- Email: [email protected]
Reduced serum levels of car☐y-terminal propeptide of human type I procollagen in a family with type I-A osteogenesis imperfecta
Reduced Serum Levels of Carboxy-Terminal Propeptide of Human Type I Procollagen in a Family With Type I-A Osteogenesis Imperfecta Salvatore
Minisola,
Anna Lina Piccioni, Rossana Rosso, Elisabetta Romagnoli, Liliana Scarnecchia, and Gianfranco Mazzuoli
Maria Teresa Pacitti,
We measured serum levels of total alkaline phosphatase activity, osteocalcin, carboxy-terminal propeptide of human type I procollagen (PICP), tartrate-resistant acid phosphatase activity (TRAP), and the fasting urinary hydroxyproline/creatinine ratio (OHPr/Cr) in seven affected members (four men, three women; age, 43.3 f 16.6 years [mean f SD]) of a family with clinically diagnosed type I-A osteogenesis imperfecta (01) and in eight (five men, three women) normal age-matched (38.2 f 10.3) relatives. Three boys with 01 and three normal girls of the same family were also studied, although they were excluded from statistical analysis. Bone mineral density was also determined at four different skeletal sites. Serum levels of PICP were measured with a radioimmunoassay (Farmos Diagnostica, Turku, Finland). There were no significant differences in mean values of the biomarkers studied between 01 patients and normal relatives, with the only exception being serum levels of PICP (35 -e 7.5 v 219 f 107.5 pg/L, P c .OOl). A significant reduction of BMD was found in 01 patients compared with normal relatives at the lumbar (L) spine (680 f 61 v 1,128 f 92 mg/cm*, P < .OOl), at the ultradistal radius ([UDR] 323 f 85 v 458 f 76, P < .006), at the femoral neck ([F] 494 f 140 v 791 f 104, P < .OOl), and at the junction of the distal and middle third of the radius ([MR] 639 f 71 v 717 -C 52, P c .029). When the patients and control subjects were combined, there were significant positive correlations between serum PICP and BMD values at various skeletal sites (L, P < .006; F, P < .05; UDR, P < .005; MR. P < .007). Our results suggest that decreased levels of serum PICP are typical of 01 type I patients, or at least of this family. This should be ascribed to a decreased amount of collagen produced, although the possibility of an abnormal sequence not recognized by the antiserum used should also be considered. Densitometric results indicate that quantitative or qualitative defects of collagen structure may contribute to the fragility of 01 bone by interfering with complete mineralization and/or normal tissue structure. Copyright 0 1994 by W.B. Saunders Company
0
IMPERFECTA (01) is the most STEOGENESIS common heritable disorder of bone. It is a heterogeneous disease characterized by bone fragility, osteopenia, progressive skeletal deformities, and often severe growth retardation. Recent biochemical, linkage, and molecular genetic studies have shown that in almost every instance 01 results from mutations in the genes that encode the chains of type I collagen. Such studies have improved our understanding of the molecular basis of this disorder, and have provided significant advances in molecular diagnosis and prognostic counseling.‘-3 However, despite these impressive breakthroughs in basic sciences, there are no detailed clinical studies that attempt to correlate biochemical and densitometric data
From the Institute of II Clinica Medica, Cattedra di Medicina Intema, Universitcidegli Studi di Roma “La Sapienza, ” Roma, Italy. Submitted July 26, 1993; accepted January 2, 1994. Supported by a grant from the Consiglio Nazionale delle Ricerche, “Progetto Finalizzato Invecchiamento, “code 96-3-407. Presented in pal? at the Fourteenth Annual Meeting of the American Sociev for Bone and Mineral Research, Minneapolis, MN, September 30 to October 4, 1992, and at the Fowih International Symposium on Osteoporosis and Consensus Development Conference, Hong Kong, March 27 to April 2, 1993. Address reprint requests to Salvatore Minisola. MD, II Clinica Medica, Policlinico Umberto I, Viale de1 Policlinico 155. 00161 Rome, Italy. Copyright 0 1994 by W B. Saunders Company 0026-0495i94l4310-0010$03.00l0
noqf-lmILL-A 12
3
1' Fig 1. ily.
Pedigree for the 01 fam-
Metabolism, Vol43,No
KEY:
lO(October),1994:pp 1261-1265
4
5
l=Q
6
affected n = d affected
7
a
9
0 = Q
unaffected
q=d unaffected
10 11
/=
12 13
14 15
deceased
#=proband
1261
MINISOLA
1262
Table 1. Anthropometric,
35.
Phenotypic, and Clinical Features and
I
Serum PICP Values in the Family With Type I-A 01 Patient 1
Age (~4
yrsl
2
3
4
5
6
7
8
9
10
12
15
20
29
35
42
53
57
67
5
Height (cm)
110
160
158
172
168
158
155
165
150
160
Weight (kg)
25
55
43
65
86
47
51
64
63
45
4
7
13
8
8
4
5
4
4
4
Fractures (no.) Ligamentous laxity Scoliosis Hearing loss Easy bruisability
+ -
vs 07
No.
-I \ 2
MMMMMFFMFM
sex
ET AL
+ + -
+ f -
+ + -
+ +
+ +
+ -
+ -
+
+
-
+
+
-
-
_
_
-
+_---_
48
40
+ + -
+ + +
E 2
+
+
8
l 8
30
15.
