardia, and anxiety–panic attacks of POTS, and the rashes, hives, and food–medication intolerances of MCAS. She had the low vitamin D levels that are common in EDS. Family history was compromised by her mother’s dementia and minimal paternal information but indicated no individuals with characteristic findings of EDS. Physical examination included a height of 160 cm (25th percentile), weight of 68 kg (75th percentile), and 13 of 40 typical EDS findings. She had an arm span equal to her height, long fingers and face, a high palate, poor balance, and was able to perform 5 of the 9 Beighton maneuvers. The patient was diagnosed with the hypermobile type of EDS, whole exome sequencing reporting a heterozygous frame-shifting DNA variant in the collagen type XVIII gene (COL18A1 c.3514_3515delCT p.Leu1172Val, fsX72, rs398122391), and the peroxisomal biogenesis factor 6 gene (PEX6 c.1841delT p.Leu614ArgfsTer5, rs863225083, a frame-shift mutation leading to termination). Both were qualified as pathogenic [35] by the testing company, the COL18A1 gene variant clinically qualified as having strong diagnostic utility and 3+ medical diagnostic utility for EDS, and the PEX6 mutation with other action that conferred 3+ diagnostic utility for the patient to be a carrier of a generalized peroxisomal disease (214100), [40].

Patient 4: A 38-year-old female requested evaluation for possible EDS because of concerns about diffuse joint pain with hypermobility and fragile skin. Early symptoms included allergies and asthma by the age of 6, glasses needed for myopia by age 9, and dental–orthodontic care by age 14 for multiple caries and crowding, early pain, and clumsiness that limited her teenage activities. Among 53 of 80 typical history findings were popping joints, TMJ issues, ankle sprains with ligament tears and fractures that have led to many surgeries or injections, menorrhagia, ovarian cysts, endometriosis, hernia, and bladder issues indicative of pelvic pooling, posterior headaches, numbness, and tingling with documented neuropathy including carpal tunnel syndrome, poor balance, Morton neuroma, gait changes, and muscle aches. She had scoliosis, spinal disk herniation, Tarlov cysts, and adhesive arachnoiditis that limited her upright posture and mobility. Findings of autonomic imbalance included the irregularity, bloating, reflux, and stomach pain of IBS that began at age 4 and progressed between ages 18 and 24. She also had brain fog, orthostatic dizziness, heat–cold sensitivity, abnormal sweating, tachycardia, and salt fancy typical of POTS, and the rashes, reactive skin, shortness of breath, food and medication intolerances of MCAS. Laboratory studies showed low vitamin D and thyroid hormone levels. She had two miscarriages and threatened ones with two of her pregnancies, her sons aged 5, 7, and 9 all having chronic joint pain. The eldest also had flexibility, IBS issues, headaches, frequent infections, and food allergies, the middle son joint flexibility, IBS issues, and allergies. Her brother aged 44 has arthritis, mast cell findings, dental issues, and many infections; her mother aged 71 with arthritis, anxiety, early menorrhagia, and 9 miscarriages; her father aged 72 with IBS issues; a maternal uncle aged 68 with flexibility, and her paternal grandfather with arthritis. A maternal cousin has severe allergies and the diagnosis of myasthenia gravis, another has severe infertility with diffuse joint pain. Her husband is tall with joint pain and multiple joint injuries. Physical examination documented a height of 173 cm (95th percentile), weight of 78 kg (70th percentile), and 24 of 40 typical EDS findings including a wingspan equal to height, long fingers with a positive thumb through fist (Steinberg) sign, and the ability to perform 8 of 9 Beighton maneuvers.

The impression was a moderate form of EDS, likely inheriting her connective tissue flexibility from her mother’s side, being most typical of hypermobile EDS although her muscle aches may suggest the newer myopathic mEDS type [3]. Whole exome sequencing showed a mutation in an intron of the collagen type XXVII alpha-1 chain gene (COL27A1 c.62+1G>A IVS1+1G>A, rs756877794, interrupting a splice site). The variant was qualified as likely pathogenic [35] by the testing company as having strong diagnostic utility (VSDU) with 4+ medical diagnostic utility for EDS by clinical interpretation [13].

