The delivery of therapeutic genes to the brain presents a significant challenge due to the blood-brain barrier (BBB). Adeno-associated viruses (AAVs) have emerged as promising vectors for gene therapy, offering the potential for non-invasive delivery to the central nervous system (CNS). However, the effectiveness of AAVs varies significantly across species. While numerous AAV variants have demonstrated efficacy in rodent models, translating this success to non-human primates (NHPs), particularly Old World primates (OWPs) which are more closely related to humans, has proven difficult. This limitation significantly hinders preclinical research and the development of effective gene therapies for neurological disorders. Current methods such as direct intraparenchymal injections are invasive and do not allow for brain-wide gene transfer. Alternatives like lumbar puncture or intra-cisterna magna injections have limited efficacy and potential for adverse effects. Therefore, the development of a safe and effective AAV vector capable of crossing the BBB in NHPs is crucial for advancing gene therapy research and its clinical translation.

The use of AAVs for gene therapy has a long history, dating back to their discovery as adenoviral contaminants. Hundreds of clinical trials have established their potential for long-term gene expression. However, recent reports of hepatotoxicity and patient deaths following high-dose systemic AAV delivery have raised concerns about safety. The inherent low therapeutic index and high effective dose of natural AAV serotypes necessitate the development of more efficient and safer vectors. This has led to a focus on engineering novel capsids with enhanced properties. Previous research has utilized techniques such as Cre-recombination-based AAV targeted evolution (CREATE) to select for AAV variants with improved BBB crossing capabilities in rodents, specifically AAV-PHP.B/eB. These variants showed enhanced neurotropism and reduced off-target transduction. However, the success achieved in rodents has not consistently translated to NHPs, particularly OWPs, highlighting the need for primate-specific vector development. While some success has been observed in marmosets (a New World primate), few AAV capsids have shown efficacy in OWPs like rhesus macaques, which are commonly used models for studying human cognition and neurodevelopment. The lack of a BBB-penetrating vector for OWPs often necessitates invasive intraparenchymal injections, limiting brain-wide gene transfer capabilities.

The study employed a multi-species screening and characterization strategy to identify AAV variants with enhanced BBB-crossing tropism in NHPs. The researchers initially screened a library of AAV variants in adult marmosets. A novel approach was used to compensate for the unavailability of Cre-transgenic marmosets, typically used to increase the stringency of selection using CREATE methodology. Instead, the researchers employed a clustering analysis based on sequence similarity from next-generation sequencing (NGS) data to identify promising candidates. Two variants, AAV.CAP-Mac (CAP-Mac) and AAV.CAP-C2, were selected for further characterization. Subsequently, a capsid pool study was conducted in newborn rhesus macaques to assess the translatability of these variants to OWPs. Eight AAV variants, including CAP-Mac and CAP-C2, along with AAV9 and other previously engineered AAVs, were pooled and administered intravenously. A unique molecular barcode was incorporated into each variant's transgene, enabling quantification of each variant's relative abundance in the brain and liver tissues. The researchers then conducted single-variant characterizations in newborn rhesus macaques and green monkeys to evaluate CAP-Mac's efficacy in different OWP species. This involved administering CAP-Mac variants carrying different fluorescent reporters under the control of a CAG promoter. Immunohistochemical analysis was performed to determine the cell-type tropism of CAP-Mac in these primate species. To demonstrate the utility of CAP-Mac, two functional applications were explored. First, a single intravenous dose of CAP-Mac was used to deliver functional GCaMP for ex vivo two-photon calcium imaging in the macaque brain. Second, a cocktail of fluorescent reporters was delivered to achieve Brainbow-like multicolor labeling for morphological tracing. Finally, the researchers investigated CAP-Mac's ability to transduce human neurons in vitro. Human-induced pluripotent stem cells (iPSCs) were differentiated into mature neurons and then transduced with CAP-Mac and AAV9, comparing transduction efficiency and expression levels. Additional ex vivo experiments were conducted using brain slices from adult rhesus macaques to investigate CAP-Mac's efficacy in adult tissues and to assess its ability to cross the BBB. The same capsid pool used in newborn macaques was used in adult macaques for in vivo testing. Lastly, CAP-Mac's tropism in adult marmosets was also analyzed.

The study identified AAV.CAP-Mac as a novel AAV9 variant with significantly improved brain delivery efficiency in multiple NHP species. In newborn OWPs (rhesus macaques and green monkeys), CAP-Mac exhibited neuron bias, efficiently transducing neurons across various brain regions after intravenous administration. Notably, the efficiency was higher than that achieved by AAV9. In adult rhesus macaques, CAP-Mac demonstrated a broad tropism, targeting both neurons and other cell types. However, in adult marmosets, CAP-Mac showed a bias towards the vasculature. In comparison to AAV9, CAP-Mac demonstrated a 45-fold increase in potency for transducing human neurons *in vitro*. In ex vivo experiments using adult rhesus macaque brain slices, CAP-Mac yielded a 3.6-fold increase in viral transcripts compared to AAV9, despite having fewer viral genomes. *In vivo* experiments in adult rhesus macaques showed a 13-fold increase in CAP-Mac-delivered genomes in the brain compared to AAV9. Functional studies showed successful delivery of GCaMP for ex vivo two-photon calcium imaging and Brainbow-like labeling in newborn macaques. Notably, none of the macaques in this study showed adverse events or abnormal liver function. The different tropism between infant and adult primates highlights that the BBB is dynamic across development. The differences in tropism across species is not surprising and highlights a challenge in AAV engineering. The study showcases how a multi-species approach can reveal a more nuanced understanding of AAV tropism and highlight that utilizing mice alone is not enough to guarantee success in primates.

The findings address the critical need for a safe and effective AAV vector capable of brain-wide gene transfer in NHPs. The successful identification and characterization of AAV.CAP-Mac represents a significant advancement in the field of AAV engineering. CAP-Mac's superior efficiency in delivering functional genetic payloads to the brains of OWPs, combined with its relatively benign safety profile, positions it as a powerful tool for neuroscience research and a promising candidate for translational gene therapies. The observed differences in tropism across species and developmental stages highlight the importance of comprehensive multi-species testing when engineering AAVs for clinical applications. Future research could explore further optimization of CAP-Mac, focusing on the tailoring of its tropism to specific cell types or brain regions and optimizing the vector's cargo capacity for larger therapeutic genes. Moreover, further research is required to investigate CAP-Mac's efficacy in various neurological disease models.

This study successfully engineered AAV.CAP-Mac, a novel AAV9 variant that enables efficient and non-invasive gene delivery to the brains of multiple NHP species. CAP-Mac's efficacy in various experimental settings, including in vivo and in vitro, establishes it as a promising tool for neuroscience research and gene therapy development. Its successful use in delivering functional genetic payloads for calcium imaging and Brainbow-like labeling demonstrates its transformative potential. Future research should focus on exploring CAP-Mac's efficacy in disease models and further optimizing its tropism and cargo capacity.

The study's focus on relatively young NHPs (newborn and infant) warrants further investigation of CAP-Mac's efficacy in older, adult primates. The observed species- and age-dependent tropism suggests the necessity of further optimization for consistent and predictable delivery across different populations. While the absence of adverse events in this study is encouraging, long-term studies are needed to fully assess CAP-Mac's safety profile.