Excessive diapho+
resis
10
Serum PICP (l&L)
128
75
100
27
26
34
35
a specific phenotype. The need for these has become more compelling, considering improvements in the methodology used to evaluate bone turnover and measure bone mineral density (BMD). We recently had the opportunity to study a very large kindred with 01 type I-A. Results obtained indicate that the in vivo measurement of procollagen molecules may be used to identify this clinical type of brittle bone disease. SUBJECTS AND METHODS
5
1
0 5 ’
300.
-
14.0
2 \
12.0
F 1
10.0
;
8.0
N
from statistical analysis; a comparison was not possible owing to the different sex, age, and pubertal status. Each patient and subject willing to participate in the study underwent an initial comprehensive health survey, a physical examination, and a multichannel autoanalyzer chemical screening. Metabolic study included a 24-hour urine collection to evaluate calcium, phosphorus, and creatinine (Cr), followed by a short-term urine collection after a 12-hour overnight fast to determine the calcium and hydroxyproline (OHPr)/Cr ratio and the renal tubular threshold of phosphate excretion. Halfway through the latter collection, a blood sample was taken for the measurement of serum levels of calcium, phosphorus, Cr. intact parathyroid hormone, osteocalcin, total alkaline phosphatase activity, tartrate-resistant
We studied seven affected members (four men, three women; age, 43.3 2 16.6 years [mean + SD]) of a family clinically diagnosed with type I-A 01 on the basis of the classification of Silence et al.4,5 Biochemical and densitometric results were compared with those of eight (five men, three women) normal age-matched (38.2 5 10.3) relatives. Three boys with 01 and three normal girls from the same family were also studied. but they were excluded yrsl
01
Fig 2. Individual values of two independent markers of bone resorption found in patients with 01 and their age-matched normal relatives (N). The values of six children studied (indicated by their respective ages) have been included, even though they were excluded from statistical analysis. Horizontal bars indicate the means.
with
400
N
01
35
yn
.a
700 600 _1 500 \ 400 0) 300 =L -
n I-
g
200.
a
I
a
vsi’
6.0
E 3 ii (/)
& 200
5 $ 4.0 v)
loo-
O-
01
N
E 2 100 Q, (I)
0
0 01
5
I
: 0
N
1:
Fig 3. Individual values of three independent markers of bone formation found in patients with 01 and their age-matched normal relatives (N). See Fig 2 legend.
REDUCED
SERUM
PICP IN TYPE
1263
I-A 01
acid phosphatase activity (TRAP), and carboxy-terminal propeptide of human type I procollagen (PICP). For the purposes of this study, biochemical markers of bone turnover were determined according to previously described methodsh.’ In particular, serum osteocalcin levels were measured using a radioimmunoassay kit (Incstar, Stillwater, MN) based on the method of Price and Nishimot0.s The limit of detectability was 0.3 ng/mL. The intraassay variation was 2%, and the interassay variation was 6.5%. Normal values in our laboratory are 3.31 ? 1.04 ng/mL. Urinary OHPr was assayed by a calorimetric method9; normal values in our laboratory for the fasting OHPriCr ratio are 30 ? 10 mg/g. Serum total alkaline phosphatase activity levels (normal values. 75 ? 18 U/L) were measured using a calorimetric method by autoanalyzer (Technicon Autoanalyzer RA 1000, Tarrytown, NY). Serum TRAP levels were measured according to the method of Lau et al.‘? normal values in our laboratory are 13 ? 2.4 U/L. Serum concentrations of PICP were measured in duplicate lOO-yL serum samples with a recently developed equilibrium radioimmunoassay supplied by Farmos Diagnostica (Turku, Finland).” The sensitivity of the test was 1.2 pg/L. The intraassay coefficient of variation was 2.8% in the concentration range of our study. The corresponding interassay variation was approximately 5.5%. The reference interval (mean ? 2 SD) for 48 normal subjects (42 to 73 years of age) was 80 to 210 kg/L. BMD was measured at the lumbar spine (L) and at the femoral neck (F) by quantitative digital radiography using a Hologic QDR-1000 x-ray bone densitometer (Hologic, Waltham, MA). Furthermore, BMD was also assessed at the junction of the distal and middle third of the radius (proximal site [MR]) and at the ultradisl.al radius ([UDR] one tenth of the forearm length, 4 2.5 cm from the styloid process of the radius) of the nondominant arm by dual-photon densitometer (NIM, Verona. Italy) as previously described.”
