378
March 1978 TheJournalofPEDIATRICS
Inherited human collagen lysyl hydroxylase deficiency: Ascorbic acid response A patient & described with congenital hypotonia, lax joints, friable skin, hemorrhagic scars, high-arched palate, and borderline microcornea. Acid hydrolyzed whole ski n collagen had a reduced hydroxylysine content of O.5 residues per 1,000 as compared to 5.1 • 0.7 in control skin. Collagen lysyl hydroxylase in dialyzed subcellular fractions o f cultured skin fibroblasts required L-ascorbate as a principal cofactor. Activity of this enzyme in cultured skin fibroblasts derived from this"patient, his father, and mother were 17%, 66%, and 39% of control values, respectively. Collagen prolyl hydroxylase activity was normal. Pharmacologic amounts of oral vitamin C (4 gin~day) produced an increase and withdrawal resulted in abrupt diminution of urinary excretion of hydroxylysine. Over a two-fear period the patient's wound healing and muscle strength improved and corneal diameter increased. Hydroxylysine content of the skin did not increase.
Louis J. E l s a s II, M.D.,* R o b e r t L. M i l l e r , B.S., A t l a n t a , Ga., and Sheldon R. Pinnell, M.D.,** D u r h a m , N . C.
INCREASED UNDERSTANDING in recent years of the structure and synthesis of collagen has been associated with the description of several inherited human disorders of collagen biosynthesis. H1 Collectively, a heterogeneous group of diseases clinically described as Ehlers-Danlos syndrome are now subclassified into at least seven different types based on genetic and phenotypic differFrom the Division of Medical Genetics, Department of Pediatrics, Emory University School o f Medicine, and the Division of Dermatology, Department of Medicine, Duke University School of Medicine. Supported by United States Public Health Service research grant CRC RR-OOO39from the National Institutes of Health. Presented in part at the National Meeting of the American Federation for Clinical Research, Atlantic City, May 1, 1976, and published in Clinical Research 24:294A, 1976. *Recipient of United States Public Health Service research grant RCDA HD 35,615from the National Institutes of Health. **Recipient of United States Public Health Service research grant AM 17,128from the National Institutes of Health and a grantfrom the Howard Hughes Medical Institute. Reprint address: Dr. L. J. Elsas lI, ,Oivisionof Medical Genetics, Box 23344, Emory University, Atlanta, GA 30322.
Vol. 92, No. 3, pp. 378-384
ences. In Types IV, V, VI, and VII different biochemical mechanisms have been proposed. '-11 The first of these to be defined biochemically was hydroxylysine (Hyl)-deficient collagen disease with impaired collagen lysyl hydroxylase (E.C.I.14.11.4) classifted as Type VI. 6, 7, lo Four affected families, including the present one, have been reported with Hyl-deficient skin collagen and impaired collagen lysyl hydroxylase?, 8-10 Reducing agents required for the assay of this enzyme included dithiothreitol and ascorbic acid. Quinn and Krane 12 demonstrated that the enzyme from cultured fibroblasts of their patient was less stable at 37 ~ than the wild type and had a lower affinity for ascorbate than the wild type. In many inherited disorders, knowledge of the cofactor requirements for the specifically impaired enzyme reactions and the subsequent administration of pharmacologic amounts of vitamin precursors of these active cofactors have resulted in dramatic clinical improvement. 1~ No specific enzyme requiring ascorbic acid has been defined which, when impaired genetically, is also augmented by supraphysiologic concentrations of this cofactor. Since Ehlers-Danlos syndrome (Type VI) involved a specifically impaired enzyme which utilized ascorbate as its principal
0022-3476/78/0392-0378500.70/0 9 1978 The C. V. Mosby Co.
Volume 92 Number 3 physiologic reductant, we considered the potential basic information and practical therapeutic alternatives to be gained from a study of the in vivo effects of ascorbate on this disease. The present work describes a patient with EhlersDanlos (Type VI) phenotype with Hyl-deficient skin collagen and a clinical response to pharmacologic amounts of vitamin C.
CASE REPORT The patient (J. D. H.), a white boy, was studied between the ages of 8 and 10 years. He was born at term with marked hypotonia. The wrists were completely subluxed and the heads of the humeri could be lifted out of the glenoid cavity bilaterally. He developed socially at a normal rate, but was unable to sit until age 16 months and did not walk until 22 months, He had multiple lacerations after minor trauma that were associated with abnormal scar formation. He had mild myopia. His skin was velvety, smooth, and hyperextensible over the elbows. There were multiple hemorrhagic, atrophic scars, particularly over the knees, elbows, and tibial area. There were molluscoid pseudotumors and spherules over knees and elbows. The joints were hyperextensible and he could touch the tips of both shoulders under his chin. Corneal diameter was below the lower limits of normal and was 10.2 mm on the right and 10.4 mm on the left. There were no retinal changes or laxity of ciliary muscles. He had a high-arched palate. Quantitative muscle testing demonstrated severe muscle weakness particularly of the chest, back, shoulder girdle, and abdomen. Pulmonary function tests indicated an increased residual volume to total lung capacity of 40% as compared to an expected value of less than 25%. Ivy method bleeding time was prolonged at 1289 minutes (normal = 1 to 9 minutes). A clinical diagnosis of Ehlers-Danlos syndrome was made and skin biopsy was obtained. The patient's skin contained only 10% of norm~tl Hyl residues, but other amino acids were present in normal amounts.
METHODS Clinical research studies. Five normal control subjects and the patient were studied through the Emory University Clinical Research Unit. Each patient was informed in detail of the investigative nature of the study and the procedures to be performed, and written consent was obtained. The patients were placed on a collagen-free diet consisting of 30 calories/kg, 1 gm protein/kg, in which 50% of total caloric intake was administered as carbohydrate. All protein was given as casein for two days before and during the studies. Preparation and analysis of urine and skin for amino acids and glycosides of hydroxylysine. Ten-milliliter aliquots of urine were acidified, lyophilized, and analyzed directly by ion exchange chromatography on a Beckman 120 amino acid analyzer using previously defined
Inherited hydroxylase deficiency
379
Table I. Collagen lysyl and prolyl hydroxylase activities in a subcellular fraction from cultured skin fibroblasts
Enzyme activity
CeH fine Patient Mother Father Control
Lysyl hydroxylase (DPM "~Hrelease/mg protein~20 min)
Prolyl hydroxyIase (DPM ~H release~rag protein~20 rain)
3,505 +_ 588 (8) 7,923 _+ 495 (9) 13,590 +_ 1,569 (8) 20,446 _+ 302 (18)
47,639 +_ 8,881 (8) 24,572 _+ 2,696 (6) 56,680 • 3,251 (6) 36,525_+ 12,006 (18)
subjects Data are presented as the mean • 1 SD with the number of observations in parentheses. Control subjects represent a total of four different cell lines, three of which were assayed at 24 to 26 population doublings a n d one at 10 population doublings.
