Human Immunodeficiency Virus-Associated Exosomes Promote Kaposi’s Sarcoma-Associated Herpesvirus Infection via the Epidermal Growth Factor Receptor

Kaposi’s sarcoma-associated herpesvirus (KSHV) is the causal agent for Kaposi’s sarcoma (KS), the most common malignancy in HIV/AIDS patients. Oral transmission through saliva is considered the most common route for spreading the virus among HIV/AIDS patients. However, the role of HIV-specific components in the cotransfection of KSHV is unclear. We demonstrate that exosomes purified from the saliva of HIV-positive patients and secreted by HIV-infected T-cell lines promote KSHV infectivity in immortalized and primary oral epithelial cells. HIV-associated exosomes promote KSHV infection, which depends on HIV trans-activation response element (TAR) RNA and EGFR of oral epithelial cells, which can be targeted for reducing KSHV infection. These results reveal that HIV-associated exosomes are a risk factor for KSHV infection in the HIV-infected population.

also presented numerous surface-exposed membrane proteins. We determined the size distribution of T-cell and saliva exosomes using a ZetaView nanoparticle tracking analyzer. The results of nanoparticle tracking analysis (NTA) indicated that over 90% of the nanoparticles were in the size range of 50 to 80 nm in diameter ( Fig. 1C; left, T-cell exosomes; right, saliva exosomes). The concentrations of saliva exosomes from 3 healthy donors ranged from 6.3 ϫ 10 10 to 2.3 ϫ 10 11 /ml, while those from 3 independently prepared culture supernatants of J1.1 cells ranged from 4 ϫ 10 9 to 8 ϫ 10 9 /ml, based on NTA. We have used the acetylcholinesterase (AChE) activities of T-cell extracellular vesicles to quantify exosomes (26). We found that the concentration of T-cell exosomes quantified using the AChE assay was the same as that determined by the NTA measurement. The total protein levels of saliva exosomes were correlated with the exosome concentrations determined with the ZetaView nanoparticle tracking analyzer. In this study, we treated cells with T-cell exosomes at a final concentration of 4 ϫ 10 9 exosomes/ml or saliva exosomes at 100 g/ml of exosomal proteins, which was equivalent to 4 ϫ 10 9 to 5 ϫ 10 9 /ml of exosomes, as quantified with the ZetaView nanoparticle tracking analyzer. T-cell EVs express proteins of the tetraspanin family (CD9, CD63, and CD81), which are markers of exosomes (26). To determine if saliva EVs also contained these markers, we performed an immunoblot assay with the total proteins of saliva exosomes from HIV-positive and HIV-negative (HIV Ϫ ) donors and found that saliva EVs contained the CD63, CD9, and CD81 proteins (Fig. 1D). These results indicate that our preparations of saliva EVs contain exosomes. As reported before, HIV RNAs were readily detected using reverse transcription-PCR (RT-PCR) in exosomes from the culture supernatants of HIV-1-infected J1.1 T cells (Fig. 1E) (26). Quantitative RT-PCR on total RNA extracted from exosomes secreted from HIV ϩ 8E5/ LAV, J1.1, and 2D10 T cells (26) demonstrated that J1.1 and 2D10 cell exosomes contained significantly higher levels of TAR RNA than Tat and Nef RNA (Fig. 1F). To determine whether HIV ϩ saliva exosomes contained HIV-specific cargo components, we performed RT-PCR on total exosomal RNA. We found that HIV ϩ saliva exosomes contained only HIV TAR, Tat, and Nef RNA and not Env RNA (Fig. 1G). Quantitative RT-PCR of the total RNA extracted from HIV ϩ saliva exosomes revealed that the expression levels of HIV TAR, Tat, and Nef RNA were similar to or higher than those in exosomes from cells of the HIV ϩ 8E5/LAV T-cell line, which is an HIV-infected CD4 ϩ human CEM T-cell line that contains a single copy of integrated HIV proviral genome (30) (Fig. 1H). Our findings indicate that the saliva of HIV ϩ patients, despite ART, contains HIV-associated exosomes.
