Corynebacterium tuberculostearicum, a human skin colonizer, induces the canonical nuclear factor‐κB inflammatory signaling pathway in human skin cells

Abstract Introduction Corynebacterium tuberculostearicum (C. t.) is a ubiquitous bacterium that colonizes human skin. In contrast to other members of the genus Corynebacterium, such as toxigenic Corynebacterium diphtheriae or the opportunistic pathogen Corynebacterium jeikeium, several studies suggest that C. t. may play a role in skin health and disease. However, the mechanisms underlying these effects remain poorly understood. Methods To investigate whether C. t. induces inflammatory pathways in primary human epidermal keratinocytes (HEKs) and human cutaneous squamous carcinoma cells (SCCs), cell culture, reverse transcription‐polymerase chain reaction (PCR), enzyme‐linked immunosorbent assay, immunofluorescence microscopy, Western blot, chromatin immunoprecipitation‐PCR, small interfering RNA knockdown and luciferase reporter expression system were used. Results Herein, we demonstrate that C. t. upregulates the messenger RNA (mRNA) and protein levels of inflammatory mediators in two human skin cell lines, HEKs and SCCs. We further show activation of the canonical nuclear factor‐κB (NF‐κB) pathway in response to C. t. infection, including phosphorylation of the inhibitor of κB (IκB), the nuclear translocation of NF‐κB subunit (NF‐κB‐P65) and the recruitment of NF‐κB‐P65 and RNA polymerase to the NF‐κB response elements at the promoter region of the inflammatory genes. Lastly, the data confirm that C. t.‐induced tumor necrosis factor mRNA expression in HEKs is toll‐like receptor 2 (TLR2) dependent. Conclusion Our results offer a mechanistic model for C. t.‐induced inflammation in human keratinocytes via TLR2 and activation of IκB kinase and downstream signaling through the canonical NF‐κB pathway. Relevance to chronic inflammatory diseases of the skin and cutaneous oncology is discussed.


| INTRODUCTION
Human skin and its constituent cells, including keratinocytes, provide the first line of defense at the interface with the environment, including against microbial pathogens. 1 In healthy individuals, a symbiotic or mutualistic relationship exists between the host and its microbial flora. Dysbiosis refers to a disequilibrium of the microbial community with resultant effects on skin health and disease. [2][3][4] Inflammatory skin diseases, such as atopic dermatitis, psoriasis, and rosacea, are postulated to be caused, at least in part, by an alteration of the normal but intricate equilibrium dictated by the environment, host genetics/immunity and the skin microbiota. [5][6][7][8][9] For example, colonization of the skin by Propionibacterium acnes is causally related to facial acne. 10,11 In turn, the inflammatory and metabolic status of the host regulate the functional impact of microbes. 12 Furthermore, host genetic factors are determinants of a skin microbiota predisposing to skin health or pathology; for example, the skin of patients with atopic dermatitis, a genetic disease related etiologically to mutations of the filaggrin (FLG) gene, demonstrates increased colonization with Staphylococcus aureus and Staphylococcus epidermidis (S. e.). 13,14 In healthy skin, commensals are unlikely to cause skin inflammation in the absence of a compromised epidermal barrier. 15 In the context of a compromised barrier, keratinocytes are effective initiators of inflammation through the production of cytokines, chemokines, and adhesion molecules. 16,17 Furthermore, an exaggerated or skewed inflammatory response is likely crucial for the development of inflammatory skin diseases, as demonstrated for atopic dermatitis, psoriasis, alopecia areata, and systemic lupus erythematosus. [18][19][20] Thus, understanding the mechanisms whereby skin microbiota impacts the keratinocytic regulation of inflammatory mediators and the induction of adhesion molecules would offer novel drug strategies for inflammatory skin diseases.