3.2. Collagen Gene Variants Found in EDS Patients

The collagen gene variants in 121 EDS patients over the age of 10 years are shown in

Table 1. DNA variants in collagen genes of various types in EDS patients.

Gene (# Pts)	VUS 1 /Path	V*DU 2 (U/M/S/E)	Additional Variants 3 in Gene:	Variants Shown by Position in Protein, Those with Additional Variants in Red, Those in Triple-Helix Regions in Bold [Nucleotide Change, Qualification in Reference]
Fibrillar Collagens
COL1-2-11 (20) 4	17/3	0/1/3/16	ABCC1x2 FBN1 KCNH2 MYH2x2 NLRP1 SCN2B/9A	COL1A1 Gly191Asp Ala340Thr Thr541Ile Arg564His   Pro982Thr Ala1032Thr  Asn1292Ser—1464aa COL1A2       Ala410Ser     Arg708Gln Pro770Leu Arg948Ser Arg1161Leu      —1366aa COL11  A1     Y386C I416V L654F L654P–-1806aa     A2 R1020X/R1551Q D1618E –-1736aa
COL3A1 (12)	12/0	0/0/2/10	MT-CO3 MT-CYB TFG	Glu29del  Arg271Gln  Arg509His Pro634Ser His1269Arg     Lys1313Arg  --1466aa His34Arg    Pro386Ser    623+6T>G    Pro830Leu      Lys1273Argx2
COL5A1 (33)	33/0	0/0/8/25	COL12A1 BSCL2 FLNAx2 MT-CO2 MYH11 TGFBR2	--K57R A114N Q123E) R193C T315M M521V P762L   T1008M   A1356V P1457L P1568L P1805H—1838aa    R65W A114N V172F(2) A232V G326E I1573V Q863V/V1140M P1083S P1371A P1508T S1769F                   V172F   A247T G358+(splice) P908S      P1302R P1436L P1566Q A1784T(2)
COL5A2 (15)	13/2	1/0/1/13	COL6A3 COL12A1 MT-ATP6 MT-CYB MYBPC3 WNT10A	--Gly126Ser Lys431Arg(2) Arg545Trp Arg836Pro Arg1106Trp/Met0001Val Asp1356Gly(2) Ile373Leu Ser451Phe Pro566Leu  Ala914Val   Gly1125Arg (2)   Gly1490Ser—1499aa
COL27A1	0/1	0/0/0/1	--	(20)+1G>A                                       --1860aa
Collagen 6-like
COL6A1-A2-3 (10)	10/0	0/0/2/8	CLCN1 CASQ1 MT-RNR1 MT-RNR2	--A1-T214M A1-T518N  A2-(605)-2dupA  A1-V972A  A3-K1418del  A1-1028aa A2-1019aa  A2-Arg317His  A3-Asp563Gly  A3-Ala876Val  A3-Thr1269Ile A3-Gly1787Arg A3-3177aa
COL7A1 (4)	0/4	0/0/0/4	MT-RNR1 NLRP12 SLC12A3 MAP3K15	---- Gly1350Trp Pro1458Leu Gly1568Ser Arg1632Ter-------------------------2944aa
Fibrillar with Interrupted Triple Helices (FACITS)
COL9A1-A3 (8)	6/2	0/0/3/5	CACNA1G POLG COL12A1 SCN10A	A1-(292)+2T>A A1-(292)+2dupT          A1-Gln570Ter  A1-921aa A3-Gly29Val A3-(61)-2A>G A3-Gly463Arg     A3-Arg555Gln A3-Gly610Glu   A3-684aa
COL12A1 (15)	15/0	0/0/4/11	FLG MT-ATP6 NLRP3x2 TYMP	---Tyr55Asp Met371Val Thr826Met Thr960Met Arg1286His Thr1798Met(2) Arg2228Trp Asn2924Ser—3063aa  Ile247Met Arg392His Thr960Met Pro1184Thr Arg1434Leu Arg1804Trp      Gly2869Arg
Membranous Collagens
COL17A1	0/1	0/0/1/0	--	Gly1274Glu                    --1497aa
Multiplexins
COL15A1	1/0	1/0/0/0	COL1A2	(4)-6T>G                                        --1388aa
COL18A1	0/1	0/0/1/0	--	Leu1172ValfsX72                                    --1754aa

