Immunological aspects of neural grafting in the mammalian central nervous system

Cellular therapy for childhood neurodegenerative disease. Part I: rationale and preclinical studies

✓ Successful cellular replacement in the diseased human central nervous system (CNS) faces numerous hurdles. In this first installment of a 2-part review, the authors report on the preclinical challenges involved in preparing for a major Phase I trial investigating the safety of human neural stem cell transplantation in a lysosomal storage disorder. Specifically, they discuss choice of the ideal disease for treatment, best donor cell type and source for implantation, the in vitro and in vivo methods used to estimate safety and efficacy, the challenges to noninvasive tracking of cells after transplantation, and the unique issues related to the immunology of CNS cellular transplantation.

Intracerebral co-transplantation of liposomal tacrolimus improves xenograft survival and reduces graft rejection in the hemiparkinsonian rat

Immunosuppression remains a key issue in neural transplantation. Systemic administration of cyclosporin-A is currently widely used but has many severe adverse side effects. Newer immunosuppressive agents, such as tacrolimus (TAC) and rapamycin (RAPA), have been investigated for their neuroprotective properties on dopaminergic neurons. These drugs have been formulated into liposomal preparations [liposomal tacrolimus (LTAC) and liposomal rapamycin (LRAPA)] which retain these neuroprotective properties. Due to the slower release of the drugs from the liposomes, we hypothesized that co-transplantation of either LTAC or LRAPA within a xenogeneic cell suspension would increase cell survival and decrease graft rejection in the hemiparkinsonian rat, and that a combination of the two drugs may have a synergistic effect. 6-hydroxydopamine-lesioned rats were divided to four groups which received intra-striatal transplants of the following: 1) a cell suspension containing 400,000 fetal mouse ventral mesencephalic cells; 2) the cell suspension containing 0.63 microM LRAPA; 3) the cell suspension containing a dose of 2.0 microM LTAC; 4) the cell suspension containing 2.0 microM LTAC and 0.63 microM LRAPA. Functional recovery was assessed by amphetamine-induced rotational behavior. Animals were killed at 4 days or 6 weeks post-transplantation, and immunohistochemistry was performed to look at the expression of tyrosine hydroxylase and major histocompatibility complex classes I and II. Only the group receiving LTAC had a decrease in rotational behavior. This observation correlated well with significantly more surviving tyrosine hydroxylase immunoreactive cells compared with the other groups and significantly lower levels of immunorejection as assessed by major histocompatibility complex class I and II staining. This study has shown the feasibility of using local immunosuppression in xenotransplantation. These findings may be useful in optimizing immunosuppression in experimental neural transplantation in the laboratory and its translation into the clinical setting.

Behavioral characterization of a unilateral 6-OHDA-lesion model of Parkinson's disease in mice

Parkinson's disease (PD) is one of the most common neurodegenerative disorders. Several toxin-induced animals models simulate the motor deficits occurring in PD. Among them, the unilateral 6-hydroxydopamine (6-OHDA) model is frequently used in rats and has the advantage of presenting side-biased motor impairments. However, the behavioral consequences of a unilateral 6-OHDA-lesion have, so far, not been described in detail in mice. The aim of this study was to characterize mice with unilateral 6-OHDA-lesions placed in the median forebrain bundle using several motor behavioral tests in order to identify the most suitable predictor of nigral cell loss. Mice underwent various drug-induced (amphetamine- and apomorphine-induced rotation) and spontaneous motor tests (cylinder, rotarod, elevated body swing, and stride length test). The amphetamine-induced rotation test, the cylinder and the rotarod test were most sensitive and reliable in detecting loss of tyrosine hydroxylase-immunoreactive cells in the substantia nigra. This study demonstrates that substantial and stable unilateral 6-OHDA-induced lesions can be established in mice, and that these lesions can be functionally assessed using several different side-bias-based behavioral tests. This mouse model offers the opportunity to use transgenic mouse strains and study the interactions between genes of interest and toxins in relation to Parkinson's disease etiology in the future.