HIV ؉ saliva exosomes promote KSHV infection in immortalized and primary human oral epithelial cells. We treated iSLK-BAC16 cells (31) with sodium butyrate and doxycycline to produce infectious KSHV virions. The recombinant KSHV contains a human elongation factor 1 alpha (EF1␣)-driven green fluorescent protein (GFP) expression cassette; thus, the KSHV-infected cells are GFP positive (GFP ϩ ) (31). KSHV can effectively infect cultured oral epithelial cells, as shown by expression of the KSHV latency-associated nuclear antigen (LANA) (32,33). To optimize the condition for KSHV infection, we infected cells of the immortalized oral epithelial cell line OKF6/ TERT2 (34) with the KSHV virion stock at 1:100, 1:50, and 1:20 dilutions for 20 h, followed by immunofluorescent staining of KSHV LANA (35) and GFP ( Fig. 2A). We estimated that about 25% of the oral epithelial cells were infected with the 1:100 dilution of the KSHV stock solution. We tested each batch of the KSHV preparations and used the dilution of KSHV stock to ensure that 15% to 25% of the cells were infected throughout all infection assays.
To determine whether HIV ϩ exosomes were able to affect KSHV infection in oral epithelial cells, we incubated OKF6/TERT2 cells with KSHV virions in the presence of exosomes isolated from the culture supernatants of latently HIV-1-infected (HIV ϩ ) J1.1 T cells or HIV-negative (HIV Ϫ ) Jurkat cells (26,36). HIV ϩ exosomes from J1.1 cells significantly enhanced KSHV infection compared to those from the control Jurkat T cells, as shown by immunofluorescence microscopy of LANA and GFP (Fig. 2B). The enhancement of KSHV infection by HIV ϩ exosomes was confirmed by flow cytometry on GFP of infected cells ( Fig. 2C and D). Our results suggest that HIV ϩ exosomes enhance KSHV infection in oral epithelial cells. KSHV infects the oral cavity and oropharynx, and the infection is more prevalent in HIV-positive persons than in the general population (9,37). We postulated that HIV-associated saliva exosomes might be responsible for the higher incidence of KSHV infection in persons living with HIV/AIDS. To test this hypothesis, we purified saliva exosomes from HIV-infected donors with ongoing ART and CD4 T-cell counts over 400 per ml and those from healthy individuals, using the differential ultracentrifugation protocol (26). Saliva exosomes from HIVpositive subjects (Fig. 3A, patient 8 [pt 8] and pt 9) significantly enhanced KSHV infection in OKF6/TERT2 cells compared to those from healthy individuals (Fig. 3A, N1 and N2), as measured by GFP flow cytometry. To determine whether KSHV was able to infect primary human oral epithelial cells (HOECs), we treated the cells with the KSHV virion stock solution. We found that KSHV was able to infect HOECs and that the infected cells expressed both KSHV LANA and GFP (Fig. 3B). To test whether HIV ϩ exosomes were able to enhance KSHV infection in primary oral epithelial cells, HOECs were treated with exosomes from HIV ϩ J1.1 and HIV Ϫ Jurkat cells as well as those purified from the saliva of HIV ϩ and HIV Ϫ donors. Exosomes isolated from HIV ϩ T cells or those from the saliva of HIV-positive donors significantly stimulated KSHV infection in HOECs compared to exosomes from control T cells or the saliva of healthy donors (Fig. 3C). To verify the stimulatory effect of HIV-associated saliva exosomes on KSHV infection, we collected unstimulated whole-saliva samples from HIV-positive donors (n ϭ 8) who were under ART with CD4 T-cell counts of over 400 per ml and a viral load of 50 to 6,000 or from healthy individuals for exosome purification. HOECs were treated with saliva exosomes for 10 min and then tested by the KSHV infection assay. KSHV infection was considerably enhanced by HIV ϩ saliva exosomes compared to HIV Ϫ saliva exosomes in primary HOECs (Fig. 3D, left, percentage of infected cells, and right, mean fluorescence intensity [MFI]). We reasoned that the elevated KSHV infectivity by HIV ϩ exosomes could result from increased viral entry or enhanced viral gene expression and replication, or both. We therefore examined the effect of HIV ϩ exosomes on viral entry by measuring