Goblet Cells Contribute to Ocular Surface Immune Tolerance-Implications for Dry Eye Disease

Abstract

Conjunctival goblet cell (GC) loss in dry eye is associated with ocular surface inflammation. This study investigated if conjunctival GCs contribute to ocular surface immune tolerance. Antigens applied to the ocular surface, imaged by confocal microscopy, passed into the conjunctival stroma through goblet cell associated passages (GAPs) in wild type C57BL/6 (WT), while ovalbumin (OVA) was retained in the epithelium of SAM pointed domain containing ETS transcription factor (Spdef) knockout mice (Spdef-/-) that lack GCs and are a novel model of dry eye. Stimulated GC degranulation increased antigen binding to GC mucins. Induction of tolerance to topically applied OVA measured by cutaneous delayed type hypersensitivity (DTH) was observed in WT, but not Spdef-/-. OTII CD4⁺ T cells primed by dendritic cells (DCs) from the conjunctival draining lymph nodes of Spdef-/- had greater IFN-γ production and lower Foxp3 positivity than those primed by WT DCs. These findings indicate that conjunctival GCs contribute to ocular surface immune tolerance by modulating antigen distribution and antigen specific immune response. GC loss may contribute to the abrogation of ocular surface immune tolerance that is observed in dry eye.

Keywords: adaptive immunity; antigen; dendritic cells; goblet cell; immune tolerance; mucins.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1

Goblet cell associated passages (GAPs) are present in the conjunctiva. Representative laser confocal microscopy images of whole mount conjunctivas taken from six mice for each antigen (n = 6/group) are shown. (A) Z-stack option from epithelium (e) to stroma (s) showing distribution of three different sized antigens 30 min after topical application to the ocular surface. Nuclei are stained with DAPI. Top to bottom—the greatest stromal passage was noted with OVA peptide 2.3 kDa (green) and Dextran 10 kDa (green) that bound to the surface epithelium and migrated in a columnar pattern into the stroma. Dextran 70 kDa (green) localized primarily to the surface epithelium, but columns extending into the superficial stroma were also noted. Negative control with 4′,6-diamidino-2-phenylindole (DAPI, blue) nuclear stain, but no antigen instilled is shown at the bottom; (B) 45 kDa OVA (red) was noted in the epithelium and in columns in the stroma of WT (top), but was localized to the epithelium only in the Spdef−/− (bottom) (n = 6); (C) Z-stack option from epithelium to stroma showing distribution of MUC2+ (green) and OVA+ (red) cells in WT mice. White squares demarcate goblet cells with co-localization of MUC2 and OVA that are shown at higher magnification below; (D) Top—surface of whole mount conjunctiva stained with WGA lectin (purple) that bound to goblet cell glycoproteins and Dextran 10 kDa (green). Minimal dextran binding was noted in WGA+ filled goblet cells (arrows), whereas dextran was noted to pass into and through empty goblet cells (WGA negative) into the underlying stroma (arrowheads indicate dextran filled goblet cells with an underlying column of dextran stromal fluorescence in Z-stack); (E) Surface of whole mount conjunctiva stained for goblet cell marker keratin 7 (in green) and nuclei in red with goblet cell openings marked with arrows. Scale bar is 20 µm; (F) Whole mount conjunctivas stained for cytokeratin 7 (K7, red) and CD11c (top, green) or CD11b c (bottom, green) cells with nuclei stained with DAPI (blue). Some CD11b+ cells had a dendritic morphology (asterisk) and others stained positively for K7 (arrow), n = 6 per antigen.

