A single intramuscular injection of rAAV-mediated mutant erythropoietin protects against MPTP-induced parkinsonism

Abstract

Erythropoietin (Epo) is neuroprotective in a number of preparations, but can lead to unacceptably high and even lethal hematocrit levels. Recent reports show that modified Epo variants confer neuroprotection in models of glaucoma and retinal degeneration without raising hematocrit. In this study, neuroprotective effects of two Epo variants (EpoR76E and EpoS71E) were assessed in a model of Parkinson's disease. The constructs were packaged in recombinant adeno-associated viral (rAAV) vectors and injected intramuscularly. After 3 weeks, mice received five daily injections of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and were killed 5 weeks later. The MPTP-lesioned mice pretreated with rAAV.eGFP (negative control) exhibited a 7- to 9-Hz tremor and slower latencies to move on a grid test (akinesia). Both of these symptomatic features were absent in mice pretreated with either modified Epo construct. The rAAV.eGFP-treated mice lesioned with MPTP exhibited a 41% reduction in tyrosine hydroxylase (TH)-positive neurons in the substantia nigra. The rAAV.EpoS71E construct did not protect nigral neurons, but neuronal loss in mice pretreated with rAAV.EpoR76E was only half that of rAAV.eGFP controls. Although dopamine levels were normal in all groups, 3,4-dihydroxyphenylacetic acid (DOPAC) was significantly reduced only in MPTP-lesioned mice pretreated with rAAV.eGFP, indicating reduced dopamine turnover. Analysis of TH-positive fibers in the striatum showed normalized density in MPTP-lesioned mice pretreated with rAAV.EpoS71E, suggesting that enhanced sprouting induced by EpoS71E may have been responsible for normal behavior and dopaminergic tone in these mice. These results show that systemically administered rAAV-generated non-erythropoietic Epo may protect against MPTP-induced parkinsonism by a combination of neuroprotection and enhanced axonal sprouting.

Figure 1. Modified Epo constructs are non-erythropoietic

Mice were injected intramuscularly with an rAAV vector harboring native Epo, eGFP control or modified variants EpoR76E or EpoS71E. Trunk blood was collected 3 weeks following the injection and hematocrit measured. Most of the mice injected with rAAV.Epo died; those that survived had significantly elevated hematocrit levels. Mice injected with rAAV.eGFP or rAAV.EpoR76E had normal hematocrit levels. The rAAV.EpoS71E group had slightly (13.9 ± 4.0%) elevated hematocrit that was statistically significant. ***P < 0.001 vs. rAAV.eGFP-treated groups (t-test).

Figure 2. Modified Epo variants protect against MPTP-induced parkinsonism

(a) Mice were placed in a FPA and allowed to explore freely for 5 min. Minute variations in force were recorded on a millisecond time scale. MPTP induced a robust 7- to 9-Hz tremor in control mice pretreated with rAAV.eGFP. Mice injected with either of the modified Epo constructs did not exhibit MPTP-induced tremor. (b) Mice in the rAAV.eGFP control group and lesioned with MPTP were slower to move all four paws when placed on a horizontal grid. Mice pretreated with rAAV.EpoS71E or rAAV.EpoR76E did not exhibit this MPTP-induced akinesia. **P < 0.01, *P < 0.05 vs. rAAV.eGFP/saline-treated group (t-test).

Figure 3. rAAV.EpoR76E protects against MPTP-induced destruction of neurons in the SNc

(a, d) MPTP destroyed approximately 40% of the TH-positive neurons in the SNc in control mice pretreated with rAAV.eGFP and induced a similar-sized lesion in the rAAV.EpoS71E group (b, e). (c, f) Mice pretreated with rAAV.EpoR76E were partially protected against the MPTP lesion, exhibiting a approximately 20% loss of TH-positive SNc cells. (g–i) Sections counterstained with cresyl violet showed that lesioned areas were largely devoid of large, Nissl-positive cell bodies. This confirms that the lesion represents neurons destroyed by MPTP and not a metabolic downregulation of TH synthesis. (j) Stereological quantification of the images presented in a–f. ***P < 0.001, **P < 0.01, *P < 0.05 vs. respective saline-treated group (t-test).

Figure 4. Modified Epo variants normalize dopamine turnover following MPTP lesion

(a) Intracellular striatal dopamine measured by HPLC 8 weeks following the last injection of saline or 18 mg/kg MPTP. There were no significant differences in dopamine across groups, possibly due in part to axon regeneration and sprouting of new terminals following MPTP lesion. (b) The dopamine metabolite DOPAC was significantly lower in MPTP-lesioned mice pretreated with rAAV.eGFP, compared with their saline-treated counterparts, suggesting reduced dopamine turnover. In contrast, mice pretreated with either of the modified Epo variants had normal DOPAC regardless of lesion status. *P < 0.05 vs. rAAV.eGFP/saline-treated group (t-test).

Figure 5. rAAV.EpoS71E pretreatment increases TH-positive optical density following MPTP lesion

(a–f) Density of TH-positive fibers was assessed in striatal sections 8 weeks following administration of MPTP or saline. MPTP-lesioned mice pretreated with rAAV.eGFP showed a characteristic loss of TH-positive fibers compared with saline-treated controls (a, d). Mice pretreated with rAAV.EpoR76E showed a similar magnitude of MPTP lesion (c, f). In contrast, mice in the rAAV.EpoS71E group exhibited no loss of TH-positive fibers after MPTP treatment, presumably due to increased axonal sprouting and regeneration (b, e). Confocal images of striatal sections show increased density of TH-positive fibers in MPTP-lesioned mice pretreated with either rAAV.eGFP (g) or rAAV.EpoS71E (h). (i) Densitometric quantification of the images presented in a–f. *P < 0.05 vs. rAAV.eGFP/saline-treated group (t-test).