Bacteria of the genus Corynebacterium account for 30% of the total bacterial inhabitants of human skin. 3 Some Corynebacterium species are opportunistic pathogens and coexist among healthy skin flora, for example, Corynebacterium jeikeium (C. j.), 21 while other species, such as toxigenic C. diphtheriae, are classified as serious and potentially life-threatening pathogens. 22,23 The species C. tuberculostearicum (C. t.), characterized by C. Feurer et al in 2004 24 is a major component of the bacterial species that colonize a variety of skin environments, including dry, moist, and sebaceous regions. 25 Several studies have associated C. t. with disease states, including inflammatory breast disease, pancreatic panniculitis, chronic rhinosinusitis, and surgical site infections. [26][27][28][29] However, the mechanisms by which C. t. plays a causative role in skin inflammation, remains unclear.
Herein, we applied the techniques of reverse transcription-polymerase chain reaction (RT-PCR), enzyme-linked immunosorbent assay (ELISA), immunofluorescence microscopy, Western blot, chromatin immunoprecipitation-PCR (ChIP-PCR), small interfering RNA (siRNA) knockdown and luciferase reporter expression system to investigate whether C. t. induces inflammatory pathways in primary human epidermal keratinocytes (HEKs) and human cutaneous squamous carcinoma cells (SCCs). Our results provide conclusive evidence that C. t. activates the canonical nuclear factor-κB (NF-κB) pathway via tolllike receptor 2 (TLR 2 ), an effect regulated by the activation of IκB kinase (IKK). In parallel, we investigated the effects of C. j. and the commensal, but occasionally pathogenic, S. e. bacteria on the messenger RNA (mRNA) expression of the corresponding inflammatory genes in the same cell lines.

| Enzyme-linked immunosorbent assay
Following each treatment, culture media were collected, centrifuged at 1000g for 5 minutes and the released cytokines were quantified using R&D Systems kits for interleukin 6 (IL6) (catalog no. Dy206), IL8 (catalog no. DY208), CSF3 (catalog no. DY214), IL1β (catalog no. DY201), CXCL10 (catalog no. DY266), and ICAM1 (catalog no. DY720), following the manufacturer's instructions. ELISA plates were read on SPECTRAmax PLUS384 Microplate spectrophotometer set to 450 and 540 nm; for wavelength correction, readings at 540 nm were subtracted from the readings at 450 nm. The concentration of cytokines was extrapolated using the third-order polynomial (cubic) equation generated using the absorbance and concentration values of each cytokine's standard (provided with the kit). Paired t tests, performed on the GraphPad Prism 6 statistics software, were used to calculate the significance between cytokine concentrations of C. t.-and TNF-treated cells, relative to control cells.

| Immunofluorescent microscopy
Cells were grown as a monolayer in an eight-well chamber slide (catalog no. 177402; Lab-Tek NALGE NUNC INTERNATIONAL). After the indicated treatments, cells were fixed in ice-cold methanol (catalog no. A412; Fisher Chemicals) for 10 minutes at −20°C. Cells were then blocked for 1 hour at room temperature in 1% BSA (catalog no. a-4503; Sigma-Aldrich) dissolved in PBS containing 0.01% Tween 20 (catalog no. P5927; Sigma-Aldrich). Cells were subsequently incubated overnight at 4°C with antibodies against phosphorylated IκBα (mouse monoclonal antibody [catalog no. 9246; Cell Signaling]), NF-κB-P 65 (mouse monoclonal antibody [catalog no. SC-293072; Santa Cruz Biotechnology]) or TLR 2 (rabbit monoclonal antibody [catalog no. 12276; Cell Signaling]) in PBS-Tween-BSA at the manufacturerrecommended dilutions. After this incubation, cells were washed three times (5 minutes each) in PBS and incubated with Alexa Fluor 488 goat anti-mouse secondary antibody (catalog no. A11029; Invitrogen) diluted in PBS-Tween-BSA (1:500) for 1 hour at room temperature, followed by three washes (5 minutes each) in PBS. For nuclear counterstain, cells were incubated for 5 minutes at room temperature in PBS containing 4′,6-diamidino-2phenylindole (catalog no. d21490; Molecular Probes) at a concentration of 300 nM and washed three times (5 minutes each) in PBS. Immunoprobed cells were mounted using prolong gold antifade reagent (catalog no. p36930; Invitrogen) and visualized with confocal microscopy (Zeiss, Oberkochen, Germany) using ZEN 2012 software. Mean fluorescence intensity was calculated using the mean gray value analysis tool in the ImageJ software.