techniques. 1~ Free Hyl and other urinary amino acids were quantitated in this manner. Hyl peaks were confirmed by their disappearance after periodate treatment. Total urinary Hyl and Hyp were quantitated after acid hydrolysis for 18 hours at 110~ Urine was analyzed for Hyl glycosides after hydrolysis i n 2 N potassium hydroxide by previously described techniques.16 4-Hyp was separated from 3-Hyp on a Jeolco polystyrene resin using citrate buffer at p H 2.5. Skin amino acids were quantitated on 4 m m lyophilized skin biopsy specimens using previously described techniquesY Ascorbic acid was measured in urine and serum after extraction with metaphosphoric acid as 2,4-dinitrophenylhydrozone derivative. TM Preparation of labeled unhydroxylated substrate and enzyme. Labeled unhydroxylated collagen substrate was prepared from 15-day-old chick embryo calvaria by a modification of previously described techniques? '~ Essentially 5-[4,5-:~H] lysine or proline was incubated with calvaria in lysine- or proline-free media, in the presence of a,a' -dipyridyl. Tissues were then homogenized, and the supernate was dialyzed, and the collagen fracfionated, pelleted, and stored frozen. Skin biopsy material was obtained by punch biopsy and grown in monolayer cultures using Dulbecco and Vogt's modified Eagle's media supplemented with 15% fetal calf serum and glutamine using previously described methods? ~ Enzyme was obtained from cells by harvesting with trypsin, freeze-thawing in Tris-HC1 buffer, and centrifuging at 15,000 x g for 10 minutes. Supernate contained the enzyme and was dialyzed for 24 hours and frozen at -20~ Activity of dialyzed enzyme was stable for four weeks at --20 ~ Protein content was determined by a modified method of Lowry et alY Enzyme assay. Lysyl hydroxylase activity was deter-
380
Elsas, Miller, and Pinnell
The Journal of Pediatrics March 1978 Controls
Patient (age 8)
Patient (age I0)
4
1 2 I
'
50
/
/7'
'
-'
'
'
0
8yr
E
z~ uJ (,9 25
~g
o E
~
9yr 12yr 34 yr 23yr
0
I
2
34
56
4
5678
DAYS Fig. 1. Effect of ascorbate on urinary free hydroxylysine excreted by control subjects and Patient J. D. H. Control values are represented by open circles before and after ascorbate and followed by the subject's age. Where two consecutive samples were analyzed the range of values is indicated by bars for both control subjects and the patient. Dosage of ascorbate is represented by stippled bars above the day administered. Free hydroxylysine was quantitated by ion exchange chromatography of lyophilized urine as described in the Methods section. mined by the tritium exchange method of Peterkofsky and DiBlasio TM with minor modifications. L-[4,5-3H] lysyl collagen substrate was incubated with fibroblast enzyme preparation at 30~ in a reaction mixture which contained 50 mM Tris-HC1 (pH 7.5), 0.5 mM a-ketoglutarate, 0.0 to l.0 mM L-ascorbate, 0.1 mg catalase, 0.05 mM Fe(NH4)~SO4, 0.75 mg bovine serum albumin, and 0.1 mM dithiothreitol. The assay was stopped by the addition of 50% trichloroacetic acid. The supernate was then applied to A G Dowex 50W-X8 columns and the eluted tritium counted by liquid scintillation spectrometry. Prolyl hydroxylase activity was measured by the same procedure except that the reaction mixture contained L[3,4-3H] prolyl collagen substrate and 0.1 mM a-ketoglutarate. In both assays enzyme activity was defined as tritium release in the presence as compared to the absence of a-ketoglutarate. Materials. L-[4,5-3H] lysine (21 Ci/mmole) and L-[3,4~H] proline (41 Ci/mmole) were purchased from Amersham/Searle Corp.; a-ketogutarate, L-ascorbic acid, and a,a' -dipyridyl were obtained from Sigma Chemical Co. Catalase was purchased from Worthington Biochemical Corp., Pentex bovine albumin fraction V, fatty acid poor, from Miles Laboratories, Inc., and dithiothreitol from Calbiochem. Dowex 50W-X8 was purchased from Bio Rad Laboratories. RESULTS Skin from the patient had 0.5 residues of Hyl per 1,000 amino acids whereas normal skin contains 5.1 _+ 0.7
residues per 1,000. The ratio of Hyl to Hyp was 0.011 as compared to normal values of 0.051 + 0.0094 Since skin collagen Hyl content is determined by collagen lysyl hydroxylase, the specificity and genetic control of this enzyme were determined using cultured skin fibroblasts from the proband, his parents, and control cell lines. Collagen lysyl hydroxylase activity in the patient's cultured skin fibroblasts was approximately 17% of control values from four different lines; his father's and mother's cell lines had 66% and 39% of control activity, respectively (Table I). By contrast, the patient's cell line had prolyl hydroxylase activity which was statistically indistinguishable from control lines. This collagen lysyl hydroxylase assay was evaluated to enable cofactor assessment. A pH optimum of 7.5 was determined at 30~ and activity was linear throughout 20 minutes of incubation. A direct relationship was seen between tritium release and protein content of the subcellular enzyme preparation between 0 and 75 /~g. At all protein concentrations, activity from the mutant cell line was markedly reduced. Following these initial experiments, prolonged dialysis of control enzyme preparations was undertaken to assure removal of aqueous-soluble cofactors such as ascorbate. When L-ascorbate alone was removed from the reaction mixture, only 4% of,complete normal" activity was present. Because of the observed role of ascorbate in collagen lysyl hydroxylase function and because of the hemorrhagic quality of the scar tissue, the patient was first evaluated for potential ascorbate deficiency. Dietary
Volume 92 Number 3 history indicated that he had been ingesting at least 50 mg of ascorbate per day over the past year. Twenty-four-hour urine and plasma ascorbate were compared to three normaL, age-matched control subjects. The patient's urinary and plasma levels of ascorbate before loading were within the normal range of 8 to 37 m g / d a y / g m of creatinine and 0.2 to 2 mg/dl, respectively. After a single oral dose Of 4 gm of ascorbate the patient excreted 786 mg/day/gm of creatinine in urine as compared to three control subjects who excreted between 285 and 1,155. Plasma levels of ascorbate after loading rose similarly to those in control subjects and reached levels between 2.0 and 4.0 mg/dl plasma. Thus ascorbate deficiency was not present. An in vivo response to pharmacologic doses of ascorbate was then investigated by measuring the urinary excretion of both total and free Hyl. As seen in Fig. 1, control subjects' 24-hour urinary free Hyl excretion fell with increasing age. After treatment with 4 gm of ascorbate daily for four days, Hyl excretion rose significantly in all control subjects (P < 0.01 using paired t test). At age 8, the patient's Hyl excretion was reduced compared to his age-matched control subject. With increasing doses of ascorbate his free Hyl excretion rose to 48 then stabilized at 25/zmoles/day/gm Of creatinine. As seen in Table II, total urinary Hyl values also rose, although not as dramatically. Although slight elevations in Hyp excretion were seen, they were not as great as those in Hyl excretion. Thus the ratio of Hyl to Hyp increased (Table II). These values for total Hyl were somewhat lower than those in the age-matched control subject, which was 196/zmoles/ day/gm of creatinine and did not change on ascorbate treatment. Urinary Hyp levels in the proband were the same as in control subjects; the Hyl to Hyp ratio in the control subjects was between 0.13 and 0.14 whereas the patient's ranged between 0.08 to 0.12. Urinary glycosides of Hyl were measured and accounted for between 66 and 85% of the total Hyl content. There was no consistent change in glucosylated Hyl or the ratio of glucosylgalactosylhydroxylysine to galactosylhydroxylysine with administration of increasing amounts of ascorbate. Since kidney basement membrane collagen is rich in Hyl and has relatively more 3-Hyp than 4-Hyp, the ratio of urinary 3-Hyp to 4-Hyp was measured. 23 The ratio in both the patient and his age-matched control subject was 0.056 before and 0.047 after ascorbate, suggesting no significant contribution by kidney basement membrane to increased Hyl excretion. The patient's urinary excretion of total and free Hyl was monitored while he was receiving ascorbate over the next two years. During this period, plasma ascorbate ranged between 2.0 and 4.0 mg/dl. At 10 years of age the patient's
Inherited hydroxylase deficiency
381
Fig. 2. Scars from skin biopsies over left buttock. Three "preascorbate" scars (arrows pointing right) were obtained in March and June, 1974. Three "postascorbate" scars (arrows pointing left) were obtained in September, 1974, and March, 1975, three and eight months after initiation of vitamin C. The photograph was taken in February, 1976. Table II. Effect of ascorbate on the urinary excretion of total hydroxylysine and 4-hydroxyproline by proband (age 8 years)
Ascorbate (gm/day) 0 0 0 1 2 3 4 4 4
Urinary amino acid* (~moles/24 hr/gm creatine) Hyl
Hyp
Ratio of Hyl/Hyp
112 116 91 113 129 132 158 163 137
1,196 1,199 1,197 1,049 1,164 1,068 1,637 1,426 1,241
0.09 0.10 0.08 0.11 0.11 0.12 0.10 0.11 0.11
*Aliquots of 24-hour urine samples were concentrated 21/2-fold,then hydrolyzedat II0~ for 18 hours in 6 N HC1. Data are single observations. maximum free Hyl excretion was approximately 13 /~moles/day/gm of creatinine (Fig. 1). Ascorbate was discontinued and urinary free Hyl was measured from four to eight days later. There was an immediate and dramatic fall in Hyl excretion to less than 3/~moles/day/ gm of creatinine, considerably lower than adult control subjects (Fig. 1). Thus the increase and fall in urinary Hyl was associated with oral administration of vitamin C. Ascorbate therapy produced no observable change in skin Hyl content, but improved the quality of scar
382
Elsas, Miller, and Pinnell
The Journal of Pediatrics March 1978
Table III. Relationship of corneal size, age, and ascorbate therapy
Patient~ age 8.5 9.0 10.0 Normal males 2~ (age 5-9)* Mean Minimum Maximum Normal males (age 10-19)t Mean Minimum Maximum
Duration of ascorbate therapy (mo) 0 8 20 0
Corneal diameter (ram) OD
[
10.2 10.3 10.5
OS 10.4 10.5 10.7
11.7 11.0 12.5 11.7 11.0 12.5
*Data from 54 corneasin 27 males. tData from 124 corneasin 62 males.
