Immune Privilege: The Microbiome and Uveitis

Abstract

Immune privilege (IP), a term introduced to explain the unpredicted acceptance of allogeneic grafts by the eye and the brain, is considered a unique property of these tissues. However, immune responses are modified by the tissue in which they occur, most of which possess IP to some degree. The eye therefore displays a spectrum of IP because it comprises several tissues. IP as originally conceived can only apply to the retina as it contains few tissue-resident bone-marrow derived myeloid cells and is immunologically shielded by a sophisticated barrier - an inner vascular and an outer epithelial barrier at the retinal pigment epithelium. The vascular barrier comprises the vascular endothelium and the glia limitans. Immune cells do not cross the blood-retinal barrier (BRB) despite two-way transport of interstitial fluid, governed by tissue oncotic pressure. The BRB, and the blood-brain barrier (BBB) mature in the neonatal period under signals from the expanding microbiome and by 18 months are fully established. However, the adult eye is susceptible to intraocular inflammation (uveitis; frequency ~200/100,000 population). Uveitis involving the retinal parenchyma (posterior uveitis, PU) breaches IP, while IP is essentially irrelevant in inflammation involving the ocular chambers, uveal tract and ocular coats (anterior/intermediate uveitis/sclerouveitis, AU). Infections cause ~50% cases of AU and PU but infection may also underlie the pathogenesis of immune-mediated "non-infectious" uveitis. Dysbiosis accompanies the commonest form, HLA-B27-associated AU, while latent infections underlie BRB breakdown in PU. This review considers the pathogenesis of uveitis in the context of IP, infection, environment, and the microbiome.

Keywords: T regulatory cells; adjuvant effect; blood retinal barrier; commensals; folate; nutritional metabolites; probiotics.

Figure 1

Hierarchy of levels of Immune Privilege (IP). IP may be considered as a property of all tissues with varying degrees of ability to influence the outcome of immune responses played out within their own microenvironment. Thus, tissues such as the skin and mucosal surfaces generate strong immune reactions to pathogens, experience extensive tissue damage and have considerable capacity for repair, while at the other extreme neural tissue (brain parenchyma, retina) temper or even prevent immune responses, induce latency rather than replication in pathogens [reviewed in (3).] but have limited ability for repair if severely damaged. Between these two extremes, different tissues have different levels of “privilege.” The outcome of pathogen challenge to any tissue depends on the level of privilege which is determined by the nature, integrity and strength of that tissue’s barrier to pathogen invasion. Barriers are complex entities comprising physical, chemical, molecular, cellular, and immunological components and are specific to each tissue. The figure illustrates what happens under normal circumstances in the top panel: here the balance of IP is set toward strong immune and repair processes in the skin at one end of the spectrum but a weak immune response in the CNS at the other end of the spectrum risks uncontrolled infection. In contrast, when this balance is shifted as shown in the lower panel, outcomes tend to reverse: in the skin, a weaker immune response might fail to clear infections or fail to promote repair while in the CNS better control of infection might be possible but still does not clear pathogens, instead promoting latent infections.

Figure 2

The Retinal Neurovascular Unit. The retinal neurovascular unit (NVU) is the seat of the blood retinal barrier (BRB), situated at the capillary/venous side of the retinal circulation. Two levels of “barrier” exist: a) at the endothelial tight junctions (1), supported by pericytes and under immunological surveillance by perivascular macrophages; and b) at the glia limitans (2), constructed by astrocyte glial cell and Müller glial cell foot processes, with contributions to the continuous membranous structure form microglial cells and neuronal dendritic processes.

Figure 3

Experimental Autoimmune Uveoretinitis (EAU): a model for sight-threatening posterior uveitis. Clinical and histological signs of EAU in mice. EAU may be induced by subcutaneous injection of interphotoreceptor retinol binding protein (IRBP) peptide 161–180 in B10.RIII mice or peptide 1–20 in C57BL/6 mice with Complete Freund’s Adjuvant (CFA) and pertussis toxin (Ptx). Disease is mediated by a Th1/IL12 and/or a Th17/IL23 mechanism with dominant effect being Th17 mediated disease. The images are from the C57/BL6 mouse model. Clinical images are shown: (A) early focal chorioretinal infiltrate (arrowhead); (B) developing retinal vasculitis shown as focal perivascular “sheathing” (arrows); (C) extensive retinal vasculitis and large granulomatous infiltrate; (D) severe retinal inflammation, damage and atrophy; (E) Section of normal mouse retina; (F) EAU showing severe inflammation with large central area of retinal necrosis and loss of photoreceptors. Modified from (48) under copyright agreement with Elsevier Publishing license number 4887001207791.

Figure 4

Experimental Autoimmune Uveoretinitis (EAU) in the rat model. EAU was induced in Lewis rats using retinal S antigen as described in (49). The image is a toluidine blue-stained thin section of the inner retina showing perivascular inflammatory cells (large arrows, black) retained within the glia limitans (arrow heads, red). Cells which have entered the retinal parenchyma have pyknotic nuclei and appear to be undergoing apoptosis (small arrows, yellow). Image provided by permission of the authors.