¹ Qualifications on commercial laboratory report—VUS, variant of uncertain significance, Path, likely pathogenic or pathogenic; ² qualifications by clinical protocol [13]—V*DU, variant of * = U uncertain, M moderate, S strong, E evidenced diagnostic utility; ³ 62 of the 131 EDS patients with collagen gene variants had additional DNA variants, representative variant genes with actions interpreted as having synergistic contribution to EDS are shown here; ⁴ not shown are COL2A1 variants Pro146Ser Lys1345Thr (1487aa). The following collagen variants in EDS patients under 10 are also not shown: COL1A1-Gly26Ser, COL1A2-Arg432Gln, Arg708Gln in sister, COL3A1-Arg491X, COL5A1-Ala1693Thr, COL5A2-Gln1307Glu, COL6A1-Lys310del COL11A2-Arg1131Trp, COL12A1-Thr960Met x2 in two sisters; A or Ala, etc.—single or three letter amino acid code depending on space; COL, collagen; Pts, patients; (605)-2dupA, notations like these indicating intronic variants interpreted to alter RNA splicing, this one near amino acid 605 of the protein.

, a subset of 566 variants interpreted as having a primary contribution to EDS [13,20]. Notice the genes for all collagen protein types except the network IV, recurring variants in 15 genes and single, novel variants in four (genes encoding collagen type XV, XVII, XVIII, and XXVII proteins from patients 1–4). Variants in black were the only ones found in the patients by DNA testing, variants in red having additional variants that are represented by their altered gene in column 4 of Table 1 (only representative genes thought to have synergistic action in causing EDS are shown for the recurring variants). The full notations, qualifications, and inheritance of these collagen and other gene variants can be found in Table S3 of the Supplementary Materials [20].

The variants are arranged by amino to carboxyl position in the respective collagen protein, the number of amino acids (aa) for each protein shown to the right (column 5 of Table 1). As with the prior demonstration of differences between most DNA variants in patients with EDS and developmental disability [13], the abundant 48 collagen type V gene variants that are most frequently associated with EDS [3,4] support contributive rather than coincidental present of these variants in EDS patients. The 12 collagen type III gene variants provide less evidence for correlation since patients with their linked and immediately recognizable disease, vascular EDS [30], were excluded from this EDS population. Like the recurring variants in collagen types I (12), VI (10), VII (4), XI (8), and XII (15) or the single novel variants in types XVII, XVII, XV, and XVIII in Table 1, the finding of collagen type III gene variants in an EDS population with expected type V variants is presented as evidence that these too contribute to EDS.

Each variant was qualified by the commercial laboratory as one of uncertain significance (VUS) or disease-causing (likely pathogenic or pathogenic—Path) in column 2. Note that only 14 of the 121 primary variants were qualified as pathogenic for EDS by the laboratory [35] while 23 were interpreted as having strong diagnostic utility (VSDU) and 83 evidenced diagnostic utility (VEDU) by the clinical protocol [13]. Consensus guidelines place more emphasis on specific variant EDS type associations while the clinical protocol emphasizes synergy of gene action with the central tissue laxity–neuro-autonomic mechanisms of EDS [13].

Note that all but five of the DNA variants are amino acid replacements, most not replacing the small glycine that is required to be present at every three residues in a collagen triple helix. Variants with triple helix regions are bolded (sequences and domains can be displayed at UniProt, www.uniprot.org) and show considerable clustering in the fibrillar collagen proteins regions with intermittent and sparse occurrences in other collagen groups. Among the 124 DNA variants (one patient in the each of the COL 1-2-11, COL5A1, and COL5A2 groups had compound heterozygous variants) were 46 (37%) in gene regions encoding triple helix domains; the other two-thirds randomly distributed across other domains [e.g., laminin-like, interrupted collagenous, disordered, nonhelical, etc. [17]].

