Materials and methods

Tissue specimens

Samples of human tissue were excised during plastic surgery operation. Three were normal skin biopsies, four were HSc with a disease duration of 4.3 1.5 y, and five were HSc with a disease duration of 3.1 1.4 y, previously treated with an intralesional administration of 0.2 ml per week of collagen-PVP if scar length was 5 cm or less, 0.4 ml per week if scar length was between 5 and 10 cm, and 0.6 ml per week if scar length was 10 cm or more, until normalization of the scar by clinical criteria (between 1 and 3 mo). A section of the samples was snap-frozen in liquid nitrogen. Then they were cut at 4–6 m and serial sections were mounted on -methacryloxypropyltrimethoxisilane (Sigma, St. Louis, MO) coated slides. Finally, sections were fixed in acetone at –20°C.

Histology and immunohistochemistry

Herovici staining was performed according to^{Herovici (1963)}. Immunohistochemical procedures were assessed by blocking with 3% egg albumin (Sigma), except for the TGF-1 assay in which the blockade was done with 3% bovine albumin (Sigma), and then the sections were incubated with a mouse anti-human E-selectin (ELAM-1) or vascular cell-adhesion molecule (VCAM-1) monoclonal IgG at 25 g (Genzyme, Cambridge, MA) per ml, or a goat anti-human IL-1, TNF-, or PDGF polyclonal IgG at 20 g (R&D Systems, Minneapolis, MN) per ml. Anti-PDGF antibodies recognize all isoforms (AA, AB, and BB). A chicken anti-human TGF- polyclonal IgG neutralizing antibody (R&D Systems), which recognizes only isoforms 1 and 1.2 and has 100 times less affinity for TGF-2, was used at 20 g per ml. Antibodies were incubated for 60 min at room temperature. Binding was detected by incubating sections for 60 min at room temperature with rabbit anti-mouse biotinylated IgG diluted 1:20 (Sigma) for the adhesion molecules assay, with rabbit anti-goat biotinylated IgG, diluted 1:100 (Sigma) for IL-1, TNF-, and PDGF detection or with rabbit anti-chicken biotinylated IgG, diluted 1:750 (Pierce, Rockford, IL) for the TGF- assay. Avidin-peroxidase complex (Sigma) was used at a 1:20 dilution for 45 min at room temperature. The sections were developed in a solution of diaminobenzidine, and counterstained with hematoxylin. Finally, sections were dehydrated with alcohol and xylene, and mounted in resin. Negative control staining was performed with normal human serum diluted 1:100, instead of primary antibody. At least two different sections were examined for each patient. Adhesion molecule and cytokine expression was assessed by estimating positively staining cells in blood vessels and cells spreading along one field below the epidermis (see Results), and it was reported as the percentage of immunoreactive cells; results were expressed as the mean SEM. Associations between quantifiable variables were determined using Student's t test and Mann–Whitney U test.

Fibroblast culture conditions

Fibroblast cultures were established by fractionating with scissors one fragment of normal skin, HSc, or HSc treated with collagen-PVP and incubating the fragments with 0.1% bacterial collagenase type II (Sigma) for 1 h at 37°C. The homogenate was centrifuged at 1250 rpm for 2 min. The supernatant was eliminated and the pellet was resuspended in Dulbecco's modified Eagle medium (DMEM) high in glucose, supplemented with heat-inactivated 10% fetal calf serum, 2 mM L-glutamine, and antibiotics (100 U penicillin per ml and 5 g streptomycin per ml). Cells were grown at 37°C in 5% CO₂ in air. Cells between passages 3 and 4 were used for the experiments. Every experimental condition was normalized by DNA content and tested at least in triplicate, except for the cytokine assay, which was tested in duplicate.