formation. Biopsies were obtained before and three months and eight months after initiation of vitamin C therapy. Hyl content was 0.5, 0.5 and 0.9 residues/I,000 amino acids in each of these samples. However, the character of biopsy scars improved dramatically (Fig. 2). Three scars produced by biopsies obtained in March and July, 1974, before ascorbate therapy was begun, were red, raised, and thin, in contrast to three biopsy scars similarly produced three months and eight months after ascorbate treatment was begun. On histologic examination, preascorbate scars contained dilated vessels with extravasated old blood, whereas postascorbate scars looked normal. Scars formed while the patient was on ascorbate therapy were not large enough to biopsy adequately for Hyl determination. In addition to improved wound healing, several other observations suggested clinical improvement. Corneal diameter increased 0.3 mm in each eye over the 20 months of ascorbate therapy (Table III). Corneal diameter normally does not increase after four years of age. 24 After ascorbate therapy, the patient's corneal diameter approached but did not reach the lower limit of his ageand sex-matched control subjects. 2' Muscle strength improved, as assessed by direct muscle testing. The increased residual volume noted on previous pulmonary function tests was normal one year after therapy. Bleeding time returned to normal. Joint laxity and friability of skin, however, remained. DISCUSSION As in most described inherited metabolic disorders, a large hiatus exists between the observed biochemical
impairment and the mechanisms producing the expressed phenotypeY 5 The patient described herein had muscle weakness, lax joints, and hemorrhagic, friable skin. His skin collagen contained only 10% of normal Hyl. Collagen lysyl hydroxylase was impaired in his cultured skin fibroblasts and an autosomal recessive mode of inheritance for this enzyme defect was observed. This constellation of phenotypic changes may be related to the biochemical alterations through a reduction of number, type, and distribution of cross-linking in dermal, tendinous, and vascular collagen. 6, ~, ~6 By contrast to previously reported phenotypes, in our patient severe vertebral and ophthalmic abnormalities were not evident. One possible explanation for these organ differences is that different isozymes of lysyl hydroxylase exist in different tissues. The expression of a mutation by a specific organ would depend, therefore, on the relative importance of a given subunit to that organ's function. For instance, Type III collagen [l(III)]:, which predominates in the collagen of fetal tissues and blood vessels, may require a different lysyl hydroxylase than Type I [1(i)]3 collagen for its production. No data on this point are available. Another possibility is that different tissues have different cofactor concentrations, particularly of Lascorbate, which we found to be the principle physiologic reductant for lysine hydroxylation. Pinnell et al 6 espoused this hypothesis to explain the variable degree of Hyl deficiency in collagen from different tissues of the same patient. Variations in tissue content of ascorbate may affect the expression of different mutations, depending on the molecular site of enzyme impairment and the relationship Of this site to L-ascorbate action. Pharmacologic amounts of ascorbate were associated with improvements in muscle strength, corneal growth, bleeding time, an d pulmonary residual volume. In addition, Urinary hydroxylysine excretion increased on administration of ascorbate and decreased on withdrawal. These in vivo observations suggested that 4 g m / d a y of ascorbate increased hydroxylation of lysyl residues in newly synthesized collagen and that its production and degradation were measurable in urine. Askenasi 2~-3~demonstrated that the amount of free and peptide-bound urinary Hyl was a physiologic index o f collagen metabolism. He also provided evidence that renal basement membrane did not contribute significantly to the urinary Hyl pool. 3i Since the Clq component of hemolytic complement represents only a ,fraction of total body Hyl, its contribution to urinary Hyl must be negligible. If this regimen of vitamin C were augmenting mutant lysyl hydroxylase, why then did skin Hyl content not increase? One explanation is that L-ascorbate enhanced collagen lysyl hydroxylase activity only where new collagen was produced, as in new scar formation or
Volume 92 Number 3 corneal growth. Gould 32 postulated in scorbutic models that there was more than a single mechanism for collagen formation: one used for normal body collagen formation and one predominating in tissue repair. He suggested that the former was independent o f ascorbate and the latter was dependent on its presence. Barnes et a133, 34 implied that in cultured L-929 cells, hydroxylation of lysine in the nonhelical portion of the developing collagen molecule was under control by a separate ascorbate-dependent lysyl hydroxylase distinct from a more active lysyl hydroxylase. Hornig 3~ found that 14C-ascorbate localized to the ground substance of the intima of guinea pig blood vessels and suggested that n o n r a n d o m distribution of ascorbate indicated differences in its utilization by various tissues. Miller 36 further demonstrated that ascorbate did not activate preformed, inactive lysyl hydroxylase in nonscorbutic cultured L-929 cells as was seen for prolyl hydroxylase. In dermal skin collagen only five of approximately 30 lysine residues are normally hydroxylated on a 1(I) and a2 chainsY In our patient only 0.5 were hydroxylated. By contrast, corneal collagen m a y have as much as 67% hydroxylation of its a l ( I V ) chain lysyl residues. 23 This unusual degree of hydroxylation in corneal collagen may account for the unique growth of this tissue during administration of ascorbate to our patient. Thus, the improved wound healing and corneal growth expressed by our patient in the absence of increased Hyl content of preformed skin collagen might be an expected result from the administration of large amounts of vitamin C in vivo, if only specialized lysyl hydroxylase, actively forming new collagen in specific organs, was augmented. The mechanisms of action of ascorbate on our mutant enzyme and the variable tissue responses require further clarification. However, the observed in vivo responses of this single gene mutation and its expression in cultured fibroblasts provides a model for further investigation. We thank Cheryl Jorgensen Vroman for typing the manuscript, and Drs. Robert Priest, Anthony Keyser, and Dean J. Danner for their support and constructive criticisms. We owe special gratitude to Dr. John Reiser for his ophthalmologic measurements, to Dr. Jean Priest for her advice and supervision regarding human fibroblast tissue culture, and to Ms. Bettye Hollins for her technical assistance.
