Biochimica et Biophysica Acta, 1049 (1990) 171-176
171
Elsevier BBAEXP 92066
Growth-dependent modulation of type I collagen production and mRNA levels in cultured human skin fibroblasts Jyrki K. MS.kela, Tuula Vuorio and Eero Vuorio Department of Medieal Bioehemistry, University of Turku, Turku (Finland)
(Received 22 February 1990)
Key words: Collagen; Fibroblast; mRNA; recombinant DNA; Polymorphism
Five human skin fibroblast lines were studied for type I collagen production and type I procoilagen mRNA levels through the different growth phases. The cells were plated at low density and followed for 11 days at daily intervals through the stages of rapid growth and visual confluency until the cultures reached stationary growth phase. Each day one culture flask was labeled with [3H]proline for 24 h, and analyzed for production of radiolabeled type I collagen into culture medium. The cell layers were counted and subjected to isolation of cytoplasmic RNA and determination of type I procollagen mRNA levels. The results revealed an approx. 2-fold increase in procollagen production and mRNA levels when the cells reached visual confluency. Thereafter the synthesis rates and mRNA levels remained relatively constant, although a decreasing tendency of both parameters was observed upon further culturing. The results confirm that determination of cell density is important when cell cultures are used for measurement of collagen synthesis or mRNA levels. For determination of proa2(l) collagen mRNA an 1193 bp cDNA clone was constructed using RNA extracted from human fetal calvaria. Sequencing of the clone revealed some nucleotide and amino acid differences between the previously published sequences. This suggests the presence of more individual variation in procollagen coding sequences than expected.
Introduction Fibroblasts cultured from human skin have been extensively used in studies on regulation of collagen production and on various disorders of collagen metabolism [1-3]. One important prerequisite for such experiments is to determine how much the rate of collagen production and m R N A levels depend on the growth cycle of the culture. Since m a n y experimental systems involve comparisons of two or more fibroblast cultures it is important to know how much variation can be caused by the culture system per se, particularly if the cultures are treated with factors affecting the growth rate. Current literature on the subject is contradictory. Detailed studies on human lung fibroblasts have shown an approx. 2-fold increase in type I and type III collagen production and m R N A s when the cells reach confluency [4,5], while others have reported unaltered procollagen m R N A levels per cell under similar conditions [6]. Studies on serum-stimulated human gingival
Correspondence: E. Vuorio, The University of Texas, M.D. Anderson Cancer Center, Department of Molecular Genetics, Box 11, 1515 Holcombe Blvd., Houston, TX 77030, U.S.A.
fibroblasts have also shown that activation of collagen synthesis is not coupled to mitogenic stimulus [7]. In another study on h u m a n skin fibroblasts confluent cells were found to produce less collagen than proliferating fibroblasts [8]. Type I collagen is the major member of the collagen gene family which consists of at least 13 different collagen types [9,10]. Since m a n y of the collagen molecules are heterotrimers of two or three different chains, more than 25 different genes are needed to code for their constituent a-chains. Molecular cloning of c D N A s for human procollagen mRNAs, and of the corresponding genes, has greatly increased our understanding of this gene family dispersed on m a n y chromosomes of the human genome. Genomic and e D N A clones have been reported for h u m a n types I - V I and XI procollagens, and for several procollagens of various other species [9,10]. The h u m a n sequences reported so far represent gene or m R N A (cDNA) sequences of only a few individuals, but even these suggest the presence of individual variation. In this paper we report on (a) the determination of type ! collagen synthesis rates and m R N A levels in human skin fibroblasts in culture, (b) the construction and sequencing of an 1193 bp e D N A clone for human
0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
172 proa2(I) collagen mRNA, and (c) the comparison of its sequence with the other nucleotide and amino acid sequences available for this region [11-15]. Materials and Methods
Cell cultures Fibroblast cell lines were started from skin biopsies of five healthy individuals using the explantation method. The fibroblast lines had undergone 7-14 passages before these experiments. From each cell line 16-20 parallel subcultures were started into 25 cm 2 Petri dishes, 12000 cells per cm 2. The cultures were fed every 24 h with Dulbecco's modification of Eagle's minimum essential medium (DMEM) supplemented with 10% fetal calf serum, streptomycin (100 /~g/ml) and penicillin (100 I U / m l ) . Starting 24 h after subculturing two plates were selected daily for labeling with [3H]proline (20/~Ci/ml) for 24 h in serum-free D M E M without glutamine. Ascorbate and /~-aminopropionitrile (50 ~ g / m l each) were added into the labeling medium. After labeling, proteinase inhibitors were added to the culture media, which were stored at - 20 o C. Cell layers were washed with phosphate-buffed saline, trypsinized and the cells were counted in a Biirker chamber, and stored in aliquots of ( 3 - 5 ) . 105 cells at - 7 0 ° C for isolation of cytoplasmic RNA. Isolation of RNA Cytoplasmic RNA was extracted from frozen aliquots of cells as described by White and Bancroft [16]. For construction of the cDNA clone, pHCAL2 total RNA was isolated from calvaria and skin of a normal human fetus (obtained at a therapeutic abortion) as described earlier [17]. The poly(A) + fraction of total RNA was prepared by two cycles of oligo(dT)-cellulose chromatography [18]. Determination of procollagen mRNA levels Cytoplasmic RNA levels were determined by slot blot analysis using a vacuum manifold (Schleicher and Schuell, F.R.G.). Cytoplasmic RNA samples corresponding to 20000 and 40000 cells were applied onto nitrocellulose filters, baked and hybridized as described earlier [19]. The recombinant plasmids p H C A L 1 U [20] and pHCAL2 (described below) were labelled with [32p]dCTP by nick-translation. After stringent washes (0.1 X SSC, 0.5% SDS, at + 55 ° C) the filters were autoradiographed the films were analyzed by laser densitometry using the area scanning method. The specificity of the probes under these conditions was confirmed by Northern blot hybridizations. Determination of collagen production rates Culture medium samples were dialyzed and lyophilized; aliquots corresponding to 3.105 cells were then
fractionated electrophoretically on 6% SDS-polyacrylamide gels after reduction as described by O'Farrell [21]. Other aliquots were digested with pepsin overnight at + 4 ° C, lyophilized and electrophoresed without reduction. The gels were fixed, dried and exposed for 1- 3 weeks with Kodak X-Omat film at - 70 o C [22]. Collagenous bands were quantified by laser densitometry as above.