| Chromatin immunoprecipitation assay
HEKs or SCC cells were grown in complete medium in 100-mm cell culture plates to 100% confluency and treated with C. t. bacteria at a ratio of 5:1 (bacteria:cells) or TNF (20 ng/mL) for 2 hours. Protein-DNA crosslinking was performed by the addition of 16% methanol-free formaldehyde (catalog no. PI28906; Thermo Fisher Scientific) directly to the culture medium to a final concentration of 1%, and ChIP was performed as previously described. 32 Purified DNA was analyzed by PCR. The quantitative (q)PCR-ChIP primer sequences are listed in the Tables S2 and S3. Relative occupancy was calculated on a log 2 scale based on comparison with the geometric mean of cycle threshold values (C t ) for two negative control regions, as described. 33 Antibodies used for ChIP were NF-κB-P 65 mouse monoclonal antibody (catalog no. SC-293072; Santa Cruz Biotechnology) and RNA polymerase II (RNA Pol-II) antibody (catalog no. 920102; BioLegend). NF-κB putative binding loci were identified using the TFBIND online tool (http://tfbind. hgc.jp/) and UCSC genome browser database.

| Transfections and luciferase assays
HEKs and SCC cells were plated in 250 μL of antibioticfree complete growth medium in 48-well plates at a density of 4 × 10 4 cells/well and incubated overnight before plasmid transfection. A complex of 1 μL Lipofectamine 2000 (catalog no. 11668-027; Invitrogen) and 400 ng DNA of pTNF3′NF-κB plasmid 34 or PGL3 promoter empty vector (E.V., catalog no. E1761; Promega) diluted in 50 μL Opti-MEM (catalog no. 31985; Gibco) was added to each well and incubated for 48 hours. Transfected cells were treated for 6 hours with C. t. bacteria or TNF as indicated and luciferase activity was then assayed as previously described. 33 For siRNA transfection, HEKs cells were plated onto 12-well plates at a density of 1.5 × 10 5 for 24 hours and transfected with 25 nM of ON-TARGETplus human TLR2 (7097) siRNA-SMARTpool (catalog no. L-005020-01; Dharmacon) or FlexiTube Lamin A/C nontargeting siRNA, ctrl-siRNA (catalog no. SI03650332; Qiagen) using Lipofectamine RNAiMAX transfection reagents (catalog no. 13778075; Thermo Fisher Scientific) following the manufacturer's protocol.

| Statistical analysis
Data were obtained from N independent biological experiments and are depicted as plots of the mean of individual values with SD error bars, or box-and-whisker plots showing the median, the 25th, and the 75th quartiles, as well as the minimum to maximum values. Technical triplicates were performed for qPCR, ELISA, Western blot densitometric analysis, ChIP-qPCR, and luciferase data unless otherwise indicated. The GraphPad Prism 6 software was used for statistical analyses. The Student t test was used to determine the significant difference between cells treated with C. t., C. j., S. e., or TNF and nonstimulated (NS) cells. One-way analysis of variance was performed with Tukey's correction for multiple comparisons. Heatmaps were generated using the Cluster 3.0 software. 35 The values of log 2 -fold change in the mRNA expression levels, referred from RT-PCR data, were filtered by removing all genes with standard deviations for observed values of less than one, centered based on the values of the mean of mRNA expression in different treatments and clustered based on the similarity metric-uncentered correlation among the tested genes.
