Advertisement

Topically Applied Vitamin C Enhances the mRNA Level of Collagens I and III, Their Processing Enzymes and Tissue Inhibitor of Matrix Metalloproteinase 1 in the Human Dermis1

      Ascorbic acid (vitamin C) is a cofactor required for the function of several hydroxylases and monooxygenases. It is not synthesized in humans and some other animal species and has to be provided by diet or pharmacologic means. Its absence is responsible for scurvy, a condition related in its initial phases to a defective synthesis of collagen by the reduced function of prolylhydroxylase and production of collagen polypeptides lacking hydroxyproline, therefore, they are unable to assemble into stable triple-helical collagen molecules. In fibroblast cultures, vitamin C also stimulates collagen production by increasing the steady-state level of mRNA of collagen types I and III through enhanced transcription and prolonged half-life of the transcripts. The aim of the experimental work has been to evaluate the effect on dermal cells of a preparation of vitamin C topically applied on one side vs placebo on the other side of the dorsal face of the upper forearm of postmenopausal women. Biopsies were collected on both sides and the level of mRNA measured by non competitive reverse transcription–polymerase chain reaction made quantitative by the simultaneous transcription and amplification of synthetic RNA used as internal standards. The mRNA of collagen type I and type III were increased to a similar extent by vitamin C and that of three post-translational enzymes, the carboxy- and amino-procollagen proteinases and lysyloxidase similarly increased. The mRNA of decorin was also stimulated, but elastin, and fibrillin 1 and 2 were not modified by the vitamin. The expression of matrix metalloproteinases 1, 2, and 9 was not significantly changed, but an increased level of tissue inhibitor of matrix metalloproteinase 1 mRNA was observed without modification of tissue inhibitor of matrix metalloproteinase 2 mRNA. The stimulating activity of topical vitamin C was most conspicuous in the women with the lowest dietary intake of the vitamin and unrelated to the level of actinic damage. The results indicate that the functional activity of the dermal cells is not maximal in postmenopausal women and can be increased.

      Key words

      Abbreviations:

      28S rRNA
      ribosomal RNA
      sRNA
      synthetic RNA
      α1 I
      collagen I α1 chain
      α1 III
      collagen III α1 chain
      N-PCP
      amino-procollagen proteinase (ADAMTS2)
      C-PCP
      carboxy-procollagen proteinase (BMP1)
      LO
      lysyloxidase
      MMP-1
      matrix metalloproteinase 1
      MMP-2
      matrix metalloproteinase 2
      MMP-9
      matrix metalloproteinase 9
      TIMP
      tissue inhibitors of matrix metalloproteinases
      L-ascorbic acid (vitamin C) is an essential nutrient for some animal species and humans that lack the last enzyme in the pathway for its synthesis from glucose (
      • Banhegyi G.
      • Braun L.
      • Csala M.
      • Puskas F.
      • Mandl J.
      Ascorbate metabolism and its regulation in animals.
      ). Vitamin C is required for the optimal functioning of several hydroxylases and monooxygenases (
      • Padh H.
      Cellular functions of ascorbic acid.
      ). Its absence is responsible for scurvy, a disease characterized by altered functions of the connective tissues, including perifollicular hemorrhages and defective healing. Scurvy was a most usual cause of morbidity and mortality in sailors during the fifteenth century and up to the end of the eighteenth century in the British Navy (
      • Thomas D.P.
      Sailors scurvy and science.
      ). It was largely prevented by the introduction of lemon juice in the diet. At the present time, scurvy still exists in the Western world (
      • Hirschmann J.V.
      • Raugi G.J.
      Adult scurvy.
      ), mainly in institutionalized patients, drug addicts, and alcoholics that consume food deprived of vitamin C. The minimal dietary intake of vitamin C to prevent scurvy is 6.5 mg per day (
      • Bartley W.
      • Krebs H.A.
      • O'Brien J.R.P.
      Vitamin C requirement of human adults.
      ) and is barely large enough to prevent a drop of the body pool of the vitamin below 50%. The recommended daily intake to keep a saturated pool of the vitamin is much larger and about 80 mg (
      • Levine M.
      • Rumsey S.C.
      • Daruwala R.
      • Park J.B.
      • Wang Y.
      Criteria and recommendations for vitamin C intake.
      ) contained in five servings of vegetables (fresh or barely cooked) or fruits per day. In the absence of chemical supplementation saturation is rarely achieved.
      It is presently known that vitamin C is required as a cofactor for the correct hydroxylation of prolyl and lysyl residues of the procollagen polypeptides (
      • Kivirikko K.I.
      • Prockop D.J.
      Enzymatic hydroxylation of proline and lysine in protocollagen.
      ) allowing their triple helical conformation in the cells (
      • Rosenbloom J.
      • Harsch M.
      • Jimenez S.
      Hydroxyproline content determines the denaturation temperature of chick tendon collagen.
      ) and the secretion, processing, and polymerization of these precursors to form the fibers ultimately conferring resistance to the tissues (
      • Levene C.I.
      Ascorbic acid and collagen synthesis in cultured fibroblasts.
      ). In fibroblast cultures, vitamin C stimulates collagen biosynthesis (
      • Peterkofsky B.
      the effect of ascorbic acid on collagen polypeptide synthesis and proline hydroxylation during the growth of cultured fibroblasts.
      ) not only by promoting the activity of the hydroxylases, but also by increasing the steady-state level of the procollagen mRNA (
      • Lyons B.L.
      • Schwarz R.I.
      Ascorbate stimulation of PAT cells causes an increase in transcription rates and a decrease in degradation rates of procollagen mRNA.
      ;
      • Schwarz R.I.
      Procollagen secretion meets the minimum requirements for the rate-controlling step in the ascorbate induction of procollagen synthesis.
      ;
      • Pinnell S.R.
      • Murad S.
      • Darr D.
      Induction of collagen synthesis by ascorbic acid.
      ). This depends on both an increased transcription of the genes and a stabilization of the transcripts to a similar extent for the mRNA of the main types of collagen (I and III) present in skin (
      • Geesin J.C.
      • Darr D.
      • Kaufman R.
      • Murad S.
      • Pinnell S.R.
      Ascorbic acid specifically increases type I and type III procollagen messenger RNA levels in human skin fibroblast.
      ).
      The effect of vitamin C on the level of elastin mRNA in cultured fibroblasts is the reverse. It reduces transcription and decreases the stability of the transcripts (
      • Davidson J.M.
      • LuValle P.A.
      • Zoia O.
      • Quaglino D.
      • Giro M.G.
      Ascorbate differentially regulates elastin and collagen biosynthesis in vascular smooth muscle cells and skin fibroblasts by pretranslational mechanisms.
      ).
      Besides its cofactor activity, vitamin C is also a free radical scavenger by its antioxidant properties (for a review see
      • Sauberlich H.E.
      Pharmacology of vitamin C.
      ). It protects keratinocytes from the damage produced by ultraviolet A (
      • Tebbe B.
      • Wu S.L.
      • Geilen C.C.
      • Eberle J.
      • Kodelja V.
      • Orfanos C.E.
      L-ascorbic acid inhibits UVA-induced lipid peroxidation and secretion of IL-1 alpha and IL-6 in cultured human keratinocytes in vitro.
      ). Its beneficial activity as a photoprotectant (
      • Darr D.
      • Dunston S.
      • Faust H.
      • Pinnell S.
      Effectiveness of antioxidants (Vitamin C and E) with and without sunscreens as topical photoprotectants.
      ) and anticancer agent (
      • Pauling L.
      Effect of ascorbic acid on incidence of spontaneous mammary tumors and UV-light-induced skin tumors in mice.
      ) has been demonstrated by dietary supplementation in humans and in animal species even in those that can synthesize the vitamin. A photoprotective effect has also been demonstrated after topical application (
      • Darr D.
      • Combs S.
      • Dunston S.
      • Manning T.
      • Pinnell S.
      Topical Vitamin-C protects porcine skin from ultraviolet radiation-induced damage.
      ;
      • Colven R.M.
      • Pinnell S.R.
      Topical vitamin C in aging.
      ). Vitamin C is also required to form competent barrier lipids in the epidermis (
      • Ponec M.
      • Weerheim A.
      • Kempenaar J.
      • Mulder A.A.
      • Gooris G.S.
      • Bouwstra J.
      • Mommaas A.M.
      The formation of competent barrier lipids in reconstructed human epidermis requires the presence of vitamin C.
      ) by stimulating the synthesis of ceramides. It has also been shown to stimulate the barrier function of the endothelial cells (
      • Utoguchi N.
      • Ikeda K.
      • Saeki K.
      • et al.
      Ascorbic acid stimulates barrier function of cultured endothelial cell monolayer.
      ).
      The aim of this study was to evaluate the effect on the dermal cells of vitamin C administrated by topical application on the skin of normal human volunteers by measuring: the steady-state level of the mRNA of procollagen I and III, their post-translational processing enzymes, carboxy-procollagen proteinase (
      • Li S.W.
      • Sieron A.L.
      • Fertala A.
      • Hojima Y.
      • Arnold W.V.
      • Prockop D.J.
      The C-proteinase that processes procollagens to fibrillar collagens is identical to the protein previously identified as bone morphogenic protein-1.
      ), amino-procollagen proteinase (
      • Colige A.
      • Li S.-W.
      • Sieron A.L.
      • Nusgens B.V.
      • Prockop D.J.
      • Lapière C.M.
      cDNA cloning and expression of bovine procollagen I N-proteinase: a new member of the superfamily of zinc-metalloproteinases with binding sites for cells and other matrix components.
      ) and lysyloxidase (LO) (
      • Hamalainen E.R.
      • Jones T.A.
      • Sheer D.
      • Taskinen K.
      • Pihlajaniemi T.
      • Kivirikko K.I.
      Molecular cloning of human lysyl oxidase and assignment of the gene to chromosome 5q23.3–31.2.
      ), decorin, a collagen fiber-associated proteoglycan (
      • Danielson K.G.
      • Baribault H.
      • Holmes D.F.
      • Graham H.
      • Kadler K.E.
      • Iozzo R.V.
      Targeted disruption of decorin leads to abnormal collagen fibril morphology and skin fragility.
      ), the main components of the elastic fibers, elastin and fibrillins 1 and 2 (
      • Ramirez F.
      • Pereira L.
      The fibrillins.
      ) as well as the metalloproteinases and their physiologic inhibitors involved in the degradation of these matrix components (
      • Nagase H.
      • Woessner J.F.
      Matrix metalloproteinases.
      ). Such investigations performed on small biopsies were made possible by the use of quantitative reverse transcription–polymerase chain reaction (RT–PCR) controlled by original newly created internal standards of synthetic RNA.

