- Email: [email protected]
In vitro biosynthesis of type I and III collagens by human dermal fibroblasts from donors of increasing age
amj Mechanisms of Ageing and Development 73 (1994) 179-187
ELSEVIER SCIENCE IRELAND
In vitro biosynthesis of type I and III collagens human dermal fibroblasts from donors of increasing age
by
Marc Dumas, Catherine Chaudagne, Fr6d6ric Bont6*, Alain Meybeck L V M H Recherche, 50, rue de Seine, 92703 Colombes, France
(Received 4 June 1993; revision received 28 November 1993; 9 December 1993)
Abstract A quantitative study of type I and type Ill collagen production was carried out on primary cultures of human dermal fibroblasts. Cultures were initiated from facial and mammary skin of 29 women aged between 19 and 68 years. Secreted and cell-associated collagen levels were determined by an enzyme linked immunosorbent assay (ELISA). We found that the secretion of type I and type III collagen decreased linearly with age (r = 0.432; P = 0.0193 and r = 0.502; P = 0.0147, respectively). There was a 29% loss in secretion ability for type I and type III collagen over the 49-year period studied. Furthermore, no significant linear age-related decrease was observed for type I and type III collagen associated with the cellular fraction. The influence of body site was also analysed. We observed a significant linear age-related decrease in type I collagen secretion by mammary skin cells (P = 0.0183 and r = 0.618) as well as facial skin cells (P = 0.0037 and r = 0.699). Furthermore, only mammary skin fibroblasts showed a significant linear age-related decrease in secreted type III collagen (P = 0.106 and r = 0.513). No age-related variations in cell-associated collagen were found. Key words: Collagens I and IlI; Human skin fibroblasts; Aging; Body site variations
1. Introduction Age-related variations in the appearance a n d mechanical properties of the skin are mainly due to modifications of the dermal extracellular matrix in which collagen * Corresponding author. Abbreviations." EDTA, ethylenediaminetetraacetate; ELISA, enzyme-linkedimmunosorbentassay; NEM,
N-ethylmaleimide;PBS, phosphate-buffered saline; PMSF, phenylmethylsulphonylfluoride. 0047-6374/94/$07.00 © 1994 Elsevier Science Ireland Ltd. All rights reserved. SSDI 0047-6374(94)01426-M
180
M. Dumas et aL /Mech. Ageing Dev. 73 (1994) 179-187
represents about 70% of the dry weight of skin [1]. Type I collagen predominates, accounting for 80-90% of the total collagen, the remaining 10-15°/,, consisting mainly of type III [2]. Both types of collagen are closely associated in a well organized fibrillar structure [3,4]. In the dermis, fibroblasts play a central role in collagen processing by synthesizing and degrading [5,6] as well as reorganizing [7] the collagen matrix; these cells are therefore essential to the study of skin aging. Quantitative changes in these two collagen types during natural or experimental skin aging have been previously reported in human skin in various biological situations: natural aging in vivo [8] or experimental aging in vitro using subcultured cells [9,10], presence or absence of ultraviolet light exposure leading respectively to accelerated or intrinsic aging [11,12]. In the present study we determined the levels of type I and III collagen production by human dermal fibroblasts in primary culture. Skin cells from two body sites were used: the breast, an area protected from the sun and the face, a more sun-exposed area. Age-related variations in the levels of secreted and intracellular or 'cell associated' type I and III collagens were evaluated for all fibroblast strains and then analysed according to body site origin. 2. Materials and methods
2.1. Fibroblast cultures Normal adult dermal fibroblast cultures were established using the explant method with skin samples obtained from plastic surgery [13]. Samples of sun protected mammary (junctional area between breast and thorax) and more exposed periauricular facial skin were collected, respectively from 12 women aged 19-63 years and 17 women aged 40-68 years. Primary fibroblasts cultures were grown to confluence in E199 medium (Gibco) supplemented with 2 mM L-glutamine (Gibco) and 10% fetal calf serum (Gibco) at 37°C in a 5% CO2 humidified atmosphere. For collagen biosynthesis experiments, these stock cultures were harvested with PBS (pH 7.2) containing 0.1% (w/v) trypsin, 0.02% EDTA. The cells were seeded in 96-well microculture plates at a density of l 0 4 fibroblasts/well in the culture medium. Twenty-four hours later, the medium was replaced by.serum-free medium supplemented with 0.15 mM L-ascorbic acid (sodium salt, Sigma) prepared extemporaneously [14]. After an additional 24 h of incubation, collagen levels were quantified. 2.2. Collagen assay The amounts of type I and type III collagens were determined by an enzymelinked immunosorbent assay [15]. For each strain, incubation media containing secreted collagen were collected and the remaining cells were homogenized by sonication in ice to determine cell-associated collagen concentrations. Both were transferred into wells of a plastic immunoplate (Nunc) containing protease inhibitors (disodium EDTA, NEM, PMSF 1 mM each, 0.01% (w/v) sodium azide, all purchased from Sigma). Type I and type III collagens were quantified in separate plates, using six wells for each assay.
