THE IMPORTANCE OF thyroid hormone (TH) to wound

0013-7227/04/$15.00/0 Endocrinology 145(5):2357 2361 Printed in U.S.A. Copyright 2004 by The Endocrine Society doi: 10.1210/en.2003-1696 A Role for Thyroid Hormone in Wound Healing through Keratin Gene Expression JOSHUA D. SAFER, TARA M. CRAWFORD, AND MICHAEL F. HOLICK Section of Endocrinology, Department of Medicine, Boston University School of Medicine, Boston, Massachusetts 02118 The importance of thyroid hormone (TH) in wound healing is not well understood. To gain insight, we evaluated the impact of TH deficiency on wound-healing genes in cultured keratinocytes. By RT-PCR, keratin 6a (K6a) and 16 (K16) gene expression in TH replete cells was 3.8- (P < 0.005) and 1.9-fold (P < 0.05) greater, respectively, than expression in TH-deficient cells. By real-time PCR, TH replete cell expression of K6a, K16, and K17 was greater than in deficient cells: 18- (P < 0.001), 10- (P < 0.001), and 4-fold (P < 0.005), respectively. To examine TH requirement for optimal wound healing, we contrasted TH-deficient vs. ip T 3 -treated mice. Four days after THE IMPORTANCE OF thyroid hormone (TH) to wound healing is not well understood. Lennox and Johnston (1) reported that exogenous T 4 improved the rate of wound healing in rats as well as the strength of the scars. They further reported that wound-healing speed was diminished in hypothyroid rats. Mehregan and Zamick (2, 3) observed that additional TH stimulated the rate and quality of wound healing in euthyroid rats. Scars were smoother in animals receiving T 3 in drinking water. Pirk et al. (4) noted no change in wound healing with euthyroid hamsters receiving ip T 4, Cannon (5) reported that hypothyroidism did not diminish wound strength in pigs, and Ladenson et al. (6) found no wound-healing deficits in hypothyroid humans. In vitro, keratinocyte proliferation is retarded in T 3 -deficient medium relative to T 3 replete medium (7). In vivo, topical T 3 stimulates epidermal proliferation (7, 8), and topical triac, the mild TH analog, thickens skin (9). The keratin genes encode the intermediate filaments, making up about 30% of the protein of the epidermis. Some associations between the keratin genes and specific phases of skin growth have been made (10 12), including the following: keratins 1 (K1) and 10 (K10) are associated with epidermal differentiation; K6a, K16, and K17 are associated with epidermal proliferation and wound repair; and K5 and K14 are expressed in the basal skin layer (their expression decreased as the skin cells differentiate). In K6a knockout mice (13), the absence of K6a resulted in diminished superficial wound healing, but no change in full thickness wound healing. When K16 was overexpressed in cultured human keratinocytes, proliferation was enhanced Abbreviations: GAPDH, Glyceraldehyde-3-phosphate dehydrogenase; HaCaTs, cultured human keratinocytes; K, keratin; TH, thyroid hormone. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community. wounding, ip T 3 -treated mice had twice the degree of wound closure as hypothyroid mice (P < 0.001). By RT-PCR, K6a and K17 gene expression from control mouse skin was greater than from hypothyroid mouse skin: 5- (P < 0.001) and 1.7-fold (P < 0.05), respectively. T 3 is necessary for the keratinocyte proliferation required for optimal wound healing. T 3 exerts influence by stimulating expression of the wound-healing keratin genes. Thus, for hypothyroid patients undergoing surgery that cannot be delayed until euthyroidism is achieved, our data support T 3 treatment for the perioperative period. (Endocrinology 145: 2357 2361, 2004) (14). However, human K16 overexpressed in mice resulted in delayed wound healing in the transgenic animals (15, 16). Ex vivo investigation suggested that the K16 overexpression inhibited keratinocyte migration. A K17 knockout mouse suffered alopecia and compensatory K16 expression (17). The alopecia was less in those animals with greater K16 expression. Although TH stimulates expression of K6a both in vivo and in vitro (18), to date only negative TH response elements