Abnormality of dermal collagen fibrils in Ehlers Danlos syndrome. Anticipation of the abnormality for the inherited hypermobile disorders

ARTICLE

Auteur(s) : Takasi KOBAYASI

Laboratory for Ultrastructural Dermatopathology Department of Dermatology University of Copenhagen, Bispebjerg Hospital, D-92 Bispebjerg Bakke 23, Kbenhavn NV DK-2400. Denmark

Article accepted on 05/04/2004

Ehlers-Danlos syndrome (EDS) is one of the inherited connective tissue disorders. The main defect of the disorder resides in the collagen fibrils. EDS is diagnosed by the criteria of the stigmata in skin, joint and vessel [1]. Individual stigma vary in severity in a wide range, from a mild to severe degree among the patients. In the daily clinic, there were many patients complaining of joint laxity with pain. Some patients showed symptoms which met the diagnostic criteria of EDS, while others, two or three times more in number, showed the stigmata of EDS, partially and to a limited extent [2-4]. Probably they were forme fruste of EDS. Ultrastructurally, the abnormality of collagen fibrils has been demonstrated in patients of EDS, however little is known about the relation of the abnormality of collagen fibrils to stigmata of EDS, nor about differentiation of EDS from the other inherited hypermobile disorders such as Marfan syndrome (MS) and osteogenesis imperfecta (OI). In spite of recent advances in the gene study of these disorders, sufficient explanation for the abnormality of collagen fibrils is missing. The present study intends to re-evaluate the significance of the abnormality of collagen fibrils for the pathology and diagnosis of EDS.

Material and methods

Description of the patients for the study

Three hundred and forty eight patients in the period from 1997 to 2000 were included in this study. The ages of the patients were from 8 months to 82 years. Most of the patients were female with inherited joint laxity, who complained of joint laxity with pain and muscle fatigue. The patients presented the various symptoms of EDS. Around 80 patients showing the symptoms filled out the diagnostic criteria of EDS, whilst the rest of them showed the stigmata of EDS, partially and in an arrested degree [1-4]. The patients seemed to be EDS types I, II and III. The other patients were not found to be suggestive clinical subtypes by the symptoms. Joint laxity, one of the symptoms of the criteria, was the commonest among the patients examined. The joint laxity could be demonstrated semi-quantitatively by Beighton’s score index (BI) with less uncertainty than the other symptoms. For these reasons, all the patients were classified by age and BI (Table Ia). Concerning the EDS symptoms other than joint laxity, 42 of 348 patients exhibited skin hyperextensibility more than 4 cm on the flexural surface of forearm. Almost all the patients complained of ecchymose, 93 patients of acrocyanosis and 2 of severe livedo reticularis. Seventy patients presented a positive Gorlin’s sign. Fifteen symptom-free members in three generations of three EDS pedigrees were also included. The patients with supposedly benign joint hypermobile syndrome (BJHS) were included in the EDS patients, because the clinical differentiation of BJHS from EDS type III was uncertain [3, 4]. In order to study ultrastructural differentiation, 7 female cases of MS and 4 female cases of OI type I were included in the study (Table Ib, Ic). These patients showed typical clinical symptoms of MS and OI type I and the diagnosis was defined beforehand. In order to find whether or not the abnormality of collagen fibrils was found among normal persons, skin specimens from 48 normal, randomly selected, adolescent females were obtained in plastic surgery operations of breast, abdomen and thigh. Furthermore, 50 skin biopsy specimens of some acquired dermatoses were randomly selected in the archives and the normal areas in the dermis were studied.

Table Ia. EDS patients.

Table Ib. MS patients.

Table Ic. OI patients.

Methods for the study

Most of the biopsy specimens were taken from normal skin in the left elbow and 10 specimens from buttocks, lower abdomen and breast, by a 3 mm punch under surface anesthesia by freezing. The skin specimens were fixed in 6% glutaraldehyde solution of cacodylate buffer pH 7.4 and prepared for routine electron microscopy. Collagen fibrils in the papillary and reticular dermis were studied. The number of patients showing abnormality of collagen fibrils in thickness, array and shape was counted and constituted the patient number with the abnormality compared to the total number of the patients in every BI group. The results were studied statistically by trend test.