Construction of recombinant plasmid pHCA L2 Poly(A) + RNA from calvarial bones was used as template to synthesize double-stranded (ds) cDNA as described previously [23]. The ds-cDNA was digested with the restriction enzymes EcoR1 and PstI and fractionated on a 0.75% agarose gel. Fragments with sizes between 950 and 1300 bp were collected by binding to DEAE membrane [24], purified and cloned between the EcoRI and PstI sites of plasmid pBR322. Transformation into E. coli strain DH-5 by the CaCI 2 method [25] produced several hundred tranformants, which were screened by colony hybridization using a 32p-labeled purified insert from clone pCAL3, containing cDNA for chick proc~2(I) collagen m R N A [26], as the probe. Several positive colonies were detected and one of those (pHCAL2) was selected for further characterization. Sequencing Clone pHCAL2 was sequenced by the Sanger method [27] using [35S]deoxy(thio)ATP as the label. For this purpose restriction fragments outlined in Fig. 1 were subcioned into phages M13mp18 and 19. The entire sequence was read from both strands. Results
Construction and sequencing of cDNA clone pHCAL2 The cloning strategy used to construct the cDNA clone for human proa2(I) collagen m R N A was essentially similar to that used previously to obtain cDNA clones for proocl(I) [20,28], proal(III) [29], and procd(II) collagen mRNAs [23]. Through available sequence information we could predict that digestion of doublestranded c D N A for human proa2(I) collagen mRNA with EcoRI and PstI should yield an approx. 1200 kb fragment [12]. Fragments in this size class were cloned into pBR322, and screened using the insert of the Pst 1
Ta.q 1
Xba 1
Sac 1
I
t
t
I
(t
t
I
~ (
g
TH _
--
~
TP I
)
EcoR1 1
I ( I
CP I
Fig. 1. Restriction map and sequencing strategy of the insert of plasmid pHCAL2. Shown below are the domains of the proa2(I) collagen chain: TH, triple helical domain; TP, telopeptide; CP, C-propeptide.
173 corresponding clone for chick proa2(I) collagen m R N A [26]. A number of positive clones were found; one of them was n a m e d p H C A L 2 and selected for further characterization. The partial restriction map and sequencing strategy of clone p H C A L 2 is shown in Fig. 1, and the c o m p l e t e
A
al a2 1
5
10
B a2
i CTGCAGGTGCACCTGGTCCTCATGGCCCCGTGGGTCCTGCTGGCAAACATGGAAACCGTG A G A P G P H G P V G P A G K H G N R G 870 61GTGAAACTGGTCCTTCTGGTCCTGTTGGTCCTGCTGGTGCTGTTGGCCC~GAGGTCCTA E T G P S G P V G P A G A V G P R G P S 890 121GTGGCCCAC~GGCATTCGTGGCGAT~GGGAGAGCCCGGTGA~.AA~GGCCCAGAGGTC G P Q G I R G D K G E P G E K G P R G L 910 C 181TTCCTGGCTTAAAGGGACAC~TGGA~GC~GGTCTGCCTGGTATCGCTGGTCACCATG P G L K G H N G L Q G L P G I A G H H G 930 F 241GTGATC~GGTGCTCCTGGCTCCGTG~TCCTGCT~TCCTA~CCCTGCTGGTCCTT D Q G A P G S V G P A G P R G P A G P S 950 301CTGGCCCTGCTGGAAAAGATGGTCGCACTGGACATCCTGGTACGGTTGGACCTGCTGGCA G P A G K D G R T G H P G T V G P A G I 970 361TTCGAGGCCCTCA~TCACC~GGCCCTGCTGGCCCCCCTGGTCCCCCTGGCCCTCCTG R G P Q G H Q G P A G P P G P P G P P G 990
1
5
10
time {days)
Fig. 3. Daily determination of cytoplasmic type I collagen mRNA levels (A) and type I collagen secretion (B) in a fibroblast culture during the phases of rapid growth (days 1-4), confluency (days 5-7) and stationary growth (days 8-10). Fibroblasts were plated onto 25 cm2 Petri dishes at 12000 cells/cm2 Each day one culture was labeled with [3H]proline and analyzed after 24 h for collagen production, procollagen mRNAs and cell number. The mRNA levels for proal(l) and proa2(I) chains were determined by hybridization of cytoplasmic RNA corresponding to 2.104 cells with [32p]dCTP-labeled cDNA probes pHCAL1U and pHCAL2, respectively. For the analysis of al and a2 chains medium samples were treated with pepsin, fractionated unreduced on a 6% SDS-polyacrylamide gel and autoradiographed. Due to the small number of cells at days 1 and 2 these samples contain half the material of the rest of the samples.
421 G A C C T C C A ~ T G T ~ G C G ~ G G T G G ~ A T G A C T ~ A C G A T ~ A G A C T T C T A C A G G G P P G V S G G G Y D F G Y D G D F Y R A i010 1014 lc 481 CTGACCAGCCTCGCTCAGCACCTTCTCTCAGACCC~ACTATG~GTTGATGCTACTC D Q P R S A P S L R P K D Y E V D A T L 16c 541 TG~GTCTCTC~C~CCAGA~GAGACCCTTCTTACTCCTG~GGCTCTAGAAAG~CC K S L N N Q I E T L L T P E G S R K N P 36c C 601 CAGCTCGCACATGCCGTGAC~GAGACTCAGCCACCCAGAGTGGAGCAGTGGTTACTACT A R T C R D L R L S H P E W S S G Y Y W 56c C AC C 661GGATTGACCCT~CC~GGATGCACTATGGATGCTATCA~LEGTATACTGTGATTTCTCTA I D P N Q G C T M D A I K V Y C D F S T 76c E P C 721CTGGCGAAACCTGTATCCGGGCCC~CCTGAAAACATCCCAGCC~G~CTGGTATAGGA G E T C I R A Q P E N I P A K N W Y R S 96c 781 GCTCC~GGAC~GAAACACGTCTGGCTA~AGAAACTATC~TGCT~CAGCCAGT~G S K D K K H V W L G E T I N A G S Q F E i16c T 841 ~ T A T ~ T G T A G ~ A G T G A C ~ C C ~ A A A T ~ C T A C C C ~ C T T G C C T T C A T G C G C C Y N V E G V T S K E M A T Q L A F M R L 136c 901 TGCTGGCC~CTATGCCTCTCAG~CATCACCTACCACTGC~G~CAGCATTGCATACA L A N Y A S Q N I T Y H C K N S I A Y M 156c 961 TGGATGAGGAGACT~C~CCTGAAAAA~CTGTC~TCTACA~CTCT~TGATGTTG D E E T G N L K K A V I L Q G S N D V E 176c 1021 ~CTTGTTGCTGAGGGC~CAGCAGG~CACTTACACTGTTCTTGTAGATGGCTGCTCTA L V E E G N S R F T Y T V L V D G C S K 196c 1081AAAAGACAAATG~TGGGGAAAGAC~TCA~G~TACAAAACAAAT~GCCATCACGCC K T N E W G K T I I E Y K T N K P S R L 216c 1141TGCCCTTCCTTGATATTGCACCT~GGACATCGGTGGTGCTGACCATG~TTC P F L D I A P L D I G G A D H E F 236c 252c
nucleotide sequence in Fig. 2. The clone contains 1193 bp of sequence c o d i n g for 144 a m i n o acids in the triple helical d o m a i n , for the 15 a m i n o acids of the C-telopeptide, and for 237 a m i n o acids of the C-propeptide. C o m p a r i s o n o f the sequence of p H C A L 2 with that previously p u b l i s h e d for the h u m a n p r o a ( I ) c D N A [ 1 1 15] revealed eight nucleotide differences, three o f which resulted in a c h a n g e in the a m i n o acid (Fig. 2). In Northern blot analysis of total calvarial R N A , the clone p H C A L 2 hybridized to two m R N A s of approx. 4.8 and 5.2 kb, whereas the probe p H C A L 1 U , for procd(I) collagen m R N A , recognized two m R N A s with estimated sizes o f 5.3 kb and 6.5 kb (data not shown).