Intracerebral cytokine profiles in adult rats grafted with neural tissue of different immunological disparity

To understand graft rejection in cell based therapies for brain repair we have quantified IL-1beta, IL-2, IL-4, IL-10, IL-12p40, IFN-gamma and TNF-alpha mRNA levels using real-time PCR, at days 4, 14, and 42 post-transplantation, in rats engrafted with syngeneic, allogeneic, concordant and discordant xenogeneic neural tissues. In addition, in the discordant xenografts immunohistochemistry and in situ hybridization were applied to detect local expression of IFN-gamma, TNF-alpha, IL-10 and TGF-beta. Allografts remained non-rejected but expressed IL-1beta, TNF-alpha and IL-4 transcripts but not IL-12p40 and IFN-gamma. Xenografts demonstrated distinct cytokine profiles that differed from syngeneic and allogeneic grafts. Non-rejected discordant xenografts contained higher levels of TNF-alpha transcripts and lower levels of IL-2 transcripts than the rejected ones at day 42. Discordant xenografts displayed a stronger and earlier expression of IL-1beta and TNF-alpha, followed by T-helper 1 and T-helper 2 associated cytokine expression. The number of cells expressing mRNA encoding TNF-alpha and TGF-beta was significantly increased over time in the discordant group. In conclusion, the immunological disparity of the implanted tissue explains survival rates and is associated with different cytokine profiles. In allografts, a chronic inflammatory reaction was detected and in xenogeneic grafts a delayed hypersensitivity like reaction may be involved in rejection.

Experimental appraisal of the lack of antitumor natural killer cell-mediated immunosurveillance in response to lymphomas growing in the mouse brain

OBJECT

Natural killer (NK) cell-mediated immunosurveillance in the brain is currently obscure, in contrast with the intracerebral immune reaction of cytotoxic T lymphocytes (CTLs) to tumor cells. The goal of this study, in which a controlled tumor model was used, was to investigate a relationship between NK cells and major histocompatibility complex (MHC) Class I gene expression in intracerebral tumor-bearing hosts.

METHODS

A matched set of two cloned tumor cell lines (lymphoma+ and lymphoma-), which differ only in MHC Class I gene expression, was established from the parental YAC-1 cell line (a target widely accepted as being sensitive to murine NK cells). An in vivo rapid elimination assay (REA) was performed using tumor cells labeled with [125I] 5-iodo-2-deoxyuridine to evaluate intracerebral NK cell-mediated defense immunity. There was no difference in the in vitro growth rate and c-myc gene expression between lymphoma+ and lymphoma- cells. An in vitro cytotoxicity assay showed that the lymphoma+ cell line was sensitive to MHC Class I-restricted CTL-mediated lysis, whereas the lymphoma- line was refractory to it. Both were susceptible to NK cell-mediated lysis, comparable to the level shown by YAC-1 cells. Flow cytometry revealed that lymphoma+ reacted positively for cell-surface MHC Class I molecules, whereas lymphoma- had no reaction. Four- to 72-hour REAs, performed using either cell line, disclosed no clearance of radiolabeled tumor cells from the brain in independent groups of untreated and T cell-depleted mice; this contrasted with eradication of radioactivity from the lungs. In NK cell-depleted mice, however, there was no elimination of radiolabeled tumor cells from the brain or lungs. The MHC Class I expression on lymphoma+ cells was enhanced after intracerebral inoculation, rendering them less sensitive to NK cells. By contrast, lymphoma- cells remained negative for cell-surface MHC expression, being sensitive to NK cells and refractory to CTLs after intracerebralinoculation. These results indicate the absence of NK cell-mediated lytic activity in the brain. This allows even NK cell-sensitive tumor cells to escape intracerebral immunosurveillance.

CONCLUSIONS

These experiments have refined the information that the brain may lack NK cell-mediated defense immunity against intracerebrally growing tumors, representing a characteristic aspect of this immunologically privileged organ.

Cell transplantation as a treatment for retinal disease

It has been shown that photoreceptor degeneration can be limited in experimental animals by transplantation of fresh RPE to the subretinal space. There is also evidence that retinal cell transplants can be used to reconstruct retinal circuitry in dystrophic animals. Here we describe and review recent developments that highlight the necessary steps that should be taken prior to embarking on clinical trials in humans.

Analysis of neural stem cells by flow cytometry: cellular differentiation modifies patterns of MHC expression

Neural stem cells are currently considered very hopeful candidates for cell replacement therapy in neurodegenerative pathologies such as Parkinson's disease. Here we show that different cell types derived from neurospheres amplified in vitro can be identified by FACS analysis relying solely on physical parameters (FSC/SSC) or autofluorescence. Additionally, after treatment with a panel of inflammatory cytokines, neurospheres and their differentiated progeny were shown to express MHC antigens which could potentially cause transplant rejection. Astrocytes expressed the highest levels of MHC. Hence removing such cells prior to transplantation could potentially optimise transplant survival.