M. Dumas et al. / Mech. Ageing Dev. 73 (1994) 179-187
181
Collagen coating was carried out by incubating the plates for 24 h at 4°C. Wells were then rinsed with PBS. Non-specific binding to the plastic plates was avoided by incubating the wells with 3% (w/v) bovine serum albumin in PBS for 24 h at 4°C. Rabbit antibodies to human type I and type III collagens (Institut Pasteur, Lyon, France) were added to each well (dilution 1:500 for each antibody). After incubation for 1.5 h at 22°C, the plates were washed and the bound antibodies reacted in identical conditions with alkaline phosphatase conjugated goat-antirabbit IgG (Immunotech, Marseille, France), dilution 1:1000. After washing, a freshly prepared solution ofp-nitrophenyl phosphate (pH 9.8) (Sigma) in PBS was added to each well. The plates were incubated in the dark for 30 min at 22°C and the absorbance of the p-nitrophenol formed was measured at 405 nm. Optical densities were converted into nanograms of collagen using standard curves established with purified human type I and III collagens (Institut Jacques Boy, Reims, France). Regression analysis shows that linearity is conserved up to 1000 ng collagen/well (r = 0.9984 and r = 0.9981 for collagen type I and III, respectively). The levels of secreted and cell-associated collagen were expressed in ng/104 fibroblasts per 24 h. Cell numbers were determined in order to verify that the cellular density at the time of the collagen assay was not modified with respect to either donor age or body site origin. 2.3. Assay specificity
Rabbit antibodies to human type I and III collagens were introduced into wells coated with human purified collagen III and I, respectively, in the above assay conditions. No cross-reactivity was observed. No modification of the estimated amount of type I and III collagen was observed when the collagen types were used alone or in combination. 2.4. Statistical analysis
For each strain, each collagen assay was performed on six independent wells. The results are expressed as the means with a standard deviation of 5% for type I and 10% for type III collagen. The results were analysed by linear regression and by analysis of variance using age as the independent variable and type I and III collagen levels as the dependent variable. In each case the equation of the line and the r and P values were determined. 3. Results
We first investigated collagen secretion and synthesis for all fibroblast strains in relation to donor age. After a 24-h cell culture period, secretion of type I and III collagens was detected in all fibroblast strains. For the 19-68-year donor-age interval, we observed a significant linear age-related decrease in the secretion of the two collagen types (Figs. l a,b) with respective P values of 0.0193 and 0.0147 and r values of 0.432 and 0.502. The equation of the regression line indicates that over the 49-year period studied, collagen secretion decreased by 32% for both collagen types (Figs. la,b). For type
M. Dumas et al./Mech. Ageing Dev. 73 (1994) 179-187
182 TYPEICOLLAGEN
TYPE IIl COLLAGEN
900
9O0 []
O
800-
800-
700-
700o
600-
O
D DD
D ~
~
o
600-
D
500-
500-
400O
5
400-
QD O
300-
300-
200100-
[]
200-100_
a
b
0
0 I0
2o
3o
' 40
' 50
' 60
70
[] D
10 10
DONORAGE (YEARS)
D
DD
[]
I
I
I
I
30
40
50
60
70
DONOR AGE (YEARS)
Fig. 1. Age-related variations in the secretion of type I (a) and type Ill (b) collagen by primary cultures of human dermal fibroblasts from facial and breast skin. For type I collagen the equation of the regression line is y = 7 6 3 - 4 . 5 x with r = 0 . 4 3 2 and P = 0 . 0 1 9 3 , and for type III collagen y = 1 9 1 - 1.0x with r = 0.502 and P = 0.0147.
I collagen, this corresponds to a decrease from 677 ng/104cells/24 h for cells of an interpolated 19-year-old donor to 457 ng/104cells/24 h for those of a 68-year-old donor (Fig. la) and for type III collagen from 172 ng/104cells/24 h to 123 ng/104cells/24 h (Fig. lb). For cell-associated type I and III collagen, no age-related variations were observed (Figs. 2a,b). Over the 49-year period studied the cell-associated type collagen levels CELL-ASSOCIATED TYPE III COLLAGEN
CELL-ASSOCIATED TYPE I COLLAGEN 9OO
900-
800-
800-
700-
700-
600-
600-
500-
500-
'-r
J m t..)
400O
300O O
8
[3 O
z
O ~ O EO
DD
[3
n
O
[]
U
200-
200100-
D O
300--
D_ ~0 []
400-
100-
8
b
0
0 10
. 20
. 30. . 40.
50
60
DONOR AGE (YEARS)
7'0
8()
20
30
' 40
' 50
' 60
;o
8(1
DONORAGE (YEARS)
Fig. 2. Age-related variations in the cell-associated type I (a) and type IIl (b) collagen by primary cultures of human dermal fibroblast from facial and breast skin. For type I collagen equation of the regression line is y = 271 - 0.Ix with r = 0.045 and P = 0.8468, and for type Ill collagen y = 361 - 0.4x with r = 0.221 and P -- 0.3792.
183
M. Dumas et al./ Mech. Ageing Dev. 73 (1994) 179-187
were all within the interval o f 2 6 9 - 2 6 4 ng/104cells/24 h, corresponding, respectively to interpolated values o f a 19- and 68-year-old donor (Fig. 2a). For type III collagen this interval was 3 5 3 - 3 3 4 ng/104cells/24 h (Fig. 2b). It was then interesting to perform an analysis of body site origin on the individual values. Mammary skin fibroblasts (Fig. 3a) exhibited a significant linear age-related decrease in type I collagen secretion over the 44 years studied (P = 0.0324, r = 0 . 6 1 8 ) . We observed a 39% decrease from 682 ng/104cells/24 h to 418 ng/104cells/24 h for cells o f interpolated 19- and 63-year-old donors, respectively. For secreted type III collagen (Fig. 3a), there was a non-significant decrease (13%) from 166 ng/104ceils/24 h to 144 ng/104cells/24 h (P = 0.376 and r = 0.281). Using these interpolated values, the ratio o f type III to type I collagen changes from 0.24 to 0.34 for cells o f a 19- and 63-year-old donor, respectively. For the facial skin (Fig. 3b), we observed a significant linear age-related decrease in the secretion o f type I collagen over the 28 years studied (P = 0.0037 and r = 0.699). A 39% loss in the ability o f fibroblasts to secrete this collagen type (from 707 ng/104cells/24 h for the 40-year-old donor to 434 ng/104 cells/24 h for the 68year-old donor) was detected. During the same 28-year period, the decrease in type III collagen secretion, was 27%, from 147 ng/104 cells/24 h to 108 ng/104cells/24 h. The linear regression plot was not statistically significant (P = 0.106 and r = 0.513). The ratio of type III to type I collagen calculated from the interpolated values was increased slightly from 0.21 to 0.25.