Statistical Analvsis Unless otherwise specified, all results are expressed as the mean ?: 1 SD. Significance between means was assessed using Student’s 1 test for unpaired data. Correlations between variables were tested by linear regression analysis. The significance level was set at P less than .05 for all tests. RESULTS The pedigree for this family is presented in Fig 1. The proband (IV-3, see arrow) was a E-year-old boy who had already had 13 fractures in his life. The main anthropometric, phenotypic, and clinical features of the 10 patients with 01 studied here are reported in Table 1. By definition, all were characterized by blue sclerae and the absence of dentinogenesis imperfecta; as for the number of fractures, its apparent low number in the older patients is most probably due to the difficulty elderly patients have in recalling exactly how many fractures they have had. The ascertainment of the number of fractures is easier for those who are presently young because they are carefully checked by their parents. Figure 2 shows individual values of two independent markers of bone resorption found in patients with 01 and in their age-matched normal relatives, respectively. No significant differences between the two groups were detected in mean values (serum TRAP, 14.6 2 2.3 v 15.8 4 2.3 U/L; fasting urinary OHPr/Cr, 29.7 + 9.1 v 24.6 + 5.6 mgig). Figure 3 represents the individual values of three indepen-
dent markers of bone formation. Although there were no significant differences between the two groups in terms of the mean values of serum total alkaline phosphatase activity (94.4 ? 25.3 v 78.6 +- 19.6 U/L) and serum osteocaltin levels (3.85 ? 1.1 v 2.93 + 0.9 ng/mL), a significant difference was found in the mean values of serum PICP (35 rt 7.5 v 219 2 107.5 kg/L, P < ,001). A significant reduction of BMD was found in patients with 01 compared with normal relatives at the L (680 f 61 II 1,128 f 92 mg/cm*, P < .OOl; Fig 4), at the UDR (323 2 85 1’ 458 % 76, P < .006), at the F (494 2 140 v 791 2 104, P < .OOl). and at the MR (639 + 71 v 717 + 52, P < .029; Fig 5). Finally, when the patients and control subjects were
pc
0.001
1300
I
0
i
l
I
0
13
8
i
1200 1100 1000
w
I
900 800 700 600 500 400
Aa/ 8 0
vsi 015; 0
121 I
0
5; I
01
N
Fig 4. Individual values of L-WAD found in patients with 01 and their age-matched normal relatives (N). Sea Fig 2 legend.
MINISOLA
800,
600
0.029. i l ; 0 I
pc
p<
; 1 /
500 -&- 700. E < F
p
I
l l
900
l
"E
I ,
1
l l
-
"E800. < F700.
s400 L
600.
loo0
0.006
ET AL
0
600'
i 300 ; m &
I
0
500.
si
8 i
E 3
0
I,
l 7
m k 500.
200
wl
400.
012;
400200-
0
01
100
5; N
01
N
combined, there were significant positive correlations between serum PICP and BMD values at the various skeletal sites measured (L, r = ,677, P < .006; F, r = S14, P < .05; UDR, r = .681, P < .005; MR, r = .662, P < ,007). Conversely, these were not found when considering other biomarkers. DISCUSSION
A number of in vitro studies have recently attempted to define the molecular basis of 01 type I more clearly. It has become evident that cells from most affected individuals with 01 type I synthesize and secrete approximately half the normal amount of type I procollagen.13.14 The decrease in production of the type I procollagen molecule results from a decreased synthesis of procul(1) chains, usually a reflection of the decreased amount of mRNA for that chain.15 Our in vivo results, which show that the affected members of the family we studied are characterized by reduced mean serum levels of PICP (with respect to their agematched normal relatives), might be considered the perfect clinical counterpart to in vitro findings. It is in fact conceivable that this serum decrease is a reflection of a generalized decrease of type I collagen synthesis; the phenotypic features and clinical findings observed in the affected patients should be ascribed to it. It is known that an altered sequence in the carboxyterminal propeptide of the procxl(1) chain, as well as mutations altering the structure of the type I procollagen by virtue of single substitutions in the triple or nontriple helical domain, result in 01 type I. Studies are now in progress to characterize the molecular defect of our family. However, even these last two defects should give rise to decreased serum PICP levels; this is because the amount of the molecule assembled will be half the normal in the first
Fig 5. Individual values of MRBMD, UDR-BMD, and F-BMD in patients with 01 and their agematched normal relatives (N). See Fig 2 legend. For technical reasons, some measurements were not performed in the young
subjects. case, or owing to the abnormal sequence, not recognized by the antiserum used in the second hypothesis. In this context, a particular point of consideration derives from the finding of very low mean serum PICP levels versus what would be expected if only one allele was functioning. The circulating level of the intact propeptide presumably reflects the balance between the rate of its synthesis and that of its incorporation into the matrix, in addition to degradation. A possible explanation should therefore be sought on a different half-life of the propeptide or on a different efficiency at which synthesized type I procollagen is incorporated into the matrix when one allele versus two functional alleles are available. Although 01 is traditionally considered a cause of osteoporosis,lJ” few data have been presented so far concerning BMD measurement in this disorder.’ Furthermore, published results usually deal with a single measurement site17; there is no extensive survey of various skeletal sites reflecting a different percentage composition of trabecular and cortical bone. Our data strongly indicate that reduced BMD is a generalized feature of the patients we investigated with 01 type I, whatever the measurement site chosen. The relevance of this finding is even greater if we consider that comparisons were made with normal relatives living in the same area. This indeed excludes the possibility that different life-styles or eating habits could have been responsible for the osteopenia. Having shown that affected members of 01 type I have not only reduced BMD values but also serum PICP levels, could we hypothesize a relationship between them? The finding of a direct correlation between serum PICP and BMD values strongly indicates, in our opinion, that quantitative (or qualitative) defects of collagen structure contrib-
1265
REDUCED SERUM PICP IN TYPE I-A 01
ute to the fragility of 01 bone by interfering with complete mineralization and/or normal tissue structure.i8 This hypothesis is further supported by the lack of significant correlations between conventional markers of bone turnover and BMD values and by the absence of significant differences in mean values of bioindices of skeletal remodeling in the two groups examined. This in fact suggests that the possible imbalance between the two remodeling processes has not been the cause of fractures in our patients. From a practical point of view, there are four main implications of our finding of reduced serum PICP in 01 type I patients: (1) this measurement may be used in making a diagnosis of 01 type I, although further studies are needed to examine serum PICP levels in diseases of structural mutations of collagen; (2) in the case of children with fractures, the measurement of this marker could be
used to differentiate between real cases of patients with mild 01 and abused childreni9; (3) this marker might be particularly useful in examining the response of drugs such as anabolic steroids, which are known to affect the rate of collagen synthesis in bone and other connective tissues; (4) finally, the measurement of serum levels of PICP might be used as a screening tool to unmask patients with idiopathic osteoporosis who present with a family history of osteoporosis, fractures with variable degrees of joint laxity, and mild scoliosis.zO-ZZ ACKNOWLEDGMENT
We are sincerely indebted to Dr David W. Rowe for his advice and critical review of the manuscript. We wish to thank Walter Rossi for technical assistance, and we are deeply grateful to Lee O’Hara for editing the manuscript.
