Our data clearly show that IVIg treatment has an impact on various immune parameters in mice, confirming the immunomodulatory action of IVIg in the periphery. Indeed, systemic administration of IVIg led to the presence of human IgG at the surface of circulating leukocytes, induced a significant decrease in the CD4+/CD8+ T-cell ratio and increased the Treg percentage. In the present study, we have also assessed the state of the brain DAergic system using a combination of validated markers. However, our results suggest that immunomodulating treatment with IVIg did not translate into neurorestoration of the denervated nigrostriatal DAergic pathway after an acute MPTP insult. Our observations rather suggest potentially negative consequences of IVIg treatment on certain components of the DAergic system, as well as on the health status of the treated animals. The vast majority of preclinical studies aiming to test new compounds for PD are tested in animal models prior to the injury (for example, MPTP treatment), thereby probing the neuroprotective properties of the potential therapeutic agents. We opted instead for a neurorestoration study design, in which the injections of IVIg began 20 hours after the last MPTP injection. MPTP-induced neurodegeneration is still ongoing at that time, as DAergic denervation stabilizes approximately 7 days after initial MPTP insult [63–65]. Nevertheless, such post-MPTP treatment paradigm is more compatible with an eventual clinical use of IVIg in human PD, which would occur after the diagnosis, when neurodegeneration processes are already engaged .
Treg cell adoptive transfer has been previously reported to protect from MPTP-induced nigrostriatal denervation [57–59] in acute MPTP mouse models. In these studies, the amount of Tregs needed to achieve neurorestoration using adoptive transfer ranged between 0.5×106 and 3.5×106 injected into the tail vein 12 hours following the last MPTP injection [57–59]. We also observed a rise in Treg percentage among the CD4+ population, reaching up to 9% in the spleen of IVIg-vehicle mice, after a 14-day treatment. However, this increase in Tregs following IVIg administration did not reach the 16 to 20% CD4+ Treg proportion previously reported [53, 54]. Nevertheless, despite the significant rise in peripheral Tregs, IVIg treatment did not translate into measurable neurorestorative effects. The lack of beneficial effects could be explained by the fact that the rise in Tregs following the initiation of IVIg treatment might have been too slow to allow a sufficient exposure to Tregs to produce any benefits. We also observed a significant increase of Treg percentage after MPTP administration with no additive effects of IVIg. This is in accordance with Rosenkranz and colleagues, who reported a higher suppressive activity of Tregs in PD and AD patients and an increased Treg number associated with aging . Finally, Ramakrishna and colleagues associated the long-term regulation of CNS inflammatory responses to the induction of iCOS+CD4+ T cells , which were left unchanged after the present IVIg treatment. The absence of neurorestorative effects of IVIg could thus also be explained by the lack of expansion of the iCOS+CD4+ T cells or the Treg population in the MPTP-treated groups.
Injections of IVIg resulted in a mild but significant decrease in the CD4+/CD8+ T-cell ratio. Such decreases are also observed in IVIg-treated patients , suggesting it may be a clinically relevant index of IVIg efficacy. Interestingly, a significant decrease in CD4+/CD8+ ratio is observed in PD patients as well [69–71], possibly accounted for by an increased susceptibility to apoptosis observed in CD4+ T cells, consequent of Fas overexpression . IVIg has been reported to modulate the level of expression of Fas and FasL and to inhibit FasL-dependent apoptosis, in both in vivo and in vitro studies [34, 73–75]. This action of IVIg on the Fas/FasL pathway could have been translated into neurorestoration. However, while being consistent with an immunoregulatory action, the effect of IVIg on the CD4+/CD8+ T-cell ratio was not associated with beneficial post-MPTP outcomes on various DAergic markers.