upregulates the mRNA and protein levels of inflammatory mediators in HEKs and SCC cells Keratinocytes contribute to the barrier functions of the epidermis through the formation of tight junctions and the stratum corneum and mediate inflammation through the secretion of cytokines, chemokines, and antibacterial peptides, and the expression of cellular adhesion molecules. 17,36-38 Accordingly, we evaluated the inflammatory response of primary HEKs and human SCC cells to in vitro infection by C. t. bacteria. Cells treated with TNF, a typical inflammatory cytokine, served as a positive control for the induction of inflammatory genes. The mRNA of 28 genes involved in the inflammatory response to bacterial infections were quantified using qRT-PCR. Data analysis shows significant upregulation by C. t. of a group of proinflammatory genes in HEKs and SCC cells ( Figure 1A,B), although the effects of C. t. on the expression of specific mRNAs varied between the two cell types. Infection with C. t. upregulated genes for IL1ra, IL17a, and IRF1 in HEKs cells only, while the corresponding treatment upregulated IL10 in SCC cells exclusively. The remaining upregulated mRNA species demonstrated increases in both HEKs and SCC cells ( Figure 1A,B). The mRNA expression profile of the tested genes was time-dependent following C. t. infection. In HEKs, genes for IL1ra, CSF3, IL1β, IL6, and IL17a were rapidly induced within 2 hours of C. t. infection, whereas the corresponding early responding genes in SCC cells were IL10, IL1β, IL1α, TNF, CXCL10, CXCL1, HBEGF, IL6, and CSF3. Enhanced transcription of the remaining upregulated genes in HEKs and SCC cells was detected 6 hours after infection ( Figure 1A,B). The release of select inflammatory proteins (IL6, IL8, CSF3, IL1β, CXCL10, and ICAM1) into the culture medium of HEKs and SCC cells following infection with C. t. or treatment with TNF for 8 hours was quantified using ELISA ( Figure 1C,D). The culture medium concentrations of the six cytokines for both cell types were significantly increased following the treatments with C. t. or TNF. Together with the mRNA data above, this provides confirmation that C. t. elicits an inflammatory response, although qualitatively and quantitatively different in HEKs and SCC cells.
3.2 | Differential effects of C. t., C. j., and S. e. on mRNA expression in HEKs and SCC cells C. j. is an opportunistic human pathogen reported to induce inflammatory cytokines, such as IL17a. 12,21,39 S. e. is a cutaneous commensal in humans, which can become an opportunistic pathogen, accounting for ∼22% of systemic infections in intensive care patients in the United States. 40 We infected HEKs and SCC cells with C. t., C. j., or S. e. for 6 hours and quantified the mRNA expression of the genes that showed significant induction by treatment with C. t. or TNF in the experiments of Figure 1. The three bacterial strains affected the mRNA expression of the tested genes differently in HEKs and SCC cells (Figure 2). Significant differences between infection of HEKs with C. t. or C. j. (P ≤ .05) were identified for CSF3, TNF, and IRF1 genes. Their corresponding mRNA expression levels showed increments greater than fivefold in response to C. t. infection, but significantly less or no expression following treatment with C. j. (Figure 2A). Infections of HEKs cells with either C. t. or C. j. increased the mRNA transcripts of IL1ra, IL8, IL1β, ICAM1, HBEGF, IL6, IL17a, CXCL1, CXCL10, LTB, and GMCSF, compared with NS controls (P ≥ .05; Figure 2A). Conversely, COX2 mRNA levels S. e.-infected HEKs showed significant (P ≤ .05) induction of mRNA levels for IL1ra, CSF3, IL8, IL1β, ICAM1, IL6, CXCL1, and IRF1 ( Figure 2A). However, the mRNA coding for IL1β and ICAM1 were significantly lower as compared to their corresponding mRNA expression levels in C. t.-infected cells (Figure 2). Notably, the expression of mRNA for TNF, HBEGF, CXCL10, LTB, and GMCSF was not affected by S. e. infection, relative to NS controls. Interestingly, Figure 2A shows that S. e. infection of HEKs increased the mRNA expression of the IL6 gene very significantly (18.78-fold; P ≤ .001), compared with C. t. infection (fourfold; P ≤ .01) or C. j. infection (4.56-fold; P ≤ .01).