      Materials and methods

      Selection and treatment of the volunteers

      Ten human volunteers were selected from a cohort of postmenopausal women 50–60 y old (mean 55.3 y). Most testers included in the study received a substitutive hormonal therapy, except tester no. 10 who received only progesterone. No vitamin supplementation was allowed during the test. The study was conducted between December and June and the testers were instructed to avoid sun exposure. No other topically applied preparation was permitted. The treatment consisted of applying, at night, on the dorsal side of the upper forearm preparation A (active) containing 5% vitamin C on one side and preparation P (placebo) on the other side. The distribution of the preparations, active and placebo, was randomly allocated and not known by the testers or by the investigators. The once-daily application was repeated for 6 mo, the tester being examined after 3 and 6 mo for clinical observation and detection of potential side-effects. The tolerance was perfect. At the termination of treatment, two 5 mm punch biopsies up to the hypodermis were collected under local anesthesia at the site of the topical application. One biopsy was used for measurement of mRNA and collagen extractability. The second biopsy was used for classical morphology and electron microscopy to evaluate sun damage. The study was conducted in agreement with the Declaration of Helsinki and approved by the ethical committee of the University Hospital Saint-Jacques in Besançon, France.

      Preparation of the stabilized w/o emulsion containing 5% vitamin C

      The active preparation (A) was a solution of 5% L-ascorbic acid (vitamin C) in glycerol at pH 6.0 emulsified in a silicone base and prepared under an atmosphere of nitrogen, commercially available (Active C, La Roche-Posay, France). The placebo (P) was the same preparation devoid of vitamin C. The preparations were kept in similar aluminum tubes preventing contact with air. The stability of the preparations was longer than 3 y at room temperature.

      Extraction and purification of the RNA

      Immediately after sampling, one biopsy from each side was freed of adhering hypodermis, wrapped in aluminum foils, placed in a small vials dropped and kept in liquid nitrogen until use. Half of the biopsy was ground in liquid nitrogen in a Microdismembrator S (Braun Biotech International, Melsungen, Germany) and the resulting powder collected in 2 ml of the lysis solution (5 M guanidium isothiocyanate, 0.1 M β-mercaptoethanol, 17 mM Na laurylsarcosyl, 25 mM Na citrate pH 7.0). After 15 min of agitation at room temperature the lysate was overlaid on a cushion of 1.4 ml 5.7 M cesium chloride, 0.1 M ethylenediamine tetraacetic acid pH 7.0 and ultracentrifuged (110,000 × g, 18 h at 20°C) in a SW 60 rotor (Beckman L70M, Palo-Alto, CA). The pellet of RNA was rinsed in 70% ethanol, centrifuged at 6800 × g (4°C, 10 min), and dissolved in 500 µl RNAse-free distilled water. The concentration of RNA and its purity were estimated by optical density at 260 nm and 260/280 nm ratio. The stock solutions of RNA were diluted to a concentration of 4 ng per µl, aliquoted, and stored at -80°C.