For the study of the population of abnormally shaped fibrils (TCF) in both papillary and reticular dermis, specimens from the female patients in the 30 to 45 age group with BI 0 - 9 were selected (The reason for the age limit of 30-45 years of age is given in the discussion). The collagen fibril bundles containing the TCF were selected in both reticular and papillary dermis and TCF were counted in areas of 0.2 square µm in 60,000 times enlarged electron micrographs. Numbers of TCF per around 900 to 1000 fibrils were counted and given as a percentage. In order to further evaluate the thickness distribution, 12 female hypermobile patients with abnormality of the collagen fibrils and 6 normal females without the abnormality were randomly selected. The age of these patients was around thirty. The thickness of collagen fibrils in the papillary dermis was measured, and the numbers per unit area counted. Regardless of BI scores, the hypermobile patients were divided in 2 groups, i.e. Group 1) 7 patients with uniformly thin fibrils. Group 2) 5 with thin and thick fibril groups. Group 3 was the patients for control (Fig. 2).

For immuno electron microscopic demonstration of collagen types I / III, 50 EDS patients with various BI from 1 to 9 in the thirties and one MS patient 19 years of age, were studied. For control, skin specimens from 7 normal females in the compatible age range without the abnormality of collagen fibrils were used. The skin specimens were fixed in 4% solution of paraformaldehyde in phosphate-buffered saline at 6 °C overnight and embedded in glycolmethacrylate (Technovit 7100, Kulzer, Germany). The ultrathin sections were stained by rabbit immunoglobulins to collagen types III and I (Biogensis, England). The immune reactants were detected by biotinylated IgG and streptavidin gold 5 and 10 nm, respectively (Amersham, England). The gold particles were counted per unit area on the pictures and calculated the ratio of collagen types I / III. Fig. 7 shows the mean values of the ratio with standard deviation for each BI group.

Results

The abnormality of dermal collagen fibrils was found in the papillary and reticular dermis. The abnormality was noticed in thickness, array and shape of collagen fibrils as described in a previous report [5]. The abnormal collagen fibrils were distributed randomly in the bundles, and the bundles with the abnormal fibrils were also found randomly between the bundles of normal collagen fibrils.

Thickness

The collagen fibrils seen were usually thinner than the normal thickness of 55-60 nm, however some showed thicker fibrils of 90-200 nm. Thin fibrils were found in either reticular or papillary or both parts of the dermis. They were distributed focally or randomly in the bundle. Fifty-five to sixty % of the patients with BI score 6 and higher, showed thin fibrils in the bundle in the papillary dermis (Fig. 1). The distribution of collagen fibril thickness in the papillary dermis presented in 3 types (Fig. 2), i.e. Group 1) A positively skewed, high and sharp spike at 33 nm. Group 2) A low asymmetrical spike at 50 nm with slight increase of 33 nm thick fibrils. Group 3 for control) A symmetrical high sharp spike at 50 nm.

Array

Disarray of collagen fibrils was found in parabola-, whirl- and wave-form and in occasional cases revealed a sharply bent form (Fig. 3,  4a,  4b). Disarrayed fibrils were normal and thinner-than-normal in thickness. More than 80% of the patients with BI score 5 or higher, showed a disarray of parabola- and whirled form and 40% showed a wavy shape in many collagen fibril bundles, either in papillary or reticular or both parts of the dermis. In addition, 20 patients showed collagen fibrils wound round each other and forming clumps in the reticular dermis. This type of disarray resembled a clumped mass of collagen fibrils in shagreen patches [6] and was included in the whirled form in this study. Trend test demonstrated that the prevalence of disarray in whirl- and wavy forms and of shape anomaly was significant for the difference between BI 6 and 5 (p = 0.000001, 0.001, 0.000001, respectively).