Fig. 2. Nucleotide sequence of clone pHCAL2. The nucleotide numbering is shown on the left. The numbering of the translated amino acid sequence follows the domains of the proa2(I) collagen molecule: the clone covers the end of the triple helical domain (amino acids 870-1014), the C-telopeptide (amino acids lc-15c), and most of the C-propeptide (amino acids 16c-152c). The amino acids are numbered under the one letter codes. The nucleotides in the previously reported sequences [11-15] which differ from the sequence of pHCAL2 are shown above the sequence, and the amino acid differences below. The sequence between nucleotides 1080 and 1193 has also been reported for the gene from a normal individual [11] and a patient with osteogenesis imperfecta [14]. The four nucleotide sequences are identical, except for a 4-bp deletion (between nucleotides 1124-1129) in the patient DNA. Asterisk (*) at position 429 shows the base, which is mutated in a family of osteogenesis imperfecta type IV variant [15].
174
A
B
.
~ 1.0
I:1
1.o
Discussion
, ~ o.5
0.5
I
I
i
1
I
I
'
i
i
5 time
< ~ 0.5 E
}
10 O
}
~, 1.o
compared with cells which had reached confluency (Fig. 4C and D.
{
,
,
i
10
1
5 time (doys)
;
. . . . . . . . 5 time (doys)
(dQyS)
ttt ttt tt I
1
I
. . . . . . . 5 time {do)s)
10
|
,
,
,
i
i
,
t
i
10
1o! ul
< z 0.5 E
; 0
Fig. 4. Compiled data of the changes in type I collagen mRNA levels and type I collagen secretion into culture medium during different growth phases of five human skin fibroblast cell lines. Individual cell lines were analyzed as shown in Fig. 3. The autoradiograms from hybridizations and gel electrophoreses were quantified by densitometry. Relative units were used with day 8 as the reference data. The mean and standard deviation for each time point are shown. Standard deviation at day 8 was obtained by changing day 7 for the reference day. Panels A and B show relative daily secretion of at and a2 chains, respectively. Panels C and D show the corresponding changes in the cytoplasmic proal(I) and proa2(I) collagen mRNA levels.
Growth-dependent modulation of collagen production and mRNA levels Five different human dermal fibroblast lines were analyzed to determine secretion of cd(I) and ~2(I) chains into culture medium and corresponding m R N A levels during different growth phases. A typical pattern of pepsin-treated culture medium collagens resolved by electrophoresis is shown in Fig. 3B. The secretion of al(I) and a2(I) chains seemed to be almost constant after the fibroblast cultures had reached visual confluency, which occurred at day 5 or 6. All the cell lines showed a similar pattern in the production of al(I) and c~2(I) chains (Fig. 4A and B), the greatest variation occurring just before visual confluency was achieved. To determine the m R N A levels for p r o a l ( I ) and proa2(I) chains in the same cells, slot blot analyses were performed using cytoplasmic R N A samples. Typical hybridization patterns for one of the cell lines are shown in Fig. 3A. A clear increase in p r o a l ( I ) and pro~2(I) m R N A levels was observed during the rapid growth phase. Densitometric analyses revealed approx. 2-fold increases in the levels of both m R N A s when
The present study addresses two questions of technical and biological interest in connective tissue research: (a) how much does the growth phase of fibroblasts in culture affect their collagen production and m R N A levels, and (b) what is the extent of individual variation observed between coding sequences of procollagen genes? Fibroblast cultures have been widely applied to studies on the biochemistry and regulation of collagen production under normal conditions [1,2] and in disease states [3,28]. Several studies have employed skin fibroblasts to determine the effects of various growth factors on collagen production (see Refs. 2 and 30). Such studies have been hampered by the conflicting information of the role of the growth phase of the culture on the rate of collagen production and on the corresponding m R N A levels [4-8]. Here we report oil studies performed on five different human skin fibroblast lines, which were all analyzed daily for 11 days through the various stages of the growth cycle. Each day collagen production into culture medium was determined and procd(I) and proc~(I) collagen m R N A levels were measured in the same cells. Both parameters were calculated per a given number of cells. A constant result of these experiments was an approx. 2-fold increase in collagen production and m R N A levels at the time the cultures reached visual confluency, i.e. when their growth rate began to slow down. This is in agreement with earlier findings on lung fibroblast cultures [4,5], but markedly different from some others [6,8]. The source of such discrepancies is difficult to explain; both the origin of the cell cultures and the experimental conditions could affect the results. In our study we used human skin fibroblasts and routine culture and labeling conditions which are widely used. The results show that careful monitoring of the growth phase of fibroblast cultures and determination of cell numbers at the end of each experiment are important when fibroblast cultures are used to study the rate of collagen production or procollagen m R N A levels. This includes studies on factors affecting the rate of fibroblast proliferation, such as epidermal growth factor which appears to stimulate collagen production due to enhanced proliferation with no direct effect on procollagen m R N A s [31]. The extent of individual variation in the human genome has received considerable attention in respect to sequences which cause restriction fragment length polymorphisms (RFLPs), since these have opened new possibilities for molecular genetics in the form of genetic linkage analysis [32]. Based on systematic studies on individual variation in the /~-globin gene family it has
175 been estimated that two individuals differ from each other approximately once per every 100 bp [33]. Most of this divergence appears in the n o n c o d i n g sequences of the genome. In the globin gene family over 300 variants have been characterized at the amino acid or nucleotide level [34]. However, very little is k n o w n about individual variation in the coding sequences of other h u m a n genes. Variation has been observed in p r o M ( I ) and proet2(I) collagen genes resulting in mutations which cause diseases like osteogenesis imperfecta and Ehlers Danlos syndrome [3,9]. The sequence information available on normal procollagen genes and c D N A s is limited to only a few individuals, and therefore the extent of variation without any detectable harmful effects for the individual has remained unknown. One of the few ways to detect such variations is to sequence c D N A or genomic clones from a n u m b e r of different individuals. Recently, three such nucleotide differences resulting in two amino acid changes in the protx2(I) collagen coding sequences of different individuals were reported [13,35]. In another comparison 17 nucleotide differences were reported between two p r o a 2 ( I ) sequences resulting in five amino acid differences [36]. Comparison of the sequence of p H C A L 2 with those of the previously published sequences [11-15] revealed eight nucleotide differences (in the 1193 bp region of overlap), three of which resulted in an amino acid change (Fig. 2). At this point it is impossible to determine whether all these differences represent true variation. I n f o r m a t i o n is also available of certain 3'-sequences from the corresponding gene [11] and from one osteogenesis imperfecta patient with a 4-bp deletion [14] and another osteogenesis imperfecta family with single amino acid substitution at the end of the triple helix [15]. A p a r t from the deletion and the point mutation, all the four sequences are identical with ours. The fact that in our translated amino acid sequence two of the three differences result in the same amino acid as seen in the corresponding chick sequence [37] might suggest that some of the differences might result from difficulties in sequence reading, or from cloning artifacts. Interestingly, however, comparison of different coding sequences for p r o M ( I ) and p r o M ( I I ) collagens have reveald approximately similar estimates (one nucleotide change per 100-200 bp of c D N A sequence) for individual variation [20,38-40]. I n f o r m a t i o n of individual variation is important for studies related to diseases believed to affect collagen genes. Genetic linkage analyses can now be used to determine with high confidence involvement of a specific collagen gene, but identification of the exact m u t a t i o n in D N A or c D N A from such individuals becomes impractical if the extent of individual variation is not known. O n the other hand, small changes in procollagen coding sequences, analogous to the globin genes, could
result not only in serious disorders, but also in changes in protein structure which only slightly affect the function of the protein. Prockop and Kivirikko [3] have suggested, and recent linkage studies support this [41], that minor mutations in collagen genes predispose to such c o m m o n 'degenerative' disorders as osteoarthrosis and osteoporosis.