have been identified for the keratin genes associated with proliferation (19 22). The following investigation was undertaken to ascertain the need for TH in optimal wound healing and to clarify the effect of TH on the expression of wound healing-associated keratin genes. Materials and Methods T 3 regulation of keratin genes in cultured human keratinocytes Cultured human keratinocytes (HaCaTs) were grown and maintained as previously described (7). Cultures were incubated overnight in medium supplemented with or absent 0.1 nm T 3 (Sigma, St. Louis, MO). Serum for the culture medium was stripped of TH with two overnight AG 1-X8 incubations (Bio-Rad, Hercules, CA) and one activated charcoal incubation (Sigma). T 3 was then supplemented in the concentration indicated. RNA was extracted from the cell cultures using a standard RNA extraction protocol (TRIzol reagent, Life Technologies, Inc./Invitrogen, Grand Island, NY). The cdna for RT-PCR was created using Superscript II RNase H Reverse Transcriptase (Life Technologies, Inc., Rockville, MD). Complete or partial sequences of human K6a, K10, K14, K16, and K17 are posted in the nucleotide section of National Library of Medicine s web site (23 26). From those sequences, the following primers were designed for the RT-PCR: K6a forward 5 -GGC TGA GGA GCG GCG TGA ACA G, K6a reverse 5 -AAG GAG GCA AAC TTG TTG TTG AG, K10 forward 5 -AAT GAA AAA GTA ACC ATG CAG AAT CTG, K10 reverse 5 -CAC GAG GCT CCC CCT GAT, K14 forward 5 -GAG TGT GGA AGC CGA CAT CAA, K14 reverse 5 -GCC TCT CAG GGC ATT CAT CTC, K16 forward 5 -GCA TGC AGT AGC GGC CTT T, K16 reverse 5 -TCC AAC AGC GAA CTG GTA CAG A, K17 forward 2357

2358 Endocrinology, May 2004, 145(5):2357 2361 Safer et al. TH and Wound Healing 5 -AGG AGA TGA CCT TGC CAT CCT, and K17 reverse 5 -GGC TGA TTG GCA GCG TGG AGG A. All expression data were quantitated relative to the glyceraldehyde- 3-phosphate dehydrogenase (GAPDH) and -actin genes that were simultaneously measured using the same techniques. At a minimum, each experiment was repeated on three independent occasions. Quantitative PCR Quantitative (real time) PCR using an ABI Prism 7700 Sequence Detector (Applied Biosystems, Foster City, CA) was performed with the following parameters: PCR amplification of 200 ng of single-stranded cdna template was carried out using SYBR Green Master Mix (Applied Biosystems). Primers described above for the RT-PCR were used for the real-time studies also. Quantification of PCR products was determined using ABI Prism Sequence Detection System software (Applied Biosystems) and controlled for total RNA. Fold change was determined by comparison of experimental data with standards of known copy number. Establishing mouse hypothyroidism All animal experimentation described was conducted in accord with accepted standards of humane animal care. Fourteen age-, sex-, and size-matched CD-1 mice (Charles River, Boston, MA) were thyroidectomized (by surgical thyroidectomy). To ascertain hypothyroidism, all mice were eye-bled at wk 6, and serum total T 4 levels measured with a standard RIA kit (ICN Biomedicals, Inc., Orangeburg, NY). Unlike other TH kits that use antimouse antibodies, the ICN kit uses antirabbit antibodies and avoids spuriously elevated readings in mice. TH levels for mice fall at the low end of the human range so the human standards included in the kit were used. Wound-healing analysis At wk 6, all 14 mice were anesthetized with 3 3-cm midline areas of their backs delineated and shaved. Four 10-mm diameter full thickness wounds were then placed on the dorsum of each mouse in the shaved area. Relative wound surface areas were compared 4 d after injury (analysis no. 1). For each animal, the four wound surface areas were averaged. Wound data reflect the mean of those averages with each animal s average representing one point in the analysis. After a 2-wk healing/washout period, all animals received 1 wk of daily 0.25- g ipt 3 injections, and wounding analysis was repeated (analysis no. 2). After a 3-wk healing/washout period allowing animals to become hypothyroid again, wounding