KSHV virions with a monoclonal antibody against the small capsid protein (ORF65) (38)  HIV ؉ saliva exosomes promote KSHV infection and transmission in organotypic culture models of oral mucosa. The oral mucosa is the first target of KSHV infection, once the virus is in the oral cavity (5,6,8,9). To evaluate the role of HIV-associated exosomes in KSHV infection through the oral mucosa, we created a 3-dimensional (3-D) organotypic culture by using OKF6/TERT2 cells, as described previously (39)  To evaluate if KSHV virions were able to infect through the fully established oral epithelium, we stained the J1.1 cell exosome-treated, KSHV-infected 3-D oral mucosal culture for the protein claudin 1 (CLDN1) (40) and the KSHV LANA. Nuclear expression of LANA expression (Fig. 4E, arrowheads) was seen in mucosal epithelial cells expressing CLDN1 (Fig. 4E, arrows), a tight junction protein that is involved in the maintenance of epithelial integrity and the function of differentiated epithelia (40). Collectively, these findings suggest that HIV ϩ exosomes promote KSHV transmission through the oral mucosa.
HIV TAR RNA is the key element for enhancement of KSHV infection. We suspected that HIV-specific exosome cargo components were responsible for HIV ϩ exosome-enhanced KSHV infection in oral epithelial cells. Latently HIV-infected J1.1 T-cell exosomes do not contain HIV-1 particles, although these exosomes have viral proteins, such as Gag and the precursor form of Env protein (p160), as shown by proteomics (24). In addition, J1.1 cells can be reactivated by tumor necrosis factor alpha (TNF-␣) to express viral proteins and assemble infectious virions (36). To determine if HIV ϩ exosomes contained viral proteins, such as Tat and Nef, that are known to contribute to cellular function (41,42), we performed immunoblot assays on total exosome proteins isolated from latently HIV-infected J1.1 cells; cells of the HIV ϩ Jurkat-based C22G cell line, which contain a disruptive HIV tat mutant and nef deletion (43); and cells of the 2D10 cell line, which lack the viral nef gene (44). While the whole-protein lysates from TNF-␣-activated J1.1 cells (26) expressed the Tat and Nef proteins, exosomes from J1.1 and C22G cells did not contain these HIV proteins (Fig.  5A). Similarly, HIV ϩ saliva exosomes did not have the Tat and Nef proteins (Fig. 5B). These results suggest that neither the Tat nor the Nef protein plays a major role in promoting KSHV infection in response to HIV ϩ exosomes. We have reported that exosomes from both the J1.1 and C22G cell lines contain HIV trans-activation response element (TAR) RNA, which is required for the HIV ϩ exosome-enhanced proliferation of cancer cells and extracellular signal-regulated kinase (ERK) signaling (26). HIV ϩ saliva exosomes also contained TAR RNA (Fig. 1B). Therefore, we postulated that the TAR RNA-bearing exosomes contributed to enhancing KSHV infection in oral epithelial cells.
To test this hypothesis, we treated OKF6/TERT2 cells with exosomes derived from C22G,  (26). The protumor effect of TAR RNA requires the bulge-loop structure of the molecule (45). To determine whether the bulge-loop region of TAR RNA was involved in enhancing KSHV infection, we transfected OKF6/TERT2 cells with a synthetic TAR RNA, consisting of a TAR RNA mutant containing 5 nucleotide replacements in the bulge and loop sequences (26), followed by KSHV infection assays. Our results indicate that while the wild-type TAR RNA increased KSHV infection in oral epithelial cells, the mutant TAR RNA failed to affect KSHV infection (Fig. 5D, TAR versus mutTAR). In addition, the RNA aptamer R06, which is complementary to the TAR RNA apical region and which blocks the TAR function without disrupting the secondary structure of the TAR (46), inhibited TAR RNA-enhanced KSHV infection in OKF6/TERT2 cells (Fig. 5D, TARϩR06). However, a scrambled aptamer (26) did not significantly affect TAR RNA-induced KSHV infectivity (Fig. 5D, TARϩScrb). These results indicate that the bulge-loop region of the TAR RNA is critical for its function associated with the enhancement of KSHV infection in oral epithelial cells.
