Title: Tonsil Epithelial Factors May Influence Oropharyngeal Human Immunodeficiency Virus Transmission
Abstract: Tonsil epithelium has been implicated in human immunodeficiency virus (HIV) pathogenesis, but its role in oral transmission remains controversial. To study characteristics of this tissue, which may influence susceptibility or resistance to HIV, we performed microarray analysis of the tonsil epithelium. Our data revealed that genes related to immune functions such as antibody production and antigen processing were increasingly expressed in tonsil compared with the epithelium of another oropharyngeal site, the gingival epithelium. Importantly, tonsil epithelium highly expressed genes associated with HIV entrapment and/or transmission, including the HIV co-receptor CXCR4 and the potential HIV-binding molecules FcRγIII, complement receptor 2, and various complement components. Immunohistochemical staining confirmed the increased presence of CXCR4 in the tonsil epithelium compared with multiple oral epithelial sites, particularly in basal and parabasal layers. This increased expression of molecules involved in viral recognition, binding, and entry may favor virus-epithelium interactions in an environment with reduced innate antiviral mechanisms. Specifically, secretory leukocyte protease inhibitor, an innate molecule with anti-HIV activity, was minimal in the tonsil epithelium, in contrast to oral mucosa. Collectively, our data suggest that increased expression of molecules associated with HIV binding and entry coupled with decreased innate antiviral factors may render the tonsil a potential site for oral transmission. Tonsil epithelium has been implicated in human immunodeficiency virus (HIV) pathogenesis, but its role in oral transmission remains controversial. To study characteristics of this tissue, which may influence susceptibility or resistance to HIV, we performed microarray analysis of the tonsil epithelium. Our data revealed that genes related to immune functions such as antibody production and antigen processing were increasingly expressed in tonsil compared with the epithelium of another oropharyngeal site, the gingival epithelium. Importantly, tonsil epithelium highly expressed genes associated with HIV entrapment and/or transmission, including the HIV co-receptor CXCR4 and the potential HIV-binding molecules FcRγIII, complement receptor 2, and various complement components. Immunohistochemical staining confirmed the increased presence of CXCR4 in the tonsil epithelium compared with multiple oral epithelial sites, particularly in basal and parabasal layers. This increased expression of molecules involved in viral recognition, binding, and entry may favor virus-epithelium interactions in an environment with reduced innate antiviral mechanisms. Specifically, secretory leukocyte protease inhibitor, an innate molecule with anti-HIV activity, was minimal in the tonsil epithelium, in contrast to oral mucosa. Collectively, our data suggest that increased expression of molecules associated with HIV binding and entry coupled with decreased innate antiviral factors may render the tonsil a potential site for oral transmission. The predominant mode of transmission for human immunodeficiency virus (HIV) is through the mucosal route,1Milman G Sharma O Mechanisms of HIV/SIV mucosal transmission.AIDS Res Hum Retroviruses. 1994; 10: 1305-1312Crossref PubMed Scopus (86) Google Scholar, 2Orenstein JM Fox C Wahl SM Macrophages as a source of HIV during opportunistic infections.Science. 1997; 276: 1857-1861Crossref PubMed Scopus (265) Google Scholar particularly genital and gastrointestinal tract mucosae. To establish infection, HIV must cross the epithelial barrier of the recipient, associate with dendritic cells to infect resident lymphocytes and macrophages, and then spread systemically. The oral cavity is considered a relatively protected mucosal site where the innate host defense molecules of saliva are capable of neutralizing HIV and the epithelium itself is not receptive for transmission.3Maher D Wu X Schacker T Larson M Southern P A model system of oral HIV exposure, using human palatine tonsil, reveals extensive binding of HIV infectivity, with limited progression to primary infection.J Infect Dis. 2004; 190: 1989-1997Crossref PubMed Scopus (23) Google Scholar, 4Crombie R Mechanism of thrombospondin-1 anti-HIV-1 activity.AIDS Patient Care STDS. 2000; 14: 211-214Crossref PubMed Scopus (9) Google Scholar, 5McNeely TB Dealy M Dripps DJ Orenstein JM Eisenberg SP Wahl SM Secretory leukocyte protease inhibitor: a human saliva protein exhibiting anti-human immunodeficiency virus 1 activity in vitro.J Clin Invest. 