analysis was repeated on all animals (analysis no. 3). Relative wound surface area calculations Upon wounding and on d 4 post wounding for each analysis, all animals were photographed. All photographs included a standard ruler to maintain size consistency. All photographs were printed on the same paper stock. Wound photographs from each animal were carefully cut from the photographs and weighed. Thus, irregular borders could be accommodated in the calculations. Each animal served as its own control with percent wound closure representing the weight of wound photographs for a specific mouse on d 4 relative to the weight of wound photographs for the same mouse on the day of wounding. In vivo gene expression analysis Tissue samples for RNA analysis were taken when the animals were killed and maintained in the RNA protecting agent, RNAlater (Ambion, Austin, TX). RNA was later extracted from the skin samples using a standard RNA extraction protocol (TRIzol reagent). Then cdna for RT-PCR was created using Superscript II Rnase H Reverse Transcriptase. Complete or partial sequences of murine K6a and K17 are posted in the nucleotide section of National Library of Medicine s web site (27). From those sequences, the following primers were designed for the RT-PCR: K6a forward 5 -GTT TGC CTC CTT CAT CGA CAA, K6a reverse 5 -CTG CGG AGG TTG CTG ATG TA, K17 forward 5 -CTT FIG. 1. A, In HaCaTs evaluated with RT-PCR, GAPDH-corrected K6a and K16 expression levels in T 3 -treated cells were 3.8- and 1.9- fold greater than levels seen in T 3 -deficient cells, respectively. No change in K17 expression could be discerned by RT-PCR. B, Representative gels show bands for keratin RNA extracted from T 3 -treated and T 3 -deficient cells with corresponding GAPDH bands below. C, In keratinocytes evaluated with real-time PCR, total RNA corrected K6a, K16, and K17 expression levels in T 3 -treated cells were 18-, 10-, and 4.4-fold greater than those of T 3 -deficient cells, respectively. ***, P 0.001; **, P 0.005; *, P 0.05.

Safer et al. TH and Wound Healing Endocrinology, May 2004, 145(5):2357 2361 2359 CCG TAC CAA GTT TGA GAC, and K17 reverse 5 -CGG TTC TTC TCC GCC ATC TTC. All expression data were quantitated relative to the GAPDH housekeeping gene, which was simultaneously measured using the same technique. Statistical analysis Statistical analysis was performed with Student s t test and ANOVA. Data are presented with sem. Results K6a, K16, and K17 gene expression was decreased in T 3 - deficient keratinocytes RNA was extracted from human keratinocytes cultured overnight in medium supplemented with or absent 0.1 nm T 3. Extracted RNA was subjected to RT-PCR with keratin gene expression quantitated relative to the housekeeping gene, GAPDH. In T 3 -treated cells, K6a gene expression was 3.8 0.6-fold greater than in T 3 -deficient cells (Fig. 1, A and B; P 0.005), and K16 gene expression was 1.9-0.3-fold greater (P 0.05). We found no change in K10, K14, or K17 gene expression in the TH-treated cells relative to the deficient cells. RNA extracted from HaCaTs treated as above was analyzed with real-time PCR. Data were corrected for total RNA loaded. In T 3 -treated cells, K6a gene expression was 18 0.8-fold greater than in T 3 -deficient cells (Fig. 1C; P 0.001), T 3 -treated K16 gene expression was 10 0.2-fold greater than the expression in T 3 -deficient cells (P 0.001), and T 3 -treated K17 gene expression was 4.4 0.6-fold greater than in deficient cells (P 0.005). T 3 -treated cells did not have FIG. 2. A, In hypothyroid mice, wound healing lagged significantly behind healing in T 3 replete animals. Four days after wounding, skin wounds in hypothyroid mice were 26% closed. In the same mice treated with ip T 3 for 1 wk, wounds closed 54% after 4 d. When the mice were permitted to become hypothyroid again, wounds closed 26% after 4 d. **, P 0.005. B, Representative hypothyroid mouse with baseline wounds. C, The same hypothyroid mouse with wounds on d 4. D, Representative ip T 3 -treated mouse with baseline wounds. E, The same ip T 3 -treated mouse with wounds on d 4.