HIV-associated exosomes promote KSHV infectivity in an EGFR-dependent fashion. We have reported that exosomes secreted by HIV-infected T cells and purified from the plasma of HIV-positive persons stimulate the proliferation of head and neck squamous cell carcinoma and lung cancer cells in an EGFR-dependent manner through the phosphorylation of ERK1/2 (26). We reasoned that HIV ϩ exosomes might promote KSHV infection in oral epithelial cells via the same mechanism. Indeed, the treatment of OKF6/TERT2 cells with cetuximab, a monoclonal antibody against EGFR that blocks ligand binding to the receptor (47), inhibited KSHV infection in OKF6/TERT2 cells in response to HIV ϩ exosome treatment, as determined by GFP flow cytometry (Fig. 6A). The KSHV LANA is a protein essential for the persistence of the virus (48-50) and does not exist in purified KSHV virions propagated in cultured cells (51). Open reading frame (ORF) K8 is a viral early gene encoding the ZIP family of proteins, which play a regulatory role in the immediate early and delayed-early stages of the viral life cycle (52)(53)(54). To correlate HIV ϩ exosome-enhanced KSHV infection with expression of viral proteins in host cells, OKF6/TERT2 cells were infected with KSHV virions for 20 h in response to HIV ϩ J1.1 cell exosomes in the presence or absence of EGFR inhibitors. Total protein lysates were extracted from the cells and subjected to immunoblotting for KSHV LANA and ORF K8 proteins. J1.1 cell exosomes apparently stimulated the expression of LANA in OKF6/TERT2 cells compared with Jurkat cell exosomes. The expression of the LANA protein induced by J1.1 cell exosomes was blocked by cetuximab or AG1478, an EGFR tyrosine kinase inhibitor (55) (Fig. 6B). J1.1 cell exosomes also induced ORF K8 expression, while Jurkat T-cell exosomes did not affect its expression upon KSHV infection. However, J1.1 cell exosome-induced ORF K8 expression was only partially reduced by cetuximab or AG1478 in OKF6/TERT2 cells (Fig. 6B). Since K8 is an early gene involved in the regulation of cellular pathways for full-blown viral lytic replication (53,56,57), these results suggest that HIV ϩ exosomes may also enhance KSHV lytic replication. To determine if cetuximab also blocked HIV ϩ exosome-enhanced KSHV infection in primary oral epithelial cells, we treated HOECs with cetuximab, followed by KSHV infection in the presence or absence of HIV ϩ J1.1 cell exosomes. HIV ϩ exosome-induced KSHV infection in HOECs was inhibited by cetuximab (Fig. 6C). To further demonstrate the involvement of EGFR in mediating KSHV infection, we also infected HOECs with KSHV in the presence or absence of recombinant EGF protein.
Interestingly, EGF increased KSHV infection in HOECs, as shown by the presence of GFP ϩ cells (Fig. 6C) and de novo KSHV infection in OKF6/TER2 cells (Fig. 6D). Our results demonstrate the involvement of EGFR in mediating HIV ϩ exosome-enhanced KSHV infection in oral epithelial cells. To determine the effect of EGFR inhibition on KSHV infection in response to HIV ϩ saliva exosomes, we infected the oral mucosal tissue with KSHV in the presence or absence of cetuximab, followed by fluorescence microscopy for GFP and LANA. Cetuximab treatment blocked HIV ϩ saliva exosome-induced LANA expression in the oral mucosal tissue (Fig. 6E). Therefore, blocking EGFR can potentially inhibit KSHV infection mediated by HIV ϩ exosomes in the oral cavity.

