Portal de Periódicos da CAPES

In Vivo Formation of Maillard Reaction Free Radicals in Mouse Skin

2001; Elsevier BV; Volume: 117; Issue: 3 Linguagem: Inglês

10.1046/j.0022-202x.2001.01448.x

1523-1747

Roger V. Lloyd, Anna J. Fong, Robert M. Sayre,

Biochemical effects in animals

The Maillard browning reaction between carbohydrates and amines is part of an extensive series of reactions that is the basis for the brown color caused by the “sunless tanning” agent dihydroxyacetone in self-tanning products. The initial stages of the reaction are quite complex, but the ultimate products are brown polymers known collectively as melanoidins. We have now used electron spin resonance to show that radicals are produced in vivo by the Maillard reaction, initiated by treating the skin of hairless mice with a solution of dihydroxyacetone in buffer. Dihydroxyacetone was used as the carbohydrate because it is simple but highly reactive and is the only USFDA approved color additive for the production of a sunless tanning response on skin. Treated skin turned brown within 24 h and showed an electron spin resonance signal after sacrifice of the animal. The control sample, consisting of untreated skin from the same animal, remained its original pink color and had no electron spin resonance signal. In corresponding ex vivo experiments in which mouse skin was soaked in dihydroxyacetone solutions, it was conclusively demonstrated that the presence of the dihydroxyacetone was required for radical formation in skin. In both the in vivo and ex vivo reactions the electron spin resonance signal consists of a broad single line with a peak-to-peak linewidth of 15 Gauss and a g value of 2.0035. We suggest that dihydroxyacetone interacts on skin through a free radical mediated reaction similar to its in vitro reactions with amines and amino acids. The Maillard browning reaction between carbohydrates and amines is part of an extensive series of reactions that is the basis for the brown color caused by the “sunless tanning” agent dihydroxyacetone in self-tanning products. The initial stages of the reaction are quite complex, but the ultimate products are brown polymers known collectively as melanoidins. We have now used electron spin resonance to show that radicals are produced in vivo by the Maillard reaction, initiated by treating the skin of hairless mice with a solution of dihydroxyacetone in buffer. Dihydroxyacetone was used as the carbohydrate because it is simple but highly reactive and is the only USFDA approved color additive for the production of a sunless tanning response on skin. Treated skin turned brown within 24 h and showed an electron spin resonance signal after sacrifice of the animal. The control sample, consisting of untreated skin from the same animal, remained its original pink color and had no electron spin resonance signal. In corresponding ex vivo experiments in which mouse skin was soaked in dihydroxyacetone solutions, it was conclusively demonstrated that the presence of the dihydroxyacetone was required for radical formation in skin. In both the in vivo and ex vivo reactions the electron spin resonance signal consists of a broad single line with a peak-to-peak linewidth of 15 Gauss and a g value of 2.0035. We suggest that dihydroxyacetone interacts on skin through a free radical mediated reaction similar to its in vitro reactions with amines and amino acids. dihydroxyacetone electron spin resonance The reaction between carbohydrates and amines is not familiar in conventional organic chemistry but to food chemists it is well known as part of an extensive series of reactions collectively called the Maillard browning reaction, or more generally the nonenzymatic browning reaction (Hodge, 1953Hodge J.E. Chemistry of the browning reaction in model systems.J Agric Food Chem. 1953; 1: 928-943Crossref Scopus (1358) Google Scholar;Namiki, 1988Namiki M. Chemistry of the Maillard reaction: recent studies on the browning reaction mechanism and the development of antioxidants and mutagens.Advan Food Rsch. 1988; 32: 115-184Crossref PubMed Scopus (246) Google Scholar;Ledl and Schleicher, 1990Ledl F. Schleicher E. New aspects of the Maillard reaction in foods and in the human body.Angew Chem Intl Ed Engl. 1990; 29: 565-594Crossref Scopus (651) Google Scholar;Mullarkey et al., 1990Mullarkey C.J. Edelstein D. Brownlee M. Free radical generation by early glycation products: a mechanism for accelerated atherogenesis in diabetes.Biochem Biophys Rsch Commun. 