Original Paper Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
Received: July 5, 2013 Accepted after revision: November 20, 2013 Published online: May 20, 2014
Effect of Induced Acute Diabetes and Insulin Therapy on Stratum Corneum Barrier Function in Rat Skin Paul A. Lehman Thomas J. Franz Department of Dermatology, University of Arkansas for Medical Sciences, John L. McCleelan Memorial Veterans Hospital, Little Rock, Ark., USA
Key Words Diabetes · Insulin · Stratum corneum · Skin · Percutaneous absorption
Abstract Background: A wide variety of cutaneous manifestations are associated with diabetes. However, there is a paucity of information on stratum corneum barrier function in diabetics, with and without insulin therapy. Methods: To assess for alteration of the stratum corneum, its barrier function was tested by evaluating the percutaneous absorption of water, ethanol, lidocaine and hydrocortisone, in vitro, on normal control, 4-week diabetic and 8-day insulin-treated diabetic Sprague-Dawley CD rats. Results: Total water penetration was not different between the 3 groups though flux profiles were different. Both total penetration and peak flux of lidocaine and hydrocortisone increased slightly in the diabetic rats over the control group. However, total penetration and peak flux (including ethanol) were significantly increased in the insulin-treated rats. Conclusion: The data indicate that diabetes modestly alters stratum corneum physiology but less so than that seen following insulin therapy. © 2014 S. Karger AG, Basel
© 2014 S. Karger AG, Basel 1660–5527/14/0275–0249$39.50/0 E-Mail [email protected] www.karger.com/spp
Introduction
A variety of cutaneous manifestations are associated with diabetes including vascular insufficiency, autonomic neuropathy, collagen metabolism anomalies, alteration in lipid components of the stratum corneum, effects on signal transduction pathways, and microangiopathy, to name a few [1–6]. However, there is a paucity of information on how diabetes and insulin therapy affect the stratum corneum barrier properties. Since the primary function of the epidermis is to produce and maintain the stratum corneum, it would be presumed that alterations to the function of the viable layer would be expressed in structural or chemical aberrations to the stratum corneum. Thus, an altered stratum corneum should demonstrate some form of abnormal barrier function. Previous investigations indicate no change in water barrier function in diabetic skin [7, 8], and one study suggesting an increase in hydrocortisone percutaneous absorption [9]. No information appears to be available on insulin-treated diabetic skin. The objective of this study was to determine if diabetes affects stratum corneum barrier function and, if so, whether insulin therapy restores its function to normal. Alterations to the barrier function, as assessed by the percutaneous absorption of water, ethanol, lidocaine and hydrocortisone, would then implicate structural anomalies of the stratum corneum as a result of abnormal viable epidermal functions in producing the stratum corneum. Paul A. Lehman QPS, LLC 1421 Rainier Street Sumner, WA 98390 (USA) E-Mail paul.lehman @ qps.com
Table 1. Body weight and blood glucose of study animals (means ±
Methods and Materials
SD) Animals Use of animals for this research was approved by the local Institutional Animal and Care Use Committees. The primary role of the animals was for diabetic research. Secondarily, the skin was harvested, following euthanasia, for this in vitro research. Eighteen adult male Sprague-Dawley CD rats (Sasco, Omaha, Nebr., USA) were maintained ad libitum on Purina Chow® diet and water for the 4-week study period. Initially 12 animals were made diabetic by a single injection of streptozotocin (Sigma Chemical Company, St. Louis, Mo., USA) at 75 mg/kg body weight through the tail vein. The streptozotocin was prepared just prior to injection in sterile 24 mM citrate buffer, pH 4.5. Control rats were injected with the citrate buffer alone. Eight days prior to sacrifice 6 of the diabetic animals received 8 units of protamine zinc insulin (Eli Lilly, Indianapolis, Ind., USA) each, by daily subcutaneous injection. Blood glucose levels were assayed by the glucose oxidase method (Sigma Chemical Company, kit No. 510). The animals were sedated with carbon dioxide inhalation followed by decapitation. Skin Preparation Prior to skin removal, the hair was closely cut using electric hair clippers while insuring that no physical damage occurred to the skin surface. Skin sections from the dorsal lumbar region, straddling the spine, were excised and cleared of subcutaneous fat and a small portion of the reticular dermis. Duplicate 1.3-cm2 skin sections from each animal were mounted onto individual static Franz diffusion cells [10]. These diffusion cells allow a 0.9-cm2 available surface area exposure to ambient laboratory conditions while the dermal side of the skin is bathed in phosphate-buffered saline (pH 7.4 ± 0.1) solution, stirred and maintained at 37 ° C. Cell samples were collected by the total removal of the receptor solution at each sample time with full replenishment with fresh solution.
Treatment
Blood glucose1, mg/dl
Body weight, g
Control Diabetic Insulin-treated diabetic
start
end1
320±9 352±16 357±13
355±13 242±30 385±25
93±15 387±712 82±76
1 Determined
on day of sacrifice. Determined after 2-hour fast.