2360 Endocrinology, May 2004, 145(5):2357 2361 Safer et al. TH and Wound Healing statistically significant changes in K10 or K14 gene expression. Hypothyroid mice had retarded wound healing To determine whether the in vitro T 3 -mediated keratin gene expression pattern could be reproduced in vivo, and whether that pattern would be associated with significantly poorer wound healing, we contrasted the impact of surgical hypothyroidism on wound healing and proliferation-associated keratin gene expression. Relative to baseline, T 4 levels in thyroidectomized mice were 84% lower (4.2 0.4 g/dl for euthyroid mice vs. 0.66 0.5 g/dl for hypothyroid animals; P 0.001) at wk 6. Fourteen thyroidectomized mice were evaluated. Ten-millimeter diameter dorsal skin wounds were established in all animals 6 wk after thyroidectomy. Relative wound surface areas were determined 4 d after injury (analysis no. 1). After a 2-wk healing/washout period, all animals received 1 wk of daily 0.25- g ipt 3 injections, and wounding analysis was repeated (analysis no. 2). After a 3-wk healing/washout period allowing animals to become hypothyroid again, wounding analysis was repeated on all animals (analysis no. 3). At d 4, percent wound closure for ip T 3 -treated mice was 208% that of hypothyroid animals (Fig. 2). In the ip T 3 -treated mice, wound surface areas were reduced 54 2.8% at d 4. In analysis no. 1, hypothyroid mouse wounds were closed 26 5.3% at d4(p 0.001 vs. ip-treated animals) and in analysis #3, hypothyroid mouse wounds were closed 26 5.6% (P 0.004 vs. ip-treated animals). Proliferation-associated keratin gene expression was diminished in hypothyroid mice RNA was extracted from the mouse epidermis samples, and keratin gene expression was determined with RT-PCR relative to GAPDH housekeeping gene control (Fig. 3). In the samples from the control mice, K6a gene expression was 5-fold greater than expression from hypothyroid mice (P 0.001), and K17 gene expression was 1.7-fold greater (P 0.05). Discussion The reported importance of TH in wound healing is contradictory. Although some authors report improved rates and quality of wound healing in response to TH (2, 3, 28 31), others report no apparent TH-mediated changes in wound healing (4 6). Studies supporting a role for TH used humans, rats, and guinea pigs. Negative studies used humans, pigs, and hamsters. To extend insight into TH action on skin, we investigated the consequence of modest hypothyroidism on skin wound healing in mice. Moderate hypothyroidism resulted in poorer wound healing. Our hypothyroid mice had T 4 concentrations 16% of control levels. Humans undergoing near-total thyroidectomy would be expected to suffer that degree of hypothyroidism 3 wk postoperatively (T 1/2 for T 4 in humans is 1 wk). Previously, we demonstrated that TH can stimulate epidermal proliferation in mice and rats (7, 8). Consistent with that finding, we demonstrated that proliferation-associated FIG. 3. K6a, K17 expression in control mouse skin was greater than that of hypothyroid animals. A, GAPDH-corrected K6a RNA expression was 5-fold greater in control mice, and GAPDH-corrected K17 RNA expression was 1.7-fold greater. ***, P 0.001; *, P 0.05. B, Representative gels show bands for keratin RNA extracted from hypothyroid and control mice with corresponding GAPDH bands below. cytokeratin 6a expression is diminished in hypothyroidism and is dramatically stimulated with supraphysiologic doses of topical TH (18). However, because K6a only mediates superficial wound healing (13), TH regulation of K6a cannot be the sole mechanism for the wound healing effect seen in vivo. Furthermore, previous investigators have reported that proliferation-associated cytokeratin genes contain inhibitory TH response elements in their promoters, suggesting that their expression would be decreased in the presence of TH. The genes most associated with epidermal proliferation are K6a, K16, and K17. Their regulation by TH seems likely to play a role in mediating TH influence over epidermal proliferation. Although TH stimulates expression of proliferation-associated keratin genes, to date only negative TH response elements have been identified for these genes (21, 22). It is not known whether the above reflects TH induction of indirect keratin gene-stimulating pathways or the existence of unidentified positive TH response elements for the keratin genes. Refuting the suggestion that the keratin genes serve only