1995; 96: 456-464Crossref PubMed Scopus (377) Google Scholar, 6Ma G Greenwell-Wild T Lei K Jin W Swisher J Hardegen N Wild CT Wahl SM Secretory leukocyte protease inhibitor binds to annexin II, a cofactor for macrophage HIV-1 infection.J Exp Med. 2004; 200: 1337-1346Crossref PubMed Scopus (175) Google Scholar, 7Chang TL Klotman ME Defensins: natural anti-HIV peptides.AIDS Rev. 2004; 6: 161-168PubMed Google Scholar, 8Lehner T Hussain L Wilson J Chapman M Mucosal transmission of HIV.Nature. 1991; 353: 709Crossref PubMed Scopus (47) Google Scholar Orogenital transmission has a very low per-contact risk of acquiring infection, with estimates that 4 of 10,000 contacts result in infection compared with heterosexual or homosexual genital contact (1 of 200 to 1000).9Vittinghoff E Douglas J Judson F McKirnan D MacQueen K Buchbinder SP Per-contact risk of human immunodeficiency virus transmission between male sexual partners.Am J Epidemiol. 1999; 150: 306-311Crossref PubMed Scopus (520) Google Scholar, 10del Romero J Marincovich B Castilla J Garcia S Campo J Hernando V Rodriguez C Evaluating the risk of HIV transmission through unprotected orogenital sex.AIDS. 2002; 16: 1296-1297Crossref PubMed Scopus (50) Google Scholar, 11Page-Shafer K Shiboski CH Osmond DH Dilley J McFarland W Shiboski SC Klausner JD Balls J Greenspan D Greenspan JS Risk of HIV infection attributable to oral sex among men who have sex with men and in the population of men who have sex with men.AIDS. 2002; 16: 2350-2352Crossref PubMed Scopus (70) Google Scholar, 12Royce RA Sena A Cates Jr, W Cohen MS Sexual transmission of HIV.N Engl J Med. 1997; 336: 1072-1078Crossref PubMed Scopus (921) Google Scholar Vertical transmission from mother to infant during breast feeding also occurs with varying rates but may be as high as 15% when breast-feeding is prolonged.13Kourtis AP Lee FK Abrams EJ Jamieson DJ Bulterys M Mother-to-child transmission of HIV-1: timing and implications for prevention.Lancet Infect Dis. 2006; 6: 726-732Abstract Full Text Full Text PDF PubMed Scopus (162) Google Scholar, 14Scarlatti G Mother-to-child transmission of HIV-1: advances and controversies of the twentieth centuries.AIDS Rev. 2004; 6: 67-78PubMed Google Scholar The most plausible portal for HIV entry in oropharyngeal transmission is thought to be the tonsil. In a primate model, when the surface of macaque palatine tonsils was exposed atraumatically with cell-free simian immunodeficiency virus (SIV), infected cells were first detected within the reticulated epithelium lining the tonsillar crypts.15Baba TW Trichel AM An L Liska V Martin LN Murphey-Corb M Ruprecht RM Infection and AIDS in adult macaques after nontraumatic oral exposure to cell-free SIV.Science. 1996; 272: 1486-1489Crossref PubMed Scopus (126) Google Scholar, 16Stahl-Hennig C Steinman RM Tenner-Racz K Pope M Stolte N Matz-Rensing K Grobschupff G Raschdorff B Hunsmann G Racz P Rapid infection of oral mucosal-associated lymphoid tissue with simian immunodeficiency virus.Science. 1999; 285: 1261-1265Crossref PubMed Scopus (209) Google Scholar This unique anatomical compartment, the tonsil crypt epithelium, is specialized to participate in antigen sampling and immune surveillance. Structurally, the surface epithelium of the palatine and lingual tonsils is a continuation of the stratified squamous epithelium of the oral mucosa, but the epithelium of the crypts becomes reticulated and highly infiltrated with lymphocytes, which together with dendritic cells and M cells facilitate transepithelial access of antigens.17Perry M Whyte A Immunology of the tonsils.Immunol Today. 