In SCC cells, the mRNA expression of IL10, IL1β, IL1α, HBEGF, ICAM1, IL8, IL6, and CSF3 was significantly higher in C. t.-as compared with C. j.-infected cells ( Figure 2B), whereas the mRNA expression for CXCL10 and CXCL1 showed essentially identical responses to C. t. or C. j. infection. In S. e.-infected SCC cells, the mRNA levels of CXCL1 and IL8 genes were significantly lower than in C. t.-infected cells, while genes encoding TNF, ICAM1, IL32, LTB, and CSF3 were significantly higher than those in C. t.-infected cells (P ≤ .05). S. e. infection of SCC cells did not affect mRNA levels of IL10, IL1β, IL1α, HBEGF, and GMCSF, compared with NS cells. The mRNA expression of EGF, TNF, IL32, GMCSF, and LTB was not affected by the treatment of SCC cells with C. t. relative to control NS-treated cells. These data collectively highlight the unique effects of the three bacterial strains on the mRNA expression of proinflammatory genes in HEKs and SCC cells.

| C. t. infection induces IκB phosphorylation and NF-κB-P 65 nuclear translocation
Inactive NF-κB, the archetypal inflammatory transcription factor, resides in the cytoplasm bound to IκB protein.
Microbial infection or exposure to proinflammatory cytokines, such as TNF, triggers the canonical NF-κB pathway via the activation of IKK which phosphorylates IκB, thereby liberating the NF-κB-P 65 subunit that translocates to the nucleus and binds to DNA at specific loci, known as NF-κB-elements. This enhances the transcription of target genes, including those encoding numerous inflammatory effector molecules. 41,42 Results presented above show the activation by infection with C. t. of the transcription of a group of inflammatory genes. We studied the potential contribution of the canonical NF-κB pathway to the C. t.-induced transcription of proinflammatory genes, including the phosphorylation of IκB and nuclear translocation of NF-κB-P 65 subunit in HEKs and SCC cells.
We treated HEKs and SCC cells with C. t. or TNF for 20 minutes, based on prior kinetic studies, which showed that the phosphorylation of IκB takes place within 1 to 20 minutes of stimulus exposure. 43 Phosphorylation of IκB (P-IκB) was assessed by fluorescent microscopy and Western blot analysis, as described in Section 2. P-IκB increased dramatically after C. t. infection or TNF treatment of HEKs ( Figure 3A) and SCC cells ( Figure 3B). Densitometric quantification of Western blot bands corresponding to P-IκB demonstrated significant increases in HEKs cells infected with C. t. (7.31-fold; P ≤ .0063) or treated with TNF (8.24-fold; P ≤ .0054), relative to control cells ( Figure 3A). Correspondingly, P-IκB levels were significantly elevated in SCC cells infected with C. t. (5.24-fold; P ≤ .0047) or treated with TNF (5.13-fold; P ≤ .0212), as compared with control cells ( Figure 3B).
As the cellular localization of NF-κB-P 65 between cytoplasmic and nuclear compartments fluctuates over short 1-hour cycles, 44,45 we assessed NF-κB-P 65 translocation 1-hour after C. t. infection or TNF treatment. Fluorescence microscopic examination of control cells revealed the homogenous distribution of NF-κB-P 65 in the cytoplasm and its absence in nuclei. In C. t.-infected or TNF-treated cells, NF-κB-P 65 was primarily localized to the nucleus ( Figure 4A,B). NF-κB-P 65 translocation was further assessed using subcellular fractionation of cytoplasmic and nuclear proteins and Western blot analysis ( Figure 4A,B). Antibodies against CREB and GAPDH were used as nuclear and cytoplasmic markers, respectively. In control cells, the nuclear content of NF-κB-P 65 was undetectable by Western blot, whereas it was detected in the cytoplasmic and the nuclear compartments of C. t.-or TNF-treated HEKs and SCC cells, confirming the translocation of NF-κB-P 65 as a result of infecting HEKs and SCC cells with C. t.