2
tent. The remainder of each sample was assayed by high pressure liquid chromatography with ultraviolet detection (HPLC/UV) for lidocaine, and by gas chromatography-mass spectrometry (GC/ MS) for ethanol. HPLC/UV was accomplished on a C18 reverse phase 0.21 × 10 cm ODS-HypersilTM column maintained at 40 ° C using a solvent solution of acetonitrile/0.1 M ammonium acetate in water (50:50 v/v, pH 7.0; J.T. Baker Chemical Co., Phillipsburg, N.J., USA). The flow rate was 0.5 ml/min, and lidocaine elution was monitored at 210 nm from a 5-μl sample injection. The HPLC/UV system used was a model 1090M manufactured by Hewlett-Packard. Ethanol was measured in the 0.5-, 1-, 2-, 4- and 6-hour samples on a Hewlett-Packard model 5890 gas chromatograph coupled to a 5988A mass spectrometer. Chromatography was accomplished on a DBWaxTM 0.2 mm × 30 m, 0.25 film thickness, capillary column (J & W, Folsom, Calif., USA) with helium carrier gas at 10 psi head pressure. The GC oven temperature was held at 55 ° C for 1 min after injection, then ramped to 200 ° C at 70 ° C/min. The injector, MS interface and MS source were maintained at 200 ° C. Samples were introduced from a 1:50 split ratio of a 1-μl injection. Ethanol was quantified by electron impact ionization at 70 eV using selected ion monitoring. Ethanol penetration was only measured in 3 rats from each group. The rate of penetration (flux) for each skin section was calculated as well as total penetration throughout the sampling period by averaging the results from the duplicate skin sections and determining the mean across the animals within each group. The data demonstrated equal variance for statistical testing allowing analysis of variance followed by Tukey’s test.
Water Penetration Study Sufficient 3H2O (NEN Products, Boston, Mass., USA; 1 mCi/g), diluted to a specific activity of 15 μCi/ml, was applied to the skin to totally cover the exposed surface for precisely 2 min, at which time it was removed by blotting. Serial receptor solution samples were collected at 3, 6, 12, 20, 30, 40, 60 and 80 min after dose removal. One milliliter of each receptor solution sample was mixed with 5 ml scintillation fluid (Opti-FluorTM, Packard Instrument Co., Downers Grove, Ill., USA) and assayed in a Packard 1900CA liquid scintillation counter. Counts per minute were converted to decays per minute using the external standard quench correction method. Ethanol, Lidocaine and Hydrocortisone Penetration Study Approximately 14 h after the water penetration experiment, and with several changes of the receptor solution to clear any residual 3H2O, the skin permeability to a finite dose of ethanol, lidocaine and hydrocortisone was evaluated. Twenty microliters of a solution mixture consisting of 5% lidocaine (w/v; Sigma Chemical Company) and a tracer amount of 3H-hydrocortisone (NEN, 55 Ci/mmol) at 6 μCi/ml (applied dose of 600 pg) in 95% ethanol were applied to the surface of the skin. Serial phosphate-buffered saline receptor samples were collected at 0.5, 1, 2, 4, 6, 9, 12 and 24 h after application. One milliliter of the receptor sample was mixed with 5 ml scintillation fluid and counted for tritium con-
250
Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
Results
As shown in table 1, the rats receiving the streptozotocin injection demonstrated the expected diabetic response characterized by a significant increase in blood glucose level and loss in body weight (both p < 0.001 compared to the control group). The diabetic rats treated with insulin maintained body weight growth, and blood glucose levels were statistically similar to those of the control group. From visible inspection, the diabetic rat skin was Lehman/Franz
Table 2. Total in vitro percutaneous absorption of water, ethanol, lidocaine and hydrocortisone (means ± SEM)
Treatment
Water, nl/cm2 (n = 6)
Ethanol, nl/cm2 (n = 3)
Lidocaine, μg/cm2 (n = 6)
Hydrocortisone, pg/cm2 (n = 6)
Control Diabetic Insulin-treated diabetic
59±18 51±8 56±5
159±22 130±6 542±9a
301±21 332±43 437±37b
49±7 85±14 144±14c, d
n = Number of animals, each with duplicate skin sections evaluated. Water: collected over 1.2 h as amount of tritium absorbed; ethanol: collected over 6 h with GC/MS quantification; lidocaine: collected over 24 h with HPLC/UV quantification; hydrocortisone: collected over 24 h as amount of tritium absorbed. a p < 0.001 to control group and diabetic group; b p < 0.05 to control group, c p < 0.001 to control group; d p < 0.05 to diabetic group.
yellowish, with small surface dandruff-like flakes and noticeably less pliable than the skin from the control animals. The insulin-treated animal skin appeared similar to that of the control animals but with the impression of being softer and more plush. Barrier Function to Water As shown in table 2 and figure 1, there was no significant difference in the total amount of water which penetrated from the 2-min topical pulse dose between groups. In addition, there was no significant difference in the peak flux of water (fig. 1). However, a clear change in flux profile can be seen with the diabetic and insulin-treated diabetic animals. Peak flux in the control animals occurred 0.6 h after application whereas in the diabetic and insulin-treated animals peak flux had clearly shifted to an earlier time (0.2–0.3 h). Barrier Function to Ethanol, Lidocaine and Hydrocortisone Table 2 and figure 1 show the total penetration and flux profiles obtained from the finite topical dose mixture of ethanol, lidocaine and hydrocortisone. Total absorption of ethanol and lidocaine in the diabetic animals was not significantly different from control animals. However, the total amount of hydrocortisone which penetrated the diabetic skin was slightly elevated (p < 0.11) over control animals. For all 3 compounds, total absorption in the insulin-treated animals was significantly greater (p < 0.05 to p < 0.001) than that of their respective control group. This change in barrier function is clearly seen in the flux profiles of lidocaine and hydrocortisone where the diabetic animals show a trend toward an earlier and slightly greater rate of penetration which becomes significantly more pronounced in the insulin-treated diabetic animals. Although the flux proSkin Barrier Function in Diabetes
file for ethanol is poorly characterized because of the extended sampling schedule necessary to accurately describe the flux patterns of lidocaine and hydrocortisone, it still demonstrates the same effect from insulin therapy as seen for the more lipophilic compounds.