1990; 173: 932-939Crossref PubMed Scopus (625) Google Scholar;Roberts and Lloyd, 1997Roberts R.L. Lloyd R.V. Free radical formation from secondary amines in the Maillard reaction.J Agric Food Chem. 1997; 45: 2413-2418https://doi.org/10.1021/jf960902cCrossref Scopus (36) Google Scholar). The initial stages of the reaction are quite complex, but the ultimate products are brown polymers known collectively as melanoidins. The Maillard reaction is the basis for the brown color caused by the “sunless tanning” agent dihydroxyacetone (DHA) in self-tanning products. The use of DHA as a topical self-tanning agent has been reviewed (Levy, 1992Levy S.B. Dihydroxyacetone-containing sunless or self-tanning lotions.J Am Acad Dermatol. 1992; 27: 989-993Abstract Full Text PDF PubMed Scopus (52) Google Scholar).Puccetti and Leblanc, 2000Puccetti G. Leblanc R. A sunscreen-tanning compromise. 3D visualization of the actions of titanium dioxide particles and dihydroxyacetone on human epiderm.Photochem Photobiol. 2000; 71: 426-430PubMed Google Scholar used photoacoustic spectroscopy with human autopsy epidermis to suggest that the action of DHA took place in the lower stratum corneum. It has been proposed that DHA when applied to skin has value in providing photoprotection (Taylor et al., 1999Taylor C.R. Kwangsukstith C. Wimberly J. Kollias N. Anderson R.R. Turbo PUVA: dihydroxyacetone-enhanced photochemotherapy for psoriasis.Arch Dermatol. 1999; 135: 540-544PubMed Google Scholar).Johnson and Fusaro, 1994Johnson J.A. Fusaro R.M. Persistence of skin color and fluorescence after treatment with dihydroxyacetone.Dermatology. 1994; 247: 188Google Scholar also point out that DHA reacts with skin to form fluorescent melanoidins, the ultimate products of the Maillard reaction, which afford ultraviolet (UV) protection. While investigating the in vitro Maillard reaction between carbohydrates and various amines or amino acids, Namiki and coworkers showed that it could proceed via a free radical pathway (Hayashi et al., 1977Hayashi T. Ohta Y. Namiki M. Electron spin resonance study on the structure of the novel free radical products formed by the reactions of sugars with amino acids or amines.J Agric Food Chem. 1977; 25: 1282-1287Crossref PubMed Scopus (63) Google Scholar;Namiki and Hayashi, 1983Namiki M. Hayashi T. A new mechanism of the Maillard reaction involving sugar fragmentation and free radical formation.ACS Symp Series. 1983; 215: 21-46Crossref Google Scholar). This was in addition to the usual reaction initiated by formation of a Schiff base, followed by the Amadori rearrangement (Hodge, 1953Hodge J.E. Chemistry of the browning reaction in model systems.J Agric Food Chem. 1953; 1: 928-943Crossref Scopus (1358) Google Scholar). The radical intermediates were identified by electron spin resonance (ESR) spectroscopy as substituted pyrazinium cation radicals, based on the N, N′-dialkyldihydropyrazine ring system. Stable free radical species are also found in the melanoidin final products (Wu et al., 1986Wu C.H. Russell G.F. Powrie W.D. Paramagnetic behavior of model system melanoidins.Dev Food Sci. 1986; 13: 135-144Google Scholar;Namiki, 1988Namiki M. Chemistry of the Maillard reaction: recent studies on the browning reaction mechanism and the development of antioxidants and mutagens.Advan Food Rsch. 1988; 32: 115-184Crossref PubMed Scopus (246) Google Scholar). The ESR spectra of the latter radicals generally consist of broad singlets, with hyperfine structure in some cases, depending on the particular amine–carbohydrate system involved. There have been a number of observations of free radicals in skin, generally after UV irradiation. Buettner and coworkers detected the ESR signal of the endogenous ascorbyl radical in mouse skin and found that it increased upon exposure of the skin to UV light (Buettner et al., 1987Buettner G.R. Motten A.G. Hall R.D. Chignell C.F. ESR detection of endogenous ascorbate free radicals in mouse skin: enhancement of radical production during UV irradiation following topical application of chlorpromazine.Photochem Photobiol. 