Collagen synthesis

The assay was achieved by seeding 10⁵ fibroblasts from normal skin, HSc, or HSc previously treated with collagen-PVP, in 0.5 ml DMEM with 10% fetal calf serum, 2 mM L-glutamine, and 50 g ascorbate per ml in 24 well plates, incubating them for 48 h at 37°C in 5% CO₂ in air. Then the medium was eliminated and the cultures were incubated for 3 h at 37°C in 5% CO₂ in air, in DMEM without fetal calf serum, and with 2 mM L-glutamine and 50 g ascorbate per ml. The medium was replaced with the same fresh medium supplemented with 0.5 Ci per ml of L-[U-¹⁴C]proline 50 Ci per ml (Amersham, U.K.) (^{Dieggelmann et al. 1990}). Collagenous and noncollagenous protein synthesis were evaluated in homogenized monolayer and supernatant fractions, and collagen content was calculated according to the Peterkofsky and Dieggelmann formula (^{Peterkofsky & Dieggelmann 1971}).

Cytokine expression

IL-1, TNF-, TGF-1, and PDGF-AB were determined in supernatants of cultures under the conditions mentioned above. Two hundred microliters were assayed by enzyme-linked immunosorbent assay according to the manufacturer's instructions (R&D Systems).

Results

Features of a group of nine patients with HSc were evaluated. After a mean treatment period of 8 wk with collagen-PVP, as mentioned in Materials and Methods, the scars got soft and were flattened until reaching normal skin border (Table 1; Figure 1). It is important to note that none of the patients presented relapse of hypertrophy after 2.5 y.

Figure 1.

HSc before and after treatment with collagen-PVP. Forty-three year old female patient with HSc in the popliteal space with 2 y of evolution (a). The scar was treated for 10 wk with a weekly intralesional administration of 0.4 ml in the collagen-PVP. Note the mark of the scar is present, although the skin border has been reached, also erythema is absent and the patient reported no pain after the first 3 wk of treatment (b).

Full figure and legend (58K)

Table 1 - Clinical parameters of HSc treated with collagen-PVP.

Full table

Herovici staining

Sections of normal skin, HSc, and HSc previously treated with collagen-PVP and clinically resolved were stained following the Herovici procedure. HSc were quite different from normal skin Figure 2a, b; the epidermal profile was flattened, whereas in normal skin it was irregular with rete ridges. In scar sections, type I collagen fibers (red in Figure 2b) were present with parallel distributions in papillary dermis with respect to epidermis and in nodular zones all along the deep dermis (Figure 2b, small arrow), and type III collagen (blue fibers) below the epidermis was reduced. On the other hand, type I collagen distribution in normal skin was characteristically reticular, and type III collagen was more abundant in the papillary dermis than in the HSc sections. Interestingly, collagen-PVP-treated HSc Figure 2c showed a type I collagen distribution and a type III collagen zone (big arrow) similar to those observed in normal skin, and in some cases partial recovering of rete ridges and the presence of cutaneous appendages were observed (Figure 2c, small arrow).

Figure 2.

Collagen-PVP effect on tissue architecture restoration. Photomicrographs of human skin and scar tissue with or without treatment, stained with Herovici technique. (a) Normal skin exhibiting rete ridges, reticular type I collagen fibers in red, type III collagen fibers in blue in papillary dermis. (b) HSc without rete ridges, type I collagen distributed in whorl-like arrangements and nodular areas (small arrow). (c) HSc treated with collagen-PVP resembles normal skin, with type III collagen present in papillary dermis (big arrow) and showing a hair follicle (small arrow). Scale bar: 1000 m.

Full figure and legend (49K)

In situ expression of proinflammatory and fibrogenic cytokines

IL-1, TNF-, TGF-1, and PDGF showed strong differences among the groups Figures 3, 4, 5 and 6. HSc sections [part (c) in each montage] revealed more cells expressing these mediators than either normal skin or collagen-PVP treated HSc [parts (b) and (d), respectively, in each montage] (IL-1, p = 0.036; TNF-, p = 0.036; and PDGF, p = 0.003, treated versus untreated HSc) Figure 7a. On the other hand, blood vessels also exhibited the same pattern as spread cells (IL-1, p = 0.002; TNF-, p < 0.001; and PDGF, p = 0.01, treated versus untreated HSc) Figure 7b. In both cases, endothelium and spread cells in treated scars had cytokine levels similar to or lower than those of normal skin, although these differences were not statistically significant. Immunoreactivity of tested cytokines (large arrows) was consistently localized in the papillary dermis, where typical HSc nodules of fibrosis were absent, except for TGF-1, which was more widely spread.