Discussion
The physiology of diabetic epidermis has garnered some investigational interest but remains less characterized than for dermatological diseases. The few published reports indicate that the corneocyte surface area in human diabetics is larger than that in age-matched nondiabetics [11] and that nonenzymatic glycosylation of keratin is significantly increased in the stratum corneum of the diabetic plantar skin [12]. In addition, Gloor et al. [13] found an increase in the percentage of free cholesterol in the surface lipids of newly diagnosed diabetics, while Huang et al. [14] found a significant restructuring in the fatty acid composition of streptozotocin diabetic rat skin when compared to control animals. The data from this study demonstrate that after 4 weeks of streptozotocin-induced diabetes there is a change in the stratum corneum. First, a pattern for the occurrence of an earlier peak flux is observed for water, lidocaine and hydrocortisone. Second, there is an increase in permeability to lipophilic compounds, evident by the increase in rate and extent of absorption, which is significantly greater for hydrocortisone. This pattern is similar to that observed in the aging of stratum corneum in normal rats fed ad libitum, including the observation that hydrocortisone permeability characteristics are more sensitive to changes in stratum corneum than those of either lidocaine or water [15]. The lack of a change in water permeability is consistent with the observation that diabetic and control groups Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
251
90
600
80
500
60
Flux (nl/cm2/h)
Flux (nl/cm2/h)
70
50 40 30 20
300 200 100
10 0
400
0
0.2
0.4
a
0.6
0.8
1.0
0
1.2
0
1
b
Mid-sample time (h)
2
3
4
5
Mid-sample time (h)
7
30
6
25 Flux (pg/cm2/h)
Flux (μg/cm2/h)
5 20 15 10
c
3 2
5 0
4
1
0
2
4
6
8
10
12
14
16
18
0
20
Mid-sample time (h)
d
0
2
4
6
8
10
12
14
16
18
20
Mid-sample time (h)
Fig. 1. Absorption profiles for water (a), ethanol (b), lidocaine (c) and hydrocortisone (d) through rat skin from the control group (◼), diabetic group (●) and the insulin-treated group (▲). Means ± SEM.
show no significant difference in stratum corneum transepidermal water loss [7, 8]. The alteration in stratum corneum permeability may be associated with changes in lipid class or fatty acid distribution in the extracellular lamella. For example, Elias and Brown [16] and Elias et al. [17] have shown that in essential fatty acid deficiency, the barrier function of the skin is compromised but repairable with topical linoleic acid. A similar situation may be occurring in the diabetic stratum corneum where Huang et al. [14] observed a significant reduction in the proportion of linoleic acid in the diabetic rat skin, and Hsia and others [18, 19] have reported that there is an impairment of lipogenesis in diabetic human skin, which may contribute to a decrease in lamellar body production [18]. 252
Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
Insulin therapy had a profound effect on the permeability of the rat skin for ethanol, lidocaine and hydrocortisone. Even though the administered insulin lowered blood glucose levels to match the control animals, it did not return the permeability of the stratum corneum to normal. This suggests that the insulin is having an independent effect on the epidermis beyond the systemic regulation of carbohydrate metabolism. Potentially, the epidermis may have experienced an overexposure of insulin, as a long-acting form was used (protamine zinc) which was administered at a relatively high dose [20]. Consequently, skin insulin levels may have been very different from what would typically be received from normal nondiabetic systemic circadian levels expected in normal roLehman/Franz
dents [21]. Insulin is well known for its stimulatory effect on cell proliferation [22] particularly for keratinocytes [23, 24]. Further, insulin will increase 14C-acetate incorporation into polyunsaturated fatty acids of skin [25] and increase fatty acid desaturase activity [20]. Thus, the response observed in the permeability kinetics of the insulin-treated diabetic rats may be from a hyperproliferation response or a further lipid imbalance in the stratum corneum lamella. Little information is available on systemic factors, separate from in situ factors, which may modify or regulate stratum corneum barrier function. It has been shown that dietary caloric intake and aging alter skin permeability characteristics [15]. In this study disruption in carbohydrate regulation by streptozotocin and the subsequent exposure to insulin to the epidermis from the systemic circulation resulted in alterations to the skin barrier function and permeability to lipophilic molecules. As it has
also been shown that blood glucose fluctuation affects skin collagen metabolism [26], one could presume that other skin physiological functions may also be affected by similar systemic effects. It would seem most profitable to closely examine the lipid composition of the stratum corneum under these different systemic conditions. The diabetic and insulin-treated diabetic animals may prove to be useful models for the further characterization of the barrier matrix organization and its influence on skin permeability. It remains to be seen whether the observations made in this study also occur in human diabetic or insulin-treated human diabetic skin. Acknowledgment The author is grateful to Dr. An Q. Dang for sharing the data in table 1 and for providing the animal skin following completion of his research.
