Materials and Methods

Biopsy samples

In a group of eight healthy male volunteers (aged 24–36 y) admitted to the Bourn Hall Clinic (LRG Biosciences, Cambridge), five full-thickness skin biopsies (at least 2 cm apart) were taken with a 4 mm disposable surgical punch from the sacrolumbar region using a local anesthetic. The wounds were sutured with a single #4.0 Nylon (Ethilon) suture and covered with an appropriate dressing. At 3, 6, 8, 12, and 16 wk after the initial biopsies, the wound sites were re-excised using a 5 mm punch under similar conditions to the initial excision: a sixth biopsy was obtained by re-excision of the 3 wk wound site 24 wk after the initial re-excision. The samples were immediately frozen and maintained at –70°C until analysis. All samples were obtained with the informed consent of the donors under protocols approved by the local ethical committee.

Preparation of tissue samples

Biopsy samples (17–40 mg wet weight) were minced with scissors and powdered under liquid nitrogen using a Spex Mill (model 6800; Glen Creston, London, UK) fitted with stainless steel micro vials and impactors. After warming to room temperature, the milled tissue was transferred to glass tubes with 40.5 ml portions of acetone:water (1:1), and the residue obtained by centrifuging at 4000 g for 5 min was defatted with acetone (1.5 ml), dried, and weighed. This procedure resulted in a homogeneous mixture of milled tissue suitable for subsampling.

Solubilization by digestion with pepsin

Weighed portions of defatted and dried tissue (0.4–0.9 mg) were equilibrated with 1 ml of 0.5 M acetic acid (adjusted to pH 2 with HCl) overnight at 4°C to allow swelling; all subsequent extraction procedures were performed at 4°C. After the addition of pepsin (Sigma, Poole, UK) to a final concentration of 1 mg per ml, extraction with gentle agitation was continued for 24 h. Following centrifugation at 13,000g, the residue was re-extracted with a similar amount of pepsin solution for a further 24 h and centrifuged at 13,000g, and the supernatants were combined. Preliminary experiments showed that further extractions with pepsin did not release significant amounts of collagen (results not shown). After removing an aliquot of the supernatant solution (about 5% of the total) for hydroxyproline determination, the solubilized material was lyophilized. The proportion of collagen solubilized by the pepsin was determined by hydroxyproline measurements. The supernatant aliquot and the whole insoluble residue were separately hydrolyzed in 6 M HCl (107°C; 20 h), and the acid was removed by evaporation to dryness. The hydroxyproline content of the fractions was determined colorimetrically, as previously described (^{Firschein and Shill, 1966}) except that 2-methoxyethanol was used as diluent. All of the commercial sources of pepsin we tested were found to contain significant amounts of hydroxyproline and, to correct for contamination in the pepsin preparation used, hydroxyproline values were determined in hydrolysates of pepsin alone. Amounts of collagen were calculated assuming 300 mol hydroxyproline per mole collagen.

Measurement of collagen types I and III

The proportions of collagen types I and III in the pepsin-solubilized tissue was determined by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (6% running gel) using the Laemmli buffer system (^{Laemmli, 1970}) with and without delayed reduction by 2-mercaptoethanol (^{Sykes et al, 1976}), necessary to effect separation of the cysteine-containing 1(III) chains from 1(I). The samples were dissolved in SDS sample buffer using the known hydroxyproline content to ensure equal collagen loading of about 0.5 mg per ml. After electrophoresis, the gels were stained for 1.5 h in Simply Blue SafeStain (Invitrogen, Paisley, UK) and destained overnight in water. The 1(III) chains in the electrophoresis gels were often detected as a doublet after the delayed reduction procedure, and this was more prominent in the sample lanes compared with the control standards run with each gel (see Figure 1). The reason for this interference is unclear but may be connected with the fact that, contrary to most reported applications of this technique, no preliminary purification of the pepsin digest was undertaken prior to gel electrophoresis. Analysis of the samples without delayed reduction confirmed that no other bands were eluted in the position of the 1(III) bands (data not shown). The relative amounts of the 1(I) and 1(III) chains in the scanned gels was determined using Phoretix v 4.0 software (Nonlinear Dynamics, Newcastle upon Tyne, UK), with reference to a standard curve. The latter was constructed by electrophoretic analysis of mixtures containing collagens I and III, purified by salt precipitation, in varying proportions from 0% to 50% collagen III. A statistical sigmoid curve fit was applied to the standard data points from which sample data were calculated using FigP software (Biosoft, Cambridge, UK). Where possible, all tissue samples from a single volunteer were run together on the same electrophoresis gel, together with three quality control standards containing 10%, 20%, and 30% collagen III with respect to collagen I: coefficients of variation for analyses of the standards (n=10) were 5.9%, 6.0%, and 4.0%, respectively. A purified collagen V standard was included on each gel to ensure that its constituent chains did not interfere with the measurement of 1(III) chains.