      Reverse transcription–PCR assay

      The specific oligonucleotide primers of the mRNA of interest were around 24 bases long with an A-T proportion close to 50%. The sequences shown in Table Iwere chosen on different exons to allow discrimination of products amplified from potential contaminating DNA. The mRNA of interest and the 28S rRNA were quantified in triplicate by reverse transcription–PCR using 10 ng of total RNA. The RNA were reverse transcribed at 70°C for 15 min and amplified (94°C for 15 s, 66°C for 20 s, 78°C for 10 s) using the Gene Amp rTth Kit (Perkin Elmer, Branchburg, NJ), except matrix metalloproteinase (MMP) -2 that was reverse-transcribed at 55°C for 30 min and amplified (94°C for 18 s, 62°C for 20 s, 68°C for 15 s) using the kit Titan (Boehringer, Mannheim, Germany). The number of amplification cycles used in this study is indicated in Table I. The efficiency of the reverse transcription and the amplification reactions was monitored by adding in each tube an appropriate number of copies of a synthetic RNA (sRNA) that can be reverse-transcribed and amplified with the primers pair used for the amplification of the mRNA of interest, but giving rise to a product slightly larger or smaller than the product amplified from the mRNA (Table I), allowing its discrimin ation after migration in a 10% polyacrylamide gel. The different sRNA were generated from linearized template plasmid containing appropriate DNA inserts by the use of SP6 RNA polymerase (SP6/T7 transcription kit; Boehringer Mannheim), purified (High Pure RNA Isolation kit; Boehringer Mannheim) and quantitated
      Lambert ChA, Colige AC, Maniglia S, Heyeres A, Munaut C, Lapière ChM. and Nusgens BV: Measurement of matrix metalloproteinases by quantitative RT-PCR assay: Application to the study of gene regulation in fibroblasts by stress relaxation. J Invest Dermatol 110:612, 1998 (abstr.)
      (manuscript in preparation). The gels were stained with SyberGreen (Biorad, Hercule, CA) or Gelstar (FMC Bioproducts, Rockland, ME) and the intensity of the fluorescent signals measured in a Fluor-S-MultiImager (Biorad).
      Table INucleotide sequence of the primers, size of the generated reverse transcription–PCR products and number of cycles for amplification
      PrimersSize of the products (bp)Cycle
      RNA
      The identity of the investigated RNA is described in the text.
      Forward (5′-3′)Reverse (5′-3′)EndogenoussRNAnumber
      28S rRNAGTTCACCCACTAATAGGGAACGTGAGGATTCTGACTTAGAGGCGTTCAGT21226918
      GAPDHCCTGGCCAAGGTCATCCATGACAGGGATGACCTTGCCCACAGCCTT18326225
      Keratin 10GGGAGCCTCGTGACTACAGCAAATCCCTACGCAGGCCGTTGATGTCA201None28
      VimentinGACAATGCGTCTCTGGCACGTCTTTCCTCCGCCTCCTGCAGGTTCTT230None30
      α1(I)CCCACCAATCACCTGCGTACAGATTCTTGGTCGGTGGGTGACTCTGA21426727
      α1(III)GAGATGTCTGGAAGCCAGAACCATGATCTCCCTTGGGGCCTTGAGGT20726527
      N-PCPGAACCATGAGGACGGCTTCTCCTGGCTGCAGCGGGACCAGTGGAA17626135
      C-PCPAAGTCCGACAACACCGTGTCCAAACCATCCCATTCCGTCCAAGGTGA20126535
      LOCCCCTACAAGTACTCTGACGACAACGCCGCGCATCTCAGGTTGTACAT20226330
      DecorinCCTGAAAGGACTGAATAATTTGGCTAGTTGCTGAAAAGACTCACACCCGAA27720130
      ElastinCCGCTAAGGCAGCCAAGTATGGAAGCTCCAACCCCGTAAGTAGGAAT27519728
      Fibrillin 1GGTGAATGTACAAACACAGTCAGCAATAGGAACAGAGCACAGCTTGTTGA27521030
      Fibrillin 2ATGGCTCTCGATGCATCGATCAGACATTGCCACTTGGGGCAAAGCCA28219935
      MMP-1GAGCAAACACATCTGAGGTACAGGATTGTCCCGATGATCTCCCCTGACA18526735
      MMP-2AGATCTTCTTCTTCAAGGACCGGTTGGCTGGTCAGTGGCTTGGGGTA22527127
      MMP-9GCGGAGATTGGGAACCAGCTGTAGACGCGCCTGTGTACACCCACA20826634
      TIMP1CATCCTGTTGTTGCTGTGGCTTGATGTCATCTTGATCTCATAACGCTGG17027030
      TIMP2CTCGCTGGACGTTGGAGGAAAGAAAGCCCATCTGGTACCTGTGGTTCA15526925
      a The identity of the investigated RNA is described in the text.
      For each of the mRNA, the optimal conditions of reaction (temperature, number of cycles and choice of the reverse transcription–PCR kit) and the amount of internal standard were determined taking into account the level of expression of the mRNA in human skin to obtain values within the linear range of measurement as recommended by
      • Freeman W.M.
      • Walker S.J.
      • Vrana K.E.
      Quantitative RT-PCR: Pitfalls and potential.
      and illustrated in Figure 1. Each measurement was normalized to the cotranscribed and coamplified internal standard. The normalized values were expressed in arbitrary units (AU) per unit of 28S rRNA measured in the same dilution of RNA to correct for RNA input of each sample.
      Figure thumbnail gr1
      Figure 1Titration of the mRNA of collagen α1 (I). The ratio of the signals obtained for the amplification products of increasing amounts of endogenous RNA with the addition of a constant amount (85,000 copies) of standard synthetic RNA increases linearly (r = 0.998) with the amount of endogenous RNA. Each value is the average of triplicate measurements ± SD.

      Measurement of collagen extractability

      The remaining half of the biopsy was used to measure collagen after sequential fractionation. The fragments of skin were lyophilized, weighed, and extracted at 4°C in 0.5 M acetic acid-HCl (HAc) at pH 2.0 for 24 h followed by extraction after pepsin digestion (200 µg per ml) in 0.5 M acetic acid-HCl pH 2.0 for a further 24 h. Aliquots of the extracts and the residual material were hydrolyzed in 6 M HCl and the concentration of hydroxyproline was determined by a colorimetric assay (
      • Bergman I.
      • Loxley R.
      Two improved and simplified methods for the spectrophotometric determination of hydroxyproline.
      ).