REFERENCES
1. Rowe DW: Osteogenesis imperfecta, in Heersche JNM, Kanis JA (eds): Bone and Mineral Research, vol 7. Amsterdam, The Netherlands, Elsevier, 1990, pp 209-241 2. Prockop DJ: Mutations in collagen genes as a cause of connective-tissue diseases. N Engl J Med 326540-546, 1992 3. Byers PH, Steiner RD: Osteogenesis imperfecta. Annu Rev Med 43:269-282,1992 4. Silence D: Osteogenesis imperfecta: An expanding panorama of variants. Clin Orthop 159:11-25, 1981 5. Silence DO, Barlow KK, Garber AP, et al: Osteogenesis imperfecta type II: Delineation of the phenotype with reference to genetic heterogeneity. Am J Med Gen 17:407-423,1984 6. Minisola S, Scarnecchia L, Scarda A, et al: Serum osteocalcin in primary hyperparathyroidism: Short-term effect of surgery. Miner Electrolyte Metab 14:201-207,198s 7. Scarnecchia L, Minisola S, Pacitti MT, et al: Clinical usefulness of serum tartrate-resistant acid phosphatase activity determination to evaluate bone turnover. Stand J Clin Lab Invest 51:517524, 1991 8. Price PA, Nishimoto SK: Radioimmunoassay for the vitamin K-dependent protein of bone and its discovery in plasma. Proc Natl Acad Sci USA 77:2234-2238,198O 9. Minisola S, Antonelli R, Mazzuoli GF: Clinical significance of free plasma hydroxyproline measurement in metabolic bone diseases. J Clin Chem Clin Biochem 23:515-519. 1985 10. Lau KHW, Onishi T, Wergedal JE, et al: Characterization and assay of tartrate-resistant acid phosphatase activity in serum: Potential use to assess bone resorption. Clin Chem 33:458-462, 1987 11. Melkko J, Niemi S, Risteli L, et al: Radioimmunoassay of the carboxyterminal propeptide of human type I procollagen. Clin Chem 36:1323-1332, 1990 12. Minisola S, Rosso R, Romagnoli E, et al: Trabecular bone
mineral density in primary hyperparathyroidism: Relationship to clinical presentation and biomarkers of skeletal turnover. Bone Miner 20:113-123,1993 13. Barsh GS, David KE, Byers PH: Type I osteogenesis imperfecta: A nonfunctional allele for proor chains of type I procollagen. Proc Nat1 Acad Sci USA 79:3838-3842, 1982 14. Rowe DW. Shapiro JR, Poirier M, et al: Diminished type I collagen synthesis and reduced alpha l(1) collagen messenger RNA in cultured fibroblasts from patients with dominantly inherited (type I) osteogenesis imperfecta. J Clin Invest 76:604-611, 1985 15. Willing MC, Prucnno CJ, Atkinson M, et al: Osteogenesis imperfecta type I is commonly due to a COLlAl null allele of type I collagen. Am J Hum Genet 51:508-515, 1992 16. Paterson CR, M&lion S, Stellman JL: Osteogenesis imperfecta after the menopause. N Engl J Med 310:1694-1696. 1984 17. Kurtz D, Morrish K, Shapiro J: Vertebral bone mineral content in osteogenesis imperfecta. Calcif Tissue Int 37:14-18, 1985 18. Vetter U, Fisher LW, Mintz KP, et al: Osteogenesis imperfecta: Changes in noncollagenous proteins in bone. J Bone Miner Res 6:501-505, 1991 19. Taitz LS: Child abuse and metabolic bone disease: Are they often confused? Br Med J 302:1244,1991 20. Shapiro JR, Rowe DW: Imperfect osteogenesis and osteoporosis. N Engl J Med 310:1738-1740,1984 21. Prockop DJ: Osteogenesis imperfecta: A model for genetic causes of osteoporosis and perhaps several other common diseases of connective tissue. Arthritis Rheum 31:1-8,1988 22. Shapiro JR, Stover ML, Burn VE, et al: An osteopenic nonfracture syndrome with features of mild osteogenesis imperfecta associated with the substitution of a cysteine for glycine at triple helix position 43 in the pro al(I) chain of type I collagen. J Clin Invest 89:567-573,1992
Minisola,
Anna Lina Piccioni, Rossana Rosso, Elisabetta Romagnoli, Liliana Scarnecchia, and Gianfranco Mazzuoli
Maria Teresa Pacitti,