Safer et al. TH and Wound Healing Endocrinology, May 2004, 145(5):2357 2361 2361 as markers of proliferation and are not proliferation factors themselves, Wojcik et al. (13) reported a wound-healing defect in K6a knockout mice. Paramio et al. (14) reported increased proliferation of cultured keratinocytes transfected with K16. Thus, TH stimulation of K16 expression provides a mechanism through which epidermal proliferation may be stimulated. We presented three analyses of gene transcription. In vivo RT-PCR studies provided the best measure but were subject to interanimal variability. In vitro RT-PCR studies provided more precise measurements but suffered from masking small differences in transcription. Real-time PCR analysis included data from each PCR cycle and provided the most precise report with the RT-PCR pictures serving as a visual complement to the real-time data. TH-deficient mice have poorer wound healing than seen with individual keratin gene knockout mice. TH acts on all three wound healing-associated keratin genes along with direct or indirect TH action on other genes required for optimal keratinocyte proliferation. With multiple genes affected, compensatory gene expression is prevented. We propose that T 3 is necessary for the keratinocyte proliferation required for optimal wound healing. T 3 exerts influence by stimulating expression of the proliferation cytokeratins 6a, 16, and 17. Thus, surgical patients should have thyroid status ascertained. For hypothyroid patients undergoing surgery that cannot be delayed until euthyroidism is achieved, our data support T 3 treatment for the perioperative period. Acknowledgments This work was supported in part by grants from the Abbott Pharmaceuticals-sponsored Thyroid Research Advisory Council (to J.D.S.) and the Evans Foundation at Boston University School of Medicine (to J.D.S.). Received December 15, 2003. Accepted January 14, 2004. Address all correspondence and requests for reprints to: Joshua D. Safer, M.D., Section of Endocrinology, Diabetes, and Nutrition, Boston University School of Medicine, 715 Albany Street, Room M-1022, Boston, Massachusetts 02118. E-mail: jsafer@bu.edu. References 1. Lennox J, Johnston ID 1973 The effect of thyroid status on nitrogen balance and the rate of wound healing after injury in rats. Br J Surg 60:309 2. Zamick P, Mehregan AH 1973 Effect of l-tri-iodothyronine on marginal scars of skin grafted burns in rats. Plast Reconstr Surg 51:71 75 3. Mehregan AH, Zamick P 1974 The effect of triiodothyronine in healing of deep dermal burns and marginal scars of skin grafts. A histologic study. J Cutan Pathol 1:113 116 4. Pirk FW, El Attar MA, Roth GD 1974 Effect of analogues of steroid and thyroxine hormones on wound healing in hamsters. J Periodontal Res 9:290 297 5. Cannon CR 1994 Hypothyroidism in head and neck cancer patients: experimental and clinical observations. Laryngoscope 104:1 22 6. Ladenson, PW, Levin AA, Ridgeway EC, Daniels GH 1984 Complications of surgery in hypothyroid patients. Am J Med 77:261 266 7. Safer JD, Crawford TM, Fraser LM, Hoa M, Ray S, Chen TC, Persons K, Holick MF 2003 Thyroid hormone action on skin: diverging effects of topical versus intraperitoneal administration. Thyroid 13:159 165 8. Safer JD, Fraser LM, Ray S, Holick MF 2001 Topical triiodothyronine stimulates epidermal proliferation, dermal thickening, and hair growth in mice and rats. Thyroid 11:717 724 9. Faergemann J, Sarnhult T, Hedner E, Carlsson B, Lavin T, Zhao X-E, Sun X-Y 2002 Dose-response effects of tri-iodothyroacetic acid (Triac) and other thyroid hormone analogues on glucocorticoid-induced skin atrophy in the