References 1 Huntley AC: The cutaneous manifestations of diabetes mellitus. J Am Acad Dermatol 1982; 7:427–455. 2 Feingolg DR, Elias PM: Endocrine-skin interactions. J Am Acad Dermatol 1987; 17: 921– 940. 3 Goodfiled MJD, Millard LG: The skin in diabetes mellitus. Diabetologia 1988;31:567–575. 4 Martinex SM, Tarres MC, Monenegro S, Revelant G, Figueroa N, Alonso D, Laudanno OMO, D’Ottavio A: Intermittent dietary restriction in eSS diabetic rats. Effects on metabolic control and skin morphology. Acta Diabetol Lat 1990;27:329–336. 5 Hanna W, Friesen D, Bombardier C, Gladman D, Hanna A: Pathologic features of diabetic thick skin. J Am Acad Dermatol 1987;16: 546–553. 6 Ge X, Shi Z, Yu N, Jiao Y, Jin L, Zhang J: The role of EGFR/ERK/ELK-1 MAP kinase pathway in the underlying damage to diabetic rat skin. Indian J Dermatol 2013;58:101–106. 7 Seirafi H, Farsinejad K, Firooz A, Davoudi SM, Robati RM, Hoseini MS, Ehsani AH, Sadr B: Biophysical characteristics of skin in diabetes: a controlled study. J Eur Acad Dermatol Venereol 2009;23:146–149. 8 Sakai S, Endo Y, Ozawa N, Sugawara T, Kusaka A, Sayo T, Tagami H, Inoue S: Characteristics of the epidermis and stratum corneum of hairless mice with experimentally induced diabetes mellitus. J Invest Dermatol 2003;120:79–85. 9 Wang L, Li GF, Hu WJ, Zhu XL, Xiong LQ, Deng ZH: The histological changes of diabetic rats’ skin and the effects on the percutaneous absorption of glucocorticoid (in Chinese). Yao Xue Xue Bao 2010;45:114–119.
Skin Barrier Function in Diabetes
10 Franz TJ: The finite dose technique as a valid in vitro model for the study of percutaneous absorption in man. Curr Probl Dermatol 1978;7:58–68. 11 Yajima Y, Sueki H, Fujisawa R: Increased corneocyte surface area in the diabetic skin. Jpn J Dermatol 1991;101:129–134. 12 Delbridge L, Ellis CS, Robertson K, Lequesne LP: Non-enzymatic glycosylation of keratin from the stratum corneum of the diabetic foot. Br J Dermatol 1985;112:547–554. 13 Gloor M, Marckardt V, Friederich HC: Biochemical and physiological particularities on the skin surface of diabetics. Arch Dermatol Res 1975;253:185–194. 14 Huang YS, Horrobin DF, Manku MS, Mitchell J, Ryan MA: Tissue phospholipid fatty acid composition in the diabetic rat. Lipids 1984; 19:367–370. 15 Lehman PA, Franz TJ: Effect of age and diet on stratum corneum barrier function in the Fischer 344 female rat. J Invest Dermatol 1993;100:200–204. 16 Elias PM, Brown BE: The mammalian cutaneous permeability barrier. Defective barrier function in essential fatty acid deficiency correlates with abnormal intercellular lipid deposition. Lab Invest 1978;39:574–583. 17 Elias PM, Brown BE, Ziboh VA: The permeability barrier in essential fatty acid deficiency: evidence for a direct role for linoleic acid in barrier function. J Invest Dermatol 1980; 74:230–233.
18 Hsia SL, Dreize MA, Marquex MC: Lipid metabolism in human skin. II. A study of lipogenesis in skin of diabetic patients. J Invest Dermatol 1966;47:443–448. 19 Park HY, Kim JH, Jung M, Chung CH, Hasham R, Park CS, Choi EH: A long-standing hyperglycaemic condition impairs skin barrier by accelerating skin ageing process. Exp Dermatol 2011;20:969–974. 20 Eck MG, Wynn JO, Carter WJ, Fass FH: Fatty acid desaturation in experimental diabetes mellitus. J Diabetes 1979;28:479–485. 21 Hunter JD, McGee J, Saldivar J, Tsai T, Feuers R, Scheving LE: Circadian variations in insulin and alloxan sensitivity noted in blood glucose alterations in normal and diabetic mice. Ann Rev Chronopharmacol 1988;5:295–298. 22 Fransson J, Hammar H: Epidermal growth. Int J Dermatol 1988;27:281–290. 23 O’Keefe EJ, Chiu MC: Stimulation of thymidine incorporation in keratinocytes by insulin, epidermal growth factor and placental extract: comparison with cell number to assess growth. J Invest Dermatol 1988;90:2–7. 24 Vaughan FL, Kass LL, Usman JA: Requirement of hydrocortisone and insulin for extended proliferation and passage of rat keratinocytes. In Vitro 1981;17:941–946. 25 Wilkinson DI: Some factors affecting C14-acetate incorporation into polyunsaturated fatty acids of skin. J Invest Dermatol 1973; 60: 188–192. 26 Ye X, Cheng X, Liu L, Zhao D, Dang Y: Blood glucose fluctuation affects skin collagen metabolism in the diabetic mouse by inhibiting the mitogen-activated protein kinase and Smad pathways. Clin Exp Dermatol 2013; 38: 530–537.
Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
253
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
Received: July 5, 2013 Accepted after revision: November 20, 2013 Published online: May 20, 2014
Effect of Induced Acute Diabetes and Insulin Therapy on Stratum Corneum Barrier Function in Rat Skin Paul A. Lehman Thomas J. Franz Department of Dermatology, University of Arkansas for Medical Sciences, John L. McCleelan Memorial Veterans Hospital, Little Rock, Ark., USA
Key Words Diabetes · Insulin · Stratum corneum · Skin · Percutaneous absorption
Abstract Background: A wide variety of cutaneous manifestations are associated with diabetes. However, there is a paucity of information on stratum corneum barrier function in diabetics, with and without insulin therapy. Methods: To assess for alteration of the stratum corneum, its barrier function was tested by evaluating the percutaneous absorption of water, ethanol, lidocaine and hydrocortisone, in vitro, on normal control, 4-week diabetic and 8-day insulin-treated diabetic Sprague-Dawley CD rats. Results: Total water penetration was not different between the 3 groups though flux profiles were different. Both total penetration and peak flux of lidocaine and hydrocortisone increased slightly in the diabetic rats over the control group. However, total penetration and peak flux (including ethanol) were significantly increased in the insulin-treated rats. Conclusion: The data indicate that diabetes modestly alters stratum corneum physiology but less so than that seen following insulin therapy. © 2014 S. Karger AG, Basel
© 2014 S. Karger AG, Basel 1660–5527/14/0275–0249$39.50/0 E-Mail [email protected] www.karger.com/spp
Introduction
A variety of cutaneous manifestations are associated with diabetes including vascular insufficiency, autonomic neuropathy, collagen metabolism anomalies, alteration in lipid components of the stratum corneum, effects on signal transduction pathways, and microangiopathy, to name a few [1–6]. However, there is a paucity of information on how diabetes and insulin therapy affect the stratum corneum barrier properties. Since the primary function of the epidermis is to produce and maintain the stratum corneum, it would be presumed that alterations to the function of the viable layer would be expressed in structural or chemical aberrations to the stratum corneum. Thus, an altered stratum corneum should demonstrate some form of abnormal barrier function. Previous investigations indicate no change in water barrier function in diabetic skin [7, 8], and one study suggesting an increase in hydrocortisone percutaneous absorption [9]. No information appears to be available on insulin-treated diabetic skin. The objective of this study was to determine if diabetes affects stratum corneum barrier function and, if so, whether insulin therapy restores its function to normal. Alterations to the barrier function, as assessed by the percutaneous absorption of water, ethanol, lidocaine and hydrocortisone, would then implicate structural anomalies of the stratum corneum as a result of abnormal viable epidermal functions in producing the stratum corneum. Paul A. Lehman QPS, LLC 1421 Rainier Street Sumner, WA 98390 (USA) E-Mail paul.lehman @ qps.com
Table 1. Body weight and blood glucose of study animals (means ±
Methods and Materials
SD) Animals Use of animals for this research was approved by the local Institutional Animal and Care Use Committees. The primary role of the animals was for diabetic research. Secondarily, the skin was harvested, following euthanasia, for this in vitro research. Eighteen adult male Sprague-Dawley CD rats (Sasco, Omaha, Nebr., USA) were maintained ad libitum on Purina Chow® diet and water for the 4-week study period. Initially 12 animals were made diabetic by a single injection of streptozotocin (Sigma Chemical Company, St. Louis, Mo., USA) at 75 mg/kg body weight through the tail vein. The streptozotocin was prepared just prior to injection in sterile 24 mM citrate buffer, pH 4.5. Control rats were injected with the citrate buffer alone. Eight days prior to sacrifice 6 of the diabetic animals received 8 units of protamine zinc insulin (Eli Lilly, Indianapolis, Ind., USA) each, by daily subcutaneous injection. Blood glucose levels were assayed by the glucose oxidase method (Sigma Chemical Company, kit No. 510). The animals were sedated with carbon dioxide inhalation followed by decapitation. Skin Preparation Prior to skin removal, the hair was closely cut using electric hair clippers while insuring that no physical damage occurred to the skin surface. Skin sections from the dorsal lumbar region, straddling the spine, were excised and cleared of subcutaneous fat and a small portion of the reticular dermis. Duplicate 1.3-cm2 skin sections from each animal were mounted onto individual static Franz diffusion cells [10]. These diffusion cells allow a 0.9-cm2 available surface area exposure to ambient laboratory conditions while the dermal side of the skin is bathed in phosphate-buffered saline (pH 7.4 ± 0.1) solution, stirred and maintained at 37 ° C. Cell samples were collected by the total removal of the receptor solution at each sample time with full replenishment with fresh solution.
Treatment
Blood glucose1, mg/dl
Body weight, g
Control Diabetic Insulin-treated diabetic
start
end1
320±9 352±16 357±13
355±13 242±30 385±25
93±15 387±712 82±76
1 Determined
on day of sacrifice. Determined after 2-hour fast.