BREAST SKIN CELLS
FACIALSKINCELLS 900
o T~
D
o'l~
Q 800-
mT~
NT~
,..e
~700-
700-
Q
D
,.2
600-
o
o
O
o
oo
o
500oo 400-
~ 300-
300-
8 200-
o
200.m
1000 30
• lo0-
a I 40
•
•
I 50
DONOR AGE (YEARS)
I 60
0 70
m • --
mm
b I
]O
mm m
20
0 3
i 40
i 50
i 60
70
DONOR AGE (YEARS)
Fig. 3. Body site age-related variations in the secretion of type I (1:1)and type III (1) collagens by primary cultures of human dermal fibroblasts from face (a) and breast (b). Equations of the regression lines are the following: for facial skin cells, y = 1094 - 9.7x with r = 0.699 and P = 0.0037 for type I collagen and y = 203 - 1.4x with r = 0.513 and P = 0.106 for type Ili collagen; for breast skin cells, y = 796 - 6.0x with r = 0.618 and P = 0.032 for type I collagen and y = 176 - 0.5x with r = 0.281 and P = 0.376 for type II! collagen.
M. Dumas et a l . / Mech. Ageing Dev. 73 (1994) 179-187
184
CELL-ASSOCIATED COLLAGEN IN FACIAL SKIN CELLS
CELL-ASSOCIATED COLLAGEN IN BREAST SKIN CELLS 900
800-
800-
700-
700-
600-
600-
500-
500-
400-
400O
300-
30013
8
200100-
a
0
100I
4O
O
50
I
60 DONOR AGE (YEARS)
7'0
80
u
O
D
[]
O
200-
b
0 10
20
30
I
I
40
50
DONOR AGE WEARS)
Fig. 4. Body site age-related variation o f cell-associated type 1 (l~) and type III ( I ) collagen in primary cultures o f h u m a n dermal fibroblasts from face (a) and breast (b). Equations o f the regression lines are as follows: for facial skin cells, y = 3 4 7 - I . l x with r = 0.31 3 and P = 0 . 3 4 9 3 for type I collagen, y = 382 - 0.8x with r = 0.326 and P = 0.3924; for breast skin cells, y = 271 - 0.1x with r = 0.052 and P = 0.8873 for type I collagen, y = 351 - 0.2x with r = 0.07 and P = 0.8570 for type III collagen.
Furthermore, as for the global analysis, we confirmed that there were no significant variations in cell-associated collagens for mammary or facial skin cells analysed separately (Figs. 4a,b). 4. Discussion In the present study, we investigated collagen biosynthesis in human dermal fibroblasts, cells known to play a central role in extracellular matrix synthesis and in aging processes [1]. In order to avoid the influence of in vitro aging due to subculturing [9,10], primary cultures of fibroblasts from 19- to 68-year-old caucasian women were used. To assess the incidence of environmental factors on aging, fibroblast cultures were initiated from facial and mammary skin, sun-exposed and sunprotected body sites, respectively. Type I and type III collagen secretion by the fibroblasts as well as cell-associated levels were measured. For the population studied, we observed an overall significant linear agedependent decrease of type I and III collagen secretion by fibroblasts. Since type I and type III collagen represent the most abundant collagen types of the dermis, the age-related decrease in both secreted collagen types by fibroblast cultures agree with studies showing in different species a progressive decrease of global skin collagen content with age [16-18]. Moreover, this could partially explain the thinning of dermal collagen fibers composed mainly of type I and type III [19], the reduction of dermal thickness [18,20] and loss of morphometric properties of this tissue [21] that occur with age.
M. Dumas et al. / Mech. Ageing Dev. 73 (1994) 179-187
185
Recently, an age-dependent decrease in expression of the genes coding for type I and type III collagen in cultured dermal fibroblasts from skin donors 16-74 years of age has been reported [22]. This could suggest that the decrease we observed in secreted type I and type III collagen is the result of a lower expression of the corresponding genes. The body site analysis confirms the significant decrease in secretion of type collagen by mammary and facial skin cells. Moreover, the negative slope obtained for secreted type III collagen by facial skin cells, significant at P = 0.106, was not detected at the same significance level for mammary skin cells. Furthermore, during the same time period, the age-related decrease in type collagen secretion was greater for the face than for the breast. This effect was even more pronounced for type III collagen. This lead us to consider that sun exposure may affect mainly the ability of fibroblasts to secrete type III collagen. We found that the type III/I collagen ratio for mammary skin cells increased much greater with age (0.24-0.36) than for facial skin cells where the increase was lower (0.21-0.25), but on a shorter period of time, i.e. 28 years, than for mammary skin (44 years). A parallel can be established between these results with skin cells and the increase in the relative proportion of type III collagen determined in vivo in the abdominal skin of older donors [23]. The pliation properties of various tissues have thought to be linked to the relative amount of type III/I collagen [24]. If such a relationship does exist, the increase in the proportion of type III collagen secreted by fibroblasts during aging could be related to an increased flexibility of the dermal extracellular matrix. According to our results, this phenomenon would be attenuated in the sun-exposed skin of the face, where the increase in the ratio of secreted type III to type I collagen is smaller than in sun-protected mammary skin. In contrast, in vivo assays in hairless mouse skin show an increased type Ill/type I + III collagen ratio after UV exposure [25]. This apparent discrepancy with our results might be explained by the fact that the literature data probably involve UV cross-linking of collagen [26] which would modify extractable collagen, as previously described [27], in addition to the fact that the species are different. Our findings describe collagen secretion ability in skin isolated from a sun-exposed tissue, which is quite different and could suggest a gradual age-related decrease in skin tonicity, as previously reported [281. For cell-associated type I and type III collagen, we did not detect any significant age-variations. If we consider these pools as mainly intracellular, these findings indicate that this internal ratio is not modified during aging. Even if greater amounts of type I and type III collagen are synthesized and then released into the medium by the cells of younger as compared with older donors, the level of these intracellular collagens is maintained. This is in accordance with the hypothesis of cellular control of the internal collagen pool [29], which apparently is established with approximately equivalent amounts of type I and type III collagen. Our study has shown, by quantitative ELISA methodology, that human skin fibroblasts from older donors have a reduced ability to produce type I and type III collagen in vitro. In the same conditions, no significant age-related modifications were observed for cell-associated pools of type I and type III collagen. These agerelated modifications, observed at two different body sites, suggest that aging does