Shape

Abnormality of shape was described as twisted collagen fibril (TCF). TCF was flower-like, zigzag-margined, polygonal and occasionally hieroglyphic in the cross sections and rope-like in the longitudinal sections (Fig. 4a, b). TCF was distributed randomly and focally in individual bundles and the TCF- containing bundles existed between the normal bundles in the papillary and reticular dermis. Flower-shaped TCF were about 60 nm to over 200 nm thick. Zigzag-margined TCF were thinner than flower shape. More than 90% of the patients with a BI score higher than 5 showed TCF in either papillary or reticular dermis or both parts. Eighty percent of the patients with BI 3 and 4 showed TCF and 60% of them with BI 2 (Fig. 5). More than 80% of the patients with BI 0 and 1 also showed TCF. Density of population of TCF did not correlate with various BI scores (Fig. 6). Rope-like TCF in longitudinal section revealed normal axial periodicity of 55 nm and the axial periodicity inclined at 10 degrees to the fibril axis (Fig. 8e). The patients with acrocyanosis and the non lax members in EDS pedigree showed TCF, as described below. The population rate of TCF in both reticular and papillary dermis of middle-aged (30-45 years of age) female patients showed almost the same rate in a range of 0.15 to 0.22% for the patients with BI 4-8. The rate was higher for the patients with BI 9. The patients with BI 0-3 showed the rates 0.2 to 0.02%. The rates for both reticular and papillary dermis were almost the same, except for the patients with BI 0 (Fig. 6).

Ratio of collagen types I to III

The ratio was reduced, if compared to the ratio 1.5 for the control of normal adolescent females (Fig. 7). The patients with BI 2-5 showed mean values between 0.56 and 0.935 with a wide range of standard deviation of ± 0.39. The patients with BI higher than 6 showed values below 0.4 in a narrow ranges of standard deviation of ± 0.12 – 0.04. Three patients with BI 1 also showed 0.32 for mean and ± 0.09 for standard deviation.
Based upon the findings in Fig. 1, 3, 5 and 6, EDS patients are probably divided into two groups by the border between BI score 5 and 6.

TCF and the other symptoms

Regardless of BI, 83 of 93 patients with acrocyanosis and livedo reticularis, showed TCF in the dermis. Among the 83 patients, 37 of acrocyanosis patients and 2 livedo reticularis patients were BI 0 – 1. For the symptoms of foot anomalies, loose joints of hips, shoulders and jaws, 24 patients were BI score 0-1 and showed TCF in the dermis. Seventy patients were positive for Gorlin’s sign and 64 of them showed TCF in the dermis. Nine of 15 symptom-free members in EDS pedigrees showed TCF in the dermis. Two of 48 non lax female skin specimens from plastic surgery demonstrated TCF. Two of the 50 biopsy specimens from the archives also showed TCF in the dermis.

Lax patients without TCF

TCF was not demonstrated in 35 patients in this study. Of the 35 patients, 17 patients were over 50 years of age, 12 patients with acrocyanosis, 2 patients with familiar Reynaud’s disease and 1 patient with each of habitual luxation, hullux valgus and pelvic distraction. One male patient showed joint laxity without pain (BI 8), mental retardation and short stature but no other criteria of EDS symptoms.

MS and OI

MS patients showed the abnormality of collagen fibrils resembling the EDS patients, however the abnormality in MS was not distinct as seen in EDS, though the MS patients showed a high BI score. Six of 7 patients showed almost uniformly about 50 nm thick collagen fibrils and showed whirl- and wave-formed disarray (Fig. 8a) and low numbers of TCF in small flower-like and zigzag margined shapes. Confusingly, one of the 7 MS patients, a 33-year-old female, showed BI 9 and abnormality of collagen fibrils, indistinguishable from EDS (Fig. 8a, b). A 17-year-old MS patient with BI 8 demonstrated 0.67 in ratio of collagen types I / III. Four female patients of OI type I showed collagen fibrils uniformly of normal size and occasional disarray. A few TCF in large flower shapes were found in 2 of them. These TCF revealed the inclined axial periodicity as in EDS (Fig. 8c, d). TCF were not found in the other 2 patients.