      Morphologic analysis

      The second biopsy from the placebo-treated side was separated in two fragments perpendicularly to the stratum corneum. One fragment was fixed in Baker's solution for 3 h and embedded in paraffin. The other fragment was further sectioned in smaller samples and fixed in 2% glutaraldehyde in cacodylate buffer for 6 h, postfixed in 1% osmium tetroxide, dehydrated in ethanol, and embedded in Epon. Ultrathin sections were counterstained with lead citrate and uranyl acetate. Five micrometer paraffin sections were stained with hematoxylin–eosin–saphran, orcein, or immunolabeled using monoclonal antibodies to vimentin (VIM 3B4, Progen, Heidelberg, Germany) and collagen IV (clone CIV22, Dako, Glostrup, Denmark), both revealed with biotin-streptavidin-alkaline phosphatase (LSAB kit, Dako). Light and electron microscopy were used to evaluate sun damage by comparison with nonexposed buttock skin of five age-matched individuals. Three parameters were analyzed: irregularity of the epidermal pigmentation and disorganization of the dendritic cells network, flattening of the epidermal–dermal junction and subepidermal hyaline bodies, and disorganization of the elastic fiber network. Each parameter was scored semiquantitatively (0, no alteration; 1, very light; 2, moderate; and 3, severe) and summed to determine the extent of solar alteration.

      Statistical analysis

      Results were expressed as the ratio between the values measured in the skin sample treated with the preparation containing vitamin C (A) and that treated with the placebo (P). The mean ratio was tested for statistical difference from 1 by using the one-sided Student's t test for paired data.

      Results

      The biopsies contain similar proportions of dermal and epidermal RNA

      The contribution of the dermis in the collected RNA was estimated by measuring the steady-state level of vimentin (VIM) mRNA, the protein of the intermediate filaments of the mesenchymal cells whereas that of the epidermis was obtained by measuring the mRNA of keratin 10 (K10) (Table II). The mean ratio VIM/K10 was similar on both sides, that treated with vitamin C (A) and the controlateral side receiving the placebo (P), indicating a similar contribution of epidermal and dermal RNA in the samples; however, the values of VIM and K10 per unit of 28S rRNA were both increased in the vitamin C-treated side, which resulted in an A/P ratio significantly higher than 1.0. The mean value for the mRNA level of the housekeeping gene GAPDH per unit of 28S rRNA was very similar on both sides resulting in an A/P ration close to 1.0 (Table II).
      Table IIThe RNA collected from the biopsies arises in a constant proportion from the epidermis (K10) and the dermis (VIM)
      K10 and VIM mRNA levels are significantly increased in vitamin C treated sides (A) as compared with placebo (P), whereas GAPDH mRNA level is not modified by the vitamin.
      VIM/K10
      TesterAPK10 A/PVIM A/PGAPDH A/P
      17.226.051.581.881.76
      26.88[14.10]
      Ectopic values excluded from mean and SD.
      [2.64]
      Ectopic values excluded from mean and SD.
      1.290.69
      35.275.630.930.870.82
      44.532.680.901.531.16
      54.824.181.171.351.51
      62.654.311.681.030.87
      73.596.341.881.061.35
      83.775.361.170.830.62
      93.214.481.230.880.64
      104.194.931.461.241.14
      Mean ± SD4.61 ± 1.50 NS
      NS, not significant.
      4.88 ± 1.12 NS1.33 ± 0.34 p < 0.011.20 ± 0.33 p < 0.051.06 ± 0.39 NS
      a K10 and VIM mRNA levels are significantly increased in vitamin C treated sides (A) as compared with placebo (P), whereas GAPDH mRNA level is not modified by the vitamin.
      b Ectopic values excluded from mean and SD.
      c NS, not significant.
      The studied population is heterogeneous in terms of level of expression of all the investigated genes. This is clearly illustrated in Figure 2 for the procollagen processing enzyme, the C-PCP ranged in the studied population from 65 to 370 AU per unit of 28S rRNA. The individual variations were, however, similar on both arms resulting in a mean value almost identical for the group of samples collected on the right (R) and the left (L) arm and an R/L ratio close to 1.0. To account for that heterogeneity the results in the tables were mostly expressed by the ratio of the values observed in the vitamin C-treated side (active: A) to that of the placebo (P). In the example illustrated in Figure 2, the A/P ratios significantly higher than 1.0 indicate an increased steady-state level of C-PCP mRNA upon vitamin C treatment.
      Figure thumbnail gr2
      Figure 2Heterogeneity of the tested population. The mRNA level (in arbitrary units per unit of 28S sRNA) for C-PCP varies among the testers but similarly on the right (R) and the left (L) sides resulting in a mean R/L ratio close to 1.0 (left part). The stimulating effect of topical vitamin C (in bold figures) is expressed by a mean ratio active/placebo (A/P) significantly higher than 1.0 (right part).

      The mRNA levels of collagen type I and type III and that of three extracellular procollagen processing enzymes are coordinately increased by vitamin C

      The steady-state level of the mRNA of the two main types of collagen in skin, types I and III, was measured by quantitative reverse transcription–PCR using specific sRNA as internal standards for the α1 chains of collagen I and III. The example of titration of the collagen polypeptides α1 (I) mRNA (Figure 1) illustrates the linearity, reproducibility, and sensitivity of the procedure.
      The variability in the level of the two collagen mRNA was lower than that observed for C-PCP and similar on both sides (not illustrated). Six of the 10 testers displayed increased values for both types of collagen on the side treated with the active preparation and the mean A/P ratio of the whole group for both collagen type I and type III supported a stimulating effect of the vitamin (Table III). This occurred to the same extent with the two types of collagen. The ratio between collagen I and collagen III was very similar on both sides in spite of the individual variability in the level of these mRNA within the group of testers.
      Table IIIThe topical application of vitamin C (A) increases the steady-state level of the mRNA of collagen I (α1 I) and collagen III (α1 III) in the same six of the 10 testers without disturbing the ratio between the two types of collagen (I/III)
      I/III
      Testerα1 I A/Pα1 III A/PAP
      11.601.342.912.45
      21.381.642.583.07
      30.920.912.842.75
      41.551.563.013.02
      51.741.592.722.49
      60.890.962.592.80
      71.241.242.552.56
      80.710.623.022.64
      90.830.782.762.57
      101.621.453.042.72
      Mean ± SD1.25 ± 0.38 p < 0.041.21 ± 0.37 p < 0.062.80 ± 0.19 NS
      NS, not significant.
      2.71 ± 0.21 NS
      a NS, not significant.
      The interindividual variability of LO was higher than that of C-PCP and N-PCP. The C-PCP mRNA enhancing activity of vitamin C is illustrated in Figure 2 by a statistically significant increase of the mean A/P ratio. The mean values of the A/P ratios were also significantly increased for the two other enzymes (N-PCP and LO) involved in the maturation of the (pro)collagen molecules (Table IV).
      Table IVThe mRNA steady-state level of the processing (N-PCP and C-PCP) and cross-linking (LO) enzymes is increased on the side treated by vitamin C (A) by comparison with the side treated with placebo (P)
      TesterN-PCP A/PC-PCP A/PLO A/P
      11.331.665.67
      21.101.222.53
      30.871.020.57
      41.961.461.18
      51.581.302.20
      61.191.141.41
      71.001.162.55
      80.861.170.24
      90.791.111.47
      101.411.331.45
      Mean ± SD1.21 ± 0.37 p < 0.061.26 ± 0.19 p < 0.011.93 ± 1.52 p < 0.05