Figure 1.

Electrophoresis of collagens solubilized by pepsin digestion of human skin biopsies. Delayed reduction with 2-mercaptoethanol was performed after 1 h and the proportions of collagen I and III were estimated by densitometric scanning of the 1(I) and 1(III) bands. The upper panel shows a typical stained gel with baseline biopsy samples run in lanes 1–5, with the 3, 6, 8, 12, and 16 wk time-points in lanes 6–10, respectively. Control standards containing 10%, 20%, and 30% collagen III relative to collagen I+III were run in lanes 11–13, respectively, and lane 14 contains a collagen V standard. The 1(III) band in the unfractionated samples often occurred as a doublet as indicated. The lower panel shows a standard curve constructed by electrophoresis of a series of collagen I:III mixtures.

Full figure and legend (86K)

Analysis of cross-links in collagen and elastin

The pyridinium cross-links of collagen, Pyd and Dpd, as well as the pyridinium cross-links of elastin, desmosine (Des) and isodesmosine (Ide), were analyzed by reversed-phase high performance liquid chromatography (HPLC) using automated sample preparation by partition chromatography on cellulose-containing extraction columns (Gilson ASPEC system; Anachem, Luton, UK) as described previously (^{Pratt et al, 1992}). Accurately weighed samples of the milled, defatted, and dried tissues were hydrolyzed in constant boiling (5.7 M) HCl in sealed tubes at 107°C for 191 h, after which time the acid was removed by evaporation and the hydrolysates were redissolved in water (1.0 ml). Appropriate aliquots, diluted to 0.5 ml, were applied to sample tubes of the ASPEC instrument, which was set up with extraction columns containing 100 mg cellulose CC31 (Whatman, Maidstone, UK) and a mobile phase of butan-1-ol:acetic acid:water (4:1:1 vol/vol/vol). The cross-links were separated by HPLC using a 4.6100 mm Microsorb C18 column (Rainin; Anachem) run with 20 mM heptafluorobutyric acid with an acetonitrile gradient. Column effluent was monitored using an FP-920 fluorescence detector (excitation 295 nm, emission 400 nm; Jasco, Great Dunmow, UK) to quantify Pyd and Dpd, and a Gilson 118 UV detector (278 nm; Anachem) connected in tandem was used to quantify Des and Ide. Under the chromatographic conditions used, Des and Ide coeluted and were quantified together using a mean extinction coefficient determined for a standard containing equimolar amounts of the cross-links; pilot experiments in which an extended chromatographic gradient was used confirmed a previous report (^{Ono and Yamauchi, 1994}) that human skin contained about equal amounts of Des and Ide. Overall recovery of both the collagen and elastin cross-links was quantified by reference to a fluorescent, semisynthetic internal standard, O-acetyl-pyridinoline (^{Pratt et al, 1992}), which was added to the hydrolysates at the initiation of sample processing. Quality control standards containing high, medium, and low amounts of the cross-links were run in duplicate with each batch of analyses. The mean coefficients of variation (n=18) for Pyd, Dpd, and Des+Ide analyses were 3.3%, 4.8%, and 6.6%, respectively.