1987; 46: 161-164Crossref PubMed Scopus (28) Google Scholar;Jurkiewicz and Buettner, 1994Jurkiewicz R.A. Buettner G.A. Ultraviolet light-induced free radical formation in skin: an electron paramagnetic resonance study.Photochem Photobiol. 1994; 59: 1-4Crossref PubMed Scopus (208) Google Scholar), suggesting that the ascorbyl radical was a response to oxidative stress. A short-lived radical has been detected in UV-irradiated white and albino skin, either directly (Pathak and Stratton, 1968Pathak M.A. Stratton K. Free radicals in human skin before and after exposure to light.Arch Biochem Biophys. 1968; 123: 468-476Crossref PubMed Scopus (146) Google Scholar) or by spin-trapping (Jurkiewicz and Buettner, 1996Jurkiewicz R.A. Buettner G.A. EPR detection of free radicals in UV-irradiated skin: mouse vs. human.Photochem Photobiol. 1996; 64: 918-922Crossref PubMed Scopus (112) Google Scholar). There have also been a number of studies of melanin free radicals in skin, again induced by UV photolysis. The melanin radical signal is reported to be a single line with a linewidth of 4–5 Gauss and a g value of 2.0041 (Collins et al., 1995Collins B. Poehler T.O. Bryden W.A. EPR persistence measurements of UV-induced melanin free radicals in whole skin.Photochem Photobiol. 1995; 62: 557-560Crossref PubMed Scopus (17) Google Scholar). In spite of the extensive literature on the Maillard reaction and on radicals in skin, there has been no direct observation of Maillard reaction free radicals in vivo. We have now used ESR to show that radicals are produced in vivo by the Maillard reaction, specifically in skin on albino hairless mice. In a corresponding ex vivo reaction, it is conclusively demonstrated that radical formation requires the presence of a carbohydrate. DHA was used for most experiments because it is a simple but highly reactive carbohydrate (Thornally et al., 1984Thornally P. Wolffe S. Crabbe J. Stern A. The autoxidation of glyceraldehyde and other simple monosaccharides under physiological conditions catalyzed by buffer ions.Biochem Biophys Acta. 1984; 797: 276-287Crossref PubMed Scopus (203) Google Scholar), and it is the only USFDA approved color additive for the production of a sunless tanning response on skin. The goal of these studies is to determine if DHA reacts with skin proteins through a free radical mechanism similar to its well-defined reactivity in solution with amino acids and proteins. DHA was obtained from Aldrich (Milwaukee, WI). All experiments were run in 500 mM phosphate buffer, pH 7.8, due to the pH dependence of the reaction. The buffer was treated with Chelex ion-exchange resin (Bio-Rad Laboratories, Richmond, CA) before use to remove adventitious trace metal ions. Hairless SKH1 mice from Charles River Laboratories (Wilmington, MA) were used in these studies. All experiments described were approved by the Institutional Animal Care and Use Committee under Protocol 9713. The solutions were buffered in all cases for consistency with ex vivo experiments. Mice used for in vivo experiments were housed individually to prevent mutual grooming. For the in vivo experiments, a mouse was “painted” once daily on one side of its body with a 2 M solution of DHA in buffer. After the brown color developed on the skin (usually within 48 h), the mouse was sacrificed (mice were sacrificed by anesthetizing (ketamine/xylazine), followed by cervical dislocation). The brown patch of skin was removed for examination by ESR, and untreated skin (normal pink in color) from the opposite side of the same animal was removed and used as a control. For an additional control, several mice were painted with buffer alone in the same manner and sacrificed after 24 h. The color of the skin treated with buffer alone appeared to be the same as that of normal untreated skin. The skin was cut to fit the 7 mm × 23 mm well of a Wilmad (Buena, NJ) WG-806 A quartz tissue cell. ESR spectra were run at room temperature with a Varian E-104 ESR spectrometer, equipped with a TM110 cavity. The spectrometer was interfaced to a PC for data acquisition and analysis. Multiple scans