      The concentration of collagen and its extractability are not significantly modified in the vitamin C treated skin

      The fraction of acetic acid-extracted collagen was very small in agreement with previously published data (
      • Legrand Y.
      • Lapière Ch M.
      • Pignaud G.
      • Caen J.
      Microméthode d'extraction du collagène à partir de biopsie de peau humaine.
      ). The proportion of extracted collagen and that remaining in the insoluble residue was similar in the two groups of samples and the mean A/P ratio did not significantly differ from 1 (Table V).
      Table VExtractability of collagen in vitamin C-treated (A) and placebo-treated (P) skin
      A (n = 10)P (n = 10)A/P (n = 10)t test
      0.5 M HAc8.3 ± 3.3
      In µg of collagen per mg of tissue, dry weight.
      8.7 ± 3.71.30 ± 1.20NS
      Pepsin55 ± 1658 ± 160.98 ± 0.27NS
      Insoluble373 ± 103349 ± 601.07 ± 0.23NS
      Total436 ± 110416 ± 641.05 ± 0.21NS
      a In µg of collagen per mg of tissue, dry weight.

      The mRNA of decorin and TIMP1 are increased by vitamin C but not that of elastin, fibrillins, MMP-1, MMP-2, MMP-9, and tissue inhibitor of matrix metalloproteinase (TIMP) 2

      The mean value of the mRNA for elastin and fibrillin 1 collected from the vitamin C-treated was not significantly increased as compared with placebo (Table VI). The mRNA of decorin was increased by the topical application of vitamin C. The difference was, however, at the limit of significance. The transcript of fibrillin 2 was not reproducibly detected.
      Table VIExpression of a series of other connective tissue related genes upon topical application of vitamin C
      A/P
      Mean value of A/P ratios.
      ± SDt test
      Elastin1.10± 0.63NS
      NS, not significant.
      Fibrillin 11.24± 0.69NS
      Decorin1.64± 1.27p < 0.10
      MMP14.50± 6.30NS
      MMP21.51± 1.09NS
      MMP91.41± 0.94NS
      TIMP11.38± 0.54p < 0.03
      TIMP21.27± 0.84NS
      a Mean value of A/P ratios.
      b NS, not significant.
      The expression of MMP-1 was very small, barely detectable in most instances, whereas a few samples displayed higher values. There was no statistically valid difference between the two sides. The interindividual expression of MMP-2 was also variable and no significant difference was observed between the side treated with the vitamin and the placebo. The same conclusions apply to MMP-9 (Table VI); however, the mRNA level of TIMP1 was increased on the side treated with the active preparation (Table VI), but not that of TIMP2.

      Relationship between the metabolic effect of topical vitamin C, the dietary intake of vitamin C and the extent of actinic damage

      The dietary intake of vitamin C has been estimated after termination of the study by a questionnaire collecting semi quantitative information to estimate the consumption of fruits and vegetables. It is expressed as the sum of servings per day. Smoking was also recorded as it increases the need for a higher intake of vitamin C to contribute to the pool. This information has been gathered in Table VII and correlated with the potential of topical vitamin C to modulate the biosynthetic activity of the cells. It is worth noting that the testers that have the lowest score of dietary vitamin C intake are those in which topically applied vitamin C displayed the most constant stimulation of the steady-state level of mRNA for collagens and their processing enzymes.
      Table VIIThe testers with the lowest intake of vitamin C are most responsive to topically applied ascorbic acid
      Dietary vitamin C
      Semiquantitative estimation of the dietary intake of vitamin C expressed as the sum of servings of fruits and vegetables per day.
      Smoking
      Cigarettes per day.
      Testersα1 Iα1 IIIN–PCPC–PCPLODecorinElastinFibrillin 1MMP2MMP9TIMP1TIMP2Actinic damage
      Semiquantitative estimation (see Materials and Methods).
      2ND10+
      +, A/P ratio > 1; –, A/P ratio < 1; =, A/P ratio = 1; ND, not determined.
      ++++++++++5
      305++++++++++++5
      4104++++++++++6
      432+++++++++6
      401+++++++++++7
      407++=+++++++++3
      506++++++++++2
      509++=+7
      508+8
      603++7
      a Semiquantitative estimation of the dietary intake of vitamin C expressed as the sum of servings of fruits and vegetables per day.
      b Cigarettes per day.
      c Semiquantitative estimation (see Materials and Methods).
      d +, A/P ratio > 1; –, A/P ratio < 1; =, A/P ratio = 1; ND, not determined.
      All 10 subjects presented histologic and ultrastructural signs of actinic damage (Table VII). There was, however, no correlation between its extent (from 2, very mild, to 8, extensive) and the responsiveness to the topically applied vitamin C.