The mature skin cross-link, HHL, was analyzed using a newly developed method in which hydrolysates of the biopsy samples were prefractionated as for the pyridinium cross-links on columns of cellulose CC31 using the Gilson ASPEC. Instead of eluting with heptafluoro-butyric acid, however, the cellulose columns were eluted with 0.75 ml 50 mM HCl and sodium citrate, pH 2.0, was added to the eluate to a final concentration of 50 mM. The samples were submitted to ion exchange chromatography on a column (2.1250 mm) of sulfonated polystyrene resin beads (5 m) 8% cross-linked with divinyl benzene (Locarte, London, UK), equilibrated with 50 mM sodium citrate, pH 2.0, and eluted at 0.25 ml per min with a linear gradient from 40% to 52% buffer B (200 mM sodium citrate, pH 6.50) over 30 min. HHL was quantified by postcolumn derivatization with ortho-phthaldialdehyde in borate buffer with fluorescence monitoring at excitation 355 nm/emission 450 nm using external standardization with known amounts of purified HHL submitted to the same prefractionation procedures. The HHL standard, isolated from ovine dermis (7-y-old animal), was fully characterized by nuclear magnetic resonance and mass spectrometry. A standard stock solution in 30 mM HCl was prepared gravimetrically from the trihydro-chloride salt (M_r=553.9). The coefficient of variation for HHL analysis (n=12) was 5.7%.

Statistical analysis

Overall significance of difference between groups at various time-points, as well as a step-wise analysis of the significance of specific time-points with successive samples, were analyzed using Genstat (Lawes Agricultural Trust, Rothamsted, UK); p-values of less than 0.05 were taken as significant.

Results

Mean values for the collagen contents, extractability by pepsin digestion, and the proportions of collagen type III in the skin biopsy samples taken at various time-points after initial biopsy are shown in Table I, together with the significance of differences between time-points. The collagen content was slightly lower in the re-excised tissues compared with the baseline values (p<0.001), but stepwise comparisons between time-points indicated that most of the collagen mass was recovered by 6 wk and no further significant changes occurred thereafter. The amounts of collagen extracted from re-excised wound tissue by pepsin digestion were significantly higher than for the initial biopsies at all time-points (p<0.001), and the proportion of collagens extracted continued to increase at each time-point up to 24 wk (p<0.001), at which time about 90% of the skin collagen was solubilized by the enzyme. The proportions of collagen III relative to collagen I in the re-excised wound tissues were significantly higher than in the initial biopsies (p<0.001), with the main increases occurring in the period up to 6 wk; there was a marginal decrease in the relative amounts of collagen III in samples taken after the 12 wk time-point (p<0.05), as shown in Table I.

Table I - Collagen extractability and composition in 5 mm timed human skin biopsy samples taken after initial 4 mm biopsies at baseline, .

Full table

The skin cross-link data are presented in Table II. For HHL analysis, the prefractionation step resulted in the removal of most other amino acids (Figure 2). The concentrations of HHL relative to collagen were higher in the baseline biopsies than in the re-excision samples, and there were significant decreases in amounts throughout most of the follow-up period (Table II).

Figure 2.

Chromatography of the mature skin collagen cross-link, HHL. Hydrolysates of skin biopsies were prefractionated as described in Materials and Methods and submitted to ion exchange chromatography with postcolumn derivatization using ortho-phthaldialdehyde and monitoring fluorescence at excitation 355 nm, emission 450 nm. The three overlaid chromatograms are of 550 pmol HHL standard (dotted line), a baseline biopsy sample (broken line), and a 24 wk re-excision biopsy (full line), the last two containing similar amounts of tissue hydroxyproline.

Full figure and legend (16K)

Table II - Cross-link concentrations in 5 mm timed human skin biopsy samples taken after initial 4 mm biopsies at baseline, .

Full table

Collagen and elastin pyridinium cross-links were analyzed concurrently in the same reversed-phase chromatographic analysis using dual monitoring and a fluorescent internal standard (Figure 3), which was used to correct for recovery from the prefractionation column. The combined Des and Ide concentrations, expressed relative to the dry weight of tissue, were significantly lower in the re-excised wounds (p<0.001), and, although the mean values were progressively lower up to the 24 wk time-point, these changes were not statistically significant (Table II). Pyd concentrations were significantly higher in the re-excised tissues (p<0.001) whereas Dpd showed the opposite trend, so that the Pyd/Dpd ratio in the biopsies showed a marked increase during the course of the experiment (Figure 4). These changes occurred predominantly in the period up to 8 wk after the initial biopsy and there were no statistically significant changes after that time.