DISCUSSION
Saliva-mediated oral transmission of KSHV is considered the most common route for viral spreading among men who have sex with men through deep kissing and by mother-to-child transmission (3)(4)(5)(6)(7)(8)(9). HIV infection appears to play an important role in KSHV coinfection, its transmission, and its progression to malignancies (64). In people living with HIV/AIDS, coinfection with KSHV is much more likely to lead to the development of KS and other KSHV-associated diseases (65)(66)(67). In addition, the incidence rates of KSHV detection were found to be more prevalent in the HIV-infected population than in the general population in a case-control study (37). However, the mechanistic connection between HIV infection and the increased incidence of KSHV infection is unclear. In this report, we demonstrate that saliva from persons living with HIV contains HIV-associated exosomes, as shown by the presence of HIV TAR, Tat, and Nef RNAs in these extracellular vesicles. Similarly, exosomes purified from the plasma of persons living with HIV and isolated from the culture supernatants of HIV ϩ T cells contained mainly HIV TAR RNA (24,26). Our results demonstrate that HIV ϩ exosomes enhance KSHV entry, as substantially higher numbers of virions were detected in cells treated with HIV ϩ exosomes during the early infection time. However, HIV ϩ exosomes also increased the expression of KSHV lytic protein K8. Therefore, HIV ϩ exosomes appear to not only enhance viral entry during early infection time but also increase KSHV lytic replication to promote infection.
In this report, HIV ϩ saliva samples were prepared from donors who were under ongoing ART and who had CD4 T-cell counts of Ͼ400 per ml as well as low-level viremia. HIV viral loads are correlated well in blood and semen but poorly in blood and saliva (68). It has been reported that HIV loads are generally lower in the saliva than in the blood (69). Our quantitative RT-PCR results showed that HIV-associated saliva exosomes contain HIV RNAs at levels similar to those in HIV-infected 8E5/LAV (30) T-cell exosomes. However, HIV ϩ saliva exosomes do not express HIV Env RNA, which encodes a protein forming the viral envelope (70). In addition, HIV ϩ saliva exosomes do not contain HIV Tat and Nef proteins, suggesting that our HIV ϩ saliva exosome samples did not contain HIV virions and that the proinfection effect of HIV ϩ exosomes was mainly due to TAR RNA in these exosomes. Since KSHV can be present in oral epithelial cells and is shed to cause de novo infection, independent of the patient's immune status (71), and since HIV ϩ exosomes enhance de novo KSHV infection in oral epithelial cells, our findings suggest that HIV-associated saliva exosomes may promote KSHV transmission by increasing both the KSHV infection rate and lytic replication in oral mucosal cells.
It has been reported that oral microbial metabolites contribute to infection and the lytic activation of KSHV (33,72,73). Supernatants of periodontopathic bacterial cultures induce KSHV replication in cells of the BCBL-1 cell line, a KSHV latently infected lymphoma-derived cell line; embryonic kidney epithelial cells; as well as human oral epithelial cells and umbilical vein endothelial cells (72,73). The saliva of patients with severe periodontal disease contains high levels of short-chain fatty acids that induce expression of KSHV lytic genes (73). These bacterial metabolic products can stimulate KSHV replication in infected cells using different mechanisms (72,73). However, it is not clear whether these microbial metabolic products are responsible for KSHV infection in the oral cavity of HIV-infected persons. Collectively, our findings and these previous reports denote that multiple microbial and viral risk factors contribute to KSHV pathogenesis in the oral cavity.
Exosomes from the plasma of people living with HIV and the culture supernatants of HIV-infected T-cell lines contain HIV TAR RNA at amounts in vast excess over those of all viral mRNAs (24,26). In patients with virtually undetectable virion levels, TAR RNA can still be found in blood exosomes (27). Our results show that HIV ϩ exosomes from saliva and T cells do not contain the HIV Tat and Nef proteins, as determined by immunoblotting. In addition, exosomes from the C22G HIV ϩ T-cell line, which contains a dysfunctional Tat mutant, which lacks the Nef gene, and which does not produce HIV virions, exhibit HIV TAR RNA and promote KSHV infection in oral epithelial cells.