Acknowledgements The expert technical assistance of Liisa Peltonen, Tuula Oivanen and Merja Lakkisto is gratefully acknowledged. This study was supported by grants from the Medical Research Council of the A c a d e m y of Finland, and from the Medical Society D u o d e c i m (J.K.M.).
References I Penttinen, R., Frey, H., Aalto, M., Vuorio, E. and Marttala, T. (1980) in Biology of Collagen (Viidik, A. and Vuust, J., eds.), pp. 87-100, Academic Press, London. 2 Bornstein, P. and Sage, H. (1989) Progr. Nucleic Acid Res. Mol. Biol. 37, 67-106. 3 Prockop, D.J. and Kivirikko, K.I. (1984) N. Engl. J. Med. 311, 376-386. 4 Tolstoshev, P., Berg, R.A., Rennard, S.I., Bradley, K.H., Trapnell, B.C. and Crystal, R.G. (1981) J. Biol. Chem. 256, 3135-3140. 5 Miskulin, M., Dalgleish, R., Kluve-Beckerman, B., Rennard, S.I., Tolstoshev, P., Brantly, M. and Crystal, R.G. (1986) Biochemistry 25, 1408-1413. 6 Voss, T. and Bornstein, P. (1986) Eur. J. Biochem. 157, 433-439. 7 Narayanan, A.S. and Page, R.C. (1987) Biochem. Biophys. Res. Commun. 145, 639-645. 8 Aumailley, M., Krieg, T., Razaka, G., Miiller, P.R. and Bricaud, H. (1982) Biochem. J. 206, 505-510. 9 Mayne, R. and Burgeson, R.E. (1987) Structure and Function of Collagen Types, pp. 1-317, Academic Press, Orlando FL. 10 Vuorio, E. and De Crombrugghe, B. (1990) Annu. Rev. Biochem. 59, 837-872. 11 Myers, J.C., Dickson, L.A., De Wet, W.J., Bernard, M.P., Chu, M.-L., Di Liberto, M., Pepe, G., Sangiorgi, F.O. and Ramirez, F. (1983) J. Biol. Chem. 258, 10128-10135. 12 Bernard, M.P., Myers, J.C., Chu, M.-L., Ramirez, F., Eikenberry, E.F. and Prockop, D.J. (1983) Biochemistry 22, 1139-1145. 13 De Wet, W., Bernard, M., Benson-Chandra, V., Chu, M.-L., Dickson, L., Weil, D. and Ramirez, F. (1987) J. Biol. Chem. 262, 16032-1636. 14 Pihlajaniemi, T., Dickson, L.A., Pope, F.M., Korhonen, V.R., Nicholls, A., Prockop, D.J. and Myers, J.C. (1984) J. Biol. Chem. 259, 12941-12944. 15 Wenstrnp, R.J., Cohn, D.H., Cohen, T. and Byers, P.H. (1988) J. Biol. Chem. 263, 7734-7740. 16 White, B.A. and Bancroft, F.C. (1982) J. Biol. Chem. 257, 85698572, 17 Rowe, D.W., Moen, R.C., Davidson, J.M., Byers, P.H., Bornstein, P. and Palmiter, R.D. (1978) Biochemistry 17, 1581-1590. 18 Aviv, H. and Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69, 1408-1412. 19 Thomas, P.S. (1980) Proc. Natl. Acad. Sci. USA 77, 5201-5205. 20 M~ikel~i, J.K., Raassina, M., Virta, A. and Vuorio, E. (1988) Nucleic Acids Res. 16, 349. 21 O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021.
176 22 Pulleyblank, D.E. and Booth, G.H. (1981) J. Biochem. Biophys. Methods 4, 339-346. 23 Elima, K., M~ikela, J.K., Vuorio, T., Kauppinen, S., Knowles, J. and Vuorio, E. (1985) Biochem. J. 229, 183-188. 24 Winberg, G. and HammarskjNd, M.-C. (1980) Nucleic Acids Res. 8, 253-264. 25 Mandel, M. and Higa, A. (1970) J. Mol. Biol. 53, 159-162. 26 M/ikel~i, J.K. and Vuorio, E. (1986) Med. Biol. 64, 15-22. 27 Sanger, F., Nicklen, S. and Coulson, E.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. 28 Vuorio, T., M~ikel~i, J.K., K~ih~iri, V-M. and Vuorio, E. (1987) Arch. Dermatol. Res. 279, 154-160. 29 Sandberg, M., M~ikela, J.K., Multim~iki, P., Vuorio, T. and Vuorio, E. (1989) Matrix 9, 82-91. 30 Sporn, M.B. and Roberts, A.B. (1986) J. Clin. Invest. 78, 329-332. 31 Laato, M., K~.hari, V.-M., Niinikoski, J. and Vuorio, E. (1987) Biochem. J. 247, 385-388. 32 Gusella, J.F. (1986) Annu. Rev. Biochem. 55, 831-854.
33 Jeffreys, A.J. (1979) Cell 18, 1-10. 34 Collins, F.S. and Weissman, S.M. (1984) Progr. Nucleic Acid Res. Mol. Biol. 31, 315 462. 35 Kuivaniemi, H., Tromp, G., Chu, M.-L. and Prockop, D.J. (1988) Biochem. J. 252, 633-640. 36 Lee, S.-T., Smith, B.D. and Greenspan, D.S. (1988) J. Biol. Chem. 263, 13414-13418. 37 Fuller, F. and Boedtker, H. (1981) Biochemistry 20, 996-1006. 38 Elima, K., Vuorio. T. and Vuorio, E. (1987) Nucleic Acids Res. 15. 9499-9504. 39 Baldwin, C.T., Reginato, A.M., Smith, C., Jimenez, S.A. and Prockop, D.J. (1989) Biochem. J. 262, 521-528. 40 Su, M.-W., Lee, B., Ramirez, F., Machado, M. and Horton, W. (1989) Nucleic Acids Res. 17, 9473. 41 Palotie, A., V~iis~inen, P., Ott, J., Ryhauen, L., Elima, K., Vikkula, M., Cheah, K., Vuorio, E. and Peltonen, k. (1989) Lancet it 924-927.