Discussion

A previous study presented various degrees of joint laxity in 500 volunteers of both sexes and various ages from 1 to 71 years of age [7]. Most of the children (more than 90%) under 5 years of age in both sexes showed hypermobility, while no person over 70 years of age of either sex showed joint laxity [7]. The numbers of the people with laxity in adolescence and adult reduced with increased age. Male lax persons fell down to 15% at 15 years of age and proceeded to fall slowly to 5% at 50 years of age, while lax females reduced in numbers gradually from 35% at 15 years of age to 5% at 50 years of age. Females of 21-60 years of age showed joint laxity at 11.3% on average, while males were 5%. In this study, 78% of the total of 348 patients were 25 – 60 years of age. And about 30% of the patients in this age group were BI 6 and higher. They were EDS. The rest of the 70% patients were BI 5 and lower (Table I). In the author’s experience, patients with BI 6 and higher always presented some stigmata of EDS other than joints. They could clinically be diagnosed EDS. But the patients with the lower scores showed uncertain other symptoms of EDS. They were not confirmed as an EDS diagnosis, however, the patients of these high and low BI groups contained TCF in the dermis, either reticular or papillary or both parts. All of them suffered from a disorder of the same nature. The patients with BI 5 and lower showed the clinical criteria of EDS symptoms to a partial and arrested degree. They were a forme fruste of EDS. The ultrastructural results were consistent with a clinical suggestion of BI 6 for differentiation of EDS from its forme fruste.
Abnormality of collagen fibrils have been found in clinical subtypes of I [5, 8, 9], II [5, 9], III [5, 9], IV [5, 9, 10], V [5], VI [5], VII B [11-13] and VIII [5, 14], however differentiation of these subtypes was impossible by the abnormality of collagen fibrils, though Subtype VIIB was possibly identified by hieroglyphic shape of TCF as discussed below. The abnormality of collagen fibrils could be recognized by abnormalities of thickness, array and shape [5]. The results presented indicate that the patients with BI 6 and higher, contained thin fibrils at same rate and the rate was not increased corresponding with the increase of BI. Thin collagen fibrils in the dermis were known to represent new formation of collagen fibrils [15]. The results shown in Fig. 2 indicated that thickening of collagen fibrils was probably arrested in these patients. Disarray of collagen fibrils in the bundle was remarkable in the dermis of the lax patients, if compared with normal dermis without TCF. Disarray seemed to be related with an increasing score of BI (Fig. 4), and also with TCF (Fig. 5). TCF was the most significant sign among the 3 categories of the abnormalities of collagen fibrils. More than 90% of the patients with BI 5 and higher, contained TCF (Fig. 5). The results for the 3 categories of the abnormality concluded that all the patients with BI 6 and higher, demonstrated all 3 categories of the abnormality. If the patients with BI 5 and lower contained TCF in the dermis, they were considered to be a forme fruste of EDS. On the other hand, as shown in Fig. 6, TCF were found at a rate of 0.1 to 0.2% at almost the same level in patients with BI 5-8 in both reticular and papillary dermis. Various BI scores did not correspond with the rate of TCF population in the dermis, except the patients with BI 9 showed a higher rate of population. The patients with BI 3 had the lowest TCF content and the content was highest in patients with BI 0. This sign was probably due to the patients presenting either for other stigma than joint hypermobility, for instance acrocyanosis or patient age.
TCF have been described by several different names by different authors such as composite fibrils, spiralled fibrils and helical fibrils in various heritable diseases. TCF is considered to be composed of abnormal twisting of protocollagen fibrils [16]. TCF in different shapes in the cross section and longitudinal section were of the same nature and seemed to be common in every sub-type. The hieroglyphic shape described in type VII B EDS [12] was presumed valid to identify subtype VII B. The author has seen this special shape admixing with ordinary TCF in occasional cases, the patients were presumably a forme fruste of sub-type VII B. Inclined cross bands of TCF at 10 degrees were also described previously [5] and indicate the twisting angle.