3.3. Comparison of EDS Patients with Different Collagen Gene Changes

Figure 1 and Figure 2 compare the findings of EDS patients or patient groups with different collagen gene changes, their totals and frequencies from the systematic evaluations detailed previously [12,13] and in Methods. These evaluations included 80 history findings grouped in 12 categories and 40 physical findings grouped in seven categories [12,20], allowing one to compare the numbers of the findings in particular categories or frequencies of individual findings. Thus, the upper panel of Figure 1 shows the greater severity of female EDS patients who averaged 30 of 80 history findings versus 21 in males (middle column, rows 5–6) and 12 versus 8.9 of 20 dysautonomia (DysA) findings (last column) that are part of that history category. The parallel Figure 2 shows the higher female frequencies of findings within the POTS part of the dysautonomia category with syncope reported in 49% of EDS females versus 26% of EDS males, with every one of the dysautonomia findings in the POTS, IBS, MCAS showing female excess (only 11 of the 20 are shown).

The key data of Figure 1 reiterate the average 7- to 14-year diagnostic delays (age seen/diagnosed minus age of severe symptoms) for all EDS patient groups and the significantly higher numbers of findings in all history–physical categories for those meeting versus not meeting (Not EDS) diagnostic criteria [29]. Although the systematic evaluation establishes a semi-quantitative cut-off of above 10 history or physical findings for the diagnosis of EDS (Not EDS patients having history/physical finding totals of 7.2/7.6 compared to 34/18 for EDS patients), it should be noted that more of the Not EDS patients were male (32 of the 64, not shown) and young (average age of 16 versus 29 years for the EDS patients). Marring their status as a control group are the facts that some of the Not EDS patients presented with early hypermobility that at younger ages would not yet have led to the other tissue laxity and dysautonomia complications of EDS. Supporting this conclusion that some Not EDS patients had the genetic disposition but not yet the distress and disease of EDS is the fact that two of them had collagen type V gene changes [20].

Given that the category numbers distinguish EDS patients from those not meeting consensus criteria, EDS patient groups can be compared to see if they fall into subtypes defined by differences in finding profiles or DNA findings. Females and males clearly have significant differences in nearly all category totals and frequencies of Figure 1 and Figure 2, but their proportionate numbers suggest similar EDS disease of different severity [see the Discussion and references [12,13]; 20 for why all the listed frequencies but those for colic—Figure 2—are higher in females].

Are those meeting criteria for the hypermobile and classical types of EDS distinctive when a holistic and systematic evaluation of all tissue laxity and neuro-autonomic findings is employed? Although the large numbers of these patients (not all 1261 because some were placed in myopathic or predominant dysautonomia categories) make most totals in Figure 1 significant, Figure 2 shows that only certain skin findings (striae, unusual scars) have significantly higher frequencies in patients with classical EDS and those of joint hypermobility (subluxations, sprains) in patients with hypermobile EDS.

The lower panels of Figure 1 show the average category totals for EDS patients with the collagen gene changes listed in Table 1, children under age 10 with EDS excluded because they had been shown to have fewer and more variable EDS findings [20]—see the 10 collagen gene variants found in these children in the legend to Table 1. The lower category and frequency averages of males in Figure 1 and Figure 2 prompted trial corrections for male proportions in the collagen gene groups, these corrections not used because they made little difference in the totals or percentages. The only exception was for those of the young male patient 1 with the synergistic collagen type XV gene change in Figure 2 (first data row for patients with DNA change).

The lower panel of Figure 1 reiterates prior evidence [12,20] for gene changes in hypermobile EDS patients, more or equal numbers of hypermobile EDS (hE) patients (column 2) in all collagen gene groups. Most striking is the fact that 31 of 48 (65%) of patients with collagen type V gene changes meet criteria for hypermobile EDS, their exclusive association with classical EDS in prior articles [3,4] not supported.

Among the patients with collagen gene changes in the lower part of Figure 1 are those with one or more variants in the collagen type V alpha-1 or alpha-2 (COL5A1 or COL5A2) genes, the former contributing two protein copies to the collagen type V triple helix, the latter gene contributing one. There seems to be little difference between the finding numbers of these COL5A1 versus COL5A2 variant patients, whether those with one collagen variant versus accessory ones are compared separately in Figure 1 or all patients with COL5A1 versus COL5A2 variants are compared in Figure 2 [20]. Note that finding numbers of collagen gene groups are compared to those with one variant in the COL5A1 gene to highlight significant differences, this being the predominant group in the present population and the gene change earliest [4] and most frequently reported in EDS studies [18,19,20,21].

