Virus-mediated delivery of MECP2 as a potential tool for the treatment of Rett syndrome

Typical Rett syndrome (RTT) is a paediatric neurological disorder caused in >95 % of cases by loss-of-function mutations in the X-linked gene, methyl-CpG-binding protein 2 (MECP2). The gene product, MeCP2, is a widely expressed nuclear protein that is especially abundant in postmitotic neurons of the central nervous system (CNS). Knocking out Mecp2 function in mice recapitulates many of the overt neurological features seen in RTT patients and provides a very useful model for testing potential therapeutic applications. The absence of a curative therapy together with the monogenecity of the disorder and established reversibility of the phenotype in mice suggest that replacement of the MECP2 gene is a potential therapeutic option worthy of exploration. In this study I used several viral vectors to test the potential of gene therapy in RTT mice. First, I generated different viral vectors which can express tagged-MeCP2 under the control of ubiquitous and cell-type specific promoters. Secondly, I assessed the ability of these vectors to deliver the Mecp2 transgene into Mecp2 knockout mice at both neonatal and adult stages of development. I then aimed to investigate the effect of exogenously delivered Mecp2 on RTT-like phenotypes. The results demonstrated that lentiviral vectors were able to effectively deliver an RFP-tagged Mecp2 minigene into neurons of Mecp2-null mice both in vitro and in vivo. Exogenous Mecp2 was targeted to the nucleus and displayed heterochromatin localization with no evidence of ectopic expression. Use of the synapsin1 (syn1, neuron-specific) promoter resulted in cellular levels of Mecp2 equal to 85 ± 0.1% of endogenous protein levels, whereas the phosphoglycerate kinase (PGK) promoter produced cellular levels of exogenous Mecp2 at a relatively high levels (210 ± 0.1 % of endogenous levels). Direct brain injection of Lentiviral vector was able to deliver exogenous Mecp2 into the CA1 region of the hippocampus and to produce high transduction efficiency around the injection sites but with limited spread. The early mortality of the injected mice precluded assessment of the functional consequences of exogenous Mecp2 expression. However, assessment of the cellular morphology was possible and this analysis revealed delivery of exogenous Mecp2 to normalise neuronal nuclear volume deficits seen in the Mecp2stop/y mice from 86 ± 0.1 % of WT values to 100 ± 0.04 % of WT levels. At the molecular level, I showed that exogenous Mecp2 3 becomes phosphorylated at serine 421 under basal conditions and that the level of phosphorylation of exogenous Mecp2 is disproportionately higher (5.5 ± 0.4 times) than that seen for endogenous Mecp2. I also showed the Mecp2 overexpression in WT neurons is associated with a reduction in the cellular levels of total histone 4 (78 ± 0.01 % of the endogenous level) and a parallel reduction in cellular levels of acetylated histone 4 (79 ± 0.01 % of the endogenous levels). In second phase of experiments, I showed that the single stranded Adeno-associated virus (ssAAV)-based vector with chicken beta actin (CBA) promoter and encapsulated with capsid of AAV serotype 9, was able to efficiently deliver exogenous MECP2 into the brain of Mecp2-null as well as WT neonatal mice after intracranial (IC) injection. In contrast to lentiviral vectors, there was widespread transduction of cells throughout the nervous system with transduction efficiency varying between 6.8 ± 2.3 % and 41.5 ± 11.3% of all cells dependent on the brain region. The transgene was mostly expressed in neurons which represented 67 ± 11.8 % to 98 ± 0.8 % of all transduced cells. ssAAV9 vector expressed exogenous MeCP2 at near-physiological levels (100-125 % of endogenous levels). At the cellular level, exogenous MeCP2 was able to rescue the neuronal nuclear volume of Mecp2-null mice (69 ± 0.02 % of WT values) to WT comparable values (97 ± 0.03 % of WT values). At the organismal levels Mecp2-null mice treated with ssAAV9/MECP2 showed extended survival (median survival of 16.7 weeks compared to 9.3 weeks for the GFP-treated control) and also displayed a modest, but significant, reduction in the RTT-like phenotype severity score compared to the GFP-treated control group. The most robust improvement reported in this study was in the locomotion activity (velocity and total distance moved in the open field test and in performance on a forced motor task). Interestingly, WT mice receiving neonatal injections of ssAAV9/CBA-MECP2 did not show any significant deficits, suggesting a tolerance for modest MeCP2 overexpression. In a further experiment, I showed that the self-complementary AAV 9 (scAAV9) vector with an Mecp2-endogenous core promoter fragment was able to deliver exogenous MECP2 into the brain of neonatal mice after intravenous (IV) or (IC) injection. Brain transduction efficiency was 8 - 12 % after IV injection and 48 - 68 % after IC injection in neonatal mice. Cellular levels of exogenous MeCP2 were between 1.4-1.8 times the endogenous levels. At the organismal level, 4 scAAV9/MECP2-injected mice displayed an overt hindlimb motor dysfunction which was observed 3 and 5 weeks post-injection after IV and IC injection respectively. The stereotyped hindlimb dysfunction suggested a toxicity issues with this vector and examination of the lumbar segment of the spinal cord confirmed evidence of axonal degeneration in the dorsal columns. IV injection of scAAV9/MECP2 into RAG-/- knockout mice (immunocompromised) displayed hindlimb motor dysfunction similar to that observed with Mecp2stop/y mice suggesting that the adaptive immune response is not likely to be involved in the pathogenesis of this phenotype. MECP2stop/y mice treated with scAAV9/MECP2 displayed higher RTT-like phenotype severity score than GFP-treated controls which is probably due to the effect produced by the hindlimb motor dysfunction on regular RTT-like phenotype (gait, mobility and hindlimb clasping). In summary, I have demonstrated the successful application of lentiviral and AAV2/9 vectors to deliver exogenous MECP2 both in vitro and in vivo. I showed that lentiviral vectors are unlikely to be useful for global brain delivery of MECP2 due to limited spread of the virus. However the lentiviral vectors I developed are potentially useful where localized brain injection is desirable. The main translational finding is the first, at the proof of concept level, demonstration of therapeutic benefits (including enhanced survival) of exogenously delivered MECP2 using the ssAAV9/CBA-MECP2 vector. I also however, identify potential toxicity issues of exogenous MECP2 delivery whereby a scAAV9 vector was found to produce overt neuromotor deficits. Overall, my data supports the potential of gene therapy in RTT but also emphasises the importance of issues including careful vector design, choice of delivery methods and the timing of treatment in any future clinical translation