Astrocyte line SVG-TH grafted in a rat model of Parkinson's disease

The present review describes gene transfer into the brain using extraneuronal cells with an ex vivo approach. The mild immunological reactions in the central nervous system to grafts provided the rationale and empirical basis for brain-transplantation, to replace dying cells, of potential clinical relevance. Fetal human astrocytes were genetically engineered to express tyrosine hydroxylase, the rate-limiting enzyme for the synthesis of catecholamines. These cells were also found to produce constitutively and secrete GDNF and interleukins. Therefore, these cells may prove as a drug-delivery system for the treatment of neurological degenerative conditions such as Parkinson's disease (PD). The field of neuronal reconstruction has reached a critical threshold and there is a need to evaluate the variables that will become critical as the field matures. One of the needs is to characterize the neurochemical alterations in the microenvironment in the context of grafted-host connectivity. This review discusses the functional effects of the pharmacologically-active construct, which consists of astrocytes producing L-DOPA and GDNF. The striatum in PD that lacks the dopaminergic projection from the substantia nigra metabolizes and releases dopamine differently from normal tissue and may react to different factors released by the grafted cells. Moreover, neurochemicals of the host tissue may effect grafted cells as well. An understanding of the way in which these neurochemicals are abnormal in PD and their role in the grafted brain is critical to the improvement of reconstructive strategies using cellular therapeutic strategies.

A novel mode of immunoprotection of neural xenotransplants: masking of donor major histocompatibility complex class I enhances transplant survival in the central nervous system

To determine the role of major histocompatibility complex (MHC) class I in immunological rejection of neural xenotransplants, F(ab')2 fragments of a monoclonal antibody to porcine MHC class I were used to mask this complex on porcine fetal striatal cells transplanted into rat striata previously lesioned with quinolinic acid. Presence of MHC class I on the surface of porcine striatal cells was confirmed by fluorescence-activated cell sorting prior to F(ab')2 treatment. At three to four months post-transplantation, survival of F(ab')2-treated xenografts was assessed by means of donor-specific immunostaining and compared to that of untreated xenografts in non-immunosuppressed rats and in rats immunosuppressed with cyclosporine A. In this study, masking of donor MHC class I by F(ab')2 treatment resulted in enhanced xenografts survival compared to the non-immunosuppressed controls (graft survival rates, 52% and 7%, respectively; P < 0.005) at survival times up to four months. While xenograft survival in F(ab')2-treated animals was not significantly different from that in cyclosporine-treated rats (74% graft survival), mean graft volume in F(ab')2-treated animals was smaller than that in cyclosporine-treated animals (1.07 +/- 0.30 mm3 versus 3.14 +/- 0.51 mm3; P < 0.005). The cytoarchitectonic organization of the xenografts was similar in F(ab')2- and cyclosporine-treated animals, and grafts in both groups exhibited long distance target-directed axonal outgrowth. The pattern of immunoreactivity to porcine MHC class I in the xenografts corresponded to the regional distribution of donor glia. In xenografts undergoing rejection, infiltration with host inflammatory cells was restricted to necrotic graft remnants and spared the nearby host structures. We conclude that MHC class-I-restricted immune mechanisms play an important role in neural xenograft rejection and that masking of this complex on donor cells may provide a useful strategy for immunoprotection of neural xenografts.

Sequential intrastriatal grafting of allogeneic embryonic dopamine-rich neuronal tissue in adult rats: will the second graft be rejected?