The patients with collagen type V gene changes highlight another factor besides variants in genes encoding different chain (e. g., alpha-1 or alpha-2) components that complicates gene group comparisons: of the 566 EDS patients with clinically relevant results from DNA testing (mostly by whole exome sequencing), 221 (39%) had additional variants that were qualified as having synergistic contribution to the patient’s EDS symptoms or as having contribution to a previous associated disease [13,20]. Questions of which DNA variant should be considered primary and which accessory were decided by the longevity of association with EDS, the subject collagen genes always considered primary.

The collagen type XV alpha-1 chain (COL15A1) gene variant in patient 1 was considered additional with synergistic contribution to EDS, its novel presence in an EDS patient rendering it secondary to a collagen type I alpha-2 chain (COL1A2) gene variant. This is because collagen type I changes have long been correlated with a related connective tissue dysplasia (osteogenesis imperfecta) and more recently with a hybrid EDS–osteogenesis disorder [41]. Figure 1 shows that patients with collagen type V alpha-1 or alpha-2 (COL5A1 or COL5A2) gene changes found alone (one) or with accessory variants (two+) had similar numbers of findings across all joint-skeletal, skin, neuromuscular, and dysautonomia categories.

Similar numbers of category findings are shown for all EDS patient groups with collagen gene changes in Figure 1, the presence of accessory DNA variants making little difference in EDS–dysautonomia profiles as shown by the COL5A1/A2 group. Only the COL3A1 group (more history and dysautonomia findings) and the COL7A1 group (more history findings) show significant differences from the established COL5A1 group.

In contrast, the category numbers for patients 1–4 with novel collagen type XV, XVII, XVIII, or XXVII alpha-1 chain (COL15A1, COL17A1, COL18A1, or COL27A1) gene variants often fall outside of the all EDS patient standard deviation, beginning with their earlier presentation (10, 10, 8, and 6 years with severe complications as opposed to 24 years for the reference COL5A1 group). Collagen XV group numbers are more concordant, different only for neuromuscular findings, but these were corrected values (increased by an average 33% because of male sex) that were likely influenced by the conjoint collagen type I alpha-2 chain (COL1A2) variant.

All three COL17-18-27 patients had significantly more joint-skeletal and fewer skin findings, discrepant in their numbers of history, physical, and neuromuscular findings as shown in Figure 1. These patients with novel variants (1–4 in the presentations) all fit consensus criteria of hypermobile EDS.

The similarity of category findings among patients with collagen gene changes is paralleled by the individual finding frequencies of Figure 2. There the percentages of all 48 COL5 patients are compared first to those for the 1064 females, none being significantly different, and then to the finding percentages of other collagen gene groups. The most significant differences are again seen in the COL3 group, scattered across categories with increases in frequencies of spinal disk issues (75% of these patients), bloating–reflux (100%), migraines (92%), and menorrhagia (92%). Not increased was the frequency of aneurysms for this gene associated with aneurysm-producing vascular EDS [30], its easily recognized facial appearance also not present in these COL3 EDS patients.

An expected increase in fractures (75%) was seen in the COL1-2-11 group plus more spinal disk issues (88%) in the COL9 group, the COL1, 9, and 11 genes previously associated with bone diseases [respectively, 166200—the major types of osteogenesis imperfecta [42]; 600969—a form of epiphyseal dysplasia [43]; and 604841—a form of Stickler syndrome [44]]. The new collagen gene changes had no significant differences from COL5 patients as a group, their small number making comparisons of low sensitivity.

4. Discussion

The results show overwhelming phenotypic similarity of EDS patients found to have collagen gene changes by whole exome sequencing, their history–physical finding profiles upon systematic evaluation similar whether the DNA variants were of different molecular type, protein position, or were qualified [13] as significantly disrupting different types of collagen proteins [17]. That 81 DNA variants would affect the types I, II, III, V, XI, or XXVII proteins classified as fibrillar types, 11 the type VI and VII classified as type VI-like, 23 the type IX and XII classified as FACIT collagens, 1 the membranous type XVII collagen, and 2 the type XV and XVIII multiplexins suggests that collagen gene mutations when broadly screened and clinically interpreted can facilitate EDS diagnosis.

