The survival of donor cells does not only depend on graft integration, allogeneic and xenogeneic donor cells will also need to evade immune rejection by if they are to suvive in the subretinal space. The subretinal space exhibit features of immune-privileged sites, characterised by the ability of foreign tissue to survive and induction of systemic immune deviation towards the foreign antigens expressed in the subretinal space (Streilein et al. 2002). Furthermore, there is evidence to suggest that the RPE (i.e. the donor tissue) may also be immune-privileged at different levels (Farrokh-Siar et al.
1999; Rezai et al. 1999; Rezai et al. 1997; Zamiri et al. 2004). The surgical procedure itself may trigger non-specific inflammatory response and the various chemokines and cytokines released during this initial response may undermine the immune-privileged features of the donor RPE. As with all differentiated cells, RPE constitutionally express class I major histocompatibility complex (MHC) antigens. However, class II MHC antigens and mRNA of various chemokines (e.g. RANTES and MCP-1), cytokines (e.g.
interleukins 6, 8 and 15) and cytokine receptors may be induced after exposure to interferon-γ and tumour necrosis factor-α (Hollborn et al. 2000; Enzmann et al. 2001b;
Hollborn et al. 2001). In the following sections, the immunological reactions to RPE grafts in different animal models or combinations of donor-host discordance will be discussed separately as they may have different levels of immune privilege and immunogenic potentials.
2.2.4.1 RPE grafts into mice and rats hosts
The extent of immune privilege in the subretinal space of transgenic mice models has been investigated previously (Wenkel and Streilein 1998). Both cell-associated and soluble antigens injected into the subretinal space can actively suppress systemic delayed-type hypersensitivity (DTH) reaction, confirmed by adoptive transfer assay of splenocytes (Wenkel and Streilein 1998). Such immune deviation was abolished when the RPE was damaged by sodium iodate (Wenkel and Streilein 1998). This suggests that submacular space is unlikely to retain its immune privilege after RPE transplantation given that the host RPE can be damaged by the disease process or the surgery itself.
Furthermore, despite evidence of persistence of immune deviation (failure to acquire DTH) to P815 tumour cells and soluble antigens in the subretinal space, the subretinal tumour graft was eliminated by an unknown mechanism after 2 weeks (Wenkel et al.
1999). This observation implies that the rejection process in subretinal space was not
induced by the T-cells that mediate donor-specific DTH (Wenkel et al. 1999). Unlike the anterior chamber, immune deviation induced by subretinal placement of antigen does not lead to immune privilege in the subretinal space. Although mice have been used for immunological studies, most transplantation experiments have been performed in rats. Below is a summary of previous works that have investigated immune reaction to allogeneic and xenogeneic RPE grafts into rat subretinal space.
RPE allograft can survive for up to 12 months in the RCS rats without immunosuppression (Li and Turner 1991). Donor RPE from Sprague-Dawley rats sustained outer nuclear layer for 6 months before age-related donor cell loss lead to gradual decline in the rescue effect (Li and Turner 1991). In contrast, Zhang and Bok (1998) demonstrated slow decline, during the first 2 months, in the ability of donor RPE allograft to rescue photoreceptors when there is mismatch in major histocompatibility complex (MHC) II or both MHC I and II between the recipient and the donor rats. The decline in rescue was accelerated when the host rat was challenged with donor spleen cells (Zhang and Bok 1998). Despite the lack of lymphocytes, Zhang and Bok (1998) concluded that systemic immunity to subretinal RPE allograft was the cause of delayed loss of RPE function.
Cow, pig and human donor RPEs have also been xenotransplanted into rat eyes.
Although cryopreserved bovine RPE was not rejected in rat subretinal space at 3 months (Durlu and Tamai 1997), fresh porcine RPE was rejected by 2 months after grafting (Grisanti et al. 2002). The rejection process was accompanied by loss of outer nuclear layer thickness and infiltration of pigment-loaded, ED1-positive cells. These were probably resident macrophages from within the retina (Grisanti et al. 2002).
Interestingly, there was no lymphocytic infiltration and the same graft was rejected when placed in subcutaneous tissue by a DTH response. Abe et al. (1999) demonstrated increased expression of interleukin 1 and 6 by xenografting cultured human RPE into rats. However, it is still not known what role these interleukins may play in immune rejection. Spontaneously immortalised human RPE cell line, ARPE-19 have also been shown to maintain outer nuclear layer and cortical visual function beyond 6 months in immunosuppressed RCS rats (Coffey et al. 2002). Inner retinal changes in the natural history of RCS rats have also been arrest for up to 10 months of age by early xenograft of ARPE-19 cells (Wang et al. 2005b). However, McGill et al. (2004) found that despite rescue of spatial vision in RCS rats by ARPE-19 cells, there is still gradual decline in vision which may be due to intrinsic properties of the graft or immune rejection. More recently, Wang et al. (2005b) demonstrated that ARPE-19 cells were
rapidly lost in the subretinal space of immunosuppressed RCS rats with only one fifth of the cells injected detectable in the subretinal space at 2 weeks. By 28 weeks, only 0.2%
of the cells remained in the subretinal space. The mismatch between the declining number of grafted ARPE-19 cell and the persistence of anatomical and functional rescue raised the possibility that diffusible factors may be primarily responsible for the action of transplanted RPE (Wang et al. 2005b). Furthermore, degree of visual rescue cannot be used as a measure of graft survival.
