ENDOGENOUS DNA OXIDATION CORRELATES TO INCREASED IRON LEVELS IN MELANOCYTES 279 NY). Human hepatocarcinoma HepG2 cells and immortalized human mammary epithe- lial MCF-10A cells were also obtained (ATCC, Manassas, VA) and cultured. ULTRAVIOLET RADIATION A bank of four Sylvania Sunlamp fs 40 bulbs, which generate UVB radiation (290–320 nm), was used to irradiate cells. Before irradiation, the culture medium was removed and replaced with 6 ml of Dulbecco’s Phosphate Buffered Saline (D-PBS). Fluences were mea- sured with an International Light IL1400A radiometer with a UVB probe attachment. ISOLATION OF CELLULAR DNA Precautions were taken to minimize artefactual oxidation (14), and DNA was isolated as previously described (15). Briefl y after cell lysis, samples were treated with Ribonuclease A and Proteinase K followed by precipitation with sodium acetate, pH 4.5. Chloroform/ isoamyl alcohol (24:1) was then used to remove proteins and lipids and the aqueous phase precipitated in isopropyl alcohol and washed in 70% ethanol. The samples were then resuspended in 10 mM Tris-HCl, pH 7.4 and quantitated. ISOLATION AND ANALYSIS OF 8-OXO-dG Purifi ed DNA was sequentially digested to the nucleoside level by DNAse 1, Nuclease P1, and alkaline phosphatase as described previously (15,16). Further, after additional purifi cation by centrifugal fi ltration, the samples were subjected to high-performance liquid chromatography (HPLC) analysis by injection onto a reverse phase column using a 100-mM lithium acetate, pH 5.2, 0.6 ml per min mobile phase, and interfaced to a UV detector for nucleoside analysis in series with an electrochemical detector for sensitive 8-oxo-dG analysis. GLUTATHIONE ANALYSIS GSH was measured by means of the GSH/GSSG-Glo Assay (Promega, Madison, WI), which is a luminescence-based assay that allows detection directly on lysed cells with a luciferin derivative sensitive to glutathione S-transferase. Concentrations of GSH were determined against a standard curve, normalized to cell number, and expressed as nmoles/ cell number. FERRITIN ANALYSIS To unify the background levels of ferritin, cells were initially grown in their respective media but subsequently changed to DMEM containing 1% FBS for at least 24 h before ferritin analysis. Ferritin was analyzed with an ELISA assay that utilized primary antibodies to human ferritin and secondary antibodies conjugated to peroxidase as described previously (17).

JOURNAL OF COSMETIC SCIENCE 280 RESULTS OXIDATIVE DNA BASE DAMAGE Primary keratinocyte and melanocyte cell cultures at approximately 50% confl uency were harvested and subjected to DNA isolation and digestion to the nucleoside level as described in the previous section. HPLC/EC analysis revealed that the endogenous level of 8-oxo-dG in NHEKs was 1.49 (±0.11 SEM)/106 dG, whereas melanoyctes were ob- served to have a higher level at 4.49 (±0.55 SEM) 8-oxo-dG/106 dG. These data represent a statistically signifi cant difference (p 0.005, two-tailed t-test) between these cell types and indicate higher oxidative DNA levels in melanocytes (Figure 1A). Further, in order to determine the effect of UVB-induced oxidative stress, melanocytes were then exposed to UVB. Interestingly, when melanocytes were exposed to 125 mJ/cm2 UVB, 8-oxo-dG levels increased nearly twofold to 8.09 (±1.22 SEM). In comparison, NHEKs increased similarly to 3.82 (±0.3 SEM) when exposed under the same conditions and indicates that Figure 1. DNA extracted from NHEKs and melanocytes were digested to the nucleoside level and analyzed by HPLC/EC for 8-oxo-dG. (A) Increased levels of 8-oxo-dG were observed in melanocytes compared to NHEKs. Data are expressed as mean ± SEM, *p 0.005. (B) Cells were exposed to 125 mJ/cm2 UVB and then analyzed. Similar UV-induced increases in 8-oxo-dG were measured for both NHEKs and melanocytes. Data are expressed as mean ± SEM.