Steinmann et al. found lysyl hydroxidase deficiency in two siblings of subtype VI and proposed TCF formation [17]. Hulmes et al. produced collagen fibrils with the inclined angle at about 10 degrees in thick fibrils. They showed pN-collagen/collagen hybrid fibrils with pleomorphism in 200 nm thick fibrils. and implied the role of N-propeptide for TCF formation [18]. Both papers presumed that abnormal twisting resulted from the assembly of abnormal collagen protofibrils. The defects of the protocollagen forming enzyme system might be responsible for TCF formation. TCF were first found in distinct large flower-shapes in the dermal lesions of pseudoxanthoma elasticum [19] and it was understood that normal twisting of collagen fibrils was altered. The large distinct flower-shaped TCF were further identified in the dermis of eruptions in some inherited non-lax connective tissue disorders such as primary amyloidosis [20], hyalinosis cutis [21] and juvenile elastoma [22]. However, the bundles of collagen fibril in these disorders scarcely showed disarray as found in EDS. The earlier ultrastructural studies of EDS found TCF in clinical subtypes I, II, III, IV, V, VI, VIIB and VIII [5, 8, 9, 11-14]. Recent in vitro studies on EDS subtypes of VII and IV demonstrated TCF which was closely related to gene anomalies [10, 13]. No findings of TCF were described in subtype IV [23] and in the patients with mild clinical symptoms and without identification [9]. Those articles probably overlooked TCF. The electron micrographs presented in these articles strongly suggested the existence of TCF in the dermis. For these reasons, it is understood that EDS patients in any subtype have TCF in the dermis and TCF was formed by the enzyme system for collagen fibril formation with multiple inherited defects.
Concerning the biochemical analysis for the ratio of collagen types I/III in the lax patients, Lovell et al. [24] proposed that by their studies of gel-electrophoresis and HPCL, the dermis of normal persons of 20-50 years of age constantly contained 18-21.5% of type III collagen. Another report demonstrated an increased value of the ratio III / I + III in the dermis and aorta of 15 hypermobile female patients at 22-57 years of age [25]. Ratio of collagen types I / III in the dermis was reduced by age over about 40 years [26, 27]. Seemingly, immuno electron microscopic evaluation for ratio of collagen types I/III was a useful method. The technique is based on ELISA. Immuno electron microscopy can not estimate the absolute amount of the content but the method can be applied for finding the ratio. The technique is favourable for studying a small, limited area of the dermis. A previous immuno electron microscopic study demonstrated that the ratio of collagen types I / III was about 1.5 in normal females at adolescent age and ratio 0.3-0.5 in scleroderma [15]. The ratio was reduced corresponding with the increase in numbers of thin collagen fibrils. The ratio presented of 0.4 for lax adolescent females was compatible with the content of thin collagen fibrils. On the other hand, an electrophoresis study demonstrated that cultivated fibroblasts from the dermis of patients of EDS subtypes I and IV secreted reduced amounts of type III collagen, though acid pepsin extraction of collagen type III was often incomplete [10, 28]. The results of the in vitro study conflict with in situ analysis. The reduced amount of type III collagen in mice skin and aorta was related to inactivation of the COL3A1 gene [29]. These studies did not consider the ratio of collagen type I/III. Seemingly, the reduced ratio of collagen types I/III was due to an inherited anomaly of collagen formation. Recently, mutation of the COL5A1 gene of type V collagen came in focus for the pathogenesis of the classical subtype [30, 31]. Type V collagen and type I collagen regulate the assembly of fibril thickness [32]. Thin fibrils and low ratio of collagen types I / III in this study were compatible with these previous reports [30-32]. Type V collagen was not demonstrated in the collagen fibrils [33]. On the above-mentioned grounds, it was presumed that type V collagen might be involved in TCF formation, however, the main process of TCF formation was presumably relying on collagen types I and III.