REFERENCES
1. Grant ME, and Prockop D J: The biosynthesis of collagen, N Engl J Med 286:194, 242, 291, 1972. 2. Nimni ME: Metabolic pathways and control mechanisms involved in the biosynthesis and turnover of collagen in normal and pathological connective tissues, J Oral Pathol 2:175, 1973. 3. McKusick VA, and Martin GR: Molecular defects in collagen, Ann Intern Med 82:585, 1975.
Inherited hydroxylase deficiency
3 83
4. Pope FM, Martin GR, Lichtenstein JR, Penttinen R, Gerson B, Rowe DW, and McKusick VA: Patients with EhlersDanlos syndrome Type IV lack Type II1 collagen, Proc Natl Acad Sci USA 72:1314, 1975. 5. Di Ferrante N, Leachman RD, Angelini P, Donnelly PV, Francis G, Almazan A, and Segni G: Lysyl oxidase deficiency in Ehlers-Danlos syndrome type V, Connect Tissue Res 3:49, 1975. 6. PinneU SR, Krane SM, Kenzora JE, and Glimcher MJ: A heritable disorder of connective tissue. Hydroxylysinedefcient collagen disease, N Engl J Med 286:1013, 1972. 7. Krane SM, Pinnell SR, and Erbe RW: Lysyl protocollagen hydroxylase deficiency in fibroblasts from siblings with hydroxylysine-deficient collagen, Proc Nail Acad Sci USA 69:2899, 1972. 8. Elsas LJ, Hollins B, and Pinnell SR: Hydroxylysine deficient collagen disease: Effect of ascorbic acid, Am J Hum Genet 26:28A, 1974. 9. Sussman M, Lichtenstein JR, Nigra TP, Martin GR, and McKusick VA: Hydroxylysine-deficient skin collagen in a patient with a form of the Ehlers-Danlos syndrome, J Bone Joint Surg 56A:1228, 1974. 10. Steinmann B, Gitzelmann R, Vogel A, Grant ME, Harwood R, and Sear CHJ: Ehlers-Danlos syndrome in two siblings with deficient lysylhydroxylase activity in cultured skin fibroblasts but only mild hydroxylysine deficit in skin, Helv Paediatr Acta 30:255, 1975. 11. Lichtenstein JR, Martin GR, Kohn LD, Byers PH, and McKusick VA: Defect in conversion of procollagen to collagen in a form of Ehlers-Danlos syndrome, Science 182:298, 1973. 12. Quinn RS, and Krane SM: Abnormal properties of collagen lysyl hydroxylase from skin fibroblasts of siblings with hydroxylysine-deficient collagen, J Clin Invest 57:83, 1975. 13. Scriver CR: Progress in endocrinology and metabolism: Vitamin-responsive inborn errors of metabolism, Metabolism 22:1319, 1973. 14. Nishikimi M, and Udenfriend S: Scurvy as an inborn error of ascorbic acid biosynthesis, Trends Biochem Sci 2:111, 1977. 15. Rosenberg LE, Durant JL, and Elsas LJ: Familial iminoglycinuria. An inborn error of renal tubular transport, N Engl J Med 2"/8:1407, 1968. 16. Pinnell SR, Rox R, and Krane SM: Human collagens: Differences in glycosylated hydroxylysines in skin and bone, Biochim Biophys Acta 229:119, 1971. 17. Miller EJ, and Piez KA: An accelerated single-column procedure for the automatic analysis of amino acids in collagen and elasfin hydrolyzates, Anal Biochem 16:320, 1966. 18. Roe JH: Appraisal of methods for the determination of Lascorbic acid, Ann NY Acad Sci 92:277, 1961. 19. Peterkofsky B, and DiBlasio R: Modification of the tritiumrelease assays for prolyl and lysyl hydroxylases using Dowex-50 columns, Anal Biochem 66:279, 1975. 20. Elsas LJ, Priest JH, Wheeler FB, Danner DJ, and Pask BA: Maple syrup urine disease: Coenzyme function and prenatal monitoring, Metabolism 23:569, 1974. 21. Priest JHi Medical cytogenetics and cell culture, ed 2, Philadelphia, 1977, Lea & Febiger, Publishers, pp 263295. 22. Lowry OH, Rosebrough N J, Farr AL, and Randall R J:
3 84
23. 24.
25.
26. 27.
28.
29.
30.
Elsas, Miller, and Pinnell
Protein measurement with the Folin phenol reagent, J Biol Chem 193:265, 1957. Kefalides NA: Structure and biosynthesis of basement membranes, Int Rev Connect Tissue Res 6:63, 1973. Smith P: On the size of the cornea in relation to age, sex, refraction, and primary glaucoma, Trans Ophthal Soc UK 10:68, ]890. Gendel BR, and Elsas L J: Medical genetics, in Sodeman WA, and Sodeman WA Jr, editors: Pathologic physiology. Mechanism of disease, ed 5, Philadelphia, 1974, WB Saunders Company, chap 3. Eyre DR, and Glimcher M J: Reducible crosslinks in connective tissue, Proc Natt Acad Sci USA 69:2594, 1972. Askenasi R: A new rapid method for measuring hydroxylysine and fis glycosides in hydrolysates and physiological fluids, Biochim Biophys Acta 304:375, 1973. Askenasi R: urinary hydroxylysine and hydroxylysyl glycoside excretions in normal and pathologic states, J Lab Clin Med 83:673, 1974. Askenasi R, and Demeester-Mirkine N: Urinary excretion of hydroxylysyl glycosides and thyroid function, J Clin Endocrinol Metab 40:342, 1975. Askenasi R: Urinary excretion of free hydroxylysine, peptide-bound hydroxylysine and hydroxylysyl glycosides in physiological conditions, Clin Chim Acta 59:87, 1975.
The Journal of Pediatrics March 1978
31. Askenasi R, Van Herweghem JL, and Ducobu J: Contribution of kidney basement membranes and collagen to the urinary excretion of hydroxylysyl glycosides, Biomedicine 22:233, 1975. 32. Gould BS: The role of certain vitamins in collagen formation, in Gould BS, editor: Treatise on collagen, New York, 1968, Academic Press, Inc, vol 2A, p 323. 33. Barnes M J, Constable BJ, Morton LF, and Royce PM: Agerelated variations in hydroxylation of lysine and proline in collagen, Biochem J 139:461, 1974. 34. Barnes MJ, Constable BJ, Morton LF, and Kodicek E: Studies on the hydroxylation of lysine occurring in the nonhelical region at the N-terminal end of the collagen molecule, Biochem J 125:16, 1971. 35. Hornig D: Distribution of aseorbic acid, metabolites and analogues in man and animals, in King CG, and Burns J J, editors: Second Conference on Vitamin C, The New York Academy of Sciences, New York, Ann NY Acad Sci 258:103, 1975. 36. Miller RL: The effect of ascorbic acid on lysyl and prolyl hydroxylase activity of cultured fibroblasts, Arch Biochem Biophys 170:341, 1975. 37. Bornstein P, and Piez KA: A biochemical study of human skin collagen and the relation between intra- and intermolecular cross-linking, J Ctin Invest 43:1813, 1964.