We measured serum levels of total alkaline phosphatase activity, osteocalcin, carboxy-terminal propeptide of human type I procollagen (PICP), tartrate-resistant acid phosphatase activity (TRAP), and the fasting urinary hydroxyproline/creatinine ratio (OHPr/Cr) in seven affected members (four men, three women; age, 43.3 f 16.6 years [mean f SD]) of a family with clinically diagnosed type I-A osteogenesis imperfecta (01) and in eight (five men, three women) normal age-matched (38.2 f 10.3) relatives. Three boys with 01 and three normal girls of the same family were also studied, although they were excluded from statistical analysis. Bone mineral density was also determined at four different skeletal sites. Serum levels of PICP were measured with a radioimmunoassay (Farmos Diagnostica, Turku, Finland). There were no significant differences in mean values of the biomarkers studied between 01 patients and normal relatives, with the only exception being serum levels of PICP (35 -e 7.5 v 219 f 107.5 pg/L, P c .OOl). A significant reduction of BMD was found in 01 patients compared with normal relatives at the lumbar (L) spine (680 f 61 v 1,128 f 92 mg/cm*, P < .OOl), at the ultradistal radius ([UDR] 323 f 85 v 458 f 76, P < .006), at the femoral neck ([F] 494 f 140 v 791 f 104, P < .OOl), and at the junction of the distal and middle third of the radius ([MR] 639 f 71 v 717 -C 52, P c .029). When the patients and control subjects were combined, there were significant positive correlations between serum PICP and BMD values at various skeletal sites (L, P < .006; F, P < .05; UDR, P < .005; MR. P < .007). Our results suggest that decreased levels of serum PICP are typical of 01 type I patients, or at least of this family. This should be ascribed to a decreased amount of collagen produced, although the possibility of an abnormal sequence not recognized by the antiserum used should also be considered. Densitometric results indicate that quantitative or qualitative defects of collagen structure may contribute to the fragility of 01 bone by interfering with complete mineralization and/or normal tissue structure. Copyright 0 1994 by W.B. Saunders Company
0
IMPERFECTA (01) is the most STEOGENESIS common heritable disorder of bone. It is a heterogeneous disease characterized by bone fragility, osteopenia, progressive skeletal deformities, and often severe growth retardation. Recent biochemical, linkage, and molecular genetic studies have shown that in almost every instance 01 results from mutations in the genes that encode the chains of type I collagen. Such studies have improved our understanding of the molecular basis of this disorder, and have provided significant advances in molecular diagnosis and prognostic counseling.‘-3 However, despite these impressive breakthroughs in basic sciences, there are no detailed clinical studies that attempt to correlate biochemical and densitometric data
From the Institute of II Clinica Medica, Cattedra di Medicina Intema, Universitcidegli Studi di Roma “La Sapienza, ” Roma, Italy. Submitted July 26, 1993; accepted January 2, 1994. Supported by a grant from the Consiglio Nazionale delle Ricerche, “Progetto Finalizzato Invecchiamento, “code 96-3-407. Presented in pal? at the Fourteenth Annual Meeting of the American Sociev for Bone and Mineral Research, Minneapolis, MN, September 30 to October 4, 1992, and at the Fowih International Symposium on Osteoporosis and Consensus Development Conference, Hong Kong, March 27 to April 2, 1993. Address reprint requests to Salvatore Minisola. MD, II Clinica Medica, Policlinico Umberto I, Viale de1 Policlinico 155. 00161 Rome, Italy. Copyright 0 1994 by W B. Saunders Company 0026-0495i94l4310-0010$03.00l0
noqf-lmILL-A 12
3
1' Fig 1. ily.
Pedigree for the 01 fam-
Metabolism, Vol43,No
KEY:
lO(October),1994:pp 1261-1265
4
5
l=Q
6
affected n = d affected
7
a
9
0 = Q
unaffected
q=d unaffected
10 11
/=
12 13
14 15
deceased
#=proband
1261
MINISOLA
1262
Table 1. Anthropometric,
35.
Phenotypic, and Clinical Features and
I
Serum PICP Values in the Family With Type I-A 01 Patient 1
Age (~4
yrsl
2
3
4
5
6
7
8
9
10
12
15
20
29
35
42
53
57
67
5
Height (cm)
110
160
158
172
168
158
155
165
150
160
Weight (kg)
25
55
43
65
86
47
51
64
63
45
4
7
13
8
8
4
5
4
4
4
Fractures (no.) Ligamentous laxity Scoliosis Hearing loss Easy bruisability
+ -
vs 07
No.
-I \ 2
MMMMMFFMFM
sex
ET AL
+ + -
+ f -
+ + -
+ +
+ +
+ -
+ -
+
+
-
+
+
-
-
_
_
-
+_---_
48
40
+ + -
+ + +
E 2
+
+
8
l 8
30
15.