1998; 19: 414-421Abstract Full Text Full Text PDF PubMed Scopus (296) Google Scholar Multiple bacteria, viruses, and their products are known to enter the body through this route,18Stenfors LE Bye HM Raisanen S Myklebust R Bacterial penetration into tonsillar surface epithelium during infectious mononucleosis.J Laryngol Otol. 2000; 114: 848-852PubMed Google Scholar including SIV in primates. Nevertheless, in humans although the tonsil is a reservoir and replication site for HIV,19Moutsopoulos NM Greenwell-Wild T Wahl SM Differential mucosal susceptibility in HIV-1 transmission and infection.Adv Dent Res. 2006; 19: 52-56Crossref PubMed Scopus (30) Google Scholar it still remains unclear whether oral transmission occurs here, because only secondary infection and viral shedding2Orenstein JM Fox C Wahl SM Macrophages as a source of HIV during opportunistic infections.Science. 1997; 276: 1857-1861Crossref PubMed Scopus (265) Google Scholar, 20Frankel SS Tenner-Racz K Racz P Wenig BM Hansen CH Heffner D Nelson AM Pope M Steinman RM Active replication of HIV-1 at the lymphoepithelial surface of the tonsil.Am J Pathol. 1997; 151: 89-96PubMed Google Scholar, 21Zuckerman RA Whittington WL Celum CL Collis T Lucchetti A Sanchez JL Hughes JP Coombs RW Factors associated with oropharyngeal human immunodeficiency virus shedding.J Infect Dis. 2003; 188: 142-145Crossref PubMed Scopus (18) Google Scholar have been clearly documented within the tonsil epithelium. To define further parameters of the tonsil epithelium that may influence its potential for HIV transmission, we isolated the epithelium by laser capture microdissection (LCM) for analysis of its gene expression profile relative to another oral site, the gingiva. The genes most highly expressed in the tonsil compared with gingiva were related to immune functions, such as antibody production and antigen presentation. Among the differentially expressed genes in the tonsil were also the viral co-receptor CXCR4 and factors with the potential to facilitate viral entrapment such as FcRIII, complement receptor 2, and complement components. Protein expression, as demonstrated by immunohistochemical staining of oral tissues, confirmed increased expression of CXCR4 in the tonsil epithelium but a decrease in expression of innate defense molecules, such as secretory leukocyte protease inhibitor (SLPI). This differential expression of factors that may influence HIV may render the tonsil more susceptible for oropharyngeal HIV transmission. Human palatine tonsils were obtained from routine therapeutic tonsillectomies (sleep apnea and nontonsillitis) performed on otherwise healthy adults at the George Washington University Hospital with informed consent (institutional review board no. 099920). Gingival tissues were collected from healthy sites with probing depths <3 mm, with no clinical evidence of inflammation during routine therapeutic periodontal surgery at the Department of Periodontics, University of Maryland, with informed consent (institutional review board no. 1201211). Tissues were rinsed in sterile saline, divided into two pieces, and immediately either snap-frozen for the microarray studies or fixed in 4% paraformaldehyde for immunohistochemistry. Archived, formalin-fixed, oral epithelium samples (Department of Diagnostic Sciences and Pathology, University of Maryland) included intact, normal epithelium of the buccal mucosa and floor of mouth with a final diagnosis of a subepithelial pathosis, ie, fibroma and ranula (institutional review board no. H27950). Five- to 10-μm frozen tissue serial sections were obtained using an RNase-treated blade at −20°C, mounted onto PEN foil slides (Leica Microsystems, Bannock Burn, IL) and stored at −80°C (maximum of 1 week). To minimize RNA degradation the tissues were thawed at room temperature (30 seconds), ethanol-fixed (30 seconds), stained with RNase-free Mayer's hematoxylin (2 minutes), diethylpyrocarbonate-treated water rinsed (5 seconds), eosin (20 seconds), and dehydrated in graded RNase-free alcohols (95 and 100%, 30 seconds each). Finally, slides were air-dried under a hood for 5 minutes and LCM performed immediately at ×10 magnification, using a Leica AS LMD system (Leica Microsystems). Areas unequivocally identified as epithelium were outlined (Figure 1A), laser-dissected, and captured onto RNase-free 200-μl polymerase chain reaction (PCR) tube caps containing RNA lysis buffer (Qiagen Inc., Valencia, CA). Multiple