C. t.-induced expression of inflammatory genes in HEKs and SCC cells
The canonical NF-κB pathway involves IκB phosphorylation and NF-κB translocation as downstream molecular events following IKK complex activation. 46 To investigate the role of IKK in C. t.-elicited transcription of inflammatory mediators, we treated HEKs and SCC cells with the IKK inhibitor PS-1145 (10 μM) or DMSO vehicle control for 1 hour before incubating the cells with C. t. or TNF for an additional 4 hours. Cells were then harvested, RNA was isolated and six representative proinflammatory and NF-κBdependent genes were quantitated by qRT-PCR, including IL6, IL8, CSF3, IL1β, CXCL10, and ICAM1. These genes are known to be dysregulated in various human skin diseases, 16,46-50 and were shown above to be upregulated in response to C. t. or TNF in both cell types. As expected, C. t. as well as TNF significantly upregulated the mRNA expression of the above-listed genes ( Figure 5A). In contrast, the corresponding mRNA expression of these genes was dramatically decreased in PS-1145-treated cells, providing direct evidence that IKK is involved in the C. t.-induced mRNA expression of these genes. The protein content of these six inflammatory mediators in the culture medium following 8 hours of culture with C. t. or TNF was quantified using ELISA ( Figure 5B). Consistent with the mRNA data, treatment with C. t. or TNF significantly elevated the levels of the respective proteins in the culture medium. PS-1145 significantly attenuated the inductive effect of C. t. and TNF on each of the released proteins. These data confirm that C. t. regulates IκB phosphorylation and NF-κB-P 65 translocation to the nucleus and supports the claim that IKK is directly involved in the modulation of select proinflammatory genes by C. t. in both HEKs and SCC cells. Poll-II to NF-κB response elements The above results addressing mRNA expression, cytokines/chemokine release, NFκB-P 65 nuclear translocation, and IκB phosphorylation collectively imply activation of the canonical NF-κB pathway following infection with C. t. bacteria. Downstream signaling further involves the binding of NFκB-P 65 subunit and RNA Pol to DNA at the transcriptionally active loci. [51][52][53] To confirm this linear process, ChIP-PCR was applied to show the recruitment of NF-κB-P 65 and RNA Pol-II to the NF-κB putative binding sequence at the promoter regions of IL1β, IL6, and CSF3 genes after 2 hours of exposure to C. t. or TNF ( Figure 6). The occupancy of NF-κB-P 65 and Pol-II was calculated relative to the nonoccupied control regions as indicated in the methods section. Occupancy data are expressed on a log 2 scale in Figure 6B,C. Data analyses indicate that in both cell types, infection with C. t. or treatment with TNF significantly increased the binding of NF-κB-P 65 and Pol-II to the promoter regions of the tested genes. Correlating the increased promoter region occupancy with the above To study the direct inducing effects of C. t. on the transcriptional activities of the proinflammatory mediators studied above, we utilized our previously cloned TNF reporter (pTNF3′NF-κB), 34 constructed by cloning the NF-κB binding elements from the 3′-untranslated region at the TNF gene to the region upstream of the luciferase gene in the pGL3 promoter vector. HEKs and SCC cells were transfected with the plasmid and treated with C. t. or TNF for 6 hours and the luciferase activities were measured as described in Section 2 ( Figure 7A,B).