2
tent. The remainder of each sample was assayed by high pressure liquid chromatography with ultraviolet detection (HPLC/UV) for lidocaine, and by gas chromatography-mass spectrometry (GC/ MS) for ethanol. HPLC/UV was accomplished on a C18 reverse phase 0.21 × 10 cm ODS-HypersilTM column maintained at 40 ° C using a solvent solution of acetonitrile/0.1 M ammonium acetate in water (50:50 v/v, pH 7.0; J.T. Baker Chemical Co., Phillipsburg, N.J., USA). The flow rate was 0.5 ml/min, and lidocaine elution was monitored at 210 nm from a 5-μl sample injection. The HPLC/UV system used was a model 1090M manufactured by Hewlett-Packard. Ethanol was measured in the 0.5-, 1-, 2-, 4- and 6-hour samples on a Hewlett-Packard model 5890 gas chromatograph coupled to a 5988A mass spectrometer. Chromatography was accomplished on a DBWaxTM 0.2 mm × 30 m, 0.25 film thickness, capillary column (J & W, Folsom, Calif., USA) with helium carrier gas at 10 psi head pressure. The GC oven temperature was held at 55 ° C for 1 min after injection, then ramped to 200 ° C at 70 ° C/min. The injector, MS interface and MS source were maintained at 200 ° C. Samples were introduced from a 1:50 split ratio of a 1-μl injection. Ethanol was quantified by electron impact ionization at 70 eV using selected ion monitoring. Ethanol penetration was only measured in 3 rats from each group. The rate of penetration (flux) for each skin section was calculated as well as total penetration throughout the sampling period by averaging the results from the duplicate skin sections and determining the mean across the animals within each group. The data demonstrated equal variance for statistical testing allowing analysis of variance followed by Tukey’s test.
Water Penetration Study Sufficient 3H2O (NEN Products, Boston, Mass., USA; 1 mCi/g), diluted to a specific activity of 15 μCi/ml, was applied to the skin to totally cover the exposed surface for precisely 2 min, at which time it was removed by blotting. Serial receptor solution samples were collected at 3, 6, 12, 20, 30, 40, 60 and 80 min after dose removal. One milliliter of each receptor solution sample was mixed with 5 ml scintillation fluid (Opti-FluorTM, Packard Instrument Co., Downers Grove, Ill., USA) and assayed in a Packard 1900CA liquid scintillation counter. Counts per minute were converted to decays per minute using the external standard quench correction method. Ethanol, Lidocaine and Hydrocortisone Penetration Study Approximately 14 h after the water penetration experiment, and with several changes of the receptor solution to clear any residual 3H2O, the skin permeability to a finite dose of ethanol, lidocaine and hydrocortisone was evaluated. Twenty microliters of a solution mixture consisting of 5% lidocaine (w/v; Sigma Chemical Company) and a tracer amount of 3H-hydrocortisone (NEN, 55 Ci/mmol) at 6 μCi/ml (applied dose of 600 pg) in 95% ethanol were applied to the surface of the skin. Serial phosphate-buffered saline receptor samples were collected at 0.5, 1, 2, 4, 6, 9, 12 and 24 h after application. One milliliter of the receptor sample was mixed with 5 ml scintillation fluid and counted for tritium con-
250
Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
Results
As shown in table 1, the rats receiving the streptozotocin injection demonstrated the expected diabetic response characterized by a significant increase in blood glucose level and loss in body weight (both p < 0.001 compared to the control group). The diabetic rats treated with insulin maintained body weight growth, and blood glucose levels were statistically similar to those of the control group. From visible inspection, the diabetic rat skin was Lehman/Franz
Table 2. Total in vitro percutaneous absorption of water, ethanol, lidocaine and hydrocortisone (means ± SEM)
Treatment
Water, nl/cm2 (n = 6)
Ethanol, nl/cm2 (n = 3)
Lidocaine, μg/cm2 (n = 6)
Hydrocortisone, pg/cm2 (n = 6)
Control Diabetic Insulin-treated diabetic
59±18 51±8 56±5
159±22 130±6 542±9a
301±21 332±43 437±37b
49±7 85±14 144±14c, d
n = Number of animals, each with duplicate skin sections evaluated. Water: collected over 1.2 h as amount of tritium absorbed; ethanol: collected over 6 h with GC/MS quantification; lidocaine: collected over 24 h with HPLC/UV quantification; hydrocortisone: collected over 24 h as amount of tritium absorbed. a p < 0.001 to control group and diabetic group; b p < 0.05 to control group, c p < 0.001 to control group; d p < 0.05 to diabetic group.
yellowish, with small surface dandruff-like flakes and noticeably less pliable than the skin from the control animals. The insulin-treated animal skin appeared similar to that of the control animals but with the impression of being softer and more plush. Barrier Function to Water As shown in table 2 and figure 1, there was no significant difference in the total amount of water which penetrated from the 2-min topical pulse dose between groups. In addition, there was no significant difference in the peak flux of water (fig. 1). However, a clear change in flux profile can be seen with the diabetic and insulin-treated diabetic animals. Peak flux in the control animals occurred 0.6 h after application whereas in the diabetic and insulin-treated animals peak flux had clearly shifted to an earlier time (0.2–0.3 h). Barrier Function to Ethanol, Lidocaine and Hydrocortisone Table 2 and figure 1 show the total penetration and flux profiles obtained from the finite topical dose mixture of ethanol, lidocaine and hydrocortisone. Total absorption of ethanol and lidocaine in the diabetic animals was not significantly different from control animals. However, the total amount of hydrocortisone which penetrated the diabetic skin was slightly elevated (p < 0.11) over control animals. For all 3 compounds, total absorption in the insulin-treated animals was significantly greater (p < 0.05 to p < 0.001) than that of their respective control group. This change in barrier function is clearly seen in the flux profiles of lidocaine and hydrocortisone where the diabetic animals show a trend toward an earlier and slightly greater rate of penetration which becomes significantly more pronounced in the insulin-treated diabetic animals. Although the flux proSkin Barrier Function in Diabetes
file for ethanol is poorly characterized because of the extended sampling schedule necessary to accurately describe the flux patterns of lidocaine and hydrocortisone, it still demonstrates the same effect from insulin therapy as seen for the more lipophilic compounds.