Patients with acrocyanosis, regardless of joint laxity, showed TCF in the dermis [34]. TCF could be a basic factor for ideopathic acrocyanosis and the patients with BI 0-3 and acrocyanosis might be a form of forme fruste of EDS. The relation of TCF to positive Gorlin’s sign is also interesting, however no explanation was made in this study. Interestingly, this study also demonstrated that TCF were found in normal subjects of the EDS pedigree and randomly selected normal people at a rate of 4%. There were considerable numbers of people in society who scarcely complained of the symptoms suggestive of EDS and showed the abnormality of collagen fibrils. Supposedly, they were carrying the genes for defected enzyme system of collagen formation and produced TCF. Some patients of EDS suffered from miscellaneous life-threatening complications. Probably the complications develop on the same basis as the abnormality of collagen fibrils in the internal organs. Cupo et al. [35] demonstrated the abnormality of collagen fibrils in he dermis, aorta and heart valves of a female patient with marked joint hypermoblity, aneurysma, infarction and pneumothorax. The diagnosis was EDS, presumed Marfan syndrome. A genetic experiment by inactivation of COL3A1 gene in the embryonic stem cells of mice demonstrated changes of collagen fibril and the mice died by aortic rupture after shortened adult lives [25]. Seemingly, the abnormality of collagen fibrils in the dermis also represents abnormality in the internal organs.
Comparing the abnormality of collagen fibrils in EDS, those with MS also presented disarray and TCF, similar to EDS, and both disorders were indistinguishable by the abnormality of collagen fibrils. The previous paper described the abnormality of collagen fibril in the dermis and aortic wall in a single case of MS and demonstrated the electron micrographs of the disarray and TCF in the aortic wall [36]. For the differentiation of OI from MS and EDS, the present findings demonstrated that the collagen fibril changes in OI were similar to EDS, however, the abnormality of collagen fibrils in OI seemed to be milder and more infrequent than in EDS and MS. The previous paper described a similar abnormality of disarray, thin thickness and zigzag margined form of TCF [37, 38].
Conclusively, EDS is diagnosed by the stigmata of joint, skin and vessels. Ultrastructurally, the normal dermis of EDS patients showed the abnormality of collagen fibrils. The abnormality was defined in the skin and internal organs by thin fibrils, disarray and shape anomaly. The abnormality in the dermis was found in either papillary, reticular or both parts of the dermis, however the intensity of the abnormality did not correspond with the severity of joint laxity. The lax patients with BI 6 and higher, constantly showed the abnormality in the dermis at various intensities and the patients with BI 5 and lower inconstantly presented the abnormality. Beighton et al. recommended the score 5 after the formula of joint laxity [1] and Holzberg et al. preferred the score 7 by their formula of the stigmata [2]. The present results preferred BI 6. The patients with the lower BI score were supposed to be a forme fruste of EDS. Regardless of the severity in joint laxity, MS and OI also revealed the abnormality. The differentiation of EDS from MS and OI was not always possible by the abnormality of collagen fibrils.
The abnormality of collagen fibrils became difficult to find in the patients aged over 60 years. The shape anomaly of the abnormality was also found in the clinically normal members in EDS pedigree and sporadically in society (about 4%). These two groups of normal persons could be carrying the genes for defective formation of abnormal collagen fibrils. The abnormality is formed by a heritably defective enzyme system for collagen fibril formation. The defects are probably multiple. The abnormality of collagen fibrils probably expresses disposition for the inherited malformation of collagen fibrils. n