Figure 3.

Immunolocalization of IL-1 in human skin sections frozen with the immunoperoxidase technique (avidin-biotin-peroxidase system). (a) Photomicrograph of negative control; (b) photomicrograph of normal skin; (c) photomicrograph of HSc, arrows indicate immunoreactive cells; (d) photomicrograph of HSc treated with collagen-PVP. Scale bar: 200 m.

Full figure and legend (97K)

Figure 4.

Immunolocalization of TNF- in human skin sections frozen with the immunoperoxidase technique (avidin-biotin-peroxidase system). (a) Photomicrograph of negative control; (b) photomicrograph of normal skin; (c) photomicrograph of HSc, arrows indicate immunoreactive cells; (d) photomicrograph of HSc treated with collagen-PVP. Scale bar: 200 m.

Full figure and legend (95K)

Figure 5.

Immunolocalization of TGF-1 in human skin sections frozen with the immunoperoxidase technique (avidin-biotin-peroxidase system). (a) Photomicrograph of negative control; (b) photomicrograph of normal skin; (c) photomicrograph of HSc, arrows indicate immunoreactive cells; (d) photomicrograph of HSc treated with collagen-PVP. Scale bar: 200 m.

Full figure and legend (96K)

Figure 6.

Immunolocalization of PDGF in human skin sections frozen with the immunoperoxidase technique (avidin-biotin-peroxidase system). (a) Photomicrograph of negative control; (b) photomicrograph of normal skin; (c) photomicrograph of HSc, arrows indicate immunoreactive cells; (d) photomicrograph of HSc treated with collagen-PVP. Scale bar: 200 m.

Full figure and legend (73K)

Figure 7.

Collagen-PVP effect on pro-inflammatory cytokine expression in HSc. The percentage of cytokine expressing immunoreactive cells was determined for normal skin, HSc, and collagen-PVP treated HSc. Cytokine values were compared between treated and untreated samples as mentioned in Materials and Methods. They were different in (a) spread cells (IL-1, p = 0.036; TNF-, p = 0.036, and PDGF, p = 0.003) and (b) blood vessels (IL-1, p = 0.002, TNF-, p < 0.001 and PDGF, p = 0.01, treated versus untreated HSc). Results represent the mean SEM of at least two sections of each tissue from normal skin (n = 3), HSc (n = 4), and HSc previously treated with collagen-PVP (n = 5).

Full figure and legend (18K)

The stratum basalic of the epidermis also showed reactivity; however, negative controls (Figure 4a, small arrow) were also positive in the basal layers of the epidermis. We attribute these differences to the intensity of skin pigmentation. This artifact did not exclude the possibility of epidermal reactivity, but obscured a quantitative interpretation, because keratinocytes are important producers of several cytokines (^{Streilein 1993}).

In situ expression of adhesion molecules

ELAM-1 and VCAM-1 were detected in normal skin and HSc (Figures 8b, c, 9b, c, respectively). In scars previously treated with collagen-PVP Figures 8d, 9d, both molecules showed slightly lower levels when compared with normal skin and HSc groups, but only the difference of VCAM-1 levels between normal skin and treated HSc were statistically significant (IL-1, p = 0.036). In contrast, levels of expression of ELAM-1 were different, although not statistically significant among the groups Figure 10. Even when we only quantitated the percentage of immunostaining in endothelium, other sites also showed reactivity, such as interstitial inflammatory infiltrated cells and some isolated labels, maybe due to soluble ELAM-1.

Figure 8.

