Male and female WT and C4 KO (C57BL/6J background) mice (24), aged 8–12 weeks, were obtained from Jackson Laboratory and maintained under pathogen-free conditions in the animal facilities of the Cleveland Clinic. All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee of the Cleveland Clinic and were in accordance with the U.S. Department of Health and Human Services Guide for the Care and Use of Laboratory Animals.

Experimental autoimmune uveitis induction was performed as described previously (20). Each mouse was injected subcutaneously at multiple sites in the back and tail base with a total of 200 μl of emulsion consisting of equal volumes of Mycobacterium tuberculosis H37Ra-supplemented complete Freund’s adjuvant (2.5 mg/ml) (Difco Laboratories, Inc., Detroit, MI, USA) and an aqueous solution of the human IRBP_651–670 peptide (LAQGAYRTAVDLESLASQLT) (2 mg/ml) (custom synthesized by GenScript USA Inc., Piscataway, NJ, USA). A single dose of 500 ng of pertussis toxin (List Biologic Laboratories, Inc., Campbell, CA, USA) was injected intraperitoneally on the same day.

After immunization, clinical manifestations were examined daily using a binocular indirect ophthalmoscope (Keeler Instruments, Inc., Broomall, PA, USA). The pupils were dilated with a mixed ophthalmic solution of 0.5% tropicamide and 1.25% phenylephrine hydrochloride and disease severity was scored on a scale of 0–4, according to published criteria (20).

On day 21 after immunization, the mice were euthanized. For histopathological evaluation, whole eyes were collected and fixed in 10% formaldehyde/PBS buffer for 24 h and the fixed tissues embedded in paraffin. Sections (5 µm) were cut through the pupil and optic nerve axis and processed for H&E staining, then retinal histopathological changes were graded on a scale of 0–4 according to previously published scoring criteria (20).

Retinal imaging was performed as described previously (23) using cSLO, SD-OCT, and TEFI under general anesthesia. A cSLO (Heidelberg Retina Angiograph II; Heidelberg Engineering, Carlsbad, CA, USA) was used to collect both reflectance and fluorescence information from the posterior segment. The SD-OCT system used was a 840 HR SDOIS (Bioptigen, Inc., Morrisville, NC, USA) with a central operating wavelength of ~840 nm and an in-depth, axial resolution of ~6 μm. Conventional visible light fundus images were collected using a custom-fabricated TEFI apparatus (25). Number of hyper-reflective foci in vitreous chamber of OCT images areas were quantitated using ImageJ software (http://imagej.nih.gov/ij/, National Institutes of Health, Bethesda, MD, USA).

T cell recall assays were performed on day 21. For each of the immunized WT and C4 KO mice, 4 × 10⁵ splenocytes were cultured in 96-well round-bottomed microtiter plates in 100 µl of complete RPMI 1640 medium in the presence or absence of 20 µg/ml of peptide IRBP_651–670 or a non-relevant peptide (ovalbumin OVA_323–339, Genscript, NJ, USA) for 72 h, then the supernatants were collected and interferon (IFN)γ and interleukin (IL)-17 levels measured using ELISA kits (BioLegend, San Diego, CA, USA).

Following a previously established protocol (26), BM cells were harvested from the femur of WT and C4 KO mice and cultured for 48 h in complete medium containing 10 ng/ml of granulocyte-macrophage colony-stimulating factor (GM-CSF) and 100 U/ml of IL-4 (Peprotech, Rocky Hill, NJ, USA), then, non-adherent cells were gently removed and the medium replaced with fresh medium containing the same concentrations of GM-CSF and IL-4. Double-positive cells (DCs) were harvested on day 5 and DC differentiation was assessed by counting MHC II and CD11c DCs using a flow cytometer.

The function of the differentiated DCs was compared by measuring their ability to activate antigen-specific T cells. In brief, T cells were enriched from splenocytes of OTII mice (Jackson Laboratory) using nylon wool and labeled with carboxyfluorescein succinimidyl ester (CFSE) according to the manufacturer’s instruction (Thermo Fisher Scientific, Waltham, MA, USA), then were cocultured with the same numbers of differentiated WT or C4 KO DCs (10:1 ratio) in the presence or absence of 2 μg/ml of peptide OVA_323–339 for 72 h. The proliferation of the activated T cells was then measured by CFSE dilution (27) using flow cytometry, while IFNγ production by the activated T cells were measured by analyzing the culture supernatants using ELISA (Biolegend) following the manufacturer’s recommended protocol.