Excessive diapho+
resis
10
Serum PICP (l&L)
128
75
100
27
26
34
35
a specific phenotype. The need for these has become more compelling, considering improvements in the methodology used to evaluate bone turnover and measure bone mineral density (BMD). We recently had the opportunity to study a very large kindred with 01 type I-A. Results obtained indicate that the in vivo measurement of procollagen molecules may be used to identify this clinical type of brittle bone disease. SUBJECTS AND METHODS
5
1
0 5 ’
300.
-
14.0
2 \
12.0
F 1
10.0
;
8.0
N
from statistical analysis; a comparison was not possible owing to the different sex, age, and pubertal status. Each patient and subject willing to participate in the study underwent an initial comprehensive health survey, a physical examination, and a multichannel autoanalyzer chemical screening. Metabolic study included a 24-hour urine collection to evaluate calcium, phosphorus, and creatinine (Cr), followed by a short-term urine collection after a 12-hour overnight fast to determine the calcium and hydroxyproline (OHPr)/Cr ratio and the renal tubular threshold of phosphate excretion. Halfway through the latter collection, a blood sample was taken for the measurement of serum levels of calcium, phosphorus, Cr. intact parathyroid hormone, osteocalcin, total alkaline phosphatase activity, tartrate-resistant
We studied seven affected members (four men, three women; age, 43.3 2 16.6 years [mean + SD]) of a family clinically diagnosed with type I-A 01 on the basis of the classification of Silence et al.4,5 Biochemical and densitometric results were compared with those of eight (five men, three women) normal age-matched (38.2 5 10.3) relatives. Three boys with 01 and three normal girls from the same family were also studied. but they were excluded yrsl
01
Fig 2. Individual values of two independent markers of bone resorption found in patients with 01 and their age-matched normal relatives (N). The values of six children studied (indicated by their respective ages) have been included, even though they were excluded from statistical analysis. Horizontal bars indicate the means.
with
400
N
01
35
yn
.a
700 600 _1 500 \ 400 0) 300 =L -
n I-
g
200.
a
I
a
vsi’
6.0
E 3 ii (/)
& 200
5 $ 4.0 v)
loo-
O-
01
N
E 2 100 Q, (I)
0
0 01
5
I
: 0
N
1:
Fig 3. Individual values of three independent markers of bone formation found in patients with 01 and their age-matched normal relatives (N). See Fig 2 legend.
REDUCED
SERUM
PICP IN TYPE
1263
I-A 01
acid phosphatase activity (TRAP), and carboxy-terminal propeptide of human type I procollagen (PICP). For the purposes of this study, biochemical markers of bone turnover were determined according to previously described methodsh.’ In particular, serum osteocalcin levels were measured using a radioimmunoassay kit (Incstar, Stillwater, MN) based on the method of Price and Nishimot0.s The limit of detectability was 0.3 ng/mL. The intraassay variation was 2%, and the interassay variation was 6.5%. Normal values in our laboratory are 3.31 ? 1.04 ng/mL. Urinary OHPr was assayed by a calorimetric method9; normal values in our laboratory for the fasting OHPriCr ratio are 30 ? 10 mg/g. Serum total alkaline phosphatase activity levels (normal values. 75 ? 18 U/L) were measured using a calorimetric method by autoanalyzer (Technicon Autoanalyzer RA 1000, Tarrytown, NY). Serum TRAP levels were measured according to the method of Lau et al.‘? normal values in our laboratory are 13 ? 2.4 U/L. Serum concentrations of PICP were measured in duplicate lOO-yL serum samples with a recently developed equilibrium radioimmunoassay supplied by Farmos Diagnostica (Turku, Finland).” The sensitivity of the test was 1.2 pg/L. The intraassay coefficient of variation was 2.8% in the concentration range of our study. The corresponding interassay variation was approximately 5.5%. The reference interval (mean ? 2 SD) for 48 normal subjects (42 to 73 years of age) was 80 to 210 kg/L. BMD was measured at the lumbar spine (L) and at the femoral neck (F) by quantitative digital radiography using a Hologic QDR-1000 x-ray bone densitometer (Hologic, Waltham, MA). Furthermore, BMD was also assessed at the junction of the distal and middle third of the radius (proximal site [MR]) and at the ultradisl.al radius ([UDR] one tenth of the forearm length, 4 2.5 cm from the styloid process of the radius) of the nondominant arm by dual-photon densitometer (NIM, Verona. Italy) as previously described.”