were usually averaged to improve the signal-to-noise ratio. For ex vivo experiments, hairless mice were sacrificed and portions of skin were incubated in a solution of carbohydrate (usually DHA) in phosphate buffer for times up to 48 h, at 37°C. The DHA concentration was varied between 0.1 and 1.0 M. Control samples consisted of skin in buffer alone for the same time and temperature. ESR spectra were run on the skin as described above. The g value of the ex vivo line was measured relative to 2,2-diphenyl-1-picrylhydrazyl (g = 2.0036). The skin of mice painted with the DHA solution developed a brown color within 24 h. Upon examination of the brown skin by ESR spectroscopy a broad single line was observed, as shown in Figure 1(a). Although weak, the signal is reproducible and is in clear contrast to the control sample from the same animal Figure 1b. The most intense signals were obtained from mice painted using the procedure described above. In several cases, mice were painted at frequent intervals, every 2 h for 8 h. After such treatment the skin did not get darker than a single treatment, nor did it show increased signal beyond that obtained by a single treatment. There were two control experiments. The first consisted of ESR examination of normal skin from animals treated with DHA solution. In all cases the brown color was confined to the skin on the treated side of the animal, indicating that grooming did not spread the DHA, and there was no ambiguity in selecting normal skin for examination. The second control experiment consisted of skin from animals treated with phosphate buffer alone. With these animals there was no apparent visual difference in the skin from the treated and untreated sides of the same animal, and no samples of skin had any detectable ESR signal (data not shown). Taken together the experiments show that the ESR signal, and thus free radicals, occurred only in skin treated with DHA. Owing to the practical difficulty of reliably measuring DHA concentrations on mouse skin, a series of ex vivo experiments was undertaken in which the effect of time and of carbohydrate concentration could be more readily assessed. Qualitatively, after 24 h at 37°C in 0.5 M DHA solution the skin is tan to brown in color and almost leathery in texture, whereas the skin in buffer alone is gray and falling apart under the same conditions. The ESR signal consists of a single line with a peak-to-peak linewidth of 14–16 Gauss Figure 1c, and g value of 2.0035 ± 0.0002. There is no detectable signal in the skin incubated without added carbohydrate Figure 1d. The intensity of the ESR spectrum is directly proportional to DHA concentration Figure 2, and the signal intensity peaks at approximately 20 h. Similar ESR signals were obtained when glyceraldehyde was used in place of DHA. In a further control experiment, samples of skin treated with DHA for 24 h were washed with deionized water and placed in fresh buffer without DHA. After another 24 h the intensity of the ESR signal was 10%-20% less than originally observed. As the ESR spectra required the presence of both skin and carbohydrate, we suggest that they are the result of the Maillard reaction between a carbohydrate, DHA, and amino groups in skin proteins, such as from lysine and arginine, leading to cation radical intermediates. The radicals are best described as melanoidin radicals, which are the end result of the Maillard reaction. The observed linewidths are similar to those reported for the radicals formed by the reaction of glycolaldehyde and bovine serum albumin (Hofmann et al., 1999Hofmann T. Bors W. Stettmaier K. Radical-assisted melanoidin formation during thermal processing of foods as well as under physiological conditions.J Agric Food Chem. 1999; 47: 391-396https://doi.org/10.1021/jf980627pCrossref PubMed Scopus (89) Google Scholar), and by methyl glyoxal bovine serum albumin (Lee et al., 1998Lee C. Yim M.B. Chock P.B. Yim H.S. Kang S.O. Oxidation-reduction properties of methylglyoxal-modified protein in relation to free radical generation.J Biol Chem. 1998; 273: 25272-25278Crossref PubMed Scopus (131) Google Scholar). In the former experiment the broad line was ascribed to radical cations that were not free to