186
M. Dumas et al./Mech. Ageing Dev. 73 (1994) 179-187
not only affect fibroblasts from more sun-exposed skin such as facial skin, but rather that other, less sun-exposed body sites such as mammary skin may also undergo agerelated decreases in type III collagen secretion, albeit to a lesser extent. All these observations confirm that fibroblasts, through their age-related decrease in type I and type III collagen production ability, are central cells in dermis aging. 5. References 1 C.M. Lapiere, The ageing dermis: the main cause for the appearance of"old" skin. Br. J. Dermatol.. 122 (1990) 5-11. 2 E.J. Kuchartz, The Collagens: Biochemistry and Pathophysiology, Springer Verlag, 1992, pp. 227-263. 3 L. Weber, E. Kirsch, P. Muller and T. Krieg, Collagen type distribution and macromolecular organization of connective tissue in different layers of human skin. J. Invest. Dermatol., 82 (1984) 156-160. 4 R. Fleischmajer, J.S. Perlish, R.E. Burgeson, F. Shaikh-Bahai and R. Timpl, Type i and type Ill collagen interactions during fibrillogenesis. Ann. N.Y. Acad. Sci.. 580 (1990) 161-175. 5 L. Weber, C. Mauch, E. Hirsch, P.K. Muller and T. Krieg, Modulation of collagen type synthesis in organ and cell culture of fibroblasts. J. Invest. Dermatol., 87 (1986) 217-220. 6 S.M. Krane, Collagenases and collagen degradation. J. Invest. Dermatol., 79 (1982) 83s-86s. 7 P. Delvoye, P. Wiliquet, J.L. Leveque, B.U. Nusgens and C.M. Lapiere, Measurement of mechanical forces generated by skin fibroblasts embedded in a three dimensional collagen gel. J. Invest. Dermatol., 97 (1991) 898-902. 8 R.H. Pearce and B.J. Grimmer, Age and the chemical constitution of normal human dermis. J. Invest. Dermatol., 58 (1972) 347-365. 9 J.C. Houck, V.K. Sharma and L. Hayflick, Functional failures of cultured diploid fibroblasts after continued population doublings. Proc. Soc. Exp. Biol. Med., 137 (1971) 331-333. 10 J.N. Hildebran, M. Absher and R.B. Low, Altered rates of collagen synthesis in in vitro aged human lung fibroblasts. In Vitro, 19 (1983) 307-314. 11 L.H. Kligrnan, Photoaging. Manifestations, prevention and treatment. Clin. Geriatr. Med., 5 (1989) 235-251. 12 A. Oikarinen and M. Kallioinen, A biochemical and immunohistochemical study of collagen in sunexposed and protected skin. Photodermatology, 6 (1989) 24-31. 13 R.I. Freshney, Culture o f Animal Cells; A Manual o f Basic Technique, A.R. Liss, New York, 1983, pp. 104-106. 14 B.A. Booth, K.L. Polak and J. Uitto, Collagen biosynthesis by human skin fibroblasts. 1. Optimization of the culture conditions for the synthesis of type I and !I i procollagens. Bioehim. Biophys. A cta, 96 (1980) 145-160. 15 S.I. Rennard, G.R. Martin and R.G. Crystal, Enzyme-linked immunoassay (EL1SA) for connective tissue proteins: Type I collagen. In H. Furthmayr (ed.), lmmunochemistry o f the Extracellular Matrix, Vol. 1, CRC Press, Boca Raton, 1982, pp. 237-252. 16 P.K. Mays, J.E. Bishop and G.J. Laurent, Age-related changes in the proportion of types 1 and 111 collagen. Mech. Ageing Dev., 45 (1988) 203-212. 17 B. Boyer, P. Kern, A. Fourtanier and J. Labat-Robert, Age dependent variations of the biosynthesis of fibronectin and fibrous collagen in mouse skin. Exp. GerontoL, 26 (1991) 375-383. 18 S. Shuster, M.M. Black and E. Mc Vitie, The influence of age and sex on skin thickness, skin collagen and density. Br. J. Dermatol., 93 (1975) 639-643. 19 L.V. Zuccarello, R. Garbelli and V.D.P. Rossi, lmmunocytochemical localisation of collagen types I, III, IV and fibronectin in the human dermis. Cell Tissue Res., 268 (1992) 505-511. 20 M.C. Branchet, S. Boisnic, C. Frances and A.M. Robert, Skin thickness changes in normal aging skin. Gerontology, 36 (1990) 28-35. 21 M.C. Branchet, S. Boisnic, C. Frances, C. Lesty and L. Robert, Morphometric analysis of dermal collagen fibers in normal human skin as a function of age. Arch. Gerontol. Geriatr., 13 (1991) 1-14. 22 Y.Q. Chen, A. Mauviel, E.M.L. Tan and J. Uitto, Age-related changes in the expression of extracellular matrix genes in human dermal fibroblasts in culture. J. Invest. Dermatol., 100 (1993) 535.
M. Dumas et al. / Mech. Ageing Dev. 73 (1994) 179-187
23
24 25 26 27
28 29
187
C.R. Lovell, K.A. Smolenski, V.C. Duance, N.D. Light, S. Young and M. Dyson, Type I and 111 collagen content and fiber distribution in normal human skin during ageing. Br. J. DermatoL, 117 (1987) 419-428. A.J. Hance and R.G. Crystal, Rigid control of synthesis of collagen types I and III by cells in culture. Nature, 268 (1977) 152-154. A. Fourtanier, Collagen metabolism in ultraviolet irradiated hairless mouse skin. J. Invest. Dermatol., 96 (1991) 994. C.M. Dalle and M.A. Pathok, Photoaging and the role of UV (290-400 nm) and reactive oxygen species in collagen cross-links. J. Invest. DermatoL, 92 (1989) 410. J. Brinkmann, M. Brey and P.K. Muller, age differences of human skin collagen. In A. Bernd, J. Bereiter-Hahn, F. Hevert and H. Holzmann (eds.), Cell and Tissue Culture Models in Dermatological Research, Springer-Verlag, Berlin, Heidelberg, 1993, pp. 278-282. G.E. Pierrard, A critical approach to in vivo mechanical testing of the skin. In J.L. Leveque (ed.), Cutaneous Investigation in Health and Disease. M. Dekker, 1989, pp. 215-240. R.I. Schwarz, R.B. Mandell and M.J. Bissell, Ascorbate induction of collagen synthesis as a mean for quantitative control of tissue specific function. Mol. Cell. Biol., 1 (1981) 843-853.