Figure 2

Goblet cells under cholinergic regulation serve as antigen conduits. (A–C) Transmission electron microscopy of conjunctival goblet cells. (A) In homeostasis, goblet cells are filled with mucus granules (asterisks); (B) 20 min after stimulation with cholinergic agonist, carbachol choline (CCh), granules in goblet cell openings appear smaller in size and darker (arrowhead); (C) Passages within and around some discharged goblet cells were observed (arrows). Magnification: 3000×. Scale bar 4 µm (in A–C); n = 4 per group; (D) Representative images of whole mount conjunctivas stained for MUC5AC. Mice received either OVA drops 30 min prior to euthanasia (homeostasis), cholinergic stimulation 20 min prior to OVA drops (+CCh), or atropine 30 min prior to CCh and OVA (+Atropine +CCh). Atropine blockade prior to cholinergic stimulation limited OVA passage into GCs and stroma, n = 3; (E) Whole mount conjunctiva stained for CD11b (green) 30 min after topical administration of OVA (red) with nuclei stained with DAPI (blue). White square demarcates OVA+CD11b+ cell that is shown in higher magnification at the upper right; (F) In conjunctival tissue sections, CD11b+ cells were found just below the basal conjunctival epithelium in WT (left), while they were present on the surface of the epithelium and within the epithelium and stoma in the Spdef−/− (KO, right); (G) Left. Mean ± SD of flow cytometry analysis of OVA+ antigen presenting cells 4 h after topical administration of OVA23–339 peptide on the ocular surface of WT C57BL/6 and Spdef−/− (KO) strains. Frequency and mean fluorescence intensity (MFI) of OVA+ cells are shown (n = 6 per strain); * p < 0.05; ** p < 0.01; comparison WT vs. Spdef−/− Right. Representative dot plots of flow cytometry analysis showing uptake of OVA323–339 peptide (OVA) after topical administration. Conjunctivas were harvested, collagenase-digested and single cell suspensions were stained with CD11b, CD11c and F4/80 antibodies.

Figure 3

Goblet cell loss in Spdef−/− abrogates induction of conjunctival immune tolerance. (A) Conjunctival immune tolerance was measured by delayed type hypersensitivity (DTH) to ovalbumin (OVA). Mice with or without pre application of OVA drops for three consecutive days received S.C. immunization (Imm) with OVA + complete Freund’s adjuvant (CFA) on day 8 and were challenged with OVA antigen by intradermal ear injection (OVA right ear and PBS left ear) on day 15. Ear swelling was measured after 48 h. Naïve mice without immunization served as controls; (B) In vivo DTH assay (ear swelling) measured 48 h after challenge. Results are means ± SD of ear swelling after calculating the difference between the antigen-injected and PBS-injected ears for each mouse (n = 5 animals/group). Induction of conjunctival tolerance was not observed in Spdef−/− (KO). Imm = immunization of OVA + CFA; B6 = C57BL/6 wild-type; KO = Spdef−/−; d = days. * p < 0.05; *** p < 0.001, **** p < 0.0001 comparison WT vs. KO.

Figure 4

Spdef−/− dendritic cells (DCs) prime Th1 cells. Cervical lymph node (CLN) cell suspensions of from either B6 WT or Spdef−/− mice (KO, n = 4 group) were pulsed with OVA323–339 peptide and co-cultured with CD4+ T cells isolated from OT II mice for 4–5 days. Cells were collected for measuring proliferation and IFN-γ was measured in supernatants by immunobead assays. Data shown in (A–E) are from representative experiments. (A) Representative CFSE histograms showing fluorescence dilution (measure of proliferation) in OT II cells after priming with either WT or KO CLN suspensions; (B) Mean fluorescence intensity (MFI) of CFSE+CD4+ cells. Means ± SD, n = 4 wells/group; (C) Representative dot-plots of CD4+Foxp3+ used to generate graph in D; (D) CD4+Foxp3+ percentage in the two groups. Means ± SD, n = 4 wells/group; (E) IFN-γ concentration in supernatants of mixed lymphocyte reactions measured by Luminex assay. Means ± SD, n = 4 wells/group. Statistical tests, Student t test comparison WT vs. SPDEFKO. * p < 0.05; ** p < 0.01; **** p < 0.001.