      Discussion

      The presented data demonstrate that daily topical application of a preparation containing 5% L-ascorbic acid enhances the steady-state level of procollagens I and III mRNA and that of their post-translational maturation enzymes. The mRNA level of the MMP responsible for the degradation of the extracellular matrix is not statistically modified, whereas the mRNA of TIMP1, a physiologic inhibitor of MMP, is increased. These modifications reflect the expression of an anabolic phenotype. This pattern of enhanced anabolic activity has been obtained by performing the mRNA measurements in a small cohort of only 10 female volunteers at the site of application of the active preparation compared with the excipient alone on the opposite parallel site. This strategy allowed us to establish a ratio between paired values in the same individual, therefore, minimizing the bearing of a marked interindividual heterogeneity in the expression of most mRNA on the statistical significance of the results.
      The reverse transcription–PCR procedure used for measuring the biosynthetic activity of the cells in small samples of tissues has been carefully controlled to generate comparative data with a high degree of confidence. Our quantification procedure of ribosomal (28S rRNA) and specific mRNA by reverse transcription–PCR under noncompetitive conditions used an internal standard of sRNA and is most adequate to demonstrate small differences (
      • Freeman W.M.
      • Walker S.J.
      • Vrana K.E.
      Quantitative RT-PCR: Pitfalls and potential.
      ). We have verified that in each biopsy the RNA arises in a similar proportion from the epidermis by measuring the mRNA of K10 and from the dermis by measuring the mRNA of VIM. This condition has to be fulfilled to permit a reliable estimation of the mRNA of interest on the basis of a unit amount of 28S rRNA. The validity of the procedure is supported by the constant ratio of collagen type I and type III mRNA level in the biopsies, similar to the ratio of these collagens in adult human skin (
      • Epstein E.H.
      [alpha1 (III)]3 Human Skin Collagen. Release by pepsin digestion and preponderance in fetal life.
      ) and that produced by human fibroblasts in culture (
      • Phillips C.L.
      • Tajima S.
      • Pinnell S.R.
      Ascorbic acid and transforming growth factor-beta1 increase collagen biosynthesis via different mechanisms—coordinate regulation of proalpha1(I) and proalpha1(III) collagens.
      ).
      In fibroblast cultures, the mRNA of the two main types of collagens are known to be coordinately upregulated by vitamin C (
      • Geesin J.C.
      • Darr D.
      • Kaufman R.
      • Murad S.
      • Pinnell S.R.
      Ascorbic acid specifically increases type I and type III procollagen messenger RNA levels in human skin fibroblast.
      ). This also seems to be the case in our in vivo study. Furthermore, the mRNA level of the three enzymes involved in the post-translational processing and cross-linking of the collagen molecules was also stimulated by the vitamin C treatment suggesting the existence of a mechanism of regulation parallel to that operating for the collagen polypeptides. An increased activity of the N-PCP is known to exist in pathologic conditions where the biosynthetic activity of the connective tissues is enhanced (
      • Lapière Ch.M.
      • Piérard G.E.
      Skin procollagen peptidase in normal and pathologic conditions.
      ). The repercussion at the tissue level of the increased expression of procollagens mRNA in vitamin C treated skin was tentatively evaluated by measuring the amount of extractable collagen at acid pH (
      • Legrand Y.
      • Lapière Ch M.
      • Pignaud G.
      • Caen J.
      Microméthode d'extraction du collagène à partir de biopsie de peau humaine.
      ) and after pepsin digestion as an indirect indicator of collagen turnover rate, the newly synthetized collagen being less cross-linked and more extractable (
      • Gross J.
      • Page I.H.
      ). In the adult human, the half-life of dermal collagen might be as long as 10 y by extrapolation of animal data (
      • Nimni M.E.
      • Bavetta L.A.
      Collagen synthesis and turnover in the growing rat under the influence of methylprednisolone.
      ; Nimni M., personal communication) and the amount of replaced collagen about 10 µg per mg dry skin during the period of treatment by the vitamin. The 30% increase induced by vitamin C (3 µg in mass of which only a small fraction can be extracted) is too small to be detected. The parameters estimating the level of cell activity are obviously more sensitive.
      The daily requirement of dietary vitamin C to prevent scurvy, i.e., to permit at least the vital function of the collagen hydroxylases and other monooxygenases, has been estimated at less than 10 mg per day (i.e., 15 ml of lemon juice). Such a low intake is compensated by a sparing mechanism of cellular recycling of the oxidized vitamin by two enzyme systems (
      • Banhegyi G.
      • Braun L.
      • Csala M.
      • Puskas F.
      • Mandl J.
      Ascorbate metabolism and its regulation in animals.
      ). A daily intake of 60–100 mg vitamin C is required to raise the serum level in humans to that of animals that synthesize L-ascorbate. This amount needs to be increased to 140 mg per day for smokers that seem to consume vitamin C to inactivate oxidants from tobacco smoke (
      • Schectman G.
      Estimating ascorbic acid requirement for cigarette smokers.
      ). The concentration of ascorbate is greater in the skin than that in plasma, probably by a mechanism of active transport (
      • Welch R.V.
      • Wang Y.
      • Crossman A. Jr
      • Park J.B.
      • Kirk K.L.
      • Levine M.
      Accumulation of vitamin C (ascorbate) and its oxidized metabolite dehydroascorbic acid occurs by separate mechanisms.
      ). Nevertheless, our data demon strate that in postmenopausal women the topical application of vitamin C is able to produce a coordinated increase in the steady-state level of the mRNA for collagen I and III as observed in six testers, and of their post-translational extracellular enzymes, as observed in at least, seven of the 10 testers. We speculated that the nonresponders might already have a concentration of the vitamin in skin great enough for maximal expression of the ascorbate-responsive mRNA. This hypothesis is supported by the ranking of the testers (Table VII) showing an inverse relationship between the responsiveness to vitamin C and the overall dietary intake of the vitamin. An indirect effect of vitamin C mediated by the epidermis, however, cannot be ruled out as the epithelial cells were also stimulated by the topical application of the vitamin. Keratinocytes are known to produce in vitro several cytokines that are modulated by vitamin C (
      • Tebbe B.
      • Wu S.L.
      • Geilen C.C.
      • Eberle J.
      • Kodelja V.
      • Orfanos C.E.
      L-ascorbic acid inhibits UVA-induced lipid peroxidation and secretion of IL-1 alpha and IL-6 in cultured human keratinocytes in vitro.
      ).
      The dermal signs of aging, photoinduced and chronologic, are to some extent comparable with scurvy, i.e., atrophy, fragility, easy bruising, and palor. In aging, the density of the dermal collagen network diminishes by reduced metabolic activity (
      • Legrand Y.
      • Lapière Ch M.
      • Pignaud G.
      • Caen J.
      Microméthode d'extraction du collagène à partir de biopsie de peau humaine.
      ), and enhanced degradation by increased production of MMP (
      • Millis A.J.
      • Hoyle T.M.
      • McCue H.M.
      • Martini H.
      Differential expression of metalloproteinase and tissue inhibitor of metalloproteinase genes in aged human fibroblasts.
      ), perhaps related to an oxidative stress (
      • Sohal R.S.
      • Weindruch R.
      Oxidative stress, caloric restriction and aging.
      ) similar to that induced by ultraviolet in fibroblasts (
      • Wlaschek M.
      • Briviba K.
      • Stricklin G.P.
      • Sies H.
      • Scharffetter-Kochanek K.
      Singlet oxygen may mediate the ultraviolet A-induced synthesis of interstitial collagenase.
      ). The level of three MMPs involved in acute (Fischer et al, 1999) or chronic (
      • Seite S.
      • Colige A.
      • Piquemal-Vivenot P.
      • et al.
      A full-UV spectrum absorbing daily use cream protects human skin against biological changes occurring in photoaging.
      ) skin alterations induced by ultraviolet irradiation was not significantly modified by the topical application of the vitamin. The activity of these proteases is known to be reduced in vivo by vitamin A (
      • Varani J.
      • Warner R.L.
      • Gharaee-Kermani M.
      • et al.
      Vitamin A: antagonizes decreased cell growth and elevated collagen-degrading matrix metalloproteinases and stimulates collagen accumulation in naturally aged human skin.
      ) through a mechanism of action probably different from that operating for the vitamin C. The expression of TIMP1, but not TIMP2, however, was increased suggesting that the MMP activity in the vitamin C-treated side may be balanced by their physiologic inhibitor. Chronic photoaging also features an increased deposition of abnormal elastic fibers. The genes of elastin and fibrillin are known to be upregulated in chronically photodamaged skin (
      • Bernstein E.F.
      • Chen Y.Q.
      • Tamai K.
      • et al.
      Enhanced elastin and fibrillin gene expression in chronically photodamaged skin.
      ). In vitro, vitamin C downregulates the biosynthesis of elastin by fibroblasts (
      • Davidson J.M.
      • LuValle P.A.
      • Zoia O.
      • Quaglino D.
      • Giro M.G.
      Ascorbate differentially regulates elastin and collagen biosynthesis in vascular smooth muscle cells and skin fibroblasts by pretranslational mechanisms.
      ) at the concentration that increases the level of mRNA for collagen in the same cells. In this study, the mRNA for elastin and fibrillins was not modified by vitamin C. Glycosaminoglycans and proteoglycans also accumulate in photoaged skin (
      • Bernstein E.F.
      • Underhill C.B.
      • Hahn P.J.
      • Brown D.B.
      • Uitto J.
      Chronic sun exposure alters both the content and distribution of dermal glycosaminoglycans.
      ). Decorin is a small proteoglycan closely associated with the collagen fibrils. The increased level of its mRNA in the vitamin C-treated skin is at the limit of significance. The level of accumulated solar damage estimated in the placebo-treated side of the testers does not correlate with the extent of responsiveness to the topically applied vitamin. This suggests that a reduced biosynthetic activity related to chronologic aging and/or a low tissue concentration of the vitamin are the likely targets of the topical treatment.
      The excellent technical assistance of M.J. Nix and T. Heyeres in the preparation of RNA samples and reverse transcription–PCR assays is gratefully acknowledged. The help of A. Albert for the statistical analysis of the data has been greatly appreciated. We thank C. Clabeck for the preparation of the manuscript. This work was supported in part by grants of the Belgian FRSM and the Faculty of Medicine of the University of Liège.