Acknowledgements. This study was supported by grants from the Danish Rheumatism Association and Aage Haensh’s foundation. The author expresses his gratitude to Drs Susanne Ullman and Poul Halberg in Department of dermatology Bispebjerg Hospital, Copenhagen Denmark for the clinical work in the clinic for connective tissue disorders, to Dr Sren Jacobsen in Department of Rheumatology for statistic analysis and to Mrs Mette Thage for her skilful technical assistance.

References

1. Beighton P, De Paepe A, Steinmann B, Tsipouras P, Wenstrup RJ. Ehlers Danlos syndrome: Revised Nosology, Villefrache, 1997. Am J Med Genet 1998; 77: 31-7.

2. Holzberg M, Hewan-Lowe KO, Olansky AJ. The Ehlers-Danlos syndrome: Recognition, characterization, and importance of a milder variant of the classic form. J Am Acad Dermatol 1988; 19: 656-66.

3. Grahame R. Heritable disorders of connective tissue. Balliere’s Clinical Rheumatology 2000; 14: 345-61.

4. Grahame R. Joint hypermobility and genetic collagen disorders: Are they related? Arch Dis Child 1999; 80: 188-91.

5. Kobayasi T, Oguchi M, Asboe-Hansen G. Dermal changes in Ehlers-Danlos syndrome. Clin Genet 1984;25: 477-81.

6. Kobayasi T, Wolf-Jurgensen P, Asboe-Hansen G. Ultrastructure of shagreen patch. Acta Dermatovene 1973; 53: 13-21.

7. Ellis,FE. Cutaneous elasticity and hyperelasticity. Arch Dermatol 1956; 74: 22-32.

8. Vogel A, Holbrook KA, Steinmann B, Gitzelmann R, Byers, PH. Abnormal collagen fibril structure in the gravis form (type 1) of Ehlers Danlos syndrome. Lab Invest 1979; 40: 201-6.

9. Hassner I, Anton Lamprecht I. Diffrerential ultrastructural abrreviations of collagen fibrils in Ehlers Danlos syndrome. Types I-IV as a means of diagnostics and classification. Hum Genet 1994; 3: 394-407.

10. Smith LT, Schwarze U, Goldstein J, ByersPH. Mutations in the COL3A1 gene result in the Ehlers Danlos syndrome type IV and alterations in the size and result in the Ehlers Danlos syndrome type IV and alterations in the size and distribution of the major collagen fibrils of the dermis. J Invest Dermatol 1997; 108: 241-7.

11. Pope FM, Nicolls AC, Palan A, Kwee ML, De Groot WP, Hausmann R. Clinical features of an affected father and daughter with Ehlers Danlos syndrome type VIIB. Brit J Dermatol 1992; 126: 77-82.

12. Watson RB, Wallis GA, Holmes DF, Viljoen D, Byers PH, Kadler KE. Ehlers danlos syndrome type VIIB. J Biol Chem 1992; 267: 9093-100.

13. Holmes DF, Watson RB, Steinmann B, Kadler KE. Ehlers Danlos syndrome type VIIB. J Biol Chem 1993; 268: 15758-65.

14. Dyne KM, Vitellaro-Zuccarello L, Bacchella L, Lanzi G, Getta G. Ehlers Danlos syndrome type VIII: Biochemical, stereological and immunocytochemical studies on dermis from a child with clinical signs of Ehlers Danlos syndrome and a family history of premature loss of permanent teeth. Brit J Dermatol 1993; 128: 458-63.

15. Kobayasi T, Karlsmark T, Ullman S. Ultrastructural aspects in scleroderma. In: Nishioka K, Krieg Th, editors. Proceedings of the international symposium on pathogenesis and management of scleroderma and connective tissue disorders. Tokyo: Med Dent Univ Press, 1993: 29-39.

16. Ghadially FN. Ultrastructural pathology of the cell and matrix. London: Butterworths 1988: 1248-51.

17. Steinmann B, Gitzelmann R, Vogel A, Grant ME, Harwood R, Seare HJ, Ehlers Danlos syndrome in two siblings with deficient lysyl hydroxylase activity in cultured skin fibroblasts but only mild hydroxylysine deficient in skin. Helv paediat Acta 1975; 30: 255-74.

18. Hulmes DJS, Kalder KE, Mould AP, Hyojima Y, Holmes DF, Cummings Ch, Chapmann JA, Prockop DJ. Pleomorphism in type I collagen fibrils produced by persistence of the procollagen N-propeptidase. J Mol Biol 1989; 210: 337-45.

19. Danielsen L, Kobayasi T, Larsen HW, Midtgaard K, Christensen HE. Pseudoxanthoma elasticum. A Clinicopathological study. Acta Dermatovener 1970; 50: 355-73.

20. Kobayasi T, Asboe-Hansen G. Ultrastructure of the skin in primary systemic amyoloidosis. Acta Dermatovene 1979; 59: 407-13.