Human skin sections frozen with the immunoperoxidase technique (avidin-biotin-peroxidase system) using an anti-ELAM-1 as primary antibody. (a) Photomicrograph of negative control; (b) photomicrograph of normal skin; (c) photomicrograph of HSc, arrows indicate immunoreactive cells; (d) photomicrograph of HSc treated with collagen-PVP. Scale bar: 200 m.

Full figure and legend (109K)

Figure 9.

Human skin sections frozen with the immunoperoxidase technique (avidin-biotin-peroxidase system) using an anti-VCAM-1 as primary antibody. (a) Photomicrograph of negative control; (b) photomicrograph of normal skin; (c) photomicrograph of HSc, arrows indicate immunoreactive cells; (d) photomicrograph of HSc treated with collagen-PVP. Scale bar: 200 m.

Full figure and legend (104K)

Figure 10.

Collagen-PVP effect on adhesion molecule expression in treated HSc. The percentage of adhesion molecule-expressing immunoreactive cells was determined in blood vessels of normal skin, HSc, and collagen-PVP-treated HSc. Adhesion molecule values were compared between treated and untreated samples as mentioned in Materials and Methods but they were not different; however, when normal skin was compared with treated HSc, these values were different with a statistical level of p = 0.036 for VCAM-1. Results represent the mean SEM of at least two sections of each tissue from normal skin (n = 3), HSc (n = 4), and HSc previously treated with collagen-PVP (n = 5).

Full figure and legend (17K)

Fibroblast culture findings

There was no difference in collagen synthesis among the different cultures assessed Figure 11a; however, there were differences in cytokine expression in the supernatants of fibroblast cultures derived from normal skin, HSc, and treated HSc Figure 11b, where only TGF-1 and PDGF-AB were quantifiable, because IL-1 and TNF- were not detected by our enzyme-linked immunosorbent assay systems. TGF-1 was 3.5-fold higher in normal skin as well as in HSc when compared with treated HSc fibroblasts, whereas PDGF-AB levels were similar between normal skin and treated HSc fibroblasts, and they were 1-fold higher in HSc.

Figure 11.

Fibroblasts derived from HSc treated with collagen-PVP were assayed for collagen synthesis and cytokine expression. The percentage of relative collagen was determined by [¹⁴C]-Proline incorporation, values were compared between treated and untreated groups, but they were not significantly different. Results represent the mean SD of triplicates of each culture (a). TGF-1 and PDGF-AB levels were evaluated by enzyme-linked immunosorbent assay. Results represent the mean SD of two different experiments performed by duplicate. Where p values for TGF-1 are 0.05 for normal skin or HSc versus HSc treated, and 0.003 for PDGF-AB for normal skin or HSc treated versus HSc (b), normal skin, HSc, and collagen-PVP-treated HSc.

Full figure and legend (17K)

Discussion

In this study we found that fibrogenic/proinflammatory cytokines were predominantly detected in the upper dermis of normal skin Figures 3, 4, 6, except for TGF-1 Figure 5, which was also found in the deep dermis in nodular zones, as previously described (^{Ghahary et al. 1995}). HSc manifested a strong parallel increase of all cytokines measured in spread cells, as well as in blood vessels Figure 7. Some of these data contrast with those of^{Castagnoli et al. (1993)}, which indicate similar levels of TNF- in HSc and normal skin.

TGF- is synthesized by many cells and tissues, including normal skin and HSc (^{Ghahary et al. 1995}). Its localization in fibroblast-like cells derived from fibrotic pathologies like hypertrophic scarring, as well as in endothelial cells adjacent to mononuclear cells, suggests a relationship between factor-producing cells and those responsible for collagen overexpression, where the connection could be either autocrine or paracrine (^{Zhang et al. 1995}). Furthermore, TGF- participates with PDGF in an orchestrated way. Here we have shown that TGF-1, as well as PDGF-AB, were diminished in fibroblast cultures derived from treated HSc when they were compared with normal skin and HSc fibroblasts Figure 11b. This is important, because generally fibroblasts are considered to be effector cells regulated by inflammatory cells. These results could indicate that the fibroblast itself has an ability for autoregulation, perhaps through collagen-PVP modulation. Because collagen synthesis by fibroblasts did not show variations among the different cultures tested Figure 11a, the results suggest that collagen synthesis is not regulated in an autocrine fashion, at least by TGF-1 and/or PDGF-AB. Therefore, it is necessary to evaluate collagen turnover in this lineage, as well as the cooperative effects between inflammatory cells and fibroblasts.