      References

        • Banhegyi G.
        • Braun L.
        • Csala M.
        • Puskas F.
        • Mandl J.
        Ascorbate metabolism and its regulation in animals.
        Free Rad Biol Med. 1997; 23: 793-803https://doi.org/10.1016/s0891-5849(97)00062-2
        • Bartley W.
        • Krebs H.A.
        • O'Brien J.R.P.
        Vitamin C requirement of human adults.
        Med Res Counc Spec Rep Ser. 1953; 280: 1-179
        • Bergman I.
        • Loxley R.
        Two improved and simplified methods for the spectrophotometric determination of hydroxyproline.
        Anal Chem. 1963; 35: 1961-1965
        • Bernstein E.F.
        • Chen Y.Q.
        • Tamai K.
        • et al.
        Enhanced elastin and fibrillin gene expression in chronically photodamaged skin.
        J Invest Dermatol. 1994; 103: 182-186
        • Bernstein E.F.
        • Underhill C.B.
        • Hahn P.J.
        • Brown D.B.
        • Uitto J.
        Chronic sun exposure alters both the content and distribution of dermal glycosaminoglycans.
        Br J Dermatol. 1996; 135: 255-262
        • Colige A.
        • Li S.-W.
        • Sieron A.L.
        • Nusgens B.V.
        • Prockop D.J.
        • Lapière C.M.
        cDNA cloning and expression of bovine procollagen I N-proteinase: a new member of the superfamily of zinc-metalloproteinases with binding sites for cells and other matrix components.
        Proc Natl Acad Sci USA. 1997; 94: 2374-2379
        • Colven R.M.
        • Pinnell S.R.
        Topical vitamin C in aging.
        Clin Dermatol. 1996; 14: 227-234https://doi.org/10.1016/0738-081x(95)00158-c
        • Danielson K.G.
        • Baribault H.
        • Holmes D.F.
        • Graham H.
        • Kadler K.E.
        • Iozzo R.V.
        Targeted disruption of decorin leads to abnormal collagen fibril morphology and skin fragility.
        J Cell Biol. 1997; 136: 729-743
        • Darr D.
        • Combs S.
        • Dunston S.
        • Manning T.
        • Pinnell S.
        Topical Vitamin-C protects porcine skin from ultraviolet radiation-induced damage.
        Br J Dermatol. 1992; 127: 247-253
        • Darr D.
        • Dunston S.
        • Faust H.
        • Pinnell S.
        Effectiveness of antioxidants (Vitamin C and E) with and without sunscreens as topical photoprotectants.
        Acta Derm Venereol (Stockh). 1996; 76: 264-268
        • Davidson J.M.
        • LuValle P.A.
        • Zoia O.
        • Quaglino D.
        • Giro M.G.
        Ascorbate differentially regulates elastin and collagen biosynthesis in vascular smooth muscle cells and skin fibroblasts by pretranslational mechanisms.
        J Biol Chem. 1997; 272: 345-352
        • Epstein E.H.
        [alpha1 (III)]3 Human Skin Collagen. Release by pepsin digestion and preponderance in fetal life.
        J Biol Chem. 1974; 249: 3225-3231
        • Fisher G.J.
        • Talwar H.S.
        • Lin J.Y.
        • Voorhees J.J.
        Molecular mechanisms of photoaging in human skin in vivo and their prevention by all-trans retinoic acid.
        Photochem Photobiol. 1999; 69: 154-187
        • Freeman W.M.
        • Walker S.J.
        • Vrana K.E.
        Quantitative RT-PCR: Pitfalls and potential.
        Biotechniques. 1999; 26: 112-125
        • Geesin J.C.
        • Darr D.
        • Kaufman R.
        • Murad S.
        • Pinnell S.R.
        Ascorbic acid specifically increases type I and type III procollagen messenger RNA levels in human skin fibroblast.
        J Invest Dermatol. 1988; 90: 420-424
        • Gross J.
        • Page I.H.
        Connective Tissue, Thrombosis and Atherosclerosis. Academic Press, New York1959: 77-95
        • Hamalainen E.R.
        • Jones T.A.
        • Sheer D.
        • Taskinen K.
        • Pihlajaniemi T.
        • Kivirikko K.I.
        Molecular cloning of human lysyl oxidase and assignment of the gene to chromosome 5q23.3–31.2.
        Genomics. 1991; 11: 508-516
        • Hirschmann J.V.
        • Raugi G.J.
        Adult scurvy.
        J Am Acad Dermatol. 1999; 41: 895-906
        • Kivirikko K.I.
        • Prockop D.J.
        Enzymatic hydroxylation of proline and lysine in protocollagen.
        Proc Nat Acad Sci USA. 1967; 57: 782-789
        • Lapière Ch.M.
        • Piérard G.E.
        Skin procollagen peptidase in normal and pathologic conditions.
        J Invest Dermatol. 1974; 62: 582-586
        • Legrand Y.
        • Lapière Ch M.
        • Pignaud G.
        • Caen J.
        Microméthode d'extraction du collagène à partir de biopsie de peau humaine.
        Pathol Biol. 1969; 17: 991-993
        • Levene C.I.
        Ascorbic acid and collagen synthesis in cultured fibroblasts.
        Ann N Y Acad Sci. 1975; 258: 288-306
        • Levine M.
        • Rumsey S.C.
        • Daruwala R.
        • Park J.B.
        • Wang Y.
        Criteria and recommendations for vitamin C intake.
        JAMA. 1999; 281: 1415-1423
        • Li S.W.
        • Sieron A.L.
        • Fertala A.
        • Hojima Y.
        • Arnold W.V.
        • Prockop D.J.