March 1978 TheJournalofPEDIATRICS
Inherited human collagen lysyl hydroxylase deficiency: Ascorbic acid response A patient & described with congenital hypotonia, lax joints, friable skin, hemorrhagic scars, high-arched palate, and borderline microcornea. Acid hydrolyzed whole ski n collagen had a reduced hydroxylysine content of O.5 residues per 1,000 as compared to 5.1 • 0.7 in control skin. Collagen lysyl hydroxylase in dialyzed subcellular fractions o f cultured skin fibroblasts required L-ascorbate as a principal cofactor. Activity of this enzyme in cultured skin fibroblasts derived from this"patient, his father, and mother were 17%, 66%, and 39% of control values, respectively. Collagen prolyl hydroxylase activity was normal. Pharmacologic amounts of oral vitamin C (4 gin~day) produced an increase and withdrawal resulted in abrupt diminution of urinary excretion of hydroxylysine. Over a two-fear period the patient's wound healing and muscle strength improved and corneal diameter increased. Hydroxylysine content of the skin did not increase.
Louis J. E l s a s II, M.D.,* R o b e r t L. M i l l e r , B.S., A t l a n t a , Ga., and Sheldon R. Pinnell, M.D.,** D u r h a m , N . C.
INCREASED UNDERSTANDING in recent years of the structure and synthesis of collagen has been associated with the description of several inherited human disorders of collagen biosynthesis. H1 Collectively, a heterogeneous group of diseases clinically described as Ehlers-Danlos syndrome are now subclassified into at least seven different types based on genetic and phenotypic differFrom the Division of Medical Genetics, Department of Pediatrics, Emory University School o f Medicine, and the Division of Dermatology, Department of Medicine, Duke University School of Medicine. Supported by United States Public Health Service research grant CRC RR-OOO39from the National Institutes of Health. Presented in part at the National Meeting of the American Federation for Clinical Research, Atlantic City, May 1, 1976, and published in Clinical Research 24:294A, 1976. *Recipient of United States Public Health Service research grant RCDA HD 35,615from the National Institutes of Health. **Recipient of United States Public Health Service research grant AM 17,128from the National Institutes of Health and a grantfrom the Howard Hughes Medical Institute. Reprint address: Dr. L. J. Elsas lI, ,Oivisionof Medical Genetics, Box 23344, Emory University, Atlanta, GA 30322.
Vol. 92, No. 3, pp. 378-384
ences. In Types IV, V, VI, and VII different biochemical mechanisms have been proposed. '-11 The first of these to be defined biochemically was hydroxylysine (Hyl)-deficient collagen disease with impaired collagen lysyl hydroxylase (E.C.I.14.11.4) classifted as Type VI. 6, 7, lo Four affected families, including the present one, have been reported with Hyl-deficient skin collagen and impaired collagen lysyl hydroxylase?, 8-10 Reducing agents required for the assay of this enzyme included dithiothreitol and ascorbic acid. Quinn and Krane 12 demonstrated that the enzyme from cultured fibroblasts of their patient was less stable at 37 ~ than the wild type and had a lower affinity for ascorbate than the wild type. In many inherited disorders, knowledge of the cofactor requirements for the specifically impaired enzyme reactions and the subsequent administration of pharmacologic amounts of vitamin precursors of these active cofactors have resulted in dramatic clinical improvement. 1~ No specific enzyme requiring ascorbic acid has been defined which, when impaired genetically, is also augmented by supraphysiologic concentrations of this cofactor. Since Ehlers-Danlos syndrome (Type VI) involved a specifically impaired enzyme which utilized ascorbate as its principal
0022-3476/78/0392-0378500.70/0 9 1978 The C. V. Mosby Co.
Volume 92 Number 3 physiologic reductant, we considered the potential basic information and practical therapeutic alternatives to be gained from a study of the in vivo effects of ascorbate on this disease. The present work describes a patient with EhlersDanlos (Type VI) phenotype with Hyl-deficient skin collagen and a clinical response to pharmacologic amounts of vitamin C.
CASE REPORT The patient (J. D. H.), a white boy, was studied between the ages of 8 and 10 years. He was born at term with marked hypotonia. The wrists were completely subluxed and the heads of the humeri could be lifted out of the glenoid cavity bilaterally. He developed socially at a normal rate, but was unable to sit until age 16 months and did not walk until 22 months, He had multiple lacerations after minor trauma that were associated with abnormal scar formation. He had mild myopia. His skin was velvety, smooth, and hyperextensible over the elbows. There were multiple hemorrhagic, atrophic scars, particularly over the knees, elbows, and tibial area. There were molluscoid pseudotumors and spherules over knees and elbows. The joints were hyperextensible and he could touch the tips of both shoulders under his chin. Corneal diameter was below the lower limits of normal and was 10.2 mm on the right and 10.4 mm on the left. There were no retinal changes or laxity of ciliary muscles. He had a high-arched palate. Quantitative muscle testing demonstrated severe muscle weakness particularly of the chest, back, shoulder girdle, and abdomen. Pulmonary function tests indicated an increased residual volume to total lung capacity of 40% as compared to an expected value of less than 25%. Ivy method bleeding time was prolonged at 1289 minutes (normal = 1 to 9 minutes). A clinical diagnosis of Ehlers-Danlos syndrome was made and skin biopsy was obtained. The patient's skin contained only 10% of norm~tl Hyl residues, but other amino acids were present in normal amounts.
METHODS Clinical research studies. Five normal control subjects and the patient were studied through the Emory University Clinical Research Unit. Each patient was informed in detail of the investigative nature of the study and the procedures to be performed, and written consent was obtained. The patients were placed on a collagen-free diet consisting of 30 calories/kg, 1 gm protein/kg, in which 50% of total caloric intake was administered as carbohydrate. All protein was given as casein for two days before and during the studies. Preparation and analysis of urine and skin for amino acids and glycosides of hydroxylysine. Ten-milliliter aliquots of urine were acidified, lyophilized, and analyzed directly by ion exchange chromatography on a Beckman 120 amino acid analyzer using previously defined
Inherited hydroxylase deficiency
379
Table I. Collagen lysyl and prolyl hydroxylase activities in a subcellular fraction from cultured skin fibroblasts
Enzyme activity
CeH fine Patient Mother Father Control
Lysyl hydroxylase (DPM "~Hrelease/mg protein~20 min)
Prolyl hydroxyIase (DPM ~H release~rag protein~20 rain)
3,505 +_ 588 (8) 7,923 _+ 495 (9) 13,590 +_ 1,569 (8) 20,446 _+ 302 (18)
47,639 +_ 8,881 (8) 24,572 _+ 2,696 (6) 56,680 • 3,251 (6) 36,525_+ 12,006 (18)
subjects Data are presented as the mean • 1 SD with the number of observations in parentheses. Control subjects represent a total of four different cell lines, three of which were assayed at 24 to 26 population doublings a n d one at 10 population doublings.