ELSEVIER SCIENCE IRELAND
In vitro biosynthesis of type I and III collagens human dermal fibroblasts from donors of increasing age
by
Marc Dumas, Catherine Chaudagne, Fr6d6ric Bont6*, Alain Meybeck L V M H Recherche, 50, rue de Seine, 92703 Colombes, France
(Received 4 June 1993; revision received 28 November 1993; 9 December 1993)
Abstract A quantitative study of type I and type Ill collagen production was carried out on primary cultures of human dermal fibroblasts. Cultures were initiated from facial and mammary skin of 29 women aged between 19 and 68 years. Secreted and cell-associated collagen levels were determined by an enzyme linked immunosorbent assay (ELISA). We found that the secretion of type I and type III collagen decreased linearly with age (r = 0.432; P = 0.0193 and r = 0.502; P = 0.0147, respectively). There was a 29% loss in secretion ability for type I and type III collagen over the 49-year period studied. Furthermore, no significant linear age-related decrease was observed for type I and type III collagen associated with the cellular fraction. The influence of body site was also analysed. We observed a significant linear age-related decrease in type I collagen secretion by mammary skin cells (P = 0.0183 and r = 0.618) as well as facial skin cells (P = 0.0037 and r = 0.699). Furthermore, only mammary skin fibroblasts showed a significant linear age-related decrease in secreted type III collagen (P = 0.106 and r = 0.513). No age-related variations in cell-associated collagen were found. Key words: Collagens I and IlI; Human skin fibroblasts; Aging; Body site variations
1. Introduction Age-related variations in the appearance a n d mechanical properties of the skin are mainly due to modifications of the dermal extracellular matrix in which collagen * Corresponding author. Abbreviations." EDTA, ethylenediaminetetraacetate; ELISA, enzyme-linkedimmunosorbentassay; NEM,
N-ethylmaleimide;PBS, phosphate-buffered saline; PMSF, phenylmethylsulphonylfluoride. 0047-6374/94/$07.00 © 1994 Elsevier Science Ireland Ltd. All rights reserved. SSDI 0047-6374(94)01426-M
180
M. Dumas et aL /Mech. Ageing Dev. 73 (1994) 179-187
represents about 70% of the dry weight of skin [1]. Type I collagen predominates, accounting for 80-90% of the total collagen, the remaining 10-15°/,, consisting mainly of type III [2]. Both types of collagen are closely associated in a well organized fibrillar structure [3,4]. In the dermis, fibroblasts play a central role in collagen processing by synthesizing and degrading [5,6] as well as reorganizing [7] the collagen matrix; these cells are therefore essential to the study of skin aging. Quantitative changes in these two collagen types during natural or experimental skin aging have been previously reported in human skin in various biological situations: natural aging in vivo [8] or experimental aging in vitro using subcultured cells [9,10], presence or absence of ultraviolet light exposure leading respectively to accelerated or intrinsic aging [11,12]. In the present study we determined the levels of type I and III collagen production by human dermal fibroblasts in primary culture. Skin cells from two body sites were used: the breast, an area protected from the sun and the face, a more sun-exposed area. Age-related variations in the levels of secreted and intracellular or 'cell associated' type I and III collagens were evaluated for all fibroblast strains and then analysed according to body site origin. 2. Materials and methods
2.1. Fibroblast cultures Normal adult dermal fibroblast cultures were established using the explant method with skin samples obtained from plastic surgery [13]. Samples of sun protected mammary (junctional area between breast and thorax) and more exposed periauricular facial skin were collected, respectively from 12 women aged 19-63 years and 17 women aged 40-68 years. Primary fibroblasts cultures were grown to confluence in E199 medium (Gibco) supplemented with 2 mM L-glutamine (Gibco) and 10% fetal calf serum (Gibco) at 37°C in a 5% CO2 humidified atmosphere. For collagen biosynthesis experiments, these stock cultures were harvested with PBS (pH 7.2) containing 0.1% (w/v) trypsin, 0.02% EDTA. The cells were seeded in 96-well microculture plates at a density of l 0 4 fibroblasts/well in the culture medium. Twenty-four hours later, the medium was replaced by.serum-free medium supplemented with 0.15 mM L-ascorbic acid (sodium salt, Sigma) prepared extemporaneously [14]. After an additional 24 h of incubation, collagen levels were quantified. 2.2. Collagen assay The amounts of type I and type III collagens were determined by an enzymelinked immunosorbent assay [15]. For each strain, incubation media containing secreted collagen were collected and the remaining cells were homogenized by sonication in ice to determine cell-associated collagen concentrations. Both were transferred into wells of a plastic immunoplate (Nunc) containing protease inhibitors (disodium EDTA, NEM, PMSF 1 mM each, 0.01% (w/v) sodium azide, all purchased from Sigma). Type I and type III collagens were quantified in separate plates, using six wells for each assay.