sections containing epithelial cells were independently captured into individual tubes within 30 minutes from the time of tissue thaw and immediately placed on dry ice. Total RNA was extracted from the microdissected samples using the Qiagen RNeasy micro kit (Qiagen), quantitated with a Nanodrop spectrometer (Nanodrop Technologies, Wilmington, DE), and RNA integrity assessed using the 2100 Bioanalyzer (Agilent, Foster City, CA). Preparation of biotin-labeled cRNA, hybridization, and scanning were performed according to the manufacturer's two-cycle protocol (Affymetrix, Santa Clara, CA). In brief, 100 ng of total RNA per sample was used to generate double-stranded cDNA using a two-cycle cDNA synthesis kit and the oligo(dT)24 primer (Affymetrix) containing a 3′ T7 RNA polymerase promoter site. Biotin-labeled cRNA probes were produced from cDNA using the IVT labeling kit (Affymetrix). The probes were purified, fragmented, and hybridized to Affymetrix Plus 2.0 microarray chips that display >47,000 transcripts for 16 hours. Chips were washed and stained using the Affymetrix Fluidics Station 450. Fluorescence intensity was measured using the Affymetrix GeneChip scanner and GeneChip Operating Software (GCOS; Affymetrix). For RT-PCR, 1 μg of amplified cRNA was reverse-transcribed using oligodeoxythymidylic acid primer (Invitrogen Inc., Carlsbad, CA), and the resulting cDNA was amplified by real-time PCR on an ABI Prism 7500 sequence detector (Applied Biosystems, Foster City, CA). Amplification was performed with TaqMan expression assays for GAPDH (assay ID: Hs9999905m1) and for CXCR4 (Hs00607978_s1). Affymetrix GCOS version 1.2 software was used to calculate signal and present call values that were stored in the NIHLIMS, a database for storage and retrieval of chip data maintained at the National Institutes of Health. Data were statistically analyzed using the MSCL analyst's toolbox,22Jison ML Munson PJ Barb JJ Suffredini AF Talwar S Logun C Raghavachari N Beigel JH Shelhamer JH Danner RL Gladwin MT Blood mononuclear cell gene expression profiles characterize the oxidant, hemolytic, and inflammatory stress of sickle cell disease.Blood. 2004; 104: 270-280Crossref PubMed Scopus (168) Google Scholar (available for download at http://abs.cit.nih.gov/MSCLtoolbox/), and the JMP statistical software package (SAS, Inc., Cary, NC). The results for 10 chips were retrieved, and the signal values were subjected to an adaptive variance-stabilizing, quantile-normalizing transformation termed “S10” (Munson, P.J., Gene Logic Workshop of Low Level Analysis of Affymetrix GeneChip Data, 2001). This transform both normalizes between chips over the full data range and makes the variance of replicates nearly uniform over expression level. A major advantage of this approach over the ordinary log-ratio is that changes in S10-transformed values have a uniform variance over the full expression scale. Visualization of the global results and detection of possible outliers among the 10 samples were facilitated by principal component analysis (data not shown) of the transformed data and presentation in bivariate plots of low-order principal components. Results from two separate types of tissues and from the five and seven independent donors were clearly separated in the first and second principal component and are uploaded in the Gene Expression Omnibus Repository (GEO accession no. GSE7224). No outlying chips were detected. To quantify the significance of gene expression differences, a one-way, two-level analysis of variance was applied comparing tonsil (n = 5) to gingival (n = 7) tissues. The P value for differences between the two tissues was collected for each probe set. To ameliorate the multiple comparison problems, the false discovery rate (FDR)23Klipper-Aurbach Y Wasserman M Braunspiegel-Weintrob N Borstein D Peleg S Assa S Karp M Benjamini Y Hochberg Y Laron Z Mathematical formulae for the prediction of the residual beta cell function during the first two years of disease in children and adolescents with insulin-dependent diabetes mellitus.Med Hypotheses. 