171
Elsevier BBAEXP 92066
Growth-dependent modulation of type I collagen production and mRNA levels in cultured human skin fibroblasts Jyrki K. MS.kela, Tuula Vuorio and Eero Vuorio Department of Medieal Bioehemistry, University of Turku, Turku (Finland)
(Received 22 February 1990)
Key words: Collagen; Fibroblast; mRNA; recombinant DNA; Polymorphism
Five human skin fibroblast lines were studied for type I collagen production and type I procoilagen mRNA levels through the different growth phases. The cells were plated at low density and followed for 11 days at daily intervals through the stages of rapid growth and visual confluency until the cultures reached stationary growth phase. Each day one culture flask was labeled with [3H]proline for 24 h, and analyzed for production of radiolabeled type I collagen into culture medium. The cell layers were counted and subjected to isolation of cytoplasmic RNA and determination of type I procollagen mRNA levels. The results revealed an approx. 2-fold increase in procollagen production and mRNA levels when the cells reached visual confluency. Thereafter the synthesis rates and mRNA levels remained relatively constant, although a decreasing tendency of both parameters was observed upon further culturing. The results confirm that determination of cell density is important when cell cultures are used for measurement of collagen synthesis or mRNA levels. For determination of proa2(l) collagen mRNA an 1193 bp cDNA clone was constructed using RNA extracted from human fetal calvaria. Sequencing of the clone revealed some nucleotide and amino acid differences between the previously published sequences. This suggests the presence of more individual variation in procollagen coding sequences than expected.
Introduction Fibroblasts cultured from human skin have been extensively used in studies on regulation of collagen production and on various disorders of collagen metabolism [1-3]. One important prerequisite for such experiments is to determine how much the rate of collagen production and m R N A levels depend on the growth cycle of the culture. Since m a n y experimental systems involve comparisons of two or more fibroblast cultures it is important to know how much variation can be caused by the culture system per se, particularly if the cultures are treated with factors affecting the growth rate. Current literature on the subject is contradictory. Detailed studies on human lung fibroblasts have shown an approx. 2-fold increase in type I and type III collagen production and m R N A s when the cells reach confluency [4,5], while others have reported unaltered procollagen m R N A levels per cell under similar conditions [6]. Studies on serum-stimulated human gingival
Correspondence: E. Vuorio, The University of Texas, M.D. Anderson Cancer Center, Department of Molecular Genetics, Box 11, 1515 Holcombe Blvd., Houston, TX 77030, U.S.A.
fibroblasts have also shown that activation of collagen synthesis is not coupled to mitogenic stimulus [7]. In another study on h u m a n skin fibroblasts confluent cells were found to produce less collagen than proliferating fibroblasts [8]. Type I collagen is the major member of the collagen gene family which consists of at least 13 different collagen types [9,10]. Since m a n y of the collagen molecules are heterotrimers of two or three different chains, more than 25 different genes are needed to code for their constituent a-chains. Molecular cloning of c D N A s for human procollagen mRNAs, and of the corresponding genes, has greatly increased our understanding of this gene family dispersed on m a n y chromosomes of the human genome. Genomic and e D N A clones have been reported for h u m a n types I - V I and XI procollagens, and for several procollagens of various other species [9,10]. The h u m a n sequences reported so far represent gene or m R N A (cDNA) sequences of only a few individuals, but even these suggest the presence of individual variation. In this paper we report on (a) the determination of type ! collagen synthesis rates and m R N A levels in human skin fibroblasts in culture, (b) the construction and sequencing of an 1193 bp e D N A clone for human
0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
172 proa2(I) collagen mRNA, and (c) the comparison of its sequence with the other nucleotide and amino acid sequences available for this region [11-15]. Materials and Methods
Cell cultures Fibroblast cell lines were started from skin biopsies of five healthy individuals using the explantation method. The fibroblast lines had undergone 7-14 passages before these experiments. From each cell line 16-20 parallel subcultures were started into 25 cm 2 Petri dishes, 12000 cells per cm 2. The cultures were fed every 24 h with Dulbecco's modification of Eagle's minimum essential medium (DMEM) supplemented with 10% fetal calf serum, streptomycin (100 /~g/ml) and penicillin (100 I U / m l ) . Starting 24 h after subculturing two plates were selected daily for labeling with [3H]proline (20/~Ci/ml) for 24 h in serum-free D M E M without glutamine. Ascorbate and /~-aminopropionitrile (50 ~ g / m l each) were added into the labeling medium. After labeling, proteinase inhibitors were added to the culture media, which were stored at - 20 o C. Cell layers were washed with phosphate-buffed saline, trypsinized and the cells were counted in a Biirker chamber, and stored in aliquots of ( 3 - 5 ) . 105 cells at - 7 0 ° C for isolation of cytoplasmic RNA. Isolation of RNA Cytoplasmic RNA was extracted from frozen aliquots of cells as described by White and Bancroft [16]. For construction of the cDNA clone, pHCAL2 total RNA was isolated from calvaria and skin of a normal human fetus (obtained at a therapeutic abortion) as described earlier [17]. The poly(A) + fraction of total RNA was prepared by two cycles of oligo(dT)-cellulose chromatography [18]. Determination of procollagen mRNA levels Cytoplasmic RNA levels were determined by slot blot analysis using a vacuum manifold (Schleicher and Schuell, F.R.G.). Cytoplasmic RNA samples corresponding to 20000 and 40000 cells were applied onto nitrocellulose filters, baked and hybridized as described earlier [19]. The recombinant plasmids p H C A L 1 U [20] and pHCAL2 (described below) were labelled with [32p]dCTP by nick-translation. After stringent washes (0.1 X SSC, 0.5% SDS, at + 55 ° C) the filters were autoradiographed the films were analyzed by laser densitometry using the area scanning method. The specificity of the probes under these conditions was confirmed by Northern blot hybridizations. Determination of collagen production rates Culture medium samples were dialyzed and lyophilized; aliquots corresponding to 3.105 cells were then
fractionated electrophoretically on 6% SDS-polyacrylamide gels after reduction as described by O'Farrell [21]. Other aliquots were digested with pepsin overnight at + 4 ° C, lyophilized and electrophoresed without reduction. The gels were fixed, dried and exposed for 1- 3 weeks with Kodak X-Omat film at - 70 o C [22]. Collagenous bands were quantified by laser densitometry as above.