Statistical Analvsis Unless otherwise specified, all results are expressed as the mean ?: 1 SD. Significance between means was assessed using Student’s 1 test for unpaired data. Correlations between variables were tested by linear regression analysis. The significance level was set at P less than .05 for all tests. RESULTS The pedigree for this family is presented in Fig 1. The proband (IV-3, see arrow) was a E-year-old boy who had already had 13 fractures in his life. The main anthropometric, phenotypic, and clinical features of the 10 patients with 01 studied here are reported in Table 1. By definition, all were characterized by blue sclerae and the absence of dentinogenesis imperfecta; as for the number of fractures, its apparent low number in the older patients is most probably due to the difficulty elderly patients have in recalling exactly how many fractures they have had. The ascertainment of the number of fractures is easier for those who are presently young because they are carefully checked by their parents. Figure 2 shows individual values of two independent markers of bone resorption found in patients with 01 and in their age-matched normal relatives, respectively. No significant differences between the two groups were detected in mean values (serum TRAP, 14.6 2 2.3 v 15.8 4 2.3 U/L; fasting urinary OHPr/Cr, 29.7 + 9.1 v 24.6 + 5.6 mgig). Figure 3 represents the individual values of three indepen-
dent markers of bone formation. Although there were no significant differences between the two groups in terms of the mean values of serum total alkaline phosphatase activity (94.4 ? 25.3 v 78.6 +- 19.6 U/L) and serum osteocaltin levels (3.85 ? 1.1 v 2.93 + 0.9 ng/mL), a significant difference was found in the mean values of serum PICP (35 rt 7.5 v 219 2 107.5 kg/L, P < ,001). A significant reduction of BMD was found in patients with 01 compared with normal relatives at the L (680 f 61 II 1,128 f 92 mg/cm*, P < .OOl; Fig 4), at the UDR (323 2 85 1’ 458 % 76, P < .006), at the F (494 2 140 v 791 2 104, P < .OOl). and at the MR (639 + 71 v 717 + 52, P < .029; Fig 5). Finally, when the patients and control subjects were
pc
0.001
1300
I
0
i
l
I
0
13
8
i
1200 1100 1000
w
I
900 800 700 600 500 400
Aa/ 8 0
vsi 015; 0
121 I
0
5; I
01
N
Fig 4. Individual values of L-WAD found in patients with 01 and their age-matched normal relatives (N). Sea Fig 2 legend.
MINISOLA
800,
600
0.029. i l ; 0 I
pc
p<
; 1 /
500 -&- 700. E < F
p
I
l l
900
l
"E
I ,
1
l l
-
"E800. < F700.
s400 L
600.
loo0
0.006
ET AL
0
600'
i 300 ; m &
I
0
500.
si
8 i
E 3
0
I,
l 7
m k 500.
200
wl
400.
012;
400200-
0
01
100
5; N
01
N
combined, there were significant positive correlations between serum PICP and BMD values at the various skeletal sites measured (L, r = ,677, P < .006; F, r = S14, P < .05; UDR, r = .681, P < .005; MR, r = .662, P < ,007). Conversely, these were not found when considering other biomarkers. DISCUSSION
A number of in vitro studies have recently attempted to define the molecular basis of 01 type I more clearly. It has become evident that cells from most affected individuals with 01 type I synthesize and secrete approximately half the normal amount of type I procollagen.13.14 The decrease in production of the type I procollagen molecule results from a decreased synthesis of procul(1) chains, usually a reflection of the decreased amount of mRNA for that chain.15 Our in vivo results, which show that the affected members of the family we studied are characterized by reduced mean serum levels of PICP (with respect to their agematched normal relatives), might be considered the perfect clinical counterpart to in vitro findings. It is in fact conceivable that this serum decrease is a reflection of a generalized decrease of type I collagen synthesis; the phenotypic features and clinical findings observed in the affected patients should be ascribed to it. It is known that an altered sequence in the carboxyterminal propeptide of the procxl(1) chain, as well as mutations altering the structure of the type I procollagen by virtue of single substitutions in the triple or nontriple helical domain, result in 01 type I. Studies are now in progress to characterize the molecular defect of our family. However, even these last two defects should give rise to decreased serum PICP levels; this is because the amount of the molecule assembled will be half the normal in the first
Fig 5. Individual values of MRBMD, UDR-BMD, and F-BMD in patients with 01 and their agematched normal relatives (N). See Fig 2 legend. For technical reasons, some measurements were not performed in the young
subjects. case, or owing to the abnormal sequence, not recognized by the antiserum used in the second hypothesis. In this context, a particular point of consideration derives from the finding of very low mean serum PICP levels versus what would be expected if only one allele was functioning. The circulating level of the intact propeptide presumably reflects the balance between the rate of its synthesis and that of its incorporation into the matrix, in addition to degradation. A possible explanation should therefore be sought on a different half-life of the propeptide or on a different efficiency at which synthesized type I procollagen is incorporated into the matrix when one allele versus two functional alleles are available. Although 01 is traditionally considered a cause of osteoporosis,lJ” few data have been presented so far concerning BMD measurement in this disorder.’ Furthermore, published results usually deal with a single measurement site17; there is no extensive survey of various skeletal sites reflecting a different percentage composition of trabecular and cortical bone. Our data strongly indicate that reduced BMD is a generalized feature of the patients we investigated with 01 type I, whatever the measurement site chosen. The relevance of this finding is even greater if we consider that comparisons were made with normal relatives living in the same area. This indeed excludes the possibility that different life-styles or eating habits could have been responsible for the osteopenia. Having shown that affected members of 01 type I have not only reduced BMD values but also serum PICP levels, could we hypothesize a relationship between them? The finding of a direct correlation between serum PICP and BMD values strongly indicates, in our opinion, that quantitative (or qualitative) defects of collagen structure contrib-
1265
REDUCED SERUM PICP IN TYPE I-A 01
ute to the fragility of 01 bone by interfering with complete mineralization and/or normal tissue structure.i8 This hypothesis is further supported by the lack of significant correlations between conventional markers of bone turnover and BMD values and by the absence of significant differences in mean values of bioindices of skeletal remodeling in the two groups examined. This in fact suggests that the possible imbalance between the two remodeling processes has not been the cause of fractures in our patients. From a practical point of view, there are four main implications of our finding of reduced serum PICP in 01 type I patients: (1) this measurement may be used in making a diagnosis of 01 type I, although further studies are needed to examine serum PICP levels in diseases of structural mutations of collagen; (2) in the case of children with fractures, the measurement of this marker could be
used to differentiate between real cases of patients with mild 01 and abused childreni9; (3) this marker might be particularly useful in examining the response of drugs such as anabolic steroids, which are known to affect the rate of collagen synthesis in bone and other connective tissues; (4) finally, the measurement of serum levels of PICP might be used as a screening tool to unmask patients with idiopathic osteoporosis who present with a family history of osteoporosis, fractures with variable degrees of joint laxity, and mild scoliosis.zO-ZZ ACKNOWLEDGMENT
We are sincerely indebted to Dr David W. Rowe for his advice and critical review of the manuscript. We wish to thank Walter Rossi for technical assistance, and we are deeply grateful to Lee O’Hara for editing the manuscript.