IL-1 and TNF- have similar fibrogenic activities. These two cytokines participate in ECM turnover and inflammation, their principal source being leukocytes; however, they have also been detected in mesenchymal cells and epidermis as a significant reservoir. Nevertheless, we were unable to detect these cytokines in fibroblast cultures.

Here we report that in collagen-PVP-treated HSc, completely resolved by clinical criteria, all the tested cytokines (IL-1, TNF-, TGF-1, PDGF) analyzed by immunohistochemistry showed diminished expression, even below normal levels, and a similar reduction in the expression of adhesion molecules was observed Figure 10. This suggests a correlation between them, although only VCAM-1 was statistically significant when normal skin versus treated HSc were compared. Because collagen-PVP treatment has been shown to modify human dermal fibrosis, as presented here Table 1, we infer that its mechanism of action can be directly associated with the inflammatory process observed in HSc, possibly by the decrease in the expression of adhesion molecules, which are upregulated in response to proinflammatory cytokines, IL-1, and TNF-. Also, low levels of these cytokines observed in endothelium from treated HSc could downregulate chemotaxis. Therefore, we suggest that the chronic inflammatory process is interrupted by collagen-PVP action, diminishing the partial production of fibrogenic soluble factors, including their self-expression by spreading cells, probably connective tissue cells and inflammatory infiltrates. This mechanism allows the recovery of tissue homeostasis, although it leads to gradual elimination of fibrosis by the increased turnover of ECM components. This is because ECM metabolism is regulated by various mechanisms, including soluble factors or cytokines (^{Postlethwaite et al. 1988;Gailit & Clark 1994}), as well as the ECM itself (^{Massagué 1990}). It could be possible that the collagen-PVP complex is recognized by cell receptors, such as integrins, and then this modifies the intracellular signals for cytokine expression and/or ECM and metalloproteases synthesis; however, it is necessary to perform the appropriate experiments to explore these possibilities.

It is evident that these cytokines are associated with the persistence of an inflammatory infiltrate, which generates new signals capable of recruiting more leukocytes and which stimulates the expression of adhesion molecules. This could perpetuate the inflammatory process present in hypertrophic scarring; however, we have found that collagen-PVP is a good choice for treatment not only in dermal fibrosis, as has been shown, but also in other fibrotic processes, such as fibrotic Achilles tendon (unpublished results). Collagen-PVP has the advantage of being a biologic drug with minimal risks, because no side-effects have been detected in healthy volunteers and HSc patients treated for long periods (even 14 y) with the drug. This was judged clinically and by laboratory tests applied before and after treatment (data not shown).

Even though it is well known that nonhuman collagen implants are capable of inducing heterologous anticollagen antibodies (^{Ellingsworth et al. 1986;Trentham 1986;Vanderveen et al. 1986;Hyder et al. 1992}), no evidence exists to associate them with autoimmune diseases related to collagen dermal implants (^{Singh & Fries 1994;Lewy 1994}). Furthermore, little is known about porcine collagen implants, as in the case of collagen-PVP. Generally, collagen implants consist of insoluble fibrilar collagens resuspended in saline solutions, in such a way that the structure of these implants is quite different from that seen in collagen-PVP. Collagen association with polyvinylpyrrolidone, and the cross-linking favored by -irradiation, confers on it various physicochemical properties, such as the impossibility of forming a gel when diluted in culture medium at 37°C and neutral pH (data not shown); therefore, collagen-PVP does not behave like other collagen implants. Moreover, electrophoretic analysis demonstrated a change in the relative mobility of collagen-PVP when compared with the mixture without -irradiation or the components alone (^{Chimal-Monroy et al. 1997}).