Construction of recombinant plasmid pHCA L2 Poly(A) + RNA from calvarial bones was used as template to synthesize double-stranded (ds) cDNA as described previously [23]. The ds-cDNA was digested with the restriction enzymes EcoR1 and PstI and fractionated on a 0.75% agarose gel. Fragments with sizes between 950 and 1300 bp were collected by binding to DEAE membrane [24], purified and cloned between the EcoRI and PstI sites of plasmid pBR322. Transformation into E. coli strain DH-5 by the CaCI 2 method [25] produced several hundred tranformants, which were screened by colony hybridization using a 32p-labeled purified insert from clone pCAL3, containing cDNA for chick proc~2(I) collagen m R N A [26], as the probe. Several positive colonies were detected and one of those (pHCAL2) was selected for further characterization. Sequencing Clone pHCAL2 was sequenced by the Sanger method [27] using [35S]deoxy(thio)ATP as the label. For this purpose restriction fragments outlined in Fig. 1 were subcioned into phages M13mp18 and 19. The entire sequence was read from both strands. Results
Construction and sequencing of cDNA clone pHCAL2 The cloning strategy used to construct the cDNA clone for human proa2(I) collagen m R N A was essentially similar to that used previously to obtain cDNA clones for proocl(I) [20,28], proal(III) [29], and procd(II) collagen mRNAs [23]. Through available sequence information we could predict that digestion of doublestranded c D N A for human proa2(I) collagen mRNA with EcoRI and PstI should yield an approx. 1200 kb fragment [12]. Fragments in this size class were cloned into pBR322, and screened using the insert of the Pst 1
Ta.q 1
Xba 1
Sac 1
I
t
t
I
(t
t
I
~ (
g
TH _
--
~
TP I
)
EcoR1 1
I ( I
CP I
Fig. 1. Restriction map and sequencing strategy of the insert of plasmid pHCAL2. Shown below are the domains of the proa2(I) collagen chain: TH, triple helical domain; TP, telopeptide; CP, C-propeptide.
173 corresponding clone for chick proa2(I) collagen m R N A [26]. A number of positive clones were found; one of them was n a m e d p H C A L 2 and selected for further characterization. The partial restriction map and sequencing strategy of clone p H C A L 2 is shown in Fig. 1, and the c o m p l e t e
A
al a2 1
5
10
B a2
i CTGCAGGTGCACCTGGTCCTCATGGCCCCGTGGGTCCTGCTGGCAAACATGGAAACCGTG A G A P G P H G P V G P A G K H G N R G 870 61GTGAAACTGGTCCTTCTGGTCCTGTTGGTCCTGCTGGTGCTGTTGGCCC~GAGGTCCTA E T G P S G P V G P A G A V G P R G P S 890 121GTGGCCCAC~GGCATTCGTGGCGAT~GGGAGAGCCCGGTGA~.AA~GGCCCAGAGGTC G P Q G I R G D K G E P G E K G P R G L 910 C 181TTCCTGGCTTAAAGGGACAC~TGGA~GC~GGTCTGCCTGGTATCGCTGGTCACCATG P G L K G H N G L Q G L P G I A G H H G 930 F 241GTGATC~GGTGCTCCTGGCTCCGTG~TCCTGCT~TCCTA~CCCTGCTGGTCCTT D Q G A P G S V G P A G P R G P A G P S 950 301CTGGCCCTGCTGGAAAAGATGGTCGCACTGGACATCCTGGTACGGTTGGACCTGCTGGCA G P A G K D G R T G H P G T V G P A G I 970 361TTCGAGGCCCTCA~TCACC~GGCCCTGCTGGCCCCCCTGGTCCCCCTGGCCCTCCTG R G P Q G H Q G P A G P P G P P G P P G 990
1
5
10
time {days)
Fig. 3. Daily determination of cytoplasmic type I collagen mRNA levels (A) and type I collagen secretion (B) in a fibroblast culture during the phases of rapid growth (days 1-4), confluency (days 5-7) and stationary growth (days 8-10). Fibroblasts were plated onto 25 cm2 Petri dishes at 12000 cells/cm2 Each day one culture was labeled with [3H]proline and analyzed after 24 h for collagen production, procollagen mRNAs and cell number. The mRNA levels for proal(l) and proa2(I) chains were determined by hybridization of cytoplasmic RNA corresponding to 2.104 cells with [32p]dCTP-labeled cDNA probes pHCAL1U and pHCAL2, respectively. For the analysis of al and a2 chains medium samples were treated with pepsin, fractionated unreduced on a 6% SDS-polyacrylamide gel and autoradiographed. Due to the small number of cells at days 1 and 2 these samples contain half the material of the rest of the samples.
421 G A C C T C C A ~ T G T ~ G C G ~ G G T G G ~ A T G A C T ~ A C G A T ~ A G A C T T C T A C A G G G P P G V S G G G Y D F G Y D G D F Y R A i010 1014 lc 481 CTGACCAGCCTCGCTCAGCACCTTCTCTCAGACCC~ACTATG~GTTGATGCTACTC D Q P R S A P S L R P K D Y E V D A T L 16c 541 TG~GTCTCTC~C~CCAGA~GAGACCCTTCTTACTCCTG~GGCTCTAGAAAG~CC K S L N N Q I E T L L T P E G S R K N P 36c C 601 CAGCTCGCACATGCCGTGAC~GAGACTCAGCCACCCAGAGTGGAGCAGTGGTTACTACT A R T C R D L R L S H P E W S S G Y Y W 56c C AC C 661GGATTGACCCT~CC~GGATGCACTATGGATGCTATCA~LEGTATACTGTGATTTCTCTA I D P N Q G C T M D A I K V Y C D F S T 76c E P C 721CTGGCGAAACCTGTATCCGGGCCC~CCTGAAAACATCCCAGCC~G~CTGGTATAGGA G E T C I R A Q P E N I P A K N W Y R S 96c 781 GCTCC~GGAC~GAAACACGTCTGGCTA~AGAAACTATC~TGCT~CAGCCAGT~G S K D K K H V W L G E T I N A G S Q F E i16c T 841 ~ T A T ~ T G T A G ~ A G T G A C ~ C C ~ A A A T ~ C T A C C C ~ C T T G C C T T C A T G C G C C Y N V E G V T S K E M A T Q L A F M R L 136c 901 TGCTGGCC~CTATGCCTCTCAG~CATCACCTACCACTGC~G~CAGCATTGCATACA L A N Y A S Q N I T Y H C K N S I A Y M 156c 961 TGGATGAGGAGACT~C~CCTGAAAAA~CTGTC~TCTACA~CTCT~TGATGTTG D E E T G N L K K A V I L Q G S N D V E 176c 1021 ~CTTGTTGCTGAGGGC~CAGCAGG~CACTTACACTGTTCTTGTAGATGGCTGCTCTA L V E E G N S R F T Y T V L V D G C S K 196c 1081AAAAGACAAATG~TGGGGAAAGAC~TCA~G~TACAAAACAAAT~GCCATCACGCC K T N E W G K T I I E Y K T N K P S R L 216c 1141TGCCCTTCCTTGATATTGCACCT~GGACATCGGTGGTGCTGACCATG~TTC P F L D I A P L D I G G A D H E F 236c 252c
nucleotide sequence in Fig. 2. The clone contains 1193 bp of sequence c o d i n g for 144 a m i n o acids in the triple helical d o m a i n , for the 15 a m i n o acids of the C-telopeptide, and for 237 a m i n o acids of the C-propeptide. C o m p a r i s o n o f the sequence of p H C A L 2 with that previously p u b l i s h e d for the h u m a n p r o a ( I ) c D N A [ 1 1 15] revealed eight nucleotide differences, three o f which resulted in a c h a n g e in the a m i n o acid (Fig. 2). In Northern blot analysis of total calvarial R N A , the clone p H C A L 2 hybridized to two m R N A s of approx. 4.8 and 5.2 kb, whereas the probe p H C A L 1 U , for procd(I) collagen m R N A , recognized two m R N A s with estimated sizes o f 5.3 kb and 6.5 kb (data not shown).