Female patients are mostly compared, comprising 1084 of 1261 (84%) EDS patients who had systematic evaluations and 114 of 121 (94%) of those with collagen gene changes. Higher percentages of females have been found in all studies of common EDS types [2,3,4,12,13,20], their higher numbers of complications in Figure 1 and Figure 2 leading to increased but still delayed diagnoses. Greater hypermobility is a major reason for their increased EDS severity (average Beighton scores of 6.8 versus 5.7 for males in Figure 1, directly implicated in the increased numbers of joint-skeletal findings (8.4 versus 5.6) and indirectly in the neuromuscular (5.7 and 4.1) and dysautonomia (12 versus 8.9) category totals that figure. Their increased double-jointedness in turn reflects lower surrounding muscle mass to constrain/support connective tissue and tissue-softening estrogen effects that undoubtedly evolved to help with parturition [2,12,13,14]. The large female majority limits conclusions of this paper to female EDS patients, the proportionately lower frequencies of most EDS findings in males suggesting that conclusions involving central EDS mechanisms will, with larger studies, apply to them.

The unbiased screening of all genes rather than a targeted few for nucleotide sequence change extends older assumptions of one gene–one disease matches to the multigene network–disease mechanism correlations emphasized by prior articles [13,20]. Adequate relation of multifaceted, common diseases to DNA influences first requires the holistic characterization of findings so that the entire EDS–dysautonomia spectrum rather than partial diagnoses like fibromyalgia [45] or chronic fatigue syndrome [46] are correlated with gene change. Especially important for EDS correlations is the inclusion of autonomic imbalance symptoms, their production to overcome cerebral deficiencies due to vessel distensibility, and lower body blood pooling as much a part of EDS as the joint–skin laxity symptoms first noted by Ehlers and Danlos [6,7,12,13,20]. Correlation of gene changes with central mechanisms of tissue laxity and dysautonomia rather than EDS types defined by extreme findings unifies EDS with other connective tissue dysplasias [1]; it also explains why certain mutations in collagen genes contribute to a congruent EDS phenotype while others contribute to different dysplasias (e.g., osteogenesis imperfecta),

The current data join with prior reports [12,13,20] and those from other groups [18,19,21] in showing that (1) patients with hypermobile EDS [2,29] have many gene changes as expected from traits affecting an ancient tissue that first joined cells and spread to many organs, (2) mutations in different part of a gene can contribute to different diseases or types of EDS per the Introduction, rendering obsolete the one gene–one EDS type matches from the era of targeted gene sequencing, (3) that heterozygous mutations as shown here can contribute to diseases related to those previously produced by two-allele (compound heterozygous or homozygous) mutations [e.g., the single copy collagen type V alpha-1 chain COL7A1 mutations contributing to typical EDS findings while their homozygous mutations contribute to a recessively inherited epidermolysis bullosae, (226,600) [47]], and (4) that a DNA change never makes a molecular diagnosis [37] without expert clinical correlation. The use of molecular and clinical correlation as in the present qualification protocol [13] is essential if the genetic contributions to common diseases and their categories are to be defined.

Preliminary data like these from whole exome sequencing have many limitations, especially given the rarity of patients with the same mutation as compared with the enormous number of mutational positions in most genes and the multiplicity of gene combinations and their positions in disease-producing networks. Even frequently observed and well-correlated changes like many of these collagen gene changes will at most provide a partial view of patient DNA diversity until the non-coding segments of the “dark genome [48]” are screened. Difficulties with correlation shown here by the combination of other gene changes with those in collagen genes will be augmented as the non-coding mutations found by whole genome sequencing are incorporated.

Although relating defined mutations with low frequencies in the general population [49] is much superior to the loosely linked polymorphisms found by whole genome association studies [50], novel and combined clinical and bioinformatic approaches are needed. One direction might involve large-language models of artificial intelligence [51], sequences like COL5-ligament connective tissue laxity-hypermobility-joint wear-joint injury-inflammation-arthritis interdigitating with others like COL3-vessel tissue laxity-aneurysm-dissection, COL3 or COL5-vessel laxity-lower body blood pooling-adrenergic stimulation-tachycardia, etc. What is crucial will be the inclusion of all symptoms and all relevant gene changes, ensuring that these gene-finding sequences are holistic and inclusive of molecular and medical expertise.