The lack of efficacy of IVIg may also reflect the poor CNS access owing to the presence of the blood–brain barrier. However, our data rather suggest that IVIg displayed significant central bioavailability after systemic administration. Indeed, a fraction of intraperitoneally administered IVIg was detected in the striatum of treated mice using a specific ELISA, consistent with a previous report where peripherally administered IVIg was also detected in APP/PS1 mouse brain using immunohistochemistry . A number of studies have reported data consistent with the penetration of a fraction of systemically administered antibodies into brain tissues leading to central therapeutic effect [38, 77, 78]. Interaction between Fc gamma receptor (FcγR) and immunoglobulins is essential for the initiation of cellular and humoral responses. In the CNS, FcγR are expressed on endothelial cells, neurons, microglia, oligodendrocytes and astrocytes (as reviewed in ) and the IVIg migration to critical regions of the brain, such as the striatum and SNpc in PD, might act as a central immunomodulating agent. A previous report showed that approximately 30% of pigmented SNpc neurons were IgG-positive  in PD patients but not in controls. This suggests that IgG can access the brain during the course of the disease. However, we found no increase in striatal IgG content in MPTP-treated animals. The amount of human IgG detected in the brain of treated mice suggests that low central bioavailability is unlikely to be the sole reason for the lack of efficacy of IVIg in restoring the DAergic pathways.
After systemic injection, MPTP produces a reproducible lesion of the nigrostriatal DAergic pathway by causing oxidative stress, mitochondrial damage and neuronal cell death, as in idiopathic PD. Validation of disease-modifying treatments before clinical trial initiation is therefore often performed in MPTP-treated rodent models [81, 82]. However, these models are not without important limitations . First, the MPTP model used here does not generate a massive degeneration (−30 to 40% of TH-positive cells), which is required for clinically detectable motor symptoms in humans (−50 to 60%) [84, 85]. This explains, at least in part, why motor symptoms in the MPTP mouse model are insufficiently reliable for systematic assessment [83, 86–89] and were not evaluated here after IVIg treatment. To investigate the symptomatic effects of IVIg, the use of the more expensive MPTP monkey model should be considered instead [83, 90]. Second, the acute mouse MPTP model does not replicate α-synucleinopathy or Lewy bodies, which are pathognomonic of PD [91–93]. The use of other models such as the chronic infusion MPTP models or transgenic mice overexpressing human α-syn might be helpful for these purposes [81, 93–95]. Third, the response of a mouse model to human IVIg may differ from humans. Indeed, the absence of positive outcome in our study might be the result of inadequate interactions between human IgG and mouse FcγR, a hypothesis only testable with the use of mouse IgG. However, given that over 2,500 mice would have been required to generate the ~5 g IgG used in this study, murine IgG is scarcely used in preclinical investigations. In a passive model of idiopathic thrombocytopenic purpura, human IVIg and purified mouse IgG shared the same kinetics to restore platelet counts, thus validating the use of human IVIg to study human therapy in mouse models . Since monomeric human IVIg is well tolerated in mice, mouse models of numerous diseases are now routinely used to investigate its efficacy as well as its mechanisms of action [35, 53, 97, 98].
The unexpected deleterious effect of IVIg on TH expression is an intriguing observation that is particularly challenging to explain. On the one hand, a plethora of compounds such as nicotine, cannabinoid agonists and progesterone receptor isoforms [99–101] have been shown to modulate TH expression without obvious harmful effects on the DAergic system. Similarly, our data suggest that IVIg regulates TH expression at the protein or RNA levels. On the other hand, the observed decrease in striatal TH protein levels associated with a trend toward decreased catecholamines, serotonin, nigral TH-positive and total neurons can also be interpreted as a deleterious effect of IVIg on the murine DAergic system. Although acute MPTP administration does not lead to α-syn-positive nigral inclusions [91–93], α-syn-deficient transgenic mouse models are more resistant to MPTP, suggesting a possible implication of α-syn in the MPTP toxicity [102–106]. Increased autoantibodies to α-syn are present in the sera of PD patients [107, 108], and stereotactic injection of human IgG purified from the sera of PD patients into mice SNpc produces a FcγR-dependent microglial activation and a 40% TH-positive cell loss in the SNpc . Since natural anti-α-syn antibodies in IVIg preparations have been recently identified , it is tempting to speculate that IVIg could have modulated the nigrostriatal toxicity of MPTP by binding to α-syn.