Fig. 2. Nucleotide sequence of clone pHCAL2. The nucleotide numbering is shown on the left. The numbering of the translated amino acid sequence follows the domains of the proa2(I) collagen molecule: the clone covers the end of the triple helical domain (amino acids 870-1014), the C-telopeptide (amino acids lc-15c), and most of the C-propeptide (amino acids 16c-152c). The amino acids are numbered under the one letter codes. The nucleotides in the previously reported sequences [11-15] which differ from the sequence of pHCAL2 are shown above the sequence, and the amino acid differences below. The sequence between nucleotides 1080 and 1193 has also been reported for the gene from a normal individual [11] and a patient with osteogenesis imperfecta [14]. The four nucleotide sequences are identical, except for a 4-bp deletion (between nucleotides 1124-1129) in the patient DNA. Asterisk (*) at position 429 shows the base, which is mutated in a family of osteogenesis imperfecta type IV variant [15].
174
A
B
.
~ 1.0
I:1
1.o
Discussion
, ~ o.5
0.5
I
I
i
1
I
I
'
i
i
5 time
< ~ 0.5 E
}
10 O
}
~, 1.o
compared with cells which had reached confluency (Fig. 4C and D.
{
,
,
i
10
1
5 time (doys)
;
. . . . . . . . 5 time (doys)
(dQyS)
ttt ttt tt I
1
I
. . . . . . . 5 time {do)s)
10
|
,
,
,
i
i
,
t
i
10
1o! ul
< z 0.5 E
; 0
Fig. 4. Compiled data of the changes in type I collagen mRNA levels and type I collagen secretion into culture medium during different growth phases of five human skin fibroblast cell lines. Individual cell lines were analyzed as shown in Fig. 3. The autoradiograms from hybridizations and gel electrophoreses were quantified by densitometry. Relative units were used with day 8 as the reference data. The mean and standard deviation for each time point are shown. Standard deviation at day 8 was obtained by changing day 7 for the reference day. Panels A and B show relative daily secretion of at and a2 chains, respectively. Panels C and D show the corresponding changes in the cytoplasmic proal(I) and proa2(I) collagen mRNA levels.
Growth-dependent modulation of collagen production and mRNA levels Five different human dermal fibroblast lines were analyzed to determine secretion of cd(I) and ~2(I) chains into culture medium and corresponding m R N A levels during different growth phases. A typical pattern of pepsin-treated culture medium collagens resolved by electrophoresis is shown in Fig. 3B. The secretion of al(I) and a2(I) chains seemed to be almost constant after the fibroblast cultures had reached visual confluency, which occurred at day 5 or 6. All the cell lines showed a similar pattern in the production of al(I) and c~2(I) chains (Fig. 4A and B), the greatest variation occurring just before visual confluency was achieved. To determine the m R N A levels for p r o a l ( I ) and proa2(I) chains in the same cells, slot blot analyses were performed using cytoplasmic R N A samples. Typical hybridization patterns for one of the cell lines are shown in Fig. 3A. A clear increase in p r o a l ( I ) and pro~2(I) m R N A levels was observed during the rapid growth phase. Densitometric analyses revealed approx. 2-fold increases in the levels of both m R N A s when
The present study addresses two questions of technical and biological interest in connective tissue research: (a) how much does the growth phase of fibroblasts in culture affect their collagen production and m R N A levels, and (b) what is the extent of individual variation observed between coding sequences of procollagen genes? Fibroblast cultures have been widely applied to studies on the biochemistry and regulation of collagen production under normal conditions [1,2] and in disease states [3,28]. Several studies have employed skin fibroblasts to determine the effects of various growth factors on collagen production (see Refs. 2 and 30). Such studies have been hampered by the conflicting information of the role of the growth phase of the culture on the rate of collagen production and on the corresponding m R N A levels [4-8]. Here we report oil studies performed on five different human skin fibroblast lines, which were all analyzed daily for 11 days through the various stages of the growth cycle. Each day collagen production into culture medium was determined and procd(I) and proc~(I) collagen m R N A levels were measured in the same cells. Both parameters were calculated per a given number of cells. A constant result of these experiments was an approx. 2-fold increase in collagen production and m R N A levels at the time the cultures reached visual confluency, i.e. when their growth rate began to slow down. This is in agreement with earlier findings on lung fibroblast cultures [4,5], but markedly different from some others [6,8]. The source of such discrepancies is difficult to explain; both the origin of the cell cultures and the experimental conditions could affect the results. In our study we used human skin fibroblasts and routine culture and labeling conditions which are widely used. The results show that careful monitoring of the growth phase of fibroblast cultures and determination of cell numbers at the end of each experiment are important when fibroblast cultures are used to study the rate of collagen production or procollagen m R N A levels. This includes studies on factors affecting the rate of fibroblast proliferation, such as epidermal growth factor which appears to stimulate collagen production due to enhanced proliferation with no direct effect on procollagen m R N A s [31]. The extent of individual variation in the human genome has received considerable attention in respect to sequences which cause restriction fragment length polymorphisms (RFLPs), since these have opened new possibilities for molecular genetics in the form of genetic linkage analysis [32]. Based on systematic studies on individual variation in the /~-globin gene family it has
175 been estimated that two individuals differ from each other approximately once per every 100 bp [33]. Most of this divergence appears in the n o n c o d i n g sequences of the genome. In the globin gene family over 300 variants have been characterized at the amino acid or nucleotide level [34]. However, very little is k n o w n about individual variation in the coding sequences of other h u m a n genes. Variation has been observed in p r o M ( I ) and proet2(I) collagen genes resulting in mutations which cause diseases like osteogenesis imperfecta and Ehlers Danlos syndrome [3,9]. The sequence information available on normal procollagen genes and c D N A s is limited to only a few individuals, and therefore the extent of variation without any detectable harmful effects for the individual has remained unknown. One of the few ways to detect such variations is to sequence c D N A or genomic clones from a n u m b e r of different individuals. Recently, three such nucleotide differences resulting in two amino acid changes in the protx2(I) collagen coding sequences of different individuals were reported [13,35]. In another comparison 17 nucleotide differences were reported between two p r o a 2 ( I ) sequences resulting in five amino acid differences [36]. Comparison of the sequence of p H C A L 2 with those of the previously published sequences [11-15] revealed eight nucleotide differences (in the 1193 bp region of overlap), three of which resulted in an amino acid change (Fig. 2). At this point it is impossible to determine whether all these differences represent true variation. I n f o r m a t i o n is also available of certain 3'-sequences from the corresponding gene [11] and from one osteogenesis imperfecta patient with a 4-bp deletion [14] and another osteogenesis imperfecta family with single amino acid substitution at the end of the triple helix [15]. A p a r t from the deletion and the point mutation, all the four sequences are identical with ours. The fact that in our translated amino acid sequence two of the three differences result in the same amino acid as seen in the corresponding chick sequence [37] might suggest that some of the differences might result from difficulties in sequence reading, or from cloning artifacts. Interestingly, however, comparison of different coding sequences for p r o M ( I ) and p r o M ( I I ) collagens have reveald approximately similar estimates (one nucleotide change per 100-200 bp of c D N A sequence) for individual variation [20,38-40]. I n f o r m a t i o n of individual variation is important for studies related to diseases believed to affect collagen genes. Genetic linkage analyses can now be used to determine with high confidence involvement of a specific collagen gene, but identification of the exact m u t a t i o n in D N A or c D N A from such individuals becomes impractical if the extent of individual variation is not known. O n the other hand, small changes in procollagen coding sequences, analogous to the globin genes, could
result not only in serious disorders, but also in changes in protein structure which only slightly affect the function of the protein. Prockop and Kivirikko [3] have suggested, and recent linkage studies support this [41], that minor mutations in collagen genes predispose to such c o m m o n 'degenerative' disorders as osteoarthrosis and osteoporosis.