M. Dumas et al. / Mech. Ageing Dev. 73 (1994) 179-187
181
Collagen coating was carried out by incubating the plates for 24 h at 4°C. Wells were then rinsed with PBS. Non-specific binding to the plastic plates was avoided by incubating the wells with 3% (w/v) bovine serum albumin in PBS for 24 h at 4°C. Rabbit antibodies to human type I and type III collagens (Institut Pasteur, Lyon, France) were added to each well (dilution 1:500 for each antibody). After incubation for 1.5 h at 22°C, the plates were washed and the bound antibodies reacted in identical conditions with alkaline phosphatase conjugated goat-antirabbit IgG (Immunotech, Marseille, France), dilution 1:1000. After washing, a freshly prepared solution ofp-nitrophenyl phosphate (pH 9.8) (Sigma) in PBS was added to each well. The plates were incubated in the dark for 30 min at 22°C and the absorbance of the p-nitrophenol formed was measured at 405 nm. Optical densities were converted into nanograms of collagen using standard curves established with purified human type I and III collagens (Institut Jacques Boy, Reims, France). Regression analysis shows that linearity is conserved up to 1000 ng collagen/well (r = 0.9984 and r = 0.9981 for collagen type I and III, respectively). The levels of secreted and cell-associated collagen were expressed in ng/104 fibroblasts per 24 h. Cell numbers were determined in order to verify that the cellular density at the time of the collagen assay was not modified with respect to either donor age or body site origin. 2.3. Assay specificity
Rabbit antibodies to human type I and III collagens were introduced into wells coated with human purified collagen III and I, respectively, in the above assay conditions. No cross-reactivity was observed. No modification of the estimated amount of type I and III collagen was observed when the collagen types were used alone or in combination. 2.4. Statistical analysis
For each strain, each collagen assay was performed on six independent wells. The results are expressed as the means with a standard deviation of 5% for type I and 10% for type III collagen. The results were analysed by linear regression and by analysis of variance using age as the independent variable and type I and III collagen levels as the dependent variable. In each case the equation of the line and the r and P values were determined. 3. Results
We first investigated collagen secretion and synthesis for all fibroblast strains in relation to donor age. After a 24-h cell culture period, secretion of type I and III collagens was detected in all fibroblast strains. For the 19-68-year donor-age interval, we observed a significant linear age-related decrease in the secretion of the two collagen types (Figs. l a,b) with respective P values of 0.0193 and 0.0147 and r values of 0.432 and 0.502. The equation of the regression line indicates that over the 49-year period studied, collagen secretion decreased by 32% for both collagen types (Figs. la,b). For type
M. Dumas et al./Mech. Ageing Dev. 73 (1994) 179-187
182 TYPEICOLLAGEN
TYPE IIl COLLAGEN
900
9O0 []
O
800-
800-
700-
700o
600-
O
D DD
D ~
~
o
600-
D
500-
500-
400O
5
400-
QD O
300-
300-
200100-
[]
200-100_
a
b
0
0 I0
2o
3o
' 40
' 50
' 60
70
[] D
10 10
DONORAGE (YEARS)
D
DD
[]
I
I
I
I
30
40
50
60
70
DONOR AGE (YEARS)
Fig. 1. Age-related variations in the secretion of type I (a) and type Ill (b) collagen by primary cultures of human dermal fibroblasts from facial and breast skin. For type I collagen the equation of the regression line is y = 7 6 3 - 4 . 5 x with r = 0 . 4 3 2 and P = 0 . 0 1 9 3 , and for type III collagen y = 1 9 1 - 1.0x with r = 0.502 and P = 0.0147.
I collagen, this corresponds to a decrease from 677 ng/104cells/24 h for cells of an interpolated 19-year-old donor to 457 ng/104cells/24 h for those of a 68-year-old donor (Fig. la) and for type III collagen from 172 ng/104cells/24 h to 123 ng/104cells/24 h (Fig. lb). For cell-associated type I and III collagen, no age-related variations were observed (Figs. 2a,b). Over the 49-year period studied the cell-associated type collagen levels CELL-ASSOCIATED TYPE III COLLAGEN
CELL-ASSOCIATED TYPE I COLLAGEN 9OO
900-
800-
800-
700-
700-
600-
600-
500-
500-
'-r
J m t..)
400O
300O O
8
[3 O
z
O ~ O EO
DD
[3
n
O
[]
U
200-
200100-
D O
300--
D_ ~0 []
400-
100-
8
b
0
0 10
. 20
. 30. . 40.
50
60
DONOR AGE (YEARS)
7'0
8()
20
30
' 40
' 50
' 60
;o
8(1
DONORAGE (YEARS)
Fig. 2. Age-related variations in the cell-associated type I (a) and type IIl (b) collagen by primary cultures of human dermal fibroblast from facial and breast skin. For type I collagen equation of the regression line is y = 271 - 0.Ix with r = 0.045 and P = 0.8468, and for type Ill collagen y = 361 - 0.4x with r = 0.221 and P -- 0.3792.
183
M. Dumas et al./ Mech. Ageing Dev. 73 (1994) 179-187
were all within the interval o f 2 6 9 - 2 6 4 ng/104cells/24 h, corresponding, respectively to interpolated values o f a 19- and 68-year-old donor (Fig. 2a). For type III collagen this interval was 3 5 3 - 3 3 4 ng/104cells/24 h (Fig. 2b). It was then interesting to perform an analysis of body site origin on the individual values. Mammary skin fibroblasts (Fig. 3a) exhibited a significant linear age-related decrease in type I collagen secretion over the 44 years studied (P = 0.0324, r = 0 . 6 1 8 ) . We observed a 39% decrease from 682 ng/104cells/24 h to 418 ng/104cells/24 h for cells o f interpolated 19- and 63-year-old donors, respectively. For secreted type III collagen (Fig. 3a), there was a non-significant decrease (13%) from 166 ng/104ceils/24 h to 144 ng/104cells/24 h (P = 0.376 and r = 0.281). Using these interpolated values, the ratio o f type III to type I collagen changes from 0.24 to 0.34 for cells o f a 19- and 63-year-old donor, respectively. For the facial skin (Fig. 3b), we observed a significant linear age-related decrease in the secretion o f type I collagen over the 28 years studied (P = 0.0037 and r = 0.699). A 39% loss in the ability o f fibroblasts to secrete this collagen type (from 707 ng/104cells/24 h for the 40-year-old donor to 434 ng/104 cells/24 h for the 68year-old donor) was detected. During the same 28-year period, the decrease in type III collagen secretion, was 27%, from 147 ng/104 cells/24 h to 108 ng/104cells/24 h. The linear regression plot was not statistically significant (P = 0.106 and r = 0.513). The ratio of type III to type I collagen calculated from the interpolated values was increased slightly from 0.21 to 0.25.
BREAST SKIN CELLS
FACIALSKINCELLS 900
o T~
D
o'l~
Q 800-
mT~
NT~
,..e
~700-
700-
Q
D
,.2
600-
o
o
O
o
oo
o
500oo 400-
~ 300-
300-
8 200-
o
200.m
1000 30
• lo0-
a I 40
•
•
I 50
DONOR AGE (YEARS)
I 60
0 70
m • --
mm
b I
]O
mm m
20
0 3
i 40
i 50
i 60
70
DONOR AGE (YEARS)
Fig. 3. Body site age-related variations in the secretion of type I (1:1)and type III (1) collagens by primary cultures of human dermal fibroblasts from face (a) and breast (b). Equations of the regression lines are the following: for facial skin cells, y = 1094 - 9.7x with r = 0.699 and P = 0.0037 for type I collagen and y = 203 - 1.4x with r = 0.513 and P = 0.106 for type Ili collagen; for breast skin cells, y = 796 - 6.0x with r = 0.618 and P = 0.032 for type I collagen and y = 176 - 0.5x with r = 0.281 and P = 0.376 for type II! collagen.