1995; 45: 486-490Abstract Full Text PDF PubMed Scopus (209) Google Scholar was controlled. Log-fold changes were computed as the difference between average values for the two groups. Probe sets with a greater than twofold change in either direction, with FDR less than 10% and with a present call in greater than 50% of samples in at least one group (four of seven and three of five) were selected for further analysis. Paraffin-embedded tissues (tonsil, n = 5; gingiva, n = 3; buccal mucosa, n = 3; and floor of mouth, n = 3) were cut into 5-μm sections, deparaffinized, and rehydrated, followed by heat-induced epitope retrieval. Methanol containing 3% hydrogen peroxide was used to block the endogenous peroxidase for 15 minutes. Sections were blocked with the corresponding preimmune serum for 30 minutes and incubated overnight at 4°C with primary antibodies to CXCR4, CCR5, CD19, CD3, defensin-β1, defensin-β4 (Abcam, Cambridge, MA), SLPI (R&D Systems, Minneapolis, MN), ICAM-3 (Novocastra, Newcastle, UK), CD4 (Invitrogen), and GalCer (Chemicon, Temecula, CA). After washing with phosphate-buffered saline three times, immunolabeling was detected using a biotinylated secondary antibody followed by visualization with an avidin-biotin horseradish peroxidase labeling kit (Invitrogen) and diaminobenzidine staining. Finally, the specimens were counterstained with Mayer's hematoxylin and mounted with Permount (Fisher Scientific, Pittsburgh, PA). Negative controls were performed by replacing primary antibody with preimmune serum. Snap-frozen tissues were sectioned (5 μm), fixed in 95% ethanol for 45 minutes, washed, incubated with 100 mmol/L glycine (MP Biomedicals, lllrich, France) for 10 minutes, and blocked with 5% bovine albumin (Sigma-Aldrich, St. Louis, MO) or with preimmune serum for 30 minutes before staining. Sections were incubated overnight with antibodies to CCR5, CXCR4, CD32, and pan-cytokeratin (Abcam) followed by visualization with an avidin-biotin horseradish peroxidase labeling kit (Invitrogen) or staining with secondary antibodies conjugated to Texas Red and fluorescein and mounted with Vectashield containing 4,6-diamidino-2-phenylindole for fluorescence microscopy (all from Vector Laboratories, Burlingame, CA). Staining was evaluated by two independent investigators (N.M., N.N.). To evaluate levels of expression in each sample, sequential electronic images from 20× fields were taken for the length of the epithelium for an average of 20 images per tonsil sample and 10 images per oral sample (smaller biopsy). The images were evaluated for the presence, localization, and intensity of staining (weak, moderate, and strong) and percentage of positively stained cells. For the latter, epithelial cells in each epithelial compartment (basal, parabasal, spinal, and keratin) were evaluated and the positive cells counted. Means and SDs of percent positive cells were calculated per sample type and a Wilcoxon signed-rank test was used to compare expression between groups. The squamous epithelia lining the oropharyngeal area share functions necessary to protect underlying tissues from injury and invasion but also exhibit location-dependent unique characteristics associated with their specific functions. Accordingly, the epithelium of the gingiva is increasingly keratinized to withstand masticatory forces, whereas the tonsil epithelium is specialized to traffic antigens to the underlying lymphoid compartment. To differentiate potential features that may influence susceptibility to HIV, we isolated the respective epithelia by LCM (Figure 1A) and subjected them to microarray gene expression analysis. Multiple probe sets showed similar expression patterns between the two groups. Based on absolute expression intensities, many of the genes most highly expressed in both tonsil and gingival epithelium were common (Figure 1B). Among them were keratin genes (6A, 13, 14), the water channel protein aquaporin-3, the keratinocyte factor stratifin, cystatins A and B, and S100 binding Ca+2 binding proteins, all of which are associated with tissues of epithelial origin.24King LS Kozono D Agre P From structure to disease: the evolving tale of aquaporin biology.Nat Rev Mol Cell Biol. 2004; 5: 687-698Crossref PubMed Scopus (786) Google Scholar, 25Dellambra E Patrone M Sparatore B Negri A Ceciliani F Bondanza S Molina F Cancedda FD De Luca M Stratifin, a keratinocyte specific 