Discussion
The physiology of diabetic epidermis has garnered some investigational interest but remains less characterized than for dermatological diseases. The few published reports indicate that the corneocyte surface area in human diabetics is larger than that in age-matched nondiabetics [11] and that nonenzymatic glycosylation of keratin is significantly increased in the stratum corneum of the diabetic plantar skin [12]. In addition, Gloor et al. [13] found an increase in the percentage of free cholesterol in the surface lipids of newly diagnosed diabetics, while Huang et al. [14] found a significant restructuring in the fatty acid composition of streptozotocin diabetic rat skin when compared to control animals. The data from this study demonstrate that after 4 weeks of streptozotocin-induced diabetes there is a change in the stratum corneum. First, a pattern for the occurrence of an earlier peak flux is observed for water, lidocaine and hydrocortisone. Second, there is an increase in permeability to lipophilic compounds, evident by the increase in rate and extent of absorption, which is significantly greater for hydrocortisone. This pattern is similar to that observed in the aging of stratum corneum in normal rats fed ad libitum, including the observation that hydrocortisone permeability characteristics are more sensitive to changes in stratum corneum than those of either lidocaine or water [15]. The lack of a change in water permeability is consistent with the observation that diabetic and control groups Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
251
90
600
80
500
60
Flux (nl/cm2/h)
Flux (nl/cm2/h)
70
50 40 30 20
300 200 100
10 0
400
0
0.2
0.4
a
0.6
0.8
1.0
0
1.2
0
1
b
Mid-sample time (h)
2
3
4
5
Mid-sample time (h)
7
30
6
25 Flux (pg/cm2/h)
Flux (μg/cm2/h)
5 20 15 10
c
3 2
5 0
4
1
0
2
4
6
8
10
12
14
16
18
0
20
Mid-sample time (h)
d
0
2
4
6
8
10
12
14
16
18
20
Mid-sample time (h)
Fig. 1. Absorption profiles for water (a), ethanol (b), lidocaine (c) and hydrocortisone (d) through rat skin from the control group (◼), diabetic group (●) and the insulin-treated group (▲). Means ± SEM.
show no significant difference in stratum corneum transepidermal water loss [7, 8]. The alteration in stratum corneum permeability may be associated with changes in lipid class or fatty acid distribution in the extracellular lamella. For example, Elias and Brown [16] and Elias et al. [17] have shown that in essential fatty acid deficiency, the barrier function of the skin is compromised but repairable with topical linoleic acid. A similar situation may be occurring in the diabetic stratum corneum where Huang et al. [14] observed a significant reduction in the proportion of linoleic acid in the diabetic rat skin, and Hsia and others [18, 19] have reported that there is an impairment of lipogenesis in diabetic human skin, which may contribute to a decrease in lamellar body production [18]. 252
Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
Insulin therapy had a profound effect on the permeability of the rat skin for ethanol, lidocaine and hydrocortisone. Even though the administered insulin lowered blood glucose levels to match the control animals, it did not return the permeability of the stratum corneum to normal. This suggests that the insulin is having an independent effect on the epidermis beyond the systemic regulation of carbohydrate metabolism. Potentially, the epidermis may have experienced an overexposure of insulin, as a long-acting form was used (protamine zinc) which was administered at a relatively high dose [20]. Consequently, skin insulin levels may have been very different from what would typically be received from normal nondiabetic systemic circadian levels expected in normal roLehman/Franz
dents [21]. Insulin is well known for its stimulatory effect on cell proliferation [22] particularly for keratinocytes [23, 24]. Further, insulin will increase 14C-acetate incorporation into polyunsaturated fatty acids of skin [25] and increase fatty acid desaturase activity [20]. Thus, the response observed in the permeability kinetics of the insulin-treated diabetic rats may be from a hyperproliferation response or a further lipid imbalance in the stratum corneum lamella. Little information is available on systemic factors, separate from in situ factors, which may modify or regulate stratum corneum barrier function. It has been shown that dietary caloric intake and aging alter skin permeability characteristics [15]. In this study disruption in carbohydrate regulation by streptozotocin and the subsequent exposure to insulin to the epidermis from the systemic circulation resulted in alterations to the skin barrier function and permeability to lipophilic molecules. As it has
also been shown that blood glucose fluctuation affects skin collagen metabolism [26], one could presume that other skin physiological functions may also be affected by similar systemic effects. It would seem most profitable to closely examine the lipid composition of the stratum corneum under these different systemic conditions. The diabetic and insulin-treated diabetic animals may prove to be useful models for the further characterization of the barrier matrix organization and its influence on skin permeability. It remains to be seen whether the observations made in this study also occur in human diabetic or insulin-treated human diabetic skin. Acknowledgment The author is grateful to Dr. An Q. Dang for sharing the data in table 1 and for providing the animal skin following completion of his research.