Acknowledgements The expert technical assistance of Liisa Peltonen, Tuula Oivanen and Merja Lakkisto is gratefully acknowledged. This study was supported by grants from the Medical Research Council of the A c a d e m y of Finland, and from the Medical Society D u o d e c i m (J.K.M.).
References I Penttinen, R., Frey, H., Aalto, M., Vuorio, E. and Marttala, T. (1980) in Biology of Collagen (Viidik, A. and Vuust, J., eds.), pp. 87-100, Academic Press, London. 2 Bornstein, P. and Sage, H. (1989) Progr. Nucleic Acid Res. Mol. Biol. 37, 67-106. 3 Prockop, D.J. and Kivirikko, K.I. (1984) N. Engl. J. Med. 311, 376-386. 4 Tolstoshev, P., Berg, R.A., Rennard, S.I., Bradley, K.H., Trapnell, B.C. and Crystal, R.G. (1981) J. Biol. Chem. 256, 3135-3140. 5 Miskulin, M., Dalgleish, R., Kluve-Beckerman, B., Rennard, S.I., Tolstoshev, P., Brantly, M. and Crystal, R.G. (1986) Biochemistry 25, 1408-1413. 6 Voss, T. and Bornstein, P. (1986) Eur. J. Biochem. 157, 433-439. 7 Narayanan, A.S. and Page, R.C. (1987) Biochem. Biophys. Res. Commun. 145, 639-645. 8 Aumailley, M., Krieg, T., Razaka, G., Miiller, P.R. and Bricaud, H. (1982) Biochem. J. 206, 505-510. 9 Mayne, R. and Burgeson, R.E. (1987) Structure and Function of Collagen Types, pp. 1-317, Academic Press, Orlando FL. 10 Vuorio, E. and De Crombrugghe, B. (1990) Annu. Rev. Biochem. 59, 837-872. 11 Myers, J.C., Dickson, L.A., De Wet, W.J., Bernard, M.P., Chu, M.-L., Di Liberto, M., Pepe, G., Sangiorgi, F.O. and Ramirez, F. (1983) J. Biol. Chem. 258, 10128-10135. 12 Bernard, M.P., Myers, J.C., Chu, M.-L., Ramirez, F., Eikenberry, E.F. and Prockop, D.J. (1983) Biochemistry 22, 1139-1145. 13 De Wet, W., Bernard, M., Benson-Chandra, V., Chu, M.-L., Dickson, L., Weil, D. and Ramirez, F. (1987) J. Biol. Chem. 262, 16032-1636. 14 Pihlajaniemi, T., Dickson, L.A., Pope, F.M., Korhonen, V.R., Nicholls, A., Prockop, D.J. and Myers, J.C. (1984) J. Biol. Chem. 259, 12941-12944. 15 Wenstrnp, R.J., Cohn, D.H., Cohen, T. and Byers, P.H. (1988) J. Biol. Chem. 263, 7734-7740. 16 White, B.A. and Bancroft, F.C. (1982) J. Biol. Chem. 257, 85698572, 17 Rowe, D.W., Moen, R.C., Davidson, J.M., Byers, P.H., Bornstein, P. and Palmiter, R.D. (1978) Biochemistry 17, 1581-1590. 18 Aviv, H. and Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69, 1408-1412. 19 Thomas, P.S. (1980) Proc. Natl. Acad. Sci. USA 77, 5201-5205. 20 M~ikel~i, J.K., Raassina, M., Virta, A. and Vuorio, E. (1988) Nucleic Acids Res. 16, 349. 21 O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021.
176 22 Pulleyblank, D.E. and Booth, G.H. (1981) J. Biochem. Biophys. Methods 4, 339-346. 23 Elima, K., M~ikela, J.K., Vuorio, T., Kauppinen, S., Knowles, J. and Vuorio, E. (1985) Biochem. J. 229, 183-188. 24 Winberg, G. and HammarskjNd, M.-C. (1980) Nucleic Acids Res. 8, 253-264. 25 Mandel, M. and Higa, A. (1970) J. Mol. Biol. 53, 159-162. 26 M/ikel~i, J.K. and Vuorio, E. (1986) Med. Biol. 64, 15-22. 27 Sanger, F., Nicklen, S. and Coulson, E.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. 28 Vuorio, T., M~ikel~i, J.K., K~ih~iri, V-M. and Vuorio, E. (1987) Arch. Dermatol. Res. 279, 154-160. 29 Sandberg, M., M~ikela, J.K., Multim~iki, P., Vuorio, T. and Vuorio, E. (1989) Matrix 9, 82-91. 30 Sporn, M.B. and Roberts, A.B. (1986) J. Clin. Invest. 78, 329-332. 31 Laato, M., K~.hari, V.-M., Niinikoski, J. and Vuorio, E. (1987) Biochem. J. 247, 385-388. 32 Gusella, J.F. (1986) Annu. Rev. Biochem. 55, 831-854.
33 Jeffreys, A.J. (1979) Cell 18, 1-10. 34 Collins, F.S. and Weissman, S.M. (1984) Progr. Nucleic Acid Res. Mol. Biol. 31, 315 462. 35 Kuivaniemi, H., Tromp, G., Chu, M.-L. and Prockop, D.J. (1988) Biochem. J. 252, 633-640. 36 Lee, S.-T., Smith, B.D. and Greenspan, D.S. (1988) J. Biol. Chem. 263, 13414-13418. 37 Fuller, F. and Boedtker, H. (1981) Biochemistry 20, 996-1006. 38 Elima, K., Vuorio. T. and Vuorio, E. (1987) Nucleic Acids Res. 15. 9499-9504. 39 Baldwin, C.T., Reginato, A.M., Smith, C., Jimenez, S.A. and Prockop, D.J. (1989) Biochem. J. 262, 521-528. 40 Su, M.-W., Lee, B., Ramirez, F., Machado, M. and Horton, W. (1989) Nucleic Acids Res. 17, 9473. 41 Palotie, A., V~iis~inen, P., Ott, J., Ryhauen, L., Elima, K., Vikkula, M., Cheah, K., Vuorio, E. and Peltonen, k. (1989) Lancet it 924-927.