Figure 3.

HPLC separation of pyridinium cross-links of elastin and collagen. Acid hydrolysates of skin biopsy samples were prefractionated on cellulose columns using an automated Gilson ASPEC system followed by reversed-phase HPLC with dual monitoring. The upper panel shows the coelution of Des and Ide detected by ultraviolet absorption and the lower panel shows the positions of Pyd and Dpd detected by fluorescence. A fluorescent internal standard (IS), O-acetyl-pyridinoline, was used to quantify both the collagen- and elastin-derived cross-links.

Full figure and legend (47K)

Figure 4.

Changes in the collagen cross-link ratio, Pyd/Dpd, in skin biopsy samples. The measured Pyd/Dpd ratios in five initial biopsies and six re-excision samples are shown as the meanSD (n=8, except for 24 wk where n=6). The significance of the difference between the relevant group and groups at all subsequent time-points is shown by asterisks (p<0.001).

Full figure and legend (12K)

There was a close inverse relationship between the extractability of collagen by pepsin and the concentration of HHL cross-links for each group of biopsy samples, and plotting individual values revealed a significant linear correlation (r²=0.89, p<0.0001; Figure 5).

Figure 5.

Changes in the extractability of skin collagen with pepsin in relation to mature cross-link content. (a) The extractability of skin collagen by pepsin digestion for five initial biopsy sites and re-excision samples at six time-points up to 24 wk expressed as meanSD (n=8, except for 24 wk where n=6) is shown in comparison with values for HHL concentrations expressed similarly for the corresponding samples. Statistically significant differences are omitted for clarity but are shown in Table I and Table II. (b) Linear regression of individual values for pepsin extractability and HHL concentration for all biopsy samples with the equation y=-0.017+1.71 and significant correlation coefficient r²=0.89 (p<0.0001).

Full figure and legend (34K)

Discussion

The results of this study suggest that the remodeling phase of repair following excisional wounds to human skin extends at least to the 6 mo period of follow-up. For many of the properties associated with skin maturity, such as the susceptibility to enzyme digestion and the proportions of collagen III, analyses of the time course revealed little sign of normalization within the study period.

Several previous studies of excisional wound repair have described the increased proportion of collagen III that occurs initially (^{Bailey et al, 1975b}), but an assumption is generally made that the collagen I/III ratio normalizes quite rapidly during the remodeling phase. This study showed that, following an initial increase up to 6 wk post injury, there was a marginal decrease after 12 wk but the ratio was still 70% above the baseline value 24 wk after the initial biopsy. These changes should be viewed in the context of the increasing solubilization of skin collagen by pepsin, so that analyses at the later time-points represented a higher proportion of the total collagen. In addition, the gel electrophoresis analyses showed that there was a higher proportion of cross-linked and components in the baseline samples (see Figure 1), and collagen I/III ratios calculated on the measured 1 chains can only be approximate. Assuming that the one-third of collagen extracted by pepsin from the baseline samples was representative, the data show that markedly higher proportions of collagen III persist in the newly synthesized skin collagen for more than 6 mo after the initial injury. These changes are likely to reflect considerable effects on the physical and mechanical properties of the tissue, particularly in terms of fibril diameter and tensile strength (^{Birk and Mayne, 1997;Brinckmann et al, 1999}).