M. Dumas et a l . / Mech. Ageing Dev. 73 (1994) 179-187
184
CELL-ASSOCIATED COLLAGEN IN FACIAL SKIN CELLS
CELL-ASSOCIATED COLLAGEN IN BREAST SKIN CELLS 900
800-
800-
700-
700-
600-
600-
500-
500-
400-
400O
300-
30013
8
200100-
a
0
100I
4O
O
50
I
60 DONOR AGE (YEARS)
7'0
80
u
O
D
[]
O
200-
b
0 10
20
30
I
I
40
50
DONOR AGE WEARS)
Fig. 4. Body site age-related variation o f cell-associated type 1 (l~) and type III ( I ) collagen in primary cultures o f h u m a n dermal fibroblasts from face (a) and breast (b). Equations o f the regression lines are as follows: for facial skin cells, y = 3 4 7 - I . l x with r = 0.31 3 and P = 0 . 3 4 9 3 for type I collagen, y = 382 - 0.8x with r = 0.326 and P = 0.3924; for breast skin cells, y = 271 - 0.1x with r = 0.052 and P = 0.8873 for type I collagen, y = 351 - 0.2x with r = 0.07 and P = 0.8570 for type III collagen.
Furthermore, as for the global analysis, we confirmed that there were no significant variations in cell-associated collagens for mammary or facial skin cells analysed separately (Figs. 4a,b). 4. Discussion In the present study, we investigated collagen biosynthesis in human dermal fibroblasts, cells known to play a central role in extracellular matrix synthesis and in aging processes [1]. In order to avoid the influence of in vitro aging due to subculturing [9,10], primary cultures of fibroblasts from 19- to 68-year-old caucasian women were used. To assess the incidence of environmental factors on aging, fibroblast cultures were initiated from facial and mammary skin, sun-exposed and sunprotected body sites, respectively. Type I and type III collagen secretion by the fibroblasts as well as cell-associated levels were measured. For the population studied, we observed an overall significant linear agedependent decrease of type I and III collagen secretion by fibroblasts. Since type I and type III collagen represent the most abundant collagen types of the dermis, the age-related decrease in both secreted collagen types by fibroblast cultures agree with studies showing in different species a progressive decrease of global skin collagen content with age [16-18]. Moreover, this could partially explain the thinning of dermal collagen fibers composed mainly of type I and type III [19], the reduction of dermal thickness [18,20] and loss of morphometric properties of this tissue [21] that occur with age.
M. Dumas et al. / Mech. Ageing Dev. 73 (1994) 179-187
185
Recently, an age-dependent decrease in expression of the genes coding for type I and type III collagen in cultured dermal fibroblasts from skin donors 16-74 years of age has been reported [22]. This could suggest that the decrease we observed in secreted type I and type III collagen is the result of a lower expression of the corresponding genes. The body site analysis confirms the significant decrease in secretion of type collagen by mammary and facial skin cells. Moreover, the negative slope obtained for secreted type III collagen by facial skin cells, significant at P = 0.106, was not detected at the same significance level for mammary skin cells. Furthermore, during the same time period, the age-related decrease in type collagen secretion was greater for the face than for the breast. This effect was even more pronounced for type III collagen. This lead us to consider that sun exposure may affect mainly the ability of fibroblasts to secrete type III collagen. We found that the type III/I collagen ratio for mammary skin cells increased much greater with age (0.24-0.36) than for facial skin cells where the increase was lower (0.21-0.25), but on a shorter period of time, i.e. 28 years, than for mammary skin (44 years). A parallel can be established between these results with skin cells and the increase in the relative proportion of type III collagen determined in vivo in the abdominal skin of older donors [23]. The pliation properties of various tissues have thought to be linked to the relative amount of type III/I collagen [24]. If such a relationship does exist, the increase in the proportion of type III collagen secreted by fibroblasts during aging could be related to an increased flexibility of the dermal extracellular matrix. According to our results, this phenomenon would be attenuated in the sun-exposed skin of the face, where the increase in the ratio of secreted type III to type I collagen is smaller than in sun-protected mammary skin. In contrast, in vivo assays in hairless mouse skin show an increased type Ill/type I + III collagen ratio after UV exposure [25]. This apparent discrepancy with our results might be explained by the fact that the literature data probably involve UV cross-linking of collagen [26] which would modify extractable collagen, as previously described [27], in addition to the fact that the species are different. Our findings describe collagen secretion ability in skin isolated from a sun-exposed tissue, which is quite different and could suggest a gradual age-related decrease in skin tonicity, as previously reported [281. For cell-associated type I and type III collagen, we did not detect any significant age-variations. If we consider these pools as mainly intracellular, these findings indicate that this internal ratio is not modified during aging. Even if greater amounts of type I and type III collagen are synthesized and then released into the medium by the cells of younger as compared with older donors, the level of these intracellular collagens is maintained. This is in accordance with the hypothesis of cellular control of the internal collagen pool [29], which apparently is established with approximately equivalent amounts of type I and type III collagen. Our study has shown, by quantitative ELISA methodology, that human skin fibroblasts from older donors have a reduced ability to produce type I and type III collagen in vitro. In the same conditions, no significant age-related modifications were observed for cell-associated pools of type I and type III collagen. These agerelated modifications, observed at two different body sites, suggest that aging does