rotate because of protein cross-linking and in the latter case the ESR spectra were run on a frozen solution, which would also lead to broad lines. Although our previous in vitro experiments (Roberts and Lloyd, 1997Roberts R.L. Lloyd R.V. Free radical formation from secondary amines in the Maillard reaction.J Agric Food Chem. 1997; 45: 2413-2418https://doi.org/10.1021/jf960902cCrossref Scopus (36) Google Scholar) exhibited the well-resolved spectra often observed for radicals in solution, broad single lines are sometimes found in solution experiments as well (Namiki and Hayashi, 1983Namiki M. Hayashi T. A new mechanism of the Maillard reaction involving sugar fragmentation and free radical formation.ACS Symp Series. 1983; 215: 21-46Crossref Google Scholar;Wu et al., 1986Wu C.H. Russell G.F. Powrie W.D. Paramagnetic behavior of model system melanoidins.Dev Food Sci. 1986; 13: 135-144Google Scholar), especially in the presence of adventitious metal ions (Roberts and Lloyd, 1997Roberts R.L. Lloyd R.V. Free radical formation from secondary amines in the Maillard reaction.J Agric Food Chem. 1997; 45: 2413-2418https://doi.org/10.1021/jf960902cCrossref Scopus (36) Google Scholar). The radicals are highly stable, in contrast to the UV-induced skin radicals previously observed (Pathak and Stratton, 1968Pathak M.A. Stratton K. Free radicals in human skin before and after exposure to light.Arch Biochem Biophys. 1968; 123: 468-476Crossref PubMed Scopus (146) Google Scholar;Jurkiewicz and Buettner, 1996Jurkiewicz R.A. Buettner G.A. EPR detection of free radicals in UV-irradiated skin: mouse vs. human.Photochem Photobiol. 1996; 64: 918-922Crossref PubMed Scopus (112) Google Scholar). Another possible source of radicals are the melanins, which exhibit both intrinsic and photo-induced semiquinone type radicals (Stratton and Pathak, 1968Stratton K. Pathak M.A. Photoenhancement of the electron spin resonance signal from melanins.Arch Biochem Biophys. 1968; 123: 477-483Crossref PubMed Scopus (17) Google Scholar;Valavanidis et al., 1995Valavanidis A. Rallis M. Papaioannou G. Xenos K. Katsarou A. Studies in vivo by electron spin resonance of free radical mechanisms implicated in UV-induced skin photocarcinogenesis.Int J Cosm Sci. 1995; 17: 157-163Crossref PubMed Scopus (8) Google Scholar) The intrinsic melanin radicals are impossible in albino hairless mice, and exposure to normal room light did not lead to any detectable ESR signals. Also, the ESR parameters that we observed are not consistent with those reported for melanin radicals in skin and other biologic systems (Blois, 1966Blois M.S. On the spectroscopic properties of some natural melanins.J Invest Dermatol. 1966; 47: 162-166Crossref PubMed Scopus (10) Google Scholar;Jahan et al., 1987Jahan M.S. Drouin T.R. Sayre R.M. Effect of humidity on photoinduced ESR signal from human hair.Photochem Photobiol. 1987; 45: 543-546Crossref PubMed Scopus (11) Google Scholar;Collins et al., 1995Collins B. Poehler T.O. Bryden W.A. EPR persistence measurements of UV-induced melanin free radicals in whole skin.Photochem Photobiol. 1995; 62: 557-560Crossref PubMed Scopus (17) Google Scholar). In summary, the formation of free radicals in mouse skin as a result of the in vivo Maillard reaction has been confirmed by use of ESR spectroscopy. We are aware that mouse skin and human skin are not identical in structure, and that the mice used in this work are specifically bred for skin studies. The epidermis of the hairless mouse consists of basal cell layers and one or two cells undergoing maturation. These cells are covered by a thin stratum corneum consisting of only two or three layers of cornified cells. It is these cells that are painted with the DHA and therefore are the first target for the Maillard reaction. In contrast, human epidermis consists of a basal cell layer with perhaps eight or 10 layers of squamous cells undergoing maturation. Covering these living cells is a stratum corneum consisting of approximately 10 layers of cornified cells. Modest tanning occurring to the stratum corneum when treated with DHA could translate into a deep tan appearance in human skin. We believe that the chemistry of the DHA reaction would be the same in both species.