JOURNAL OF COSMETIC SCIENCE 246 MATERIALS AND METHODS SAMPLING Twenty lipstick samples used by women between the ages of 20 and 50 were tested for microbial contamination. Before opening the sample cap, the lipstick surface was cleaned with 70% ethanol. The surface of the lipstick was then swabbed using a sterile cotton swab. Samples of the swab surface were suspended in 1 ml of sterile distilled water. GENOMIC DNA EXTRACTION For extraction of bacterial DNA from each of the 20 lipstick samples, G-spin Genomic DNA Extraction Kit for bacteria (Intron Biotechnology Inc., Seongnam-si, Korea) was used, and DNA was extracted according to the manufacturer’s instructions. Pyrosequencing analysis was performed once with the target sample consisting of the 20 DNA samples mixed in equal ratios. PYROSEQUENCING Polymerase chain reaction (PCR) amplifi cation was performed using primers targeting the V3 to V4 regions of the 16S rRNA gene found in the extracted DNA. For bacterial amplifi cation, primers 341F (5′-TCGTCGGCAGCGTC-AGATGTGTATAAGAGACAG- CCTACGGGNGGCWGCAG-3′ underlining sequence indicates the complimentary region of the primer) and 805R (5′-GTCTCGTGGGCTCGG-AGATGTGTATAAGAGACAG- GACTACHVGGGTATCTAATCC-3′) were used. The amplifi cations were carried out under the following conditions: initial denaturation at 95°C for 3 min, followed by 25 cycles of denaturation at 95°C for 30 s, primer annealing at 55°C for 30 s, and extension at 72°C for 30 s, with a fi nal elongation at 72°C for 5 min. Secondary amplifi cation for attaching the Illumina NexTera barcode was performed with i5 forward primer (5′-AAT- GATACGGCGACCACCGAGATCTACAC-XXXXXXXX-TCGTCGGCAGCGTC-3′ X indicates the barcode region) and i7 reverse primer (5′-CAAGCAGAAGACGGCATAC- GAGAT-XXXXXXXX-AGTCTCGTGGGCTCGG-3′). Conditions for secondary amplifi cation were similar to the previous one except with eight cycles of amplifi cation. The correct PCR product was confi rmed using electrophoresis on a 2% agarose gel followed by visualization under a Gel Doc system (BioRad, Hercules, CA). The amplifi ed products were purifi ed using the QIAquick PCR purifi cation kit (Qiagen, Valencia, CA). Equal concentrations of purifi ed products were pooled together and short fragments corresponding to nontarget products were removed with the Ampure bead kit (Agencourt Bioscience, Beverly, MA). Sample quality and product size were assessed using a Bioanalyzer 2100 (Agilent, Palo Alto, CA) and a DNA 7500 chip (Agilent). Mixed amplicons were pooled and the sequencing was carried out at Chunlab, Inc. (Seoul, Korea) using an Illumina MiSeq Sequencing system (Illumina, San Diego, CA) according to the manufacturer’s instructions. SEQUENCING DATA ANALYSIS Obtained reads were sorted using the unique barcodes of each PCR product. The barcode, linker, and primer sequences were removed from the original reads. Any reads containing

BACTERIAL CONTAMINATION OF LIPSTICK 247 two or more ambiguous nucleotides, a low-quality score (average score 25), or reads shorter than 300 bp were discarded. Potential chimerical sequences were detected using the Bellerophon method consisting of comparing the BLASTN search results between the forward and reverse half sequences (7). After removing chimerical sequences, the taxonomic classifi cation of each read was assigned against the EzTaxon-e database (http://eztaxon-e. ezbiocloud.net) (8). Briefl y, this database contains the 16S rRNA gene sequences of type strains along with valid published names and representative species phylotypes of cul- tured and uncultured entries in the GenBank database. Complete hierarchical taxonomic classifi cations from phylum to species are also included. PATHOGENIC BACTERIA ANALYSIS Cases within the last 10 year reporting the pathogenicity of each identifi ed bacterial genera identifi ed were searched for using PubMed. Pyrosequencing results were matched to the genus level, with the limitation that the analysis could not be completed up to the species level. Therefore, a genus was considered pathogenic when any one of the species belonging to it had a reported case of pathogenicity. RESULTS A total of 19,863 sequence reads were obtained and 105 genera of bacteria were identifi ed (Table I). The bacteria identifi ed included those found not only on the skin, but also in saliva and water. Leifsonia (65.86%), Methylobacterium (14.95%), Streptococcus (7.51%), and Haemophilus (3.58%) were predominant among all identifi ed genera (Figure 1). Patho- genic bacteria such as Staphylococcus, Pseudomonas, Escherichia, Salmonella, Corynebacterium, Mycobacterium, and Neisseria were also found. These potentially pathogenic bacteria represented 27.6% of the 105 genera, whereas the four most dominant genera comprised 92% of the 19,863 total reads. Actinomyces, Bacteroides, Porphyromonas, Prevotella, Capnocy- tophaga, Lactobacillus, Streptococcus, Veillonella, and Fusobacterium were among the oral bac- teria identifi ed. The most commonly identifi ed oral bacteria belonged to the Streptococcus genus. DISCUSSION The most commonly identifi ed bacteria overall belonged to the Leifsonia, a genus of aquatic bacteria commonly found in water. Although the particular species was not identi- fi ed, Leifsonia aquatica, a bacterium belonging to the Leifsonia genus, causes catheter- related disease and, in rare cases, acute sepsis in immunocompromised patients (9,10). The next most commonly identifi ed genus, Methylobacterium, comprises opportunistic pathogenic bacteria that cause infections in immunocompromised individuals (11). Streptococcus, the third most commonly identifi ed genus, is composed of gram-positive bacteria found in large numbers in the oral cavity and saliva, attaching to the oral mucosa and the surfaces of teeth (12). The fourth most common, Haemophilus, includes life- threatening microorganisms that cause respiratory infection and are known to have wide pathogenicity (13).