21. Hashimoto K, Daniel WP, Wang PW, Etoh H. Late onset hyalinosis cutis et mucosae. J Dermatol 2000; 22: 425-33.

22. Kobayasi T, Bartosik J, Ullman S. Elastic fibers in dermis of juvenile elastoma. J Dermatol 1998; 5: 5-9.

23. Akar A, Gl D, sarathchandra P, Tysoe C, Pope M. Acrogeric Ehlers Danlos syndrome type IV: report of a new patient with additional findings. Eur J Dermatol 2002; 12: 428-31.

24. Lovell CR, Smolensk KA, Duance VC, Light ND, Young S, Dyson M. Type I and III collagen content and fibre distribution in normal human skin during ageing. Brit J Dermatol 1987;17: 419-28.

25. Handler CE, Child A, Light ND, Dorrance DE. Mitral valve prolaps, aortic compliance, and skin collagen in joint hypermobility syndrome. Br Heart J 1985; 54: 501-8.

26. Kohn RR, Schnider SL. Collagen changes in aging skin. In: Balin AJ, Kligman AM, editors. Aging and the skin. New York. Raven Press. 1988: 121-39.

27. De Paepe A, Nicolls A, Nardicfici P, De Keyser F, Qutatacker J, van Staey M, Matton M, Pope FM. Ehlers Danlos syndrome type I: a clinical and ultrastructural study of a family with reduced amounts of collagen type III. Brit J Dermatol 1987; 117: 89-97.

28. Bailey AG, Paul RG, Knott L. Review. Mechanisms of maturtion and ageing of collagen. Mechanisms of ageing and development 1998; 106: 1-56.

29. Liu X, Wu H, Byrne M, Krane S, Jaenisch R. Type III collagen is crucial for collagen I fibrogenesis and for normal cardiovascular development. Proc Natl Acad Sci USA 1997; 94: 1852-6.

30. Schwarze U, Atkinson M, Hoffman GG, Greenspan DS, Byers PH. Null Adelles of the COL5A1 gene of type V collagen. Are a cause of the classical forms of Ehlers Danlos syndrome (Types I and II). Am J Hum Genet 2000; 66: 1757-65.

31. Nicolls AC, Oliver JE, McCarron S, Harrison JB, Greenspan DS, Pope FM. An exon skipping mutation of a type V collagen gene (COL5A) in Ehlers-Danlos syndrome. J Med Genet 1996; 33: 940-6.

32. Birk DE, Fitch JM, Babiarz JP, Doane KJ, Linsenmeyer TF. Collagen fibrillogenesis in vitro: Interaction of type I and V collagen regulates fibril diameter. J Cell Sci 1990; 95: 649-57.

33. Kobayasi,T. Probable cohesion of basal lamina to dermis by type V collagen. J Invest Dermatol 2003; 21: 964 (Abstract).

34. Kobayasi T, Ullman S. Twisted collagen fibrils in acrocyanosis. Eur J Dermatol 1999; 9: 285-8.

35. Cupo LN, Pyeritz RE, Olson JL, McPhee SJ, Hutchins GM, McKusick VA: Ehlers-Dsanlos syndrome with abnormal colloagen fibrils, sinus of valsalva aneurysms, myocardial infarction, panacinar emphysema and cerebral heterotopias. Am J Med 1981; 71: 1051-8.

36. Scheck M, Siegel RC, PerkerJ, Chang Y-H, Fu JCC. Aortic aneurysm in Marfan’s syndrome: Changes in the ultrastructure and composition of collagen. J Anat 1979; 129: 645-57.

37. Viterallo-Zuccarello L, Garino Canina, G., DeBiasi S, Cocchini F, Vitellaro-Zuccarello P. Ultrastructural study on the stapes and skin in a case of type I Osteogenesis Imperfecta. J Submicrosc Cytol 1984; 16: 779-86.

38. Stöss H. Pathologische Anatomie der Osteogenesis imperfecta. Licht und elektronenmikroskopische Untersuchungen an Sttzgewebe und Haut. Veröffentlichungen aus der Pathologie. Progress in Pathologie 1990; 134: 1-78.