References 1 Huntley AC: The cutaneous manifestations of diabetes mellitus. J Am Acad Dermatol 1982; 7:427–455. 2 Feingolg DR, Elias PM: Endocrine-skin interactions. J Am Acad Dermatol 1987; 17: 921– 940. 3 Goodfiled MJD, Millard LG: The skin in diabetes mellitus. Diabetologia 1988;31:567–575. 4 Martinex SM, Tarres MC, Monenegro S, Revelant G, Figueroa N, Alonso D, Laudanno OMO, D’Ottavio A: Intermittent dietary restriction in eSS diabetic rats. Effects on metabolic control and skin morphology. Acta Diabetol Lat 1990;27:329–336. 5 Hanna W, Friesen D, Bombardier C, Gladman D, Hanna A: Pathologic features of diabetic thick skin. J Am Acad Dermatol 1987;16: 546–553. 6 Ge X, Shi Z, Yu N, Jiao Y, Jin L, Zhang J: The role of EGFR/ERK/ELK-1 MAP kinase pathway in the underlying damage to diabetic rat skin. Indian J Dermatol 2013;58:101–106. 7 Seirafi H, Farsinejad K, Firooz A, Davoudi SM, Robati RM, Hoseini MS, Ehsani AH, Sadr B: Biophysical characteristics of skin in diabetes: a controlled study. J Eur Acad Dermatol Venereol 2009;23:146–149. 8 Sakai S, Endo Y, Ozawa N, Sugawara T, Kusaka A, Sayo T, Tagami H, Inoue S: Characteristics of the epidermis and stratum corneum of hairless mice with experimentally induced diabetes mellitus. J Invest Dermatol 2003;120:79–85. 9 Wang L, Li GF, Hu WJ, Zhu XL, Xiong LQ, Deng ZH: The histological changes of diabetic rats’ skin and the effects on the percutaneous absorption of glucocorticoid (in Chinese). Yao Xue Xue Bao 2010;45:114–119.
Skin Barrier Function in Diabetes
10 Franz TJ: The finite dose technique as a valid in vitro model for the study of percutaneous absorption in man. Curr Probl Dermatol 1978;7:58–68. 11 Yajima Y, Sueki H, Fujisawa R: Increased corneocyte surface area in the diabetic skin. Jpn J Dermatol 1991;101:129–134. 12 Delbridge L, Ellis CS, Robertson K, Lequesne LP: Non-enzymatic glycosylation of keratin from the stratum corneum of the diabetic foot. Br J Dermatol 1985;112:547–554. 13 Gloor M, Marckardt V, Friederich HC: Biochemical and physiological particularities on the skin surface of diabetics. Arch Dermatol Res 1975;253:185–194. 14 Huang YS, Horrobin DF, Manku MS, Mitchell J, Ryan MA: Tissue phospholipid fatty acid composition in the diabetic rat. Lipids 1984; 19:367–370. 15 Lehman PA, Franz TJ: Effect of age and diet on stratum corneum barrier function in the Fischer 344 female rat. J Invest Dermatol 1993;100:200–204. 16 Elias PM, Brown BE: The mammalian cutaneous permeability barrier. Defective barrier function in essential fatty acid deficiency correlates with abnormal intercellular lipid deposition. Lab Invest 1978;39:574–583. 17 Elias PM, Brown BE, Ziboh VA: The permeability barrier in essential fatty acid deficiency: evidence for a direct role for linoleic acid in barrier function. J Invest Dermatol 1980; 74:230–233.
18 Hsia SL, Dreize MA, Marquex MC: Lipid metabolism in human skin. II. A study of lipogenesis in skin of diabetic patients. J Invest Dermatol 1966;47:443–448. 19 Park HY, Kim JH, Jung M, Chung CH, Hasham R, Park CS, Choi EH: A long-standing hyperglycaemic condition impairs skin barrier by accelerating skin ageing process. Exp Dermatol 2011;20:969–974. 20 Eck MG, Wynn JO, Carter WJ, Fass FH: Fatty acid desaturation in experimental diabetes mellitus. J Diabetes 1979;28:479–485. 21 Hunter JD, McGee J, Saldivar J, Tsai T, Feuers R, Scheving LE: Circadian variations in insulin and alloxan sensitivity noted in blood glucose alterations in normal and diabetic mice. Ann Rev Chronopharmacol 1988;5:295–298. 22 Fransson J, Hammar H: Epidermal growth. Int J Dermatol 1988;27:281–290. 23 O’Keefe EJ, Chiu MC: Stimulation of thymidine incorporation in keratinocytes by insulin, epidermal growth factor and placental extract: comparison with cell number to assess growth. J Invest Dermatol 1988;90:2–7. 24 Vaughan FL, Kass LL, Usman JA: Requirement of hydrocortisone and insulin for extended proliferation and passage of rat keratinocytes. In Vitro 1981;17:941–946. 25 Wilkinson DI: Some factors affecting C14-acetate incorporation into polyunsaturated fatty acids of skin. J Invest Dermatol 1973; 60: 188–192. 26 Ye X, Cheng X, Liu L, Zhao D, Dang Y: Blood glucose fluctuation affects skin collagen metabolism in the diabetic mouse by inhibiting the mitogen-activated protein kinase and Smad pathways. Clin Exp Dermatol 2013; 38: 530–537.
Skin Pharmacol Physiol 2014;27:249–253 DOI: 10.1159/000357478
253
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