        The C-proteinase that processes procollagens to fibrillar collagens is identical to the protein previously identified as bone morphogenic protein-1.
        Proc Natl Acad Sci USA. 1996; 93: 5127-5130
        • Lyons B.L.
        • Schwarz R.I.
        Ascorbate stimulation of PAT cells causes an increase in transcription rates and a decrease in degradation rates of procollagen mRNA.
        Nucleic Acids Res. 1984; 12: 2569-2579
        • Millis A.J.
        • Hoyle T.M.
        • McCue H.M.
        • Martini H.
        Differential expression of metalloproteinase and tissue inhibitor of metalloproteinase genes in aged human fibroblasts.
        Exp Cell Res. 1992; 201: 373-379
        • Nagase H.
        • Woessner J.F.
        Matrix metalloproteinases.
        J Biol Chem. 1999; 274: 21491-21494
        • Nimni M.E.
        • Bavetta L.A.
        Collagen synthesis and turnover in the growing rat under the influence of methylprednisolone.
        Proc Soc Exp Biol Med. 1964; 117: 618-624
        • Padh H.
        Cellular functions of ascorbic acid.
        Biochem Cell Biol. 1990; 68: 1166-1173
        • Pauling L.
        Effect of ascorbic acid on incidence of spontaneous mammary tumors and UV-light-induced skin tumors in mice.
        Am J Clin Nutr. 1991; 54: 1252S-1255S
        • Peterkofsky B.
        the effect of ascorbic acid on collagen polypeptide synthesis and proline hydroxylation during the growth of cultured fibroblasts.
        Arch Biochem Biophys. 1972; 152: 318-328
        • Phillips C.L.
        • Tajima S.
        • Pinnell S.R.
        Ascorbic acid and transforming growth factor-beta1 increase collagen biosynthesis via different mechanisms—coordinate regulation of proalpha1(I) and proalpha1(III) collagens.
        Arch Biochem Biophys. 1992; 295: 397-403
        • Pinnell S.R.
        • Murad S.
        • Darr D.
        Induction of collagen synthesis by ascorbic acid.
        Arch Dermatol. 1987; 123: 1684-1686
        • Ponec M.
        • Weerheim A.
        • Kempenaar J.
        • Mulder A.A.
        • Gooris G.S.
        • Bouwstra J.
        • Mommaas A.M.
        The formation of competent barrier lipids in reconstructed human epidermis requires the presence of vitamin C.
        J Invest Dermatol. 1997; 109: 348-355
        • Ramirez F.
        • Pereira L.
        The fibrillins.
        Int J Biochem Cell Biol. 1999; 31: 255-259https://doi.org/10.1016/s1357-2725(98)00109-5
        • Rosenbloom J.
        • Harsch M.
        • Jimenez S.
        Hydroxyproline content determines the denaturation temperature of chick tendon collagen.
        Arch Biochem Biophys. 1973; 158: 478-484
        • Sauberlich H.E.
        Pharmacology of vitamin C.
        Annu Rev Nutr. 1994; 14: 371-391
        • Schectman G.
        Estimating ascorbic acid requirement for cigarette smokers.
        Ann N Y Acad Sci. 1993; 686: 335-345
        • Schwarz R.I.
        Procollagen secretion meets the minimum requirements for the rate-controlling step in the ascorbate induction of procollagen synthesis.
        J Biol Chem. 1985; 260: 3045-3049
        • Seite S.
        • Colige A.
        • Piquemal-Vivenot P.
        • et al.
        A full-UV spectrum absorbing daily use cream protects human skin against biological changes occurring in photoaging.
        Photodermatol Photoimmunol Photomed. 2000; 16: 147-155
        • Sohal R.S.
        • Weindruch R.
        Oxidative stress, caloric restriction and aging.
        Science. 1996; 273: 59-63
        • Tebbe B.
        • Wu S.L.
        • Geilen C.C.
        • Eberle J.
        • Kodelja V.
        • Orfanos C.E.
        L-ascorbic acid inhibits UVA-induced lipid peroxidation and secretion of IL-1 alpha and IL-6 in cultured human keratinocytes in vitro.
        J Invest Dermatol. 1997; 108: 302-306
        • Thomas D.P.
        Sailors scurvy and science.
        J R Soc Med. 1997; 90: 50-54
        • Utoguchi N.
        • Ikeda K.
        • Saeki K.
        • et al.
        Ascorbic acid stimulates barrier function of cultured endothelial cell monolayer.
        J Cell Physiol. 1995; 163: 393-399
        • Varani J.
        • Warner R.L.
        • Gharaee-Kermani M.
        • et al.
        Vitamin A: antagonizes decreased cell growth and elevated collagen-degrading matrix metalloproteinases and stimulates collagen accumulation in naturally aged human skin.
        J Invest Dermatol. 2000; 114: 480-486https://doi.org/10.1046/j.1523-1747.2000.00902.x
        • Welch R.V.
        • Wang Y.
        • Crossman A. Jr
        • Park J.B.
        • Kirk K.L.
        • Levine M.
        Accumulation of vitamin C (ascorbate) and its oxidized metabolite dehydroascorbic acid occurs by separate mechanisms.
        J Biol Chem. 1995; 270: 12584-12592
        • Wlaschek M.
        • Briviba K.
        • Stricklin G.P.
        • Sies H.
        • Scharffetter-Kochanek K.
        Singlet oxygen may mediate the ultraviolet A-induced synthesis of interstitial collagenase.
        J Invest Dermatol. 1995; 104: 194-198