REFERENCES
1. Rowe DW: Osteogenesis imperfecta, in Heersche JNM, Kanis JA (eds): Bone and Mineral Research, vol 7. Amsterdam, The Netherlands, Elsevier, 1990, pp 209-241 2. Prockop DJ: Mutations in collagen genes as a cause of connective-tissue diseases. N Engl J Med 326540-546, 1992 3. Byers PH, Steiner RD: Osteogenesis imperfecta. Annu Rev Med 43:269-282,1992 4. Silence D: Osteogenesis imperfecta: An expanding panorama of variants. Clin Orthop 159:11-25, 1981 5. Silence DO, Barlow KK, Garber AP, et al: Osteogenesis imperfecta type II: Delineation of the phenotype with reference to genetic heterogeneity. Am J Med Gen 17:407-423,1984 6. Minisola S, Scarnecchia L, Scarda A, et al: Serum osteocalcin in primary hyperparathyroidism: Short-term effect of surgery. Miner Electrolyte Metab 14:201-207,198s 7. Scarnecchia L, Minisola S, Pacitti MT, et al: Clinical usefulness of serum tartrate-resistant acid phosphatase activity determination to evaluate bone turnover. Stand J Clin Lab Invest 51:517524, 1991 8. Price PA, Nishimoto SK: Radioimmunoassay for the vitamin K-dependent protein of bone and its discovery in plasma. Proc Natl Acad Sci USA 77:2234-2238,198O 9. Minisola S, Antonelli R, Mazzuoli GF: Clinical significance of free plasma hydroxyproline measurement in metabolic bone diseases. J Clin Chem Clin Biochem 23:515-519. 1985 10. Lau KHW, Onishi T, Wergedal JE, et al: Characterization and assay of tartrate-resistant acid phosphatase activity in serum: Potential use to assess bone resorption. Clin Chem 33:458-462, 1987 11. Melkko J, Niemi S, Risteli L, et al: Radioimmunoassay of the carboxyterminal propeptide of human type I procollagen. Clin Chem 36:1323-1332, 1990 12. Minisola S, Rosso R, Romagnoli E, et al: Trabecular bone
mineral density in primary hyperparathyroidism: Relationship to clinical presentation and biomarkers of skeletal turnover. Bone Miner 20:113-123,1993 13. Barsh GS, David KE, Byers PH: Type I osteogenesis imperfecta: A nonfunctional allele for proor chains of type I procollagen. Proc Nat1 Acad Sci USA 79:3838-3842, 1982 14. Rowe DW. Shapiro JR, Poirier M, et al: Diminished type I collagen synthesis and reduced alpha l(1) collagen messenger RNA in cultured fibroblasts from patients with dominantly inherited (type I) osteogenesis imperfecta. J Clin Invest 76:604-611, 1985 15. Willing MC, Prucnno CJ, Atkinson M, et al: Osteogenesis imperfecta type I is commonly due to a COLlAl null allele of type I collagen. Am J Hum Genet 51:508-515, 1992 16. Paterson CR, M&lion S, Stellman JL: Osteogenesis imperfecta after the menopause. N Engl J Med 310:1694-1696. 1984 17. Kurtz D, Morrish K, Shapiro J: Vertebral bone mineral content in osteogenesis imperfecta. Calcif Tissue Int 37:14-18, 1985 18. Vetter U, Fisher LW, Mintz KP, et al: Osteogenesis imperfecta: Changes in noncollagenous proteins in bone. J Bone Miner Res 6:501-505, 1991 19. Taitz LS: Child abuse and metabolic bone disease: Are they often confused? Br Med J 302:1244,1991 20. Shapiro JR, Rowe DW: Imperfect osteogenesis and osteoporosis. N Engl J Med 310:1738-1740,1984 21. Prockop DJ: Osteogenesis imperfecta: A model for genetic causes of osteoporosis and perhaps several other common diseases of connective tissue. Arthritis Rheum 31:1-8,1988 22. Shapiro JR, Stover ML, Burn VE, et al: An osteopenic nonfracture syndrome with features of mild osteogenesis imperfecta associated with the substitution of a cysteine for glycine at triple helix position 43 in the pro al(I) chain of type I collagen. J Clin Invest 89:567-573,1992