A surprising finding in this study was that the susceptibility of the wound tissue collagen to degradation by pepsin continued to increase throughout the 24 wk follow-up period, whereas many of the other parameters measured showed little change after the 6 wk time-point. The measured concentrations of the mature collagen cross-link, HHL, however, provided a possible explanation for these changes. The close negative correlation between pepsin solubility and HHL concentration suggests the possibility of a causal relationship, and consideration of the structural restrictions imposed by the mature cross-link supports such a contention. Thus, as the predominant lysine-derived collagen cross-link in mature skin (^{Yamauchi et al, 1987}), HHL is thought to link primarily a single 1(I) chain C-telopeptide to two helical sites, which are likely to be on different molecules (^{Mechanic et al, 1987}). This is in contrast to the pyridinium cross-links that stabilize the quarter-stagger overlap by linking two telopeptides to a single residue in the helix. The formation of HHL cross-links therefore introduces interhelical bonds within the collagen fibrils and a substantial network of this form of intermolecular bond would result in resistance to degradation by pepsin, as this enzyme cleaves native collagen only at telopeptide-helix junctions. The amounts of HHL in bovine and human skin have been shown to increase with age suggesting a link with the solubility of these tissues (^{Yamauchi et al, 1988}). The concentrations of HHL in human skin have been shown to vary with sampling site (^{Yamauchi et al, 1991}), but this variable was avoided in this study as all biopsies were taken from the same region of the lower back. It is possible that a tetra-functional cross-link detected after borohydride reduction as histidinohydroxymerodesmosine in young skin (^{Robins and Bailey, 1973;Tanzer et al, 1973}) may mature to give interhelical cross-links; the amounts of this intramolecular aldol-derived compound decrease with age (^{Robins et al, 1973}) but the mature form has not yet been identified. The amounts of Pyd and Dpd cross-links in skin were low, consistent with the lack of hydroxylated telopeptide lysine residues in normal dermal tissue (^{Eyre, 1984}). These cross-links are likely to reflect predominantly the presence of blood vessels in the tissue but, as the facia has been shown to have high concentrations of Pyd (^{Istok et al, 2001}), some contamination of dermal tissue with facia may also contribute to the lack of consistency in reported values (^{Moriguchi and Fujimoto, 1979;Pasquali et al, 1995;Wan and Chow, 1999}). In fibrotic skin or in scar tissue, there is an increase in Pyd concentration associated with an increase in the hydroxylation of lysine residues (^{Brinckmann et al, 1999}). In this study, the concentrations of Pyd in collagen almost doubled and the changes in Pyd/Dpd ratio indicate that these data reflect altered lysine hydroxylation in the newly formed tissue. The maximal values for Pyd concentrations were still very low, representing about one Pyd cross-link per 14 collagen molecules, and the functional consequences of these changes are likely to be minor. As a marker of tissue normalization, however, the Pyd/Dpd ratio may have some application, although longer term studies need to be performed.

The concentrations of desmosine cross-links have been used as an indicator of the elastin content of tissues (^{Starcher, 1977}). The time course of changes in desmosine cross-link concentrations in this study indicated that there was little replacement of the elastic tissue in the wound area over the full 6 mo of follow-up. This observation appears to be consistent with results from skin transplantation studies, which showed that 4–5 y elapsed before elastic tissue could be detected (^{Compton et al, 1989;Putland et al, 1995}).

At the earlier stages after injury, much of the re-excision tissue contained the surrounding normal tissue (which had been brought together with a single suture) and accumulation of new collagenous tissue in the 3–6 wk samples probably contributes to the marked changes in most of the measured parameters. Indeed, previous studies have shown that water solubility of the tissue collagen (80°C, 30 min) is not increased until at least 2 wk after injury (S. P. Robins and I. James, unpublished results), suggesting that newly formed collagen is not a significant proportion of the total until after this time. The fact that the pepsin solubility increased throughout the 24 wk follow-up indicates that the area of remodeling continued to expand during this period, presumably as the tissue adjusted to the mechanical stimuli emanating from the necessary reorganization around the wound site. The observed plateau in the amounts of collagen III is therefore likely to be a result of the combined, opposing effects of a continuous normalization of fibrillar collagen characteristics together with the expanding area of remodeling.

In summary, this time course study of excisional wound healing in human skin has demonstrated that the marked increases in the proportion of collagen III relative to collagen I persist up to 6 mo after injury. Analyses of desmosine cross-links indicated that elastic tissue was lost rapidly and not regenerated within the study period, whereas the pyridinium cross-links of collagen revealed a rapid increase in Pyd/Dpd ratio that did not normalize within 6 mo. The most striking changes in collagen properties during healing were the large increases in extractability by pepsin, mirrored by inverse changes in the concentrations of the mature skin cross-link, HHL. The results provide the clearest evidence to date of a causal relationship between the content of HHL in skin collagen and its susceptibility to digestion by pepsin.