14-3-3 protein, harbors a pleckstrin homology (PH) domain and enhances protein kinase C activity.J Cell Sci. 1995; 108: 3569-3579Crossref PubMed Google Scholar, 26Bánki Z Kacani L Rusert P Pruenster M Wilflingseder D Falkensammer B Stellbrink HJ van Lunzen J Trkola A Dierich MP Stoiber H Complement dependent trapping of infectious HIV in human lymphoid tissues.AIDS. 2005; 19: 481-486Crossref PubMed Scopus (46) Google Scholar, 27Robinson NA Lapic S Welter JF Eckert RL S100A11, S100A10, annexin I, desmosomal proteins, small proline-rich proteins, plasminogen activator inhibitor-2, and involucrin are components of the cornified envelope of cultured human epidermal keratinocytes.J Biol Chem. 1997; 272: 12035-12046Crossref PubMed Scopus (208) Google Scholar Combined with our selective dissection of the epithelium, these results confirm the epithelial cell specificity of the dissected specimens. Further analysis of the variability of gene expression in the tonsil and gingival data set revealed a clear separation between the two sample groups, indicating that each tissue type has a distinct gene expression profile. Statistical analysis of the microarray data yielded 660 probe sets that were significantly differentially expressed between the two groups. These data demonstrate that surface epithelia from different locations exhibit both distinct and shared characteristics. We next focused on the significantly differentially expressed genes and their functions. With GO-scan analysis, a computerized process that links probe set annotations with their listed function, we categorized the differentially expressed genes by their known functions. As evident in Figure 2, in the tonsil the genes most highly expressed were categorized as defense response, response to biotic stimuli, immune response, and response to pathogens, consistent with an immunological role for this compartment. Among these were genes associated with antibody production such as the J chain of dimeric IgJ (up-regulated 96-fold) and the Ig κ light chain (29-fold), genes linked to major histocompatibility complex antigen presentation such as MHCII and cathepsin S28Kitamura H Kamon H Sawa S Park SJ Katunuma N Ishihara K Murakami M Hirano T IL-6-STAT3 controls intracellular MHC class II alphabeta dimer level through cathepsin S activity in dendritic cells.Immunity. 2005; 23: 491-502Abstract Full Text Full Text PDF PubMed Scopus (178) Google Scholar (2.5- to 3.5-fold), and genes related to the activation of the complement cascade (CR2, 4.6; C7, 3.7; and C1q, 3) (Figure 2). Consistent with this pattern, lymphocyte and particularly T- and B-cell-related genes were also pronounced in the tonsil epithelium as seen by the significant up-regulation of antibody components and B-cell markers (CD19, CD79) as well as T-cell markers (CD3, CD2, CD69) (Figure 2, Figure 3) and by immunohistochemical staining for CD19 and CD3 (Figure 3, D and E; and Table 1).Figure 3Differential expression of HIV-related genes. A: Parallel plot showing gene expression intensity for B-cell markers (CD19, CD79), antibody-related genes (Ig J polypeptide), and T-cell marker (CD3) in gingival (n = 7) and tonsil (n = 5) samples. B: Parallel plot showing expression levels/intensity for genes associated with HIV entrapment/transmission in the tonsil (n = 5) and gingival (n = 7) epithelium. C: Parallel plot showing gene expression intensity for the HIV receptor CD4 and co-receptors (GalCer, CXCR4, and CCR5) in the tonsil and gingival epithelium, *FDR < 10% and P < 0.05 of differential gene expression between sample groups. Immunohistochemical staining for CD19 in the tonsil (D) and CD3 (E). Immunohistochemical staining for ICAM-3 (F), for CD4 (H), and GalCer (I) in the tonsil epithelium. Immunofluorescence FcR-CD32 (red), pan-cytokeratin (green) and 4,6-diamidino-2-phenylindole (G). Arrows indicate positive staining. Original magnifications: ×20 (D–F, H, and I); ×63 (G).