Therefore, our results reveal that HIV proteins and/or Tat/Nef RNA is not involved in the proinfection effect of HIV ϩ exosomes. Several reports have shown that HIV TAR RNA is a functional component of the HIV ϩ exosome cargo and induces the expression of proinflammatory cytokines and proto-oncogenes in primary human macrophages and cancer cells, respectively (24,26,27). Synthetic TAR RNA alone can stimulate the proliferation and migration of head and neck cancer cells (26). The mutant TAR RNA with 5 nucleotide substitutions in the bulge and loop sequences fails to induce gene expression in cancer cells (26). Similarly, our results demonstrate that, while the wild-type synthetic TAR RNA stimulates KSHV infection, the TAR RNA mutant cannot enhance KSHV infection in oral epithelial cells. In addition, the R06 nucleotide aptamer, which creates an imperfect hairpin to complement the entire TAR loop and which blocks TAR RNA function (46), attenuates TAR RNA-induced KSHV infection. The R06 aptamer and its derivatives are able to reduce HIV-1 infection and inhibit viral transcription (74,75). Kolb et al. have reported that the replication of HIV-1 and that the activity of ␤-galactosidase under the control of the HIV-1 5= long terminal repeat were reduced in cells expressing the nuclear R06 transcript (74), suggesting antiviral activity for the nucleotide aptamer. Our results suggest that the R06 RNA aptamer and its functional derivatives can potentially be developed as treatments for controlling coinfection with herpesviruses in the HIV-infected population.
We have reported that HIV ϩ exosomes stimulate the proliferation and protooncogene expression of squamous cell carcinoma cells in an EGFR-dependent manner (26). Similarly, EGFR is critical for HIV ϩ exosome-enhanced KSHV infectivity; blocking the receptor with cetuximab or the receptor kinase inhibitor effectively inhibits KSHV infection. Concordant with the role of EGRF in mediating KSHV infection, EGF alone increases KSHV infectivity in oral epithelial cells, a process that can be inhibited by cetuximab. In head and neck as well as lung cancer cells, EGFR mediates HIV ϩ exosome entry into target cells and participates in exosome-induced cellular signaling, including ERK1/2 phosphorylation, without activating the receptor (26). However, B-cell lymphoma cells, which lack EGFR, do not respond to the protumor effect of HIV ϩ exosomes (26). In this report, we demonstrate that HIV ϩ T-cell exosomes specifically activate MAPK p38, but not ERK1/2, through EGFR without inducing canonical receptor phosphorylation. The ERK1/2 pathway is primarily driven by EGFR in vitro and in vivo in various types of cancers and is a potential target in cancer therapy (76,77). EGFR mediates the entry of HIV ϩ exosomes into cancer cells, possibly enriched in endosomes (26). HIV ϩ exosome-induced activation of the EGFR-ERK axis requires Toll-like receptor 3 (TLR3), an endosomal receptor that has been shown to bind to EGFR and the kinase Src residing in endosomes for TLR3 activation in response to double-stranded RNA (dsRNA) (78). Interestingly, EGF alone can promote KSHV infection in a fashion similar to that for HIV ϩ exosomes. In addition, EGF induces the phosphorylation of p38 and ERK1/2 MAPKs, suggesting that the EGFR signaling is critical for enhanced KSHV infection in oral epithelial cells. The p38 signaling may play a key role in KSHV virology. Inhibition of the catalytic activity of phosphorylated p38 blocks KSHV reactivation, possibly through a reduction in global H3 acetylation and phosphorylation (79). The MAP kinase pathways are well-known to control KSHV infection (63). In addition, several host and environmental factors have been considered to play roles in KSHV infection and reactivation. Various chromatin-silencing mechanisms, including histone deacetylation, repressive histone methylation, and DNA methylation, lead to silencing of the genomes of herpesviruses and HIV during latency. Inhibition of class 1/2 histone deacetylases (HDACs) by multiple short-chain fatty acids results in the expression of genes associated with the fate of KSHV infection and viral reactivation (81)(82)(83). The epigenetic modifications, particularly the acetylation of histones, are required for the maintenance of KSHV latency in classic and AIDS-associated KS tissues (82). Taken together, our findings provide insight into the mechanisms underlying HIV-specific components and coinfection with KSHV through the oral cavity in people living with HIV/AIDS. In addition, targeting HIV TAR RNA and EGFR of oral epithelial cells may serve as novel approaches to control KSHV infection in the HIV-infected population. staining on a FACSAria flow cytometer (BD Biosciences). Fluorescence-activated cell sorting data were analyzed with FlowJo software (TreeStar Inc.).