Based on the results obtained in this study, in particular the low levels of fibrogenic cytokines and adhesion molecules in treated versus untreated HSc, we suggest that collagen-PVP can modulate extracellular matrix turnover, mainly of collagen, and thus block the progression of the fibrotic process.

Notes

² Díaz de León HL, Krötzsch-Gómez FE, Guerrero-Padilla E, Cervantes-Viramontes R, Reyes-Márquez R: A novel approach for the treatment of tissular fibrosis. Matrix Biol 145:401, 1994 (abstr.)

References

1. Ahn, ST, Monafo, WW, Mustoe, TA Topical silicone gel for the prevention and treatment of hypertrophic scar. Arch Surg, 1991 126, 499–504,
2. Almazán Díaz, A, la Cruz García, JC, Lira Romero, JM, et al. Investigación experimental de la regeneración ósea en fémures de rata después de la aplicación de colágena I polimerizada: Estudio radiológico, histológico e histoquímico. Rev Mex Ortop Traum, 1996 10, 142–152,
3. Castagnoli, C, Stella, M, Berthod, Ch, Magliacani, G, Momigliano Richiardi, P TNF production and hypertrophic scarring. Cell Immunol, 1993 147, 51–63,
4. Castagnoli, C, Stella, M, Magliacani, G, Ferrone, S, Momogliano Richiardi, P Similar ectopic expression of ICAM-1 and HLA Class II molecules in HSc following thermal injury. Burns, 1994 20, 430–433,  | Article | PubMed | ISI | ChemPort |
5. Chapman, PT & Haskard, DO Leukocyte adhesion molecules. Br Med Bull, 1995 51, 296–311,
6. Chimal-Monroy, J, Bravo-Ruíz, T, Krötzsch-Gómez, FE, y Díaz León, L Implantes de Fibroquel^MR aceleran la formación de hueso nuevo en defectos óseos inducidos experimentalmente en cráneo de rata: Un estudio histológico. Rev Biomed, 1997 8, 81–88,
7. Darzi, MA & Chowdri, NA Evaluation of various methods of treating keloids and hypertrophic scars, a 10 year follow up study. Br J Plast Surg, 1992 45: 374,
8. Dieggelmann, RF, Graham, MF, Bryson, GR & An. in vitro microassay to quantitate collagen synthesis and its utility for analyzing the effects of defined inflammatory mediators. Ann New York Acad Sci, 1990 580, 514–517,
9. Eckes, B, Mauch, C, Hüppe, G, Krieg, T Downregulation of collagen synthesis in fibroblast within three-dimensional collagen lattices involves transcriptional and posttranscriptional mechanisms. Febs, 1993 318, 129–133,
10. Ellingsworth, LR, DeLustro, F, Brennan, JE, Sawamura, S, McPherson, J The human immune response to reconstituted bovine collagen. J Immunol, 1986 136, 877–882,
11. Gailit, J & Clark, RaF Wound repair in the context of extracellular matrix. Curr Op Cell Biol, 1994 6, 717–725,  | PubMed |
12. Ghahary, A, Shen, YJ, Scott, PG, Tredget, EE Immunolocalization of TGF-1 in human hypertrophic scar and normal dermal tissues. Cytokine, 1995 7, 184–190,
13. Herovici, C Polychrome stain for differentiating precollagen from collagen. Stain Technol, 1963 38, 204–205,
14. Hyder, P, Singh, G, Adam, S Humoral responses to type I collagen after surgical curettage procedures employing bovine collagen implants. Biomaterials, 1992 13, 693–696,
15. Kawaguchi, Y, Harigai, M, Kitani, A, et al. Effect of prolyl 4-hydroxylase inhibitor on fibroblast collagen production in vitro: an approach to the treatment of systemic sclerosis. J Rheumatol, 1992b 19, 1710–1715,