techniques. 1~ Free Hyl and other urinary amino acids were quantitated in this manner. Hyl peaks were confirmed by their disappearance after periodate treatment. Total urinary Hyl and Hyp were quantitated after acid hydrolysis for 18 hours at 110~ Urine was analyzed for Hyl glycosides after hydrolysis i n 2 N potassium hydroxide by previously described techniques.16 4-Hyp was separated from 3-Hyp on a Jeolco polystyrene resin using citrate buffer at p H 2.5. Skin amino acids were quantitated on 4 m m lyophilized skin biopsy specimens using previously described techniquesY Ascorbic acid was measured in urine and serum after extraction with metaphosphoric acid as 2,4-dinitrophenylhydrozone derivative. TM Preparation of labeled unhydroxylated substrate and enzyme. Labeled unhydroxylated collagen substrate was prepared from 15-day-old chick embryo calvaria by a modification of previously described techniques? '~ Essentially 5-[4,5-:~H] lysine or proline was incubated with calvaria in lysine- or proline-free media, in the presence of a,a' -dipyridyl. Tissues were then homogenized, and the supernate was dialyzed, and the collagen fracfionated, pelleted, and stored frozen. Skin biopsy material was obtained by punch biopsy and grown in monolayer cultures using Dulbecco and Vogt's modified Eagle's media supplemented with 15% fetal calf serum and glutamine using previously described methods? ~ Enzyme was obtained from cells by harvesting with trypsin, freeze-thawing in Tris-HC1 buffer, and centrifuging at 15,000 x g for 10 minutes. Supernate contained the enzyme and was dialyzed for 24 hours and frozen at -20~ Activity of dialyzed enzyme was stable for four weeks at --20 ~ Protein content was determined by a modified method of Lowry et alY Enzyme assay. Lysyl hydroxylase activity was deter-
380
Elsas, Miller, and Pinnell
The Journal of Pediatrics March 1978 Controls
Patient (age 8)
Patient (age I0)
4
1 2 I
'
50
/
/7'
'
-'
'
'
0
8yr
E
z~ uJ (,9 25
~g
o E
~
9yr 12yr 34 yr 23yr
0
I
2
34
56
4
5678
DAYS Fig. 1. Effect of ascorbate on urinary free hydroxylysine excreted by control subjects and Patient J. D. H. Control values are represented by open circles before and after ascorbate and followed by the subject's age. Where two consecutive samples were analyzed the range of values is indicated by bars for both control subjects and the patient. Dosage of ascorbate is represented by stippled bars above the day administered. Free hydroxylysine was quantitated by ion exchange chromatography of lyophilized urine as described in the Methods section. mined by the tritium exchange method of Peterkofsky and DiBlasio TM with minor modifications. L-[4,5-3H] lysyl collagen substrate was incubated with fibroblast enzyme preparation at 30~ in a reaction mixture which contained 50 mM Tris-HC1 (pH 7.5), 0.5 mM a-ketoglutarate, 0.0 to l.0 mM L-ascorbate, 0.1 mg catalase, 0.05 mM Fe(NH4)~SO4, 0.75 mg bovine serum albumin, and 0.1 mM dithiothreitol. The assay was stopped by the addition of 50% trichloroacetic acid. The supernate was then applied to A G Dowex 50W-X8 columns and the eluted tritium counted by liquid scintillation spectrometry. Prolyl hydroxylase activity was measured by the same procedure except that the reaction mixture contained L[3,4-3H] prolyl collagen substrate and 0.1 mM a-ketoglutarate. In both assays enzyme activity was defined as tritium release in the presence as compared to the absence of a-ketoglutarate. Materials. L-[4,5-3H] lysine (21 Ci/mmole) and L-[3,4~H] proline (41 Ci/mmole) were purchased from Amersham/Searle Corp.; a-ketogutarate, L-ascorbic acid, and a,a' -dipyridyl were obtained from Sigma Chemical Co. Catalase was purchased from Worthington Biochemical Corp., Pentex bovine albumin fraction V, fatty acid poor, from Miles Laboratories, Inc., and dithiothreitol from Calbiochem. Dowex 50W-X8 was purchased from Bio Rad Laboratories. RESULTS Skin from the patient had 0.5 residues of Hyl per 1,000 amino acids whereas normal skin contains 5.1 _+ 0.7
residues per 1,000. The ratio of Hyl to Hyp was 0.011 as compared to normal values of 0.051 + 0.0094 Since skin collagen Hyl content is determined by collagen lysyl hydroxylase, the specificity and genetic control of this enzyme were determined using cultured skin fibroblasts from the proband, his parents, and control cell lines. Collagen lysyl hydroxylase activity in the patient's cultured skin fibroblasts was approximately 17% of control values from four different lines; his father's and mother's cell lines had 66% and 39% of control activity, respectively (Table I). By contrast, the patient's cell line had prolyl hydroxylase activity which was statistically indistinguishable from control lines. This collagen lysyl hydroxylase assay was evaluated to enable cofactor assessment. A pH optimum of 7.5 was determined at 30~ and activity was linear throughout 20 minutes of incubation. A direct relationship was seen between tritium release and protein content of the subcellular enzyme preparation between 0 and 75 /~g. At all protein concentrations, activity from the mutant cell line was markedly reduced. Following these initial experiments, prolonged dialysis of control enzyme preparations was undertaken to assure removal of aqueous-soluble cofactors such as ascorbate. When L-ascorbate alone was removed from the reaction mixture, only 4% of,complete normal" activity was present. Because of the observed role of ascorbate in collagen lysyl hydroxylase function and because of the hemorrhagic quality of the scar tissue, the patient was first evaluated for potential ascorbate deficiency. Dietary
Volume 92 Number 3 history indicated that he had been ingesting at least 50 mg of ascorbate per day over the past year. Twenty-four-hour urine and plasma ascorbate were compared to three normaL, age-matched control subjects. The patient's urinary and plasma levels of ascorbate before loading were within the normal range of 8 to 37 m g / d a y / g m of creatinine and 0.2 to 2 mg/dl, respectively. After a single oral dose Of 4 gm of ascorbate the patient excreted 786 mg/day/gm of creatinine in urine as compared to three control subjects who excreted between 285 and 1,155. Plasma levels of ascorbate after loading rose similarly to those in control subjects and reached levels between 2.0 and 4.0 mg/dl plasma. Thus ascorbate deficiency was not present. An in vivo response to pharmacologic doses of ascorbate was then investigated by measuring the urinary excretion of both total and free Hyl. As seen in Fig. 1, control subjects' 24-hour urinary free Hyl excretion fell with increasing age. After treatment with 4 gm of ascorbate daily for four days, Hyl excretion rose significantly in all control subjects (P < 0.01 using paired t test). At age 8, the patient's Hyl excretion was reduced compared to his age-matched control subject. With increasing doses of ascorbate his free Hyl excretion rose to 48 then stabilized at 25/zmoles/day/gm Of creatinine. As seen in Table II, total urinary Hyl values also rose, although not as dramatically. Although slight elevations in Hyp excretion were seen, they were not as great as those in Hyl excretion. Thus the ratio of Hyl to Hyp increased (Table II). These values for total Hyl were somewhat lower than those in the age-matched control subject, which was 196/zmoles/ day/gm of creatinine and did not change on ascorbate treatment. Urinary Hyp levels in the proband were the same as in control subjects; the Hyl to Hyp ratio in the control subjects was between 0.13 and 0.14 whereas the patient's ranged between 0.08 to 0.12. Urinary glycosides of Hyl were measured and accounted for between 66 and 85% of the total Hyl content. There was no consistent change in glucosylated Hyl or the ratio of glucosylgalactosylhydroxylysine to galactosylhydroxylysine with administration of increasing amounts of ascorbate. Since kidney basement membrane collagen is rich in Hyl and has relatively more 3-Hyp than 4-Hyp, the ratio of urinary 3-Hyp to 4-Hyp was measured. 23 The ratio in both the patient and his age-matched control subject was 0.056 before and 0.047 after ascorbate, suggesting no significant contribution by kidney basement membrane to increased Hyl excretion. The patient's urinary excretion of total and free Hyl was monitored while he was receiving ascorbate over the next two years. During this period, plasma ascorbate ranged between 2.0 and 4.0 mg/dl. At 10 years of age the patient's
Inherited hydroxylase deficiency
381
Fig. 2. Scars from skin biopsies over left buttock. Three "preascorbate" scars (arrows pointing right) were obtained in March and June, 1974. Three "postascorbate" scars (arrows pointing left) were obtained in September, 1974, and March, 1975, three and eight months after initiation of vitamin C. The photograph was taken in February, 1976. Table II. Effect of ascorbate on the urinary excretion of total hydroxylysine and 4-hydroxyproline by proband (age 8 years)
Ascorbate (gm/day) 0 0 0 1 2 3 4 4 4
Urinary amino acid* (~moles/24 hr/gm creatine) Hyl
Hyp
Ratio of Hyl/Hyp
112 116 91 113 129 132 158 163 137
1,196 1,199 1,197 1,049 1,164 1,068 1,637 1,426 1,241
0.09 0.10 0.08 0.11 0.11 0.12 0.10 0.11 0.11
*Aliquots of 24-hour urine samples were concentrated 21/2-fold,then hydrolyzedat II0~ for 18 hours in 6 N HC1. Data are single observations. maximum free Hyl excretion was approximately 13 /~moles/day/gm of creatinine (Fig. 1). Ascorbate was discontinued and urinary free Hyl was measured from four to eight days later. There was an immediate and dramatic fall in Hyl excretion to less than 3/~moles/day/ gm of creatinine, considerably lower than adult control subjects (Fig. 1). Thus the increase and fall in urinary Hyl was associated with oral administration of vitamin C. Ascorbate therapy produced no observable change in skin Hyl content, but improved the quality of scar
382
Elsas, Miller, and Pinnell
The Journal of Pediatrics March 1978
Table III. Relationship of corneal size, age, and ascorbate therapy
Patient~ age 8.5 9.0 10.0 Normal males 2~ (age 5-9)* Mean Minimum Maximum Normal males (age 10-19)t Mean Minimum Maximum
Duration of ascorbate therapy (mo) 0 8 20 0
Corneal diameter (ram) OD
[
10.2 10.3 10.5
OS 10.4 10.5 10.7
11.7 11.0 12.5 11.7 11.0 12.5
*Data from 54 corneasin 27 males. tData from 124 corneasin 62 males.