186
M. Dumas et al./Mech. Ageing Dev. 73 (1994) 179-187
not only affect fibroblasts from more sun-exposed skin such as facial skin, but rather that other, less sun-exposed body sites such as mammary skin may also undergo agerelated decreases in type III collagen secretion, albeit to a lesser extent. All these observations confirm that fibroblasts, through their age-related decrease in type I and type III collagen production ability, are central cells in dermis aging. 5. References 1 C.M. Lapiere, The ageing dermis: the main cause for the appearance of"old" skin. Br. J. Dermatol.. 122 (1990) 5-11. 2 E.J. Kuchartz, The Collagens: Biochemistry and Pathophysiology, Springer Verlag, 1992, pp. 227-263. 3 L. Weber, E. Kirsch, P. Muller and T. Krieg, Collagen type distribution and macromolecular organization of connective tissue in different layers of human skin. J. Invest. Dermatol., 82 (1984) 156-160. 4 R. Fleischmajer, J.S. Perlish, R.E. Burgeson, F. Shaikh-Bahai and R. Timpl, Type i and type Ill collagen interactions during fibrillogenesis. Ann. N.Y. Acad. Sci.. 580 (1990) 161-175. 5 L. Weber, C. Mauch, E. Hirsch, P.K. Muller and T. Krieg, Modulation of collagen type synthesis in organ and cell culture of fibroblasts. J. Invest. Dermatol., 87 (1986) 217-220. 6 S.M. Krane, Collagenases and collagen degradation. J. Invest. Dermatol., 79 (1982) 83s-86s. 7 P. Delvoye, P. Wiliquet, J.L. Leveque, B.U. Nusgens and C.M. Lapiere, Measurement of mechanical forces generated by skin fibroblasts embedded in a three dimensional collagen gel. J. Invest. Dermatol., 97 (1991) 898-902. 8 R.H. Pearce and B.J. Grimmer, Age and the chemical constitution of normal human dermis. J. Invest. Dermatol., 58 (1972) 347-365. 9 J.C. Houck, V.K. Sharma and L. Hayflick, Functional failures of cultured diploid fibroblasts after continued population doublings. Proc. Soc. Exp. Biol. Med., 137 (1971) 331-333. 10 J.N. Hildebran, M. Absher and R.B. Low, Altered rates of collagen synthesis in in vitro aged human lung fibroblasts. In Vitro, 19 (1983) 307-314. 11 L.H. Kligrnan, Photoaging. Manifestations, prevention and treatment. Clin. Geriatr. Med., 5 (1989) 235-251. 12 A. Oikarinen and M. Kallioinen, A biochemical and immunohistochemical study of collagen in sunexposed and protected skin. Photodermatology, 6 (1989) 24-31. 13 R.I. Freshney, Culture o f Animal Cells; A Manual o f Basic Technique, A.R. Liss, New York, 1983, pp. 104-106. 14 B.A. Booth, K.L. Polak and J. Uitto, Collagen biosynthesis by human skin fibroblasts. 1. Optimization of the culture conditions for the synthesis of type I and !I i procollagens. Bioehim. Biophys. A cta, 96 (1980) 145-160. 15 S.I. Rennard, G.R. Martin and R.G. Crystal, Enzyme-linked immunoassay (EL1SA) for connective tissue proteins: Type I collagen. In H. Furthmayr (ed.), lmmunochemistry o f the Extracellular Matrix, Vol. 1, CRC Press, Boca Raton, 1982, pp. 237-252. 16 P.K. Mays, J.E. Bishop and G.J. Laurent, Age-related changes in the proportion of types 1 and 111 collagen. Mech. Ageing Dev., 45 (1988) 203-212. 17 B. Boyer, P. Kern, A. Fourtanier and J. Labat-Robert, Age dependent variations of the biosynthesis of fibronectin and fibrous collagen in mouse skin. Exp. GerontoL, 26 (1991) 375-383. 18 S. Shuster, M.M. Black and E. Mc Vitie, The influence of age and sex on skin thickness, skin collagen and density. Br. J. Dermatol., 93 (1975) 639-643. 19 L.V. Zuccarello, R. Garbelli and V.D.P. Rossi, lmmunocytochemical localisation of collagen types I, III, IV and fibronectin in the human dermis. Cell Tissue Res., 268 (1992) 505-511. 20 M.C. Branchet, S. Boisnic, C. Frances and A.M. Robert, Skin thickness changes in normal aging skin. Gerontology, 36 (1990) 28-35. 21 M.C. Branchet, S. Boisnic, C. Frances, C. Lesty and L. Robert, Morphometric analysis of dermal collagen fibers in normal human skin as a function of age. Arch. Gerontol. Geriatr., 13 (1991) 1-14. 22 Y.Q. Chen, A. Mauviel, E.M.L. Tan and J. Uitto, Age-related changes in the expression of extracellular matrix genes in human dermal fibroblasts in culture. J. Invest. Dermatol., 100 (1993) 535.
M. Dumas et al. / Mech. Ageing Dev. 73 (1994) 179-187
23
24 25 26 27
28 29
187
C.R. Lovell, K.A. Smolenski, V.C. Duance, N.D. Light, S. Young and M. Dyson, Type I and 111 collagen content and fiber distribution in normal human skin during ageing. Br. J. DermatoL, 117 (1987) 419-428. A.J. Hance and R.G. Crystal, Rigid control of synthesis of collagen types I and III by cells in culture. Nature, 268 (1977) 152-154. A. Fourtanier, Collagen metabolism in ultraviolet irradiated hairless mouse skin. J. Invest. Dermatol., 96 (1991) 994. C.M. Dalle and M.A. Pathok, Photoaging and the role of UV (290-400 nm) and reactive oxygen species in collagen cross-links. J. Invest. DermatoL, 92 (1989) 410. J. Brinkmann, M. Brey and P.K. Muller, age differences of human skin collagen. In A. Bernd, J. Bereiter-Hahn, F. Hevert and H. Holzmann (eds.), Cell and Tissue Culture Models in Dermatological Research, Springer-Verlag, Berlin, Heidelberg, 1993, pp. 278-282. G.E. Pierrard, A critical approach to in vivo mechanical testing of the skin. In J.L. Leveque (ed.), Cutaneous Investigation in Health and Disease. M. Dekker, 1989, pp. 215-240. R.I. Schwarz, R.B. Mandell and M.J. Bissell, Ascorbate induction of collagen synthesis as a mean for quantitative control of tissue specific function. Mol. Cell. Biol., 1 (1981) 843-853.