16. Kawaguchi, Y, Kitani, A, Hara, M, et al. Cytokine regulation of prolyl 4-hydoxylase production in skin fibroblast cultures from patients with systemic sclerosis: Contribution to collagen synthesis and fibrosis. J Rheumatol, 1992a 19, 1195–1201,
17. Kovacs, EJ Fibrogenic cytokines: the role of immune mediators in the development of scar tissue. Immunol Today, 1991 12, 17–23,  | Article | PubMed | ISI | ChemPort |
18. Lawrence, WT In: Search of the Optimal Treatment of Keloids: Report of a Series and a Review of the Literature. Ann Plast Surg 1991.27: 164,
19. Lewy, RI Autoimmune disease and collagen dermal implants (see comments). Ann Intern Med, 1994 120: 525,
20. Massagué, J The transforming growth factor- family. Ann Rev Cell Biol, 1990 6, 597–641,  | PubMed | ISI | ChemPort |
21. McPherson, JM & Piez, KA Collagen in dermal wound repair In: Clark Raf, Henson PM, (eds). The Molecular and Cellular Biology of Wound Repair. 1988, New York: Plenum Press, pp. 471–496,
22. Peterkofsky, B & Dieggelmann, R Use of a mixture of proteinase-free collagenase for the specific assay of radioactive collagen in the presence of other proteins. Biochem, 1971 10, 988–993,  | ISI | ChemPort |
23. Postlethwaite, AE, Raghow, R, Stricklin, GP, Poppleton, H, Seyer, JM, Kang, AH Modulation of fibroblast functions by interleukin 1: increased steady-state accumulation of type I procollagen messenger RNAs and stimulation of other functions but not chemotaxis by human recombinant interleukin 1 and . J Cell Biol, 1988 106, 311–318,  | PubMed |
24. Rockwell, WB, Cohen, IK, Ehrlich, HP Keloids and hypertrophic scars: a comprehensive review. Plast Reconstr Surg, 1989 84, 827–837,  | PubMed | ISI | ChemPort |
25. Rosenbloom, J, Feldman, G, Freundlich, B, Jimenez, SA Inhibition of excessive scleroderma fibroblast collagen production by recombinant -interferon. Arthritis Rheum, 1986 29, 851–856,
26. Singh, G & Fries, JF Autoimmune disease and collagen dermal implants (see comments). Ann Intern Med, 1994 120: 524,
27. Soderberg, T, Hallimans, G, Bartholdson, L Treatment of keloids and hypertrophic scars with adhesive zinc tape. Scand J Plast Reconstr Surg, 1982 16: 261,
28. Streilein, JW Dermal cells: underappreciated components of skin-associated lymphoid tissues (SALT) In: Nickoloff BJ, (ed.). Dermal Immune System. 1993, Ann Arbor, Michigan: CRC Press, pp. 26–37,
29. Tang, YW Intra and postoperative steroid injections for keloids and hypertrophic scars. Br J Plast Surg, 1992 45: 371,
30. Trentham, DE Adverse reactions to bovine collagen implants. Arch Dermatol, 1986 122, 643–644,
31. Vanderveen, EE, McCoy, JP, Schade, W, et al. The association of HLA and immune responses to bovine collagen implants. Arch Dermatol, 1986 122, 650–654,
32. Zhang, K, Garner, W, Cohen, L, Rodriguez, J, Phan, S Increased types I and III collagen and transforming growth factor-1 mRNA and protein in hypertrophic burn scar. J Invest Dermatol, 1995 104, 750–754,

Acknowledgments

We acknowledge Dr. Jeffrey Davidson, Dr. Carlos Rosales, and Dr. Philip Maini for their critical review of the manuscript, Dr. Miguel Morales for microscopical and cryostat assistance, and Isabel Pérez-Monfort and Ana Luisa Weckmann for correcting the English version of the manuscript. This work was partially supported by grant LDL-94 provided by Aspid S.A. de C.V. and PADEP 030336 (UNAM).