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table 1Semiquantitative Evaluation for Immunological MarkersMarkersTonsil epitheliumOral epitheliumCD19+−CD3+−ICAM-3+−FcR+−CD4+−GalCer++++Semiquantitative analysis of immunohistochemical staining. Symbols shown represent percent positive staining; −, <5%; +, 5 to 25%; ++, 25 to 50%; +++, 50 to 75%; ++++, 75 to 100%. Open table in a new tab Semiquantitative analysis of immunohistochemical staining. Symbols shown represent percent positive staining; −, <5%; +, 5 to 25%; ++, 25 to 50%; +++, 50 to 75%; ++++, 75 to 100%. In the gingiva the more highly expressed genes belonged to the categories of organ development and cytoskeleton (Figure 2). Genes of the keratin family dominated, including keratins 6, 14, 16, and 17, and cytokeratin 2 (up-regulated up to 500-fold), reflecting the increased potential for keratinization in the gingival sites, although the highest layers of keratin were not included in our dissection. Although the majority of highly expressed genes were not linked to immune functions, some genes encoded factors that may also participate in the immune response such as the kallikrein proteases 5, 6, and 7 (5- to 20-fold), which may be involved in matrix reorganization during inflammation,29Paliouras M Diamandis EP The kallikrein world: an update on the human tissue kallikreins.Biol Chem. 2006; 387: 643-652Crossref PubMed Scopus (62) Google Scholar and spondin 2 (fivefold), a pattern recognition molecule for microbial pathogens that participates in innate defenses.30Yeaman GR Asin S Weldon S Demian DJ Collins JE Gonzalez JL Wira CR Fanger MW Howell AL Chemokine receptor expression in the human ectocervix: implications for infection by the human immunodeficiency virus-type I.Immunology. 2004; 113: 524-533Crossref PubMed Scopus (75) Google Scholar We next investigated whether the differential gene expression patterns between the two tissue types might influence HIV susceptibility. In this regard, the increased keratinization of the gingival tissue, reflected by enhanced regulation of keratin genes, may contribute to a barrier against HIV. Conversely, an increased presence of immune cells in the tonsil epithelium may favor HIV entry and access to target cells.26Bánki Z Kacani L Rusert P Pruenster M Wilflingseder D Falkensammer B Stellbrink HJ van Lunzen J Trkola A Dierich MP Stoiber H Complement dependent trapping of infectious HIV in human lymphoid tissues.AIDS. 2005; 19: 481-486Crossref PubMed Scopus (46) Google Scholar Within the differentially overexpressed tonsil genes were Fcγ receptor III,31van Montfort T Nabatov AA Geijtenbeek TB Pollakis G Paxton WA Efficient capture of antibody neutralized HIV-1 by cells expressing DC-SIGN and transfer to CD4+ T lymphocytes.J Immunol. 2007; 178: 3177-3185PubMed Google Scholar complement receptor CR2, and various complement components26Bánki Z Kacani L Rusert P Pruenster M Wilflingseder D Falkensammer B Stellbrink HJ van Lunzen J Trkola A Dierich MP Stoiber H Complement dependent trapping of infectious HIV in human lymphoid tissues.AIDS. 2005; 19: 481-486Crossref PubMed Scopus (46) Google Scholar (FDR <10%). Other key HIV-binding molecules such as the adhesion molecule ICAM-3,32Barat C Tremblay MJ Adhesion molecules as costimulators: ICAM-3 signaling potentiates CD3-mediated HIV-1 stimulation.Ann NY Acad Sci. 2002; 973: 590-593Crossref PubMed Scopus (1) Google Scholar the dendritic cell-specific C-type lectin DC-SIGN,33Su SV Hong P Baik S Negrete OA Gurney KB Lee B DC-SIGN binds to HIV-1 glycoprotein 120 in a distinct but overlapping fashion compared with ICAM-2 and ICAM-3.J Biol Chem. 2004; 279: 19122-19132Crossref PubMed Scopus (57) Google Scholar and syndecan-134Bobardt MD Saphire AC Hung HC Yu X Van der Schueren B Zhang Z David G Gallay PA Syndecan captures, protects, and transmits HIV to T lymphocytes.Immunity. 2003; 18: 27-39Abstract Full Text Full Text PDF PubMed Scopus (186) Google Scholar were not significantly differentially expressed (Figure 3B). Gene expression for the viral co-receptor CXCR4, although variable (2- to 10-fold difference in expression between donors, as seen by microarray and RT-PCR analysis data not shown), was significantly higher in the tonsil. CCR5 and galactosylceramide (GalCer),35Fantini J Yahi N Delezay O Gonzalez-Scarano F GalCer, CD26 and HIV infection of intestinal epithelial cells.AIDS. 1994; 8: 1347-1348Crossref PubMed Scopus (19) Google Scholar as well as the primary HIV receptor CD4, were not significantly differentially expressed (Figure 3C). By immuno