formation. Biopsies were obtained before and three months and eight months after initiation of vitamin C therapy. Hyl content was 0.5, 0.5 and 0.9 residues/I,000 amino acids in each of these samples. However, the character of biopsy scars improved dramatically (Fig. 2). Three scars produced by biopsies obtained in March and July, 1974, before ascorbate therapy was begun, were red, raised, and thin, in contrast to three biopsy scars similarly produced three months and eight months after ascorbate treatment was begun. On histologic examination, preascorbate scars contained dilated vessels with extravasated old blood, whereas postascorbate scars looked normal. Scars formed while the patient was on ascorbate therapy were not large enough to biopsy adequately for Hyl determination. In addition to improved wound healing, several other observations suggested clinical improvement. Corneal diameter increased 0.3 mm in each eye over the 20 months of ascorbate therapy (Table III). Corneal diameter normally does not increase after four years of age. 24 After ascorbate therapy, the patient's corneal diameter approached but did not reach the lower limit of his ageand sex-matched control subjects. 2' Muscle strength improved, as assessed by direct muscle testing. The increased residual volume noted on previous pulmonary function tests was normal one year after therapy. Bleeding time returned to normal. Joint laxity and friability of skin, however, remained. DISCUSSION As in most described inherited metabolic disorders, a large hiatus exists between the observed biochemical
impairment and the mechanisms producing the expressed phenotypeY 5 The patient described herein had muscle weakness, lax joints, and hemorrhagic, friable skin. His skin collagen contained only 10% of normal Hyl. Collagen lysyl hydroxylase was impaired in his cultured skin fibroblasts and an autosomal recessive mode of inheritance for this enzyme defect was observed. This constellation of phenotypic changes may be related to the biochemical alterations through a reduction of number, type, and distribution of cross-linking in dermal, tendinous, and vascular collagen. 6, ~, ~6 By contrast to previously reported phenotypes, in our patient severe vertebral and ophthalmic abnormalities were not evident. One possible explanation for these organ differences is that different isozymes of lysyl hydroxylase exist in different tissues. The expression of a mutation by a specific organ would depend, therefore, on the relative importance of a given subunit to that organ's function. For instance, Type III collagen [l(III)]:, which predominates in the collagen of fetal tissues and blood vessels, may require a different lysyl hydroxylase than Type I [1(i)]3 collagen for its production. No data on this point are available. Another possibility is that different tissues have different cofactor concentrations, particularly of Lascorbate, which we found to be the principle physiologic reductant for lysine hydroxylation. Pinnell et al 6 espoused this hypothesis to explain the variable degree of Hyl deficiency in collagen from different tissues of the same patient. Variations in tissue content of ascorbate may affect the expression of different mutations, depending on the molecular site of enzyme impairment and the relationship Of this site to L-ascorbate action. Pharmacologic amounts of ascorbate were associated with improvements in muscle strength, corneal growth, bleeding time, an d pulmonary residual volume. In addition, Urinary hydroxylysine excretion increased on administration of ascorbate and decreased on withdrawal. These in vivo observations suggested that 4 g m / d a y of ascorbate increased hydroxylation of lysyl residues in newly synthesized collagen and that its production and degradation were measurable in urine. Askenasi 2~-3~demonstrated that the amount of free and peptide-bound urinary Hyl was a physiologic index o f collagen metabolism. He also provided evidence that renal basement membrane did not contribute significantly to the urinary Hyl pool. 3i Since the Clq component of hemolytic complement represents only a ,fraction of total body Hyl, its contribution to urinary Hyl must be negligible. If this regimen of vitamin C were augmenting mutant lysyl hydroxylase, why then did skin Hyl content not increase? One explanation is that L-ascorbate enhanced collagen lysyl hydroxylase activity only where new collagen was produced, as in new scar formation or
Volume 92 Number 3 corneal growth. Gould 32 postulated in scorbutic models that there was more than a single mechanism for collagen formation: one used for normal body collagen formation and one predominating in tissue repair. He suggested that the former was independent o f ascorbate and the latter was dependent on its presence. Barnes et a133, 34 implied that in cultured L-929 cells, hydroxylation of lysine in the nonhelical portion of the developing collagen molecule was under control by a separate ascorbate-dependent lysyl hydroxylase distinct from a more active lysyl hydroxylase. Hornig 3~ found that 14C-ascorbate localized to the ground substance of the intima of guinea pig blood vessels and suggested that n o n r a n d o m distribution of ascorbate indicated differences in its utilization by various tissues. Miller 36 further demonstrated that ascorbate did not activate preformed, inactive lysyl hydroxylase in nonscorbutic cultured L-929 cells as was seen for prolyl hydroxylase. In dermal skin collagen only five of approximately 30 lysine residues are normally hydroxylated on a 1(I) and a2 chainsY In our patient only 0.5 were hydroxylated. By contrast, corneal collagen m a y have as much as 67% hydroxylation of its a l ( I V ) chain lysyl residues. 23 This unusual degree of hydroxylation in corneal collagen may account for the unique growth of this tissue during administration of ascorbate to our patient. Thus, the improved wound healing and corneal growth expressed by our patient in the absence of increased Hyl content of preformed skin collagen might be an expected result from the administration of large amounts of vitamin C in vivo, if only specialized lysyl hydroxylase, actively forming new collagen in specific organs, was augmented. The mechanisms of action of ascorbate on our mutant enzyme and the variable tissue responses require further clarification. However, the observed in vivo responses of this single gene mutation and its expression in cultured fibroblasts provides a model for further investigation. We thank Cheryl Jorgensen Vroman for typing the manuscript, and Drs. Robert Priest, Anthony Keyser, and Dean J. Danner for their support and constructive criticisms. We owe special gratitude to Dr. John Reiser for his ophthalmologic measurements, to Dr. Jean Priest for her advice and supervision regarding human fibroblast tissue culture, and to Ms. Bettye Hollins for her technical assistance.
