Parkinson's disease (PD), the second most prevalent neurodegenerative disease, is characterized by motor symptoms of parkinsonism associated with Lewy bodies (LBs) and selective loss of dopaminergic (DA) neurons in the substantia nigra pars compacta (SNc) [1]. The etiology of PD is multifactorial, involving both genetic and environmental factors, and approximately 5 % to 10 % of PD cases are caused by genetic abnormalities [2]. Although most PD occurs sporadically with several possible cellular mechanisms, such as oxidative stress and mitochondrial dysfunction, both sporadic and familial PD have the same pathological hallmark: the toxic oligomers and insoluble aggregates formed by misfolding of α-synuclein (α-syn) are the main component of LBs, leading to the DA neurons degeneration in the SNc [3], [4], [5]. Multiple factors caused either by genetic mutations, such as A53T, A30P, E46K, H50Q, G51D, and A53E mutations in α-syn (SNCA), or post-translational modifications such as the phosphorylation of α-syn at the Ser129 site, induce the transformation of the native state of α-syn into the pathogenic aggregates [6,7]. Meanwhile, the elevated levels of the wild-type of α-syn alone have been widely reported to be sufficient to cause Parkinson's-like symptoms [8]. Currently, the main strategy of PD treatment remains to replenish the level of dopamine through the exogenous prodrug levodopa (L-DOPA) decarboxylation, which alleviates clinical symptoms but does not delay PD development [9]. Based on the current findings, developing a strategy to reduce the expression of α-syn in the brain may well be a potential therapeutic avenue that could slow the progress of PD.

Indeed, two major types of oligonucleotide drugs are currently being developed as therapeutic platforms for the reduction of SNCA expression: RNA interference (RNAi) and RNase H-dependent single-stranded antisense oligonucleotides (ASOs), either in preclinical studies in rodents and nonhuman primates or in drug development [10], [11], [12], [13]. ASO, a single-stranded gapmer that mediates sequence-specific cleavage of different types of RNAs, including mRNA and non-coding RNAs, has been a reliable and well-established method for treating neurodegenerative diseases like spinal muscular atrophy [14]. As one of the most promising treatments to reduce α-syn production, ASO-based PD gene therapy remains an active area of research. In the previous study, we participated in the design and screening of a specific ASO sequence targeting human SNCA, and then to achieve a more stable and efficient potential, we upgraded the ASO to a heteroduplex oligonucleotide (HDO) consisting of DNA/locked nucleic acid (LNA) gapmer and its complementary RNA (coRNA) conjugated to α-tocopherol (Toc-HDO), accompanied by in vivo assessment to rescue PD-like symptoms [15,16]. Although both the ASO and Toc-HDO we designed have demonstrated an ability to decrease the expression of α-syn and ameliorate DA neuron degeneration in A53T mice, they were administered by traumatic intracerebroventricular (ICV) injection, which is not conducive to clinical application. Meanwhile, in ICV injections, side effects caused by diffuse silencing of whole-brain SCNA are almost inevitable. Thus, there is an urgent need to develop α-syn HDO delivery strategies that are simultaneously safe and effective in targeting DA neurons in SNc through peripheral administration.

Artificial non-viral biomimetic nanoparticles (NPs) offer a variety of design, synthesis, and formulation possibilities, which can be used to protect and deliver ASOs into the brain [17], [18], [19]. In the recent decade, cell membranes have been continuously imitated and utilized to modify NPs to improve their biological properties, especially for treating neurological diseases [20,21]. Cerebrovascular endothelial cells (CECs), a constitutive component of the blood-brain barrier (BBB), can serve as a raw material for the fabrication of cell-membrane-coated nanoparticles [22,23]. Engineering CEC membranes with targeting ligands, while simultaneously crossing the BBB and precisely binding to DA neurons, may enhance HDO accumulation for PD therapy.

Herein, to further optimize the potency and accumulation of HDO, we replaced the complementary RNA strand in the Toc-HDO with phosphodiester (PO) bonded DNA with cholesterol-conjugated at 5′-terminus, creating a DNA/DNA double-stranded molecule, named Chol-HDO (coDNA). In vitro study showed that Chol-HDO (coDNA) silenced α-syn mRNA more powerfully compared to the parent ASO and Toc-HDO (coRNA), accompanied by the attenuation of α-syn aggregation. To avoid traumatic ICV injection while achieving targeted DA neuronal delivery, we developed a brain-targeted biomimetic nanodecoy for the treatment of PD. In this biomimetic system, Chol-HDO (coDNA) nanoparticles (Chol-HDO@NPs) were prepared by the nanoprecipitation method, and the natural CECm was used as the carriers and modified with the DSPE-PEG2000-levodopa that was synthesized through an amide formation reaction of DSPE-PEG2000 and L-DOPA to form engineering CEC membranes (L-DOPA-CECm) (Scheme 1). The L-DOPA in the carriers is decarboxylated into dopamine in the brain which can specifically bind to dopamine receptors expressed in DA neurons. Subsequently, Chol-HDO@NPs were loaded into the carriers to generate DA-targeted biomimetic nanodecoy Chol-HDO@LMNPs with extremely high HDO loading yields. Chol-HDO@LMNPs performed better than free Chol-HDO and Chol-HDO@NPs in drug release, neuroprotective effects, and BBB penetration. Importantly, the in vivo results demonstrated that tail vein injection of Chol-HDO@LMNPs could target DA neurons in the brain and attenuate the loss of DA neurons in SNc by inhibiting abnormal α-syn aggregation, as well as ameliorate motor deficits in a PD mouse model. Meanwhile, Chol-HDO@LMNPs exhibited no toxicity in A53T mice. Our results provide a promising peripheral delivery platform of L-DOPA-CECm nanodecoy loaded with Chol-HDO (coDNA) targeting DA neurons in PD therapy.