2.2.4.2 RPE grafts into rabbits and cats hosts
Rabbit models have also been used extensively to study immune rejection of RPE grafts. Activated and non-activated RPE allograft from pigmented to albino New Zealand rabbits are both associated with increased level of interleukin 6 in the vitreous during the first 2 weeks (Enzmann et al. 2000). Although the relationship between interleukin 6 and RPE graft rejection is unknown, serum interleukin 6 levels have been measured for identifying early local inflammatory reactions after liver transplantation (Kunz et al. 1998). Crafoord et al. (1999) showed that the RPE allograft rejection at 6 months in New Zealand white rabbits was characterised by dispersion of the melanin pigment in the subretinal space and throughout the neuroretina, phagocytosis of pigment by macrophage, Muller glial cells and host RPE, absence of lymphocyte, loss of the adjacent photoreceptor outer and inner segments, and reduced thickness of outer nuclear layer. Immunosuppression of these rabbits with cyclosporine did not prevent or reduce this inflammatory response (Crafoord et al. 2000). A more rapid rejection of porcine fetal RPE xenograft in rabbits was described by Del Priore et al. (2003a) with around 10% of the grafted cells detectable at 12 weeks even with triple immunosuppression using prednisone, cyclosporine-A and azathioprine. Similar to allograft rejection, porcine xenograft elicited a macrophage response in the absence of lymphocytes. There was also no vascular leakage on fluorescein angiography during rejection. Human fetal RPE xenograft transplanted as a monolayer sheet (Sheng et al. 1995) or cell suspension (He et al. 1993; Gabrielian et al. 1999b; Lai et al. 1999) into rabbits also induced macrophage migration within 1 to 3 months. Gabrielian et al. (1999b) demonstrated initial retinal glial cell response within the first 2 weeks at the site of human fetal xenograft and around the retinotomy. By 3 weeks, donor cell number has significantly reduced and by 4 weeks, both retina and choroid were involved in inflammatory response dominated by macrophages. In green fluorescent protein (GFP)-labelled human RPE xenograft to rabbits, local immunosuppression with intravitreal
cyclosporine delayed rejection (Lai et al. 2000). The use of intravitreal tacrolimus resulted in survival of cultured adult human RPE xenograft in rabbits for up to 1 year (Wang et al. 2002). Human fetal RPE on collagen, fibrinogen or biodegradable polymer as substrates induced choroidal inflammatory response when xenografted into rabbits.
(Bhatt et al. 1994; Oganesian et al. 1999; Rezai et al. 2000). A short-term follow up study (7 days) in feline RPE allograft did not reveal any rejection (Wang et al. 2004).
2.2.4.3 RPE grafts into pigs and monkeys hosts
Porcine RPE allograft induced infiltration of macrophage in the subretinal space of non-immunosuppressed pigs by 9 days but the graft was not rejected at 3 months (Del Priore et al. 2004). The authors suggested that these macrophages migrated into the subretinal space to clear the cell debris from degenerated RPE which did not have the opportunity to attach to the BM. Human fetal RPE xenograft to non-immunosuppressed monkeys have revealed 30% to 60% rate of rejection depending on the site of graft (Berglin et al.
1997). Interestingly, foveal grafts were rejected more often than extramacular grafts and rejection occurred in one subretinal location and not in another of the same or contralateral eye (Berglin et al. 1997).
2.2.4.4 Approaches to reduce immune rejection
The role of the host immune system in limiting the success of allogeneic or xenogeneic transplants cannot be overemphasized and graft rejection remains a major obstacle to successful RPE transplantation. Modification of the donor cell to reduce its immunogenicity and immune modulation of the host animal may minimise graft loss due to rejection. Factors that may influence rejection includes histocompatibility matching (Zhang and Bok 1998), culture and cryopreservation of donor cells (Durlu and Tamai 1997; Valtink et al. 1999a), cytokine-induced immune activation of the donor RPE (Kohen et al. 1997; Enzmann et al. 2000), donor RPE cell transfection with genes expressing immunosuppressive cytokines (Enzmann et al. 2001b), graft tissue architecture as a monolayer sheet or a cell suspension (Wenkel and Streilein 2000), type of substrate used to support the tissue (Bhatt et al. 1994; Gabrielian et al. 1999a; Rezai et al. 2000; Kiilgaard et al. 2002), break down in blood-retinal barrier due to disease and surgery (Wenkel and Streilein 1998), local or systemic immunosuppression (Crafoord et al. 1999; Lai et al. 2000; Lund et al. 2001a; Wang et al. 2002; Del Priore et al. 2003a) and donor-specific challenge (Jiang et al. 1994).