haired mouse. Acta Derm Venereol 82:179 183 10. Fisher C, Blumenberg M, Tomic-Canic M 1995 Retinoid receptors and keratinocytes. Crit Rev Oral Biol Med 6:284 301 11. Paladini RD, Takahashi K, Bravo NS, Coulombe PA 1996 Onset of re-epithelialization after skin injury correlates with a reorganization of keratin filaments in wound edge keratinocytes defining a potential role for keratin 16. J Cell Biol 132:381 397 12. Freedberg IM, Tomic-Canic M, Komine M, Blumenberg M 2001 Keratins and the keratinocyte activation cycle. J Invest Dermatol 116:633 640 13. Wojcik SM, Bundman DS, Roop DR 2000 Delayed wound healing in keratin 6a knockout mice. Mol Cell Biol 20:5248 5255 14. Paramio JM, Casanova ML, Segrelles C, Mittnacht S, Lane EB, Jorcano JL 1999 Modulation of cell proliferation by cytokeratins K10 and K16. Mol Cell Biol 19:3086 3094 15. Takahashi K, Folmer J, Coulombe PA 1994 Increased expression of keratin 16 causes anomalies in cytoarchitecture and keratinization in transgenic mouse skin. J Cell Biol 127:505 520 16. Wawersik MJ, Mazzalupo S, Nguyen D, Coulombe PA 2001 Increased levels of keratin 16 alter epithelialization potential of mouse skin keratinocytes in vivo and ex vivo. Mol Biol Cell 12:3439 3450 17. McGowan KM, Tong X, Colucci-Guyon E, Langa F, Babinet C, Coulombe PA 2002 Keratin 17 null mice exhibit age- and strain-dependent alopecia. Genes Dev 16:1412 1422 18. Safer JD, Crawford TM, Holick TM 2003 Thyroid hormone stimulates gene and protein expression of proliferation-associated keratin 6a. Thyroid 13:694 19. Tomic M, Jiang CK, Epstein HS, Freedberg IM, Samuels HH, Blumenberg M 1990 Nuclear receptors for retinoic acid and thyroid hormone regulate transcription of keratin genes. Cell Regul 1:965 973 20. Ohtsuki M, Tomic-Canic M, Freedberg IM, Blumenberg M 1992 Regulation of epidermal keratin expression by retinoic acid and thyroid hormone. J Dermatol 19:774 780 21. Tomic-Canic M, Freedberg IM, Blumenberg M 1996 Codominant regulation of keratin gene expression by cell surface receptors and nuclear receptors. Exp Cell Res 224:96 102 22. Radoja N, Diaz DV, Minars TJ, Freedberg IM, Blumenberg M, Tomic-Canic M 1997 Specific organization of the negative response elements for retinoic acid and thyroid hormone receptors in keratin gene family. J Invest Dermatol 109:566 572 23. Marchuk D, McCrohon S, Fuchs E 1985 Complete sequence of a gene encoding a human type I keratin: sequences homologous to enhancer elements in the regulatory region of the gene. Proc Natl Acad Sci USA 82:1609 1613 24. RayChaudhury A, Marchuk D, Lindhurst M, Fuchs E 1985 Three tightly linked genes encoding human type I keratins: conservation of sequence in the 5 -untranslated leader and 5 -upstream regions of coexpressed keratin genes. Mol Cell Biol 6:539 548 25. Rieger M, Franke WW 1988 Identification of an orthologous mammalian cytokeratin gene. High degree of intron sequence conservation during evolution of human cytokeratin 10. J Mol Biol 204:841 856 26. Rosenberg M, Fuchs E, Le Beau MM, Eddy RL, Shows TB 1991 Three epidermal and one simple epithelial type II keratin genes map to human chromosome 12. Cytogenet Cell Genet 57:33 38 27. Takahashi K, Yan B, Yamanishi K Imamura S, Coulombe PA 1998 The two functional keratin 6 genes of mouse are differentially regulated and evolved independently from their human orthologs. Genomics 53:170 183 28. Herndon DN, Wilmore DW, Mason AD, Curreri PW 1979 Increased rates of wound healing in burned guinea pigs treated with L-thyroxine. Surg Forum 30:95 97 29. Alexander MV, Zajtchuk JT, Henderson RL 1982 Hypothyroidism and wound healing: occurrence after head and neck radiation and surgery. Arch Otolaryngol 108:289 291 30. Talmi YP, Finkelstein Y, Zohar Y 1989 Pharyngeal fistulas in postoperative hypothyroid patients. Ann Otol Rhinol Laryngol 98:267 268 31. Erdogan M, Ilhan YS, Akkus MA, Caboglu SA, Ozercan I, Ilhan N, Yaman M 1999 Effects of L-thyroxine and zinc therapy on wound healing in hypothyroid rats. Acta Chir Belg 99:72 77 Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.