C. t. infection increased the basal luciferase activity 2.85fold (P ≤ .032) and 2.55-fold (P ≤ .001) in HEKs and SCC cells, respectively, indicating that C. t. is directly involved in the activation of an NF-κB-dependent transcription system. The effects of TNF were more pronounced and increased luciferase activity 3.68-fold (P ≤ .0003) and 4.18-fold (P ≤ .008) in HEKs and SCC cells, respectively. These data are consistent with our previous findings demonstrating the TNF-induction of pTNF3′NF-κB in bronchial epithelial cells. 34  To identify the cellular receptor involved in the activation of the NF-κB pathway, we measured the mRNA expression of all known human TLR (1)(2)(3)(4)(5)(6)(7)(8)(9)(10) in HEKs in response to C. t. or S. e. infection. After 6 hours of treatment of HEKs with C. t., TLR 2 mRNA was increased by 2.28 PCR cycles or 4.86-fold (P ≤ .001), whereas mRNA for other TLR genes were not significantly increased ( Figure 8A). Furthermore, mRNA of all TLR genes including TLR 2 remained unaffected by exposure to S. e. (Figure 8A) Figure 8B, left panel). TNF mRNA expression was measured in HEKs cells treated for 2 hours with C. t. or FSL1 (Pam2CGDPKHPKSF), a synthetic lipopeptide and a specific TLR 2/6 agonist. As expected, C. t. increased the mRNA expression of TNF 2.81-fold (P ≤ .001) and 2.59-fold (P ≤ .001) in naïve and control siRNA-treated cells, respectively, an effect that was abolished in TLR 2 knockdown cells. Treatment with FSL1 increased levels of TNF mRNA 11.03-fold (P ≤ .001) and 9.91-fold (P ≤ .001) in naïve and control siRNAtreated cells, respectively, and TLR 2 knockdown significantly reduced the levels of TNF mRNA in FSL1-treated cells to 6.65-fold (P ≤ .001) ( Figure 8B, right panel). These results suggest that TLR 2 plays a role in C. t.-induced inflammation in human keratinocytes. Moreover, the upregulation of TLR 2 mRNA and protein in C. t.-but not in S. e.-infected cells, is postulated to explain, at least partly, the relatively greater induction of NF-κB inflammatory signaling pathway by C. t. as compared with S. e. (Figure 2A).

| DISCUSSION
Through shotgun metagenomics, C. t. has been identified as an important colonizer of many skin or treatment with TNF. After 6 hours, luciferase activities were measured and collected data, N = 5, was analyzed with the GraphPad Prism 6 software. A two-way analysis of variance was performed with Tukey's correction for multiple comparisons. The Student t test was used to determine the significant difference between two treatment groups, *P < .05; **P < .01; ***P < .001. Data are expressed as relative luciferase activity and depicted as box-and-whiskers plots showing the median, the 25th, and 75th quartiles, as well as a minimum to maximum values. DMSO, dimethyl sulfoxide; HEK, human epidermal keratinocyte; NF-κB, nuclear factor-κB; SCC, squamous carcinoma cell; TNF, tumor necrosis factor microenvironments, whether dry, moist or sebaceous. 25 C. t. disease associations have included inflammatory breast disease, pancreatic panniculitis, chronic rhinosinusitis, and surgical site infections. [26][27][28][29] Using 16S ribosomal RNA sequencing, our laboratory has also shown that C. t. is more commonly identified in keratinocyte carcinomas (ie, basal and squamous cell carcinomas) as compared with matched, perilesional skin controls (data presented at the Canadian Dermatology Association's 93rd The Student t test was used to determine the significant difference between the two treatment groups. DAPI, 4′,6-diamidino-2-phenylindole; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HEK, human epidermal keratinocyte; MFI, mean fluorescence intensity; mRNA, messenger RNA; RT-PCR, reverse transcription-polymerase chain reaction; siRNA, small interfering RNA; TLR 2 , toll-like receptor 2; TNF, tumor necrosis factor. ***P < .001