PROPERTIES OF ANTHOCYANIN-PIGMENTED LIPSTICK FORMULATIONS 67 DPPH FREE RADICAL SCAVENGING ASSAY The DPPH free radical scavenging assay was used as a measure of potential antioxidant ca- pacity as described previously (28,29), with modifi cations described by Montanari et al. (11) for use with cosmetics. Methanolic solutions containing the pigments recovered from the lipsticks, the dried pigment extracts at concentrations equal to their contents in the lipstick formulations, and also BHT were used for the testing of antioxidant capacity. DPPH dissolved in methanol with serial dilutions was used to obtain a standard curve for DPPH based on concentration. To the DPPH solution, methanol solutions containing ei- ther BHT, the lipstick base, the dried extract, or the ACN formulas were added in tripli- cate. Samples were equilibrated for 30 min in the dark at 20°C, and absorbance at 515 nm was measured by using a SpectraMax190 plate reader (Molecular Devices) with 30-min equilibration time. Then the samples were blanked against wells containing only methanol. The average of the absorbance readings was then used to determine the DPPH inhibitory percentage and 50% inhibitory concentration (IC50) of each sample. The inhibitory per- centage of each sample was determined based on the following equation:  ¬ ­q100 ž ­ ž ® DPPH Sample DPPH Abs Abs % Abs I (3) IC50 values are defi ned as the concentration necessary to inhibit 50% of the free radical (11). The IC50 values of each sample were determined using linear regression of the ab- sorbances at different concentrations. The readings for the ACN formulas were corrected for the absorbance of their respective dried extract in the methanol solution. Samples were tested again after 4 weeks of storage at 4°C, and new IC50 values were calculated and compared with the original values. ANTITYROSINASE ASSAY Tyrosinase inhibition assay was performed as previously reported (20). Methanolic lip- stick extracts and mushroom tyrosinase (5 U) were gently mixed in phosphate buffer (pH 6.5, 50 mM) and incubated for 10 min in a 96-well plate. Absorbance was measured at 475 nm on a SpectraMax190 plate reader (Molecular Devices). Results were compared with the negative control (phosphate buffer) and positive control (kojic acid). The percentage tyrosinase inhibition was calculated as follows:  ¬ ­q100 ž ­ ž ® Control Sample Control Abs Abs % Abs I (4) Results are presented as mean (n = 8) ± SD. IC50 values were predicted using linear regression and are expressed in μg/mL. STATISTICAL ANALYSIS Results of the Folin–Ciocalteu method, DPPH free radical scavenging assay, and antity- rosinase activity were analyzed using two-way ANOVA and regression modeling (α = 0.05)

JOURNAL OF COSMETIC SCIENCE 68 using Minitab statistical software version 16 (State College, PA) and GraphPad Prism version 6 (La Jolla, CA). RESULTS AND DISCUSSION CHARACTERIZATION OF ACNS IN LIPSTICK FORMULATIONS Sources of ACNs used in lipstick formulations were selected to represent a variety of natural structures and commercial availability. The main characteristics of each extract and their concentrations used in these lipstick formulations were characterized by Westfall and Giusti (21) and are summarized in Figure 1. Several sources of cyanidin, such as elderberry (Sambucus nigra L.), purple carrot (Daucus carota L.), purple corn (Zea mays L.), and purple sweet potato (Ipomoea batatas L.), were selected because of the high predomi- nance in nature and commercial availability. Sources of acylated ACNs were also included as they have high reported stability (30). Hibiscus (Hibiscus sabdariffa L.) is a source of nonacylated delphinidin and was evaluated because of its reported high antioxidant activ- ity (31). Red radish (Raphanus sativus L.), a source of acylated pelargonidin, was included as it is reported to be highly stable and as an alternative to synthetic red colorants (32). Red grape (Vitis vinifera L.), which contains predominantly malvidin and also all six ma- jor aglycones, was also investigated to better understand the effect of chemical structure on color stability. As the lipstick formulations were developed based on inclusion of 8% dried plant extract, ACN concentrations in the lipstick formulations ranged 2.2–16.5 μg/mg of lipstick (Figure 1). TOTAL PHENOLIC CONTENT The total phenolic content in the methanolic extracts was expressed as micrograms of GAEs per milligram lipstick (Figure 2). The total phenolic content ranged from 85.8 ± 3.3 μg GAE/mg lipstick for purple sweet potato to 46.1 ± 5.5 μg GAE/mg lipstick for purple carrot. The other lipstick formulations contained 51.6 ± 8.3 (red grape), 56.6 ± 8.1 (red radish), 66.7 ± 3.5 (elderberry), and 76.6 ± 1.8 (purple corn) μg GAE/mg lip- stick. The phenolic content was compared with the total ACN content determined in a previous study (21). A high correlation (r = 0.97) was found between the total phenolic and total ACN contents for formulas containing acylated ACNs: purple carrot, purple corn, purple sweet potato, and red radish. These results are expected as the assay also measures the cinnamic and malonic acid acylating groups within the formulas (24). As phenolics are often associated with antioxidant activity, it is an important variable to consider when interpreting the antioxidant activity results of the formulas (33). UV ABSORPTION AND SPF CALCULATIONS The UV light absorbance of each formula compared with that of the lipstick base alone, from 290 to 400 nm, is shown in Figure 3. All formulas showed greater absorbance than the lipstick base alone along the entire UV spectrum, although absorption was higher in