An important issue in clinical neural grafting is whether a second instriatial allograft can survive well in a patient who has received an allograft before. In this study, the survival, immunogenicity and function of intrastriatal grafts of allogeneic or syngeneic embryonic dopamine-rich tissue in rats which had previously received either an intrastriatal allo- or syn-graft or sham injections were examined. The first graft tissue was taken from inbred Lewis or Sprague-Dawley rat embryos and grafted into an intact striatum of adult Sprague-Dawley rats subjected to a unilateral 6-hydroxydopamine lesion on the contralateral side. Eight weeks after the first transplantation, either allogeneic or syngeneic tissue was grafted as dissociated tissue into the dopamine depleted striatum. The function of the second grafts was assessed by rotational asymmetry at two different time points, i.e. eight and 14 weeks after the second transplantation. There were significant reductions of rotational asymmetry in all groups over time, but no significant difference between groups. Tyrosine hydroxylase immunocytochemistry was used to assess dopamine cell survival and graft size. Statistical analysis revealed no significant differnce in the mean number of tyrosine hydroxylase immunoreactive cells or the mean volume of the second grafts placed on the right side (lesioned side) between groups. Monoclonal antibodies were used to evaluate cellular immune reactions and the major histocompatibility complex class I and class II expression in and around grafts. No major histocompatibility complex class I expression was seen in any of the graft combinations. The expression of the major histocompatibility complex class II antigens was generally higher in patches in and around the second allograft of rats which had previously received an allograft than that in and around any other type of grafts. However, the expression of the major histocompatibility complex class II antigens was low throughout the grafts and did not appear as marked perivascular infiltrates. All the major histocompatibility complex class II positive cells displayed a microglia-like morphology, supported by the parallel microglia and macrophage-specific OX-42 immunostaining. The results show that there is no marked on-going immune reactions in or around the implantation site in any group fourteen weeks after a second transplantation. It may be concluded, therefore, that sequential allografting, using stereotaxic implantation of dissociated embryonic neural tissue into the striatal parenchyma, is possible to perform without a major risk of graft rejection, provided that an atraumatic technique is used.

Transient and long-lasting beneficial behavioral effects of grafts in the damaged hippocampus of rat

One week after receiving bilateral neurotoxic lesions of their dorsal hippocampi, adult Wistar rats were bilaterally grafted with fetal hippocampal tissue suspensions. The behavior of the animals was tested during a period of 5 months after grafting to determine changes in lesion-induced deficits. The transplants caused various behavioral effects with different time courses. Grafted animals showed an early, however, transient amelioration of behavioral deficits in a T-maze alternation task and they performed with a long-lasting improvement in the alcove-test. Transplant histology demonstrated high levels of AChE-activity in patches correlating with clusters or rudimentary layers of pyramidal neurons.

Human fetal dopamine neurons grafted in a rat model of Parkinson's disease: immunological aspects, spontaneous and drug-induced behaviour, and dopamine release

We have used a rat model of Parkinson's disease (PD) to address issues of importance for a future clinical application of dopamine (DA) neuron grafting in patients with PD. Human mesencephalic DA neurons, obtained from 6.5-8 week old fetuses, were found to survive intracerebral cell suspension xenografting to the striatum of rats immunosuppressed with Cyclosporin A. The grafts produced an extensive new DA-containing terminal network in the previously denervated caudate-putamen, and they normalized amphetamine-induced, apomorphine-induced and spontaneous motor asymmetry in rats with unilateral lesions of the mesostriatal DA pathway. Grafts from an 11.5-week old donor exhibited a lower survival rate and smaller functional effects. As assessed with the intracerebral dialysis technique the grafted DA neurons were found to restore spontaneous DA release in the reinnervated host striatum to normal levels. The neurons responded with large increases in extracellular striatal DA levels after the intrastriatal administration of the DA-releasing agent d-amphetamine and the DA-reuptake blocker nomifensine, although not to the same extent as seen in striata with an intact mesostriatal DA system. DA fiber outgrowth from the grafts was dependent on the localization of the graft tissue. Thus, grafts located within the striatum gave rise to an extensive axonal network throughout the whole host striatum, whereas grafted DA neurons localized in the neocortex had their outgrowing fibers confined within the grafts themselves. In contrast to the good graft survival and behavioural effects obtained in immunosuppressed rats, there was no survival, or behavioural effects, of human DA neurons implanted in rats that did not receive immunosuppression. In addition, we found that all the graft recipients were immunized, having formed antibodies against antigens present on human T-cells. This supports the notion that the human neurons grafted to the non-immunosuppressed rats underwent immunological rejection. Based on an estimation of the survival rate and extent of fiber outgrowth from the grafted human fetal DA neurons, we suggest that DA neurons that can be obtained from one fetus may be sufficient to restore significant DA neurotransmission unilaterally, in one putamen, in an immunosuppressed PD patient.