RT-PCR, immunoblotting, and immunofluorescence microscopy. Total RNA was isolated and purified from exosomes using a High Pure RNA isolation kit (Roche) according to the manufacturer's instructions. Extracted RNA was reverse transcribed to cDNA (High-Capacity cDNA reverse transcription; Applied Biosystems). PCR was performed with the Q5 Hot Start high-fidelity 2ϫ master mix (New England Biolabs) on the T100 thermal cycler (Bio-Rad) using the primers described previously (26).
For immunoblotting, total exosome proteins were purified using a Total Exosome RNA & Protein isolation kit (Thermo Fisher Scientific) following the manufacturer's instructions. To prepare total cellular proteins, cells were washed with PBS, and then the cellular lysates were obtained by adding 300 l of radioimmunoprecipitation assay lysis and extraction buffer (Thermo Fisher Scientific). Protein lysates were separated by SDS-PAGE and then transferred onto polyvinylidene fluoride (PVDF) membranes (Merck Millipore) for immunoblot analysis. Protein detection was performed by chemiluminescence using an enhance chemiluminescence kit (Thermo Fisher Scientific) with a ChemiDoc XRSϩ imaging system (Bio-Rad).
Immunofluorescence microscopy was performed as previously described (86), with minor modifications. Briefly, each section (5 m) of organotypic cultures or the oral mucosal tissues was deparaffinized 3 times in Clear-Rite 3 reagent (Thermo Fisher Scientific), hydrated with 100% alcohol followed by 95% alcohol, and washed with PBS. Samples were blocked with 10% donkey serum at room temperature for 1 h. Each section was incubated with the primary antibody at 4°C overnight. After washing in PBS, sections were stained with Alexa Fluor-conjugated secondary antibody appropriate to the species of the primary antibody. Sections were then mounted with Vectashield fluorescent mounting medium (Vector Laboratories Inc., Burlingame, CA) containing DAPI (4=,6diamidino-2-phenylindole) to visualize the nuclei. Immunofluorescent images were generated using an AMG Evos FL digital inverted fluorescence microscope (AMG, Mill Creek, WA). Confocal images were acquired with a Leica TCS SP8 system (Leica Microsystems) using a 63ϫ (numerical aperture, 1.4) objective at a pixel size of 90 nm. For fluorescence immunocytochemistry, cells on an 8-well culture slide were fixed with 100% methanol at Ϫ20°C for 20 min, followed by permeabilization with 0.3% Triton X-100 in PBS. The cells were then stained with the primary antibody, followed by incubation with appropriate secondary antibodies. Fluorescent images were taken as described above. Micrographic images were processed using ImageJ software.
Statistics. The results obtained with the treatments were compared with those obtained for the respective controls. Data are represented as the mean Ϯ standard deviation (SD). Flow cytometry data were subjected to one-way analysis of variance (ANOVA). Statistical significance was considered at P values of Ͻ0.05. For donors n Յ 5, the F test was applied. Data analyses were performed and graphs were generated using Prism (GraphPad Software, La Jolla, CA) and Excel 2013 (Microsoft Inc. Redmond, WA) software.