Splenic DCs from WT and C4 KO mice were isolated using a mouse pan-DC enrichment kit (Stemcell Technologies Inc., Cambridge, MA, USA) following the manufacturer’s instructions. The isolated splenic DCs were matured with 2 µg/ml LPS 4 h before being cocultured with CFSE-labeled CD4+ T cells isolated from OTII mouse splenocytes using a mouse CD4+ T cell enrichment kits (Stemcell Technologies Inc.) at a ratio of 10:1 (T cells: DCs) in the absence or presence of 2 µg/ml of OVA_323–339 peptide for 72 h. CD4+ T cells isolated by this protocol usually reaches >97% purity based on flow cytometric analysis. The proliferation of the DC-activated T cells was analyzed by flow cytometry, and levels of IFNγ produced by the activated T cells in the culture supernatants were measured using ELISA.

CD4+ T cells were isolated as described above according to the manufacturer’s instructions, then, 4 × 10⁵ of the purified T cells were seeded in 96-well round-bottomed microtiter plates precoated overnight with 5 µg/ml of anti-mouse CD3 mAb (Biolegend, CA, USA) in the presence of 1 µg/ml of anti-mouse CD28 mAb and 1,000 U/ml of mouse IL-2 (Peprotech, Rocky Hill, NJ, USA).

For the activation assay, T cells were collected and stained with a PE-labeled anti-CD69 mAb for 5 h. For the T cell proliferation assay, 10 nM BrdU was added to the plates for 48 h, then incorporated BrdU was detected using a BrdU ELISA kit (BD Biosciences, San Jose, CA, USA) after DNase treatment according to the manufacturer’s protocol. For the apoptosis assay, cells were collected and stained with Annexin V and PI as recommended by the manufacturer (BD Biosciences, San Jose, CA, USA).

GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA, USA) was used for data analysis. Data were expressed as means ± SEM unless labeled otherwise. The EAU clinical scores were assessed by two-way ANOVA. Two sets of data were compared by Student’s t-test and three sets of data were analyzed by one-way ANOVA. p-Values less than 0.05 were considered significant.

To explore the potential role of C4 in the pathogenesis of EAU, we first immunized age- and sex-matched WT and C4 KO mice with IRBP peptide in CFA to induce EAU following a previously published protocol. We followed the development of EAU daily by indirect ophthalmoscopy from day 7 to day 20 after immunization, then euthanized the mice to prepare ocular sections for histopathological scoring and other assays. We found that, compared to clinical scores of ~1.5–1.8 in WT mice at day 14 after EAU induction, those in the C4 KO mice were significantly lower, with an average score of ~1.0 (Figure 1A) and that histopathological scores for the ocular sections on day 21 were also significantly lower in the C4 KO mice (0.41 ± 0.15 vs 0.85 ± 0.15) (Figures 1B,C).

In addition to the above tests, we used three ocular imaging techniques, TEFI, cSLO, and SD-OCT, to evaluate the severity of EAU in the WT and C4 KO mice on day 15 after disease induction. These studies showed multiple retinal pathologic changes in the WT mice, including focal and linear chorioretinal lesions observed by TEFI (Figure 2A) and increased hyper-reflective features adjacent to the retinal vessels in the inner retina and around the optic nerve in the outer retina observed by cSLO (Figures 2C–F). SD-OCT also revealed inflammatory cell infiltration in the vitreous chamber, retinal infoldings, and a disrupted inner segment/outer segment band in the WT mice (Figure 2K). All of these features were much less obvious in the C4 KO mice (Figures 2B,G–J,L). Hyper-reflective foci in vitreous cavity are likely cells and/or cellular aggregates that have infiltrated into the vitreous. We quantitated these foci and found that C4 KO mice showed signficantly lower numbers of these foci in the vitreous cavity than WT mice in EAU (p = 0.003, Figure 2M).