REFERENCES
1. Grant ME, and Prockop D J: The biosynthesis of collagen, N Engl J Med 286:194, 242, 291, 1972. 2. Nimni ME: Metabolic pathways and control mechanisms involved in the biosynthesis and turnover of collagen in normal and pathological connective tissues, J Oral Pathol 2:175, 1973. 3. McKusick VA, and Martin GR: Molecular defects in collagen, Ann Intern Med 82:585, 1975.
Inherited hydroxylase deficiency
3 83
4. Pope FM, Martin GR, Lichtenstein JR, Penttinen R, Gerson B, Rowe DW, and McKusick VA: Patients with EhlersDanlos syndrome Type IV lack Type II1 collagen, Proc Natl Acad Sci USA 72:1314, 1975. 5. Di Ferrante N, Leachman RD, Angelini P, Donnelly PV, Francis G, Almazan A, and Segni G: Lysyl oxidase deficiency in Ehlers-Danlos syndrome type V, Connect Tissue Res 3:49, 1975. 6. PinneU SR, Krane SM, Kenzora JE, and Glimcher MJ: A heritable disorder of connective tissue. Hydroxylysinedefcient collagen disease, N Engl J Med 286:1013, 1972. 7. Krane SM, Pinnell SR, and Erbe RW: Lysyl protocollagen hydroxylase deficiency in fibroblasts from siblings with hydroxylysine-deficient collagen, Proc Nail Acad Sci USA 69:2899, 1972. 8. Elsas LJ, Hollins B, and Pinnell SR: Hydroxylysine deficient collagen disease: Effect of ascorbic acid, Am J Hum Genet 26:28A, 1974. 9. Sussman M, Lichtenstein JR, Nigra TP, Martin GR, and McKusick VA: Hydroxylysine-deficient skin collagen in a patient with a form of the Ehlers-Danlos syndrome, J Bone Joint Surg 56A:1228, 1974. 10. Steinmann B, Gitzelmann R, Vogel A, Grant ME, Harwood R, and Sear CHJ: Ehlers-Danlos syndrome in two siblings with deficient lysylhydroxylase activity in cultured skin fibroblasts but only mild hydroxylysine deficit in skin, Helv Paediatr Acta 30:255, 1975. 11. Lichtenstein JR, Martin GR, Kohn LD, Byers PH, and McKusick VA: Defect in conversion of procollagen to collagen in a form of Ehlers-Danlos syndrome, Science 182:298, 1973. 12. Quinn RS, and Krane SM: Abnormal properties of collagen lysyl hydroxylase from skin fibroblasts of siblings with hydroxylysine-deficient collagen, J Clin Invest 57:83, 1975. 13. Scriver CR: Progress in endocrinology and metabolism: Vitamin-responsive inborn errors of metabolism, Metabolism 22:1319, 1973. 14. Nishikimi M, and Udenfriend S: Scurvy as an inborn error of ascorbic acid biosynthesis, Trends Biochem Sci 2:111, 1977. 15. Rosenberg LE, Durant JL, and Elsas LJ: Familial iminoglycinuria. An inborn error of renal tubular transport, N Engl J Med 2"/8:1407, 1968. 16. Pinnell SR, Rox R, and Krane SM: Human collagens: Differences in glycosylated hydroxylysines in skin and bone, Biochim Biophys Acta 229:119, 1971. 17. Miller EJ, and Piez KA: An accelerated single-column procedure for the automatic analysis of amino acids in collagen and elasfin hydrolyzates, Anal Biochem 16:320, 1966. 18. Roe JH: Appraisal of methods for the determination of Lascorbic acid, Ann NY Acad Sci 92:277, 1961. 19. Peterkofsky B, and DiBlasio R: Modification of the tritiumrelease assays for prolyl and lysyl hydroxylases using Dowex-50 columns, Anal Biochem 66:279, 1975. 20. Elsas LJ, Priest JH, Wheeler FB, Danner DJ, and Pask BA: Maple syrup urine disease: Coenzyme function and prenatal monitoring, Metabolism 23:569, 1974. 21. Priest JHi Medical cytogenetics and cell culture, ed 2, Philadelphia, 1977, Lea & Febiger, Publishers, pp 263295. 22. Lowry OH, Rosebrough N J, Farr AL, and Randall R J:
3 84
23. 24.
25.
26. 27.
28.
29.
30.
Elsas, Miller, and Pinnell
Protein measurement with the Folin phenol reagent, J Biol Chem 193:265, 1957. Kefalides NA: Structure and biosynthesis of basement membranes, Int Rev Connect Tissue Res 6:63, 1973. Smith P: On the size of the cornea in relation to age, sex, refraction, and primary glaucoma, Trans Ophthal Soc UK 10:68, ]890. Gendel BR, and Elsas L J: Medical genetics, in Sodeman WA, and Sodeman WA Jr, editors: Pathologic physiology. Mechanism of disease, ed 5, Philadelphia, 1974, WB Saunders Company, chap 3. Eyre DR, and Glimcher M J: Reducible crosslinks in connective tissue, Proc Natt Acad Sci USA 69:2594, 1972. Askenasi R: A new rapid method for measuring hydroxylysine and fis glycosides in hydrolysates and physiological fluids, Biochim Biophys Acta 304:375, 1973. Askenasi R: urinary hydroxylysine and hydroxylysyl glycoside excretions in normal and pathologic states, J Lab Clin Med 83:673, 1974. Askenasi R, and Demeester-Mirkine N: Urinary excretion of hydroxylysyl glycosides and thyroid function, J Clin Endocrinol Metab 40:342, 1975. Askenasi R: Urinary excretion of free hydroxylysine, peptide-bound hydroxylysine and hydroxylysyl glycosides in physiological conditions, Clin Chim Acta 59:87, 1975.
The Journal of Pediatrics March 1978
31. Askenasi R, Van Herweghem JL, and Ducobu J: Contribution of kidney basement membranes and collagen to the urinary excretion of hydroxylysyl glycosides, Biomedicine 22:233, 1975. 32. Gould BS: The role of certain vitamins in collagen formation, in Gould BS, editor: Treatise on collagen, New York, 1968, Academic Press, Inc, vol 2A, p 323. 33. Barnes M J, Constable BJ, Morton LF, and Royce PM: Agerelated variations in hydroxylation of lysine and proline in collagen, Biochem J 139:461, 1974. 34. Barnes MJ, Constable BJ, Morton LF, and Kodicek E: Studies on the hydroxylation of lysine occurring in the nonhelical region at the N-terminal end of the collagen molecule, Biochem J 125:16, 1971. 35. Hornig D: Distribution of aseorbic acid, metabolites and analogues in man and animals, in King CG, and Burns J J, editors: Second Conference on Vitamin C, The New York Academy of Sciences, New York, Ann NY Acad Sci 258:103, 1975. 36. Miller RL: The effect of ascorbic acid on lysyl and prolyl hydroxylase activity of cultured fibroblasts, Arch Biochem Biophys 170:341, 1975. 37. Bornstein P, and Piez KA: A biochemical study of human skin collagen and the relation between intra- and intermolecular cross-linking, J Ctin Invest 43:1813, 1964.