References

1. Bailey, AJ, Bazin, S, Sims, TJ, Le Lous, M, Nicoletis, C, Delaunay, A: Characterization of the collagen of human hypertrophic and normal scars. Biochim Biophys Acta 1975a 405: 412–421,  | PubMed | ISI | ChemPort |
2. Bailey, AJ, Sims, TJ, Le Lous, M, Bazin, S: Collagen polymorphism in experimental granulation tissue. Biochem Biophys Res Commun 1975b 66: 1160–1165,  | Article | PubMed | ISI | ChemPort |
3. Berthod, F, Germain, L, Li, H, Xu, W, Damour, O, Auger, FA: Collagen fibril network and elastic system remodeling in a reconstructed skin transplanted on nude mice. Matrix Biol 2001 20: 463–473,  | Article | PubMed | ISI | ChemPort |
4. Birk, DE, Mayne, R: Localization of collagen types I, III and V during tendon development. Changes in collagen types I and III are correlated with changes in fibril diameter. Eur J Cell Biol 1997 72: 352–361,  | PubMed | ISI | ChemPort |
5. Brinckmann, J, Notbohm, H, Tronnier, M, et al: Overhydroxylation of lysyl residues is the initial step for altered collagen cross-links and fibril architecture in fibrotic skin. J Invest Dermatol 1999 113: 617–621,  | Article | PubMed | ISI | ChemPort |
6. Compton, CC, Gill, JM, Bradford, DA, Regauer, S, Gallico, GG, O'Connor, NE: Skin regenerated from cultured epithelial autografts on full-thickness burn wounds from 6 days to 5 years after grafting. A light, electron microscopic and immunohistochemical study. Lab Invest 1989 60: 600–612,  | PubMed | ISI | ChemPort |
7. Ehrlich, HP, Krummel, TM: Regulation of wound healing from a connective tissue perspective. Wound Repair Regen 1996 6: 186–193,
8. Epstein, EH Jr, Munderloh, NH: Human skin collagen. Presence of type I and type III at all levels of the dermis. J Biol Chem 1978 253: 1336–1337,  | PubMed | ISI | ChemPort |
9. Eyre, DR: Cross-linking in collagen and elastin. Ann Rev Biochem 1984 53: 717–748,  | PubMed | ISI | ChemPort |
10. Firschein, HE, Shill, JP: The determination of total hydroxyproline in urine and bone extracts. Anal Biochem 1966 14: 296–304,  | Article | PubMed | ISI | ChemPort |
11. Garrone, R, Lethias, C, LeGuellec, D: Distribution of minor collagens during skin development. Microsc Res Tech 1997 38: 407–412,  | Article | PubMed | ISI | ChemPort |
12. Istok, R, Bely, M, Stancikova, M, Rovensky, J: Evidence for increased pyridinoline concentration in fibrotic tissues in diffuse systemic sclerosis. Clin Exp Dermatol 2001 26: 545–547,  | Article | PubMed | ISI | ChemPort |
13. Laemmli, UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970 227: 680–685,  | Article | PubMed | ISI | ChemPort |
14. Lamme, E, deVries, H, vanVeen, H, Gabbiani, G, Westerhof, W, Middelkoop, E: Extracellular matrix characterization during healing of full-thickness wounds treated with a collagen/elastin dermal substitute shows improved skin regeneration in pigs. J Histochem Cytochem 1996 44: 1311–1322,  | PubMed | ISI | ChemPort |
15. Lindblad, WJ: Perspective article: Collagen expression by novel cell populations in the dermal wound environment. Wound Repair Regen 1998 6: 186–193,  | Article | PubMed | ChemPort |
16. Mechanic, GL, Katz, EP, Henmi, M, Noyes, C, Yamauchi, M: Locus of a histidine-based, stable trifunctional, helix to helix collagen crosslink: Stereospecific collagen structure of type I skin fibrils. Biochemistry 1987 26: 3500–3509,  | Article | PubMed | ISI | ChemPort |