We used ex vivo antigen-specific recall assays to determine pathogenic Th1 and Th17 responses after EAU induction. We collected splenocytes from WT and C4 KO mice on day 21, then incubated the cells with or without 20 µg/ml of IRBP peptide or a non-relevant peptide, OVA_323–339, for 72 h, then measured levels of IFNγ and IL-17 in the culture supernatants by ELISA. These ex vivo antigen-specific T cell recall assays showed that IRBP-specific T cells from immunized C4 KO mice produced only 54.21 ± 8.58 ng/ml of IFNγ and 9.10 ± 1.81 ng/ml of IL-17 compared to 94.63 ± 15.96 ng/ml of IFNγ and 14.72 ± 2.05 ng/ml of IL-17 produced by cells from WT mice (Figures 3A,B). These data suggest that pathogenic Th1 and Th17 responses were significantly reduced in C4 KO mice compared to WT mice in EAU.

Since ocular infiltration of lymphocytes and other inflammatory cells is a hallmark of EAU, we compared numbers of ocular-infiltrating CD4+ cells in WT and CD4 KO mice with EAU. In brief, at day 21 after immunization, we treated enucleated eyes from five mice from each group with 1 mg/ml collagenase, then pooled the resultant single cells together and incubated them with an anti-CD4 mAb and analyzed the stained cells by flow cytometry. As shown in Figures 3C,D, 0.2% of the cells from the eyes of WT mice with EAU were CD4+, but this value was only 0.06% in the eyes of CD4 KO mice with EAU, showing that C4 KO mice have reduced ocular T cell infiltration in EAU.

Our in vivo and ex vivo studies above show that pathogenic T cell responses were reduced in EAU C4 KO vs EAU WT mice. Since DCs are known to locally produce C4 (28), it is possible that absence of C4 in DCs could impair DC differentiation and/or the function of the differentiated DCs. To explore the mechanism underlying attenuated pathogenic T cell responses in C4 KO mice with EAU, we generated DCs from the BM of naïve WT or C4 KO mice and assessed the differentiation of DCs by quantifying the resultant MHC II/CD11c DCs using flow cytometry and found no significant difference (Figures 4A,B). We then incubated the same numbers of the differentiated WT or C4 KO DCs with the CFSE-labeled T cells from naïve OTII mice in the presence or absence of peptide OVA_323–339 for 72 h, then measured T cell proliferation by flow cytometry and IFNγ levels in the culture supernatants by ELISA. As shown in Figure 4C, OTII T cells cocultured with WT or C4 KO DCs showed similar proliferation after OVA stimulation. In addition, Figure 4D shows that, while IFNγ was barely detectable in DC/OTII T cell cocultures in the absence of OVA peptide, comparable IFNγ levels was produced by T cells stimulated with either WT or C4 KO DCs. In addition to the BM-derived DCs, we also studied the splenic DCs from WT and C4 KO mice and got similar results (Figures 4E,F). These data suggest that C4 deficiency in DCs is not causally involved in the reduced antigen-specific T cell response seen in EAU C4 KO mice.

To assess the impact of C4 deficiency from the T cell side, we purified CD4+ T cells from naïve WT or C4 KO mice, then activated them with anti-CD3 and anti-CD28 mAbs, together with IL-2, then evaluated T cell activation after 5 h by comparing levels of the T cell activation marker, CD69, and investigated the proliferation of the activated T cells after 24 h (BrdU incorporation) and their survival after 48 h (Annexin V and PI staining). Finally, we measured IFNγ levels in the culture supernatants after 48 and 72 h by ELISA. Figures 5A,B show that, after 5 h of activation, CD69 levels were significantly lower on C4 KO T cells than on WT T cells. In addition, the proliferation of activated C4 KO T cells was reduced, as indicated by a significantly lower percentage of BrdU+ cells in the C4 KO T cells than in the activated WT T cells at 24 h (Figures 6A,B), and there were significantly more apoptotic (Annexin V+) and dead (PI+) C4 KO T cells than WT T cells at 48 h after activation (Figure 7). Finally, activated WT T cells produced higher levels of IFNγ than activated C4 KO T cells at both 48 and 72 h (48 h: 21.56 ± 1.39 vs 14.63 ± 1.36, 72 h: 54.33 ± 1.73 vs 37.94 ± 3.34 ng/ml) (Figure 8). These results demonstrate that T cells from naïve C4 KO mice show reduced activation, proliferation, survival, and inflammatory cytokine production, which could explain the decreased IRBP-specific T cell responses in C4 KO mice in EAU.