17. Moriguchi, T, Fujimoto, D: Crosslink of collagen in hypertrophic scar. J Invest Dermatol 1979 72: 143–145,  | Article | PubMed | ISI | ChemPort |
18. Ono, S, Yamauchi, M: Elastin cross-linking in the skin from patients with amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 1994 57: 94–96,  | PubMed | ISI | ChemPort |
19. Pasquali, M, Still, MJ, Dembure, PP, Elsas, LJ: Pyridinium cross-links in heritable disorders of collagen. Am J Hum Genet 1995 57: 1508–1510,  | PubMed | ISI |
20. Pratt, DA, Daniloff, Y, Duncan, A, Robins, SP: Automated analysis of the pyridinium crosslinks of collagen in tissue and urine using solid-phase extraction and reversed-phase high-performance liquid chromatography. Anal Biochem 1992 207: 168–175,  | Article | PubMed | ISI | ChemPort |
21. Putland, M, Snelling, CF, Macdonald, I, Tron, VA: Histologic comparison of cultured epithelial autograft and meshed expanded split-thickness skin graft. J Burn Care Rehabil 1995 16: 627–640,  | Article | PubMed | ChemPort |
22. Reiser, K, McCormick, RJ, Rucker, RB: Enzymatic and nonenzymatic cross-linking of collagen and elastin. FASEB J 1992 6: 2439–2449,  | PubMed | ISI | ChemPort |
23. Robins, SP: Turnover and crosslinking of collagen. In: Weiss JB, Jayson MIV (eds). Collagen in Health and Disease. 1982 Edinburgh: Churchill Livingstone, p 160–178,
24. Robins, SP, Bailey, AJ: The characterisation of fraction C (histidinohydroxymerodesmosine), a possible artifact produced during the reduction of collagen fibres with borohydride. Biochem J 1973 135: 657–665,  | PubMed | ISI | ChemPort |
25. Robins, SP, Brady, JD: Collagen cross-linking and metabolism. In: Bilezikian JP, Raisz LG, Rodan GA (eds.) Principles of Bone Biology. 2002 San Diego: Academic Press, p 211–223,
26. Robins, SP, Shimokomaki, M, Bailey, AJ: The chemistry of the collagen crosslinks: Age-related changes in the reducible components of intact bovine collagen fibres. Biochem J 1973 131: 771–780,  | PubMed | ISI | ChemPort |
27. Starcher, BC: Determination of the elastin content of tissues by measuring desmosine and isodesmosine. Anal Biochem 1977 79: 11–15,  | Article | PubMed | ISI | ChemPort |
28. Sykes, B, Puddle, B, Francis, M, Smith, R: The estimation of two collagens from human dermis by interrupted gel electrophoresis. Biochem Biophys Res Commun 1976 72: 1472–1480,  | Article | PubMed | ISI | ChemPort |
29. Tanzer, ML, Housley, T, Berube, L, Fairweather, R, Franzblau, C, Gallop, PM: Structure of two histidine-containing cross-links from collagen. J Biol Chem 1973 248: 393–402,  | PubMed | ISI | ChemPort |
30. Wan, KC, Chow, TC: Age-related changes in the concentration of pyridinoline cross-links in human skin. Ann Clin Biochem 1999 36: 666–668,  | PubMed | ISI | ChemPort |
31. Yamauchi, M, London, RE, Guenat, C, Hashimoto, F, Mechanic, GL: Structure and formation of a stable histidine-based trifunctional crosslink in skin collagen. J Biol Chem 1987 262: 11428–11434,  | PubMed | ISI | ChemPort |
32. Yamauchi, M, Woodley, DT, Mechanic, GL: Aging and cross-linking of skin collagen. Biochem Biophys Rescommun 1988 152: 898–903,  | ChemPort |
33. Yamauchi, M, Prisayanh, P, Haque, Z, Woodley, DT: Collagen cross-linking in sun-exposed and unexposed sites of aged human skin. J Invest Dermatol 1991 97: 938–941,  | PubMed | ISI | ChemPort |

Acknowledgments

We are grateful to Dr Graham Horgan of BioSS for the statistical analyses and to the Scottish Executive Environment and Rural Affairs Department for support.