f/8:

How Apertura’s Platforms Work Together to Create Next-Generation Gene Therapies

By: Sisi Chen, PhD, Senior Director of Platform Technology at Apertura Gene Therapy

I lead Apertura’s Platform Technology group, whose mission is to engineer best-in-class AAV capsids and regulatory elements to better patients’ lives through genetic medicine. We combine high-throughput screening, in vitro experiments and in vivo animal models with cutting-edge machine learning (ML) to develop next-generation gene therapies.

My educational background is in electrical engineering and bioengineering. I completed my PhD at UC Berkeley focusing on single-cell genomics, and during my postdoctoral training at UCSF, I developed a single-cell profiling platform to study stem cell differentiation. As Director of the Beckman Single-cell Profiling and Engineering Center at Caltech, I then applied single-cell genomics platforms to screen highly complex data sets to analyze massive genomic data at scale. Bridging the worlds of ‘wet lab’ biomedical research and computational biology was exciting, but I wanted to translate innovations from academic research directly to the patients who need them.

Joining Apertura provided me the opportunity to leverage my expertise in genomics, cell engineering and computational data analysis to develop genetic medicines to help restore quality of life to patients with devastating diseases.

Machine Learning-Based Platforms for Engineering Inside and Outside the Capsid

At Apertura, we take a holistic approach to creating next-generation gene therapies through integrated capsid and payload engineering. We are not deterred by challenges that have faced the industry in the past; rather, it is these obstacles that propel us forward.

We screen millions of sequence variants to identify capsids and payload elements for gene therapy applications using powerful machine learning (ML) tools. When searching for viral capsids with ideal properties for gene therapy, it is impossible to empirically screen every theoretical sequence combination using traditional laboratory experiments.

Machine learning methods allow us to expand the search space, learning from the capsids that we do screen to make predictions about the remainder of the search space. While every capsid we nominate to move forward follows a series of successively rigorous experimental characterizations, ML approaches help triage our initial search.

Capsid Engineering for Greater Translational Potential

Canonical approaches to identifying novel capsids for gene therapy delivery rely on screening randomly constructed capsid libraries with millions of variants in vivo using animal models, and the capsids selected in animals by this process are not always specific to humans.

By first screening capsid libraries for binding with human receptors expressed on target cells or tissues, we can enrich libraries with human-specific capsids before moving forward. We then use machine learning to optimize capsid tropism, testing variants in silico to predict binding and fitness properties. This target-specific selection strategy enables data-driven capsid design with increased confidence for greater translational potential for patients.

Pioneering Next-Generation Capsids and Payloads for Gene Therapy

Importantly, our technology and approach also extend to payload engineering. Apertura combines multiple cutting-edge ML platforms to engineer inside and outside the capsid, allowing us to design genetic medicines with the potential to impact more patients.

Outside of work, I fuel my creative passion as a ceramicist, molding clay to create pottery and showcase my point of view as an artist. The ability to embrace multiple perspectives also propels me as a scientist. Working together, the platform team meets challenges head-on and combines disparate ideas to give us the confidence and creativity to push the boundaries of what has been possible.

Working within Apertura’s gene therapy startup environment in NYC allows my creativity, entrepreneurship and scientific innovation to flourish, and I can’t wait to see how our team works together to make the impossible possible for patients through novel genetic medicines.

We are engineering next-generation gene therapies with patients in mind. Interested in joining our team? Learn more about our technology and approach at aperturagtx.com.

1. Eid F et al. (2022) Systematic multi-trait AAV capsid engineering for efficient gene delivery. bioRxiv. doi: https://doi.org/10.1101/2022.12.22.521680
2. Huang Q et al. (2022) Targeting AAV vectors to the CNS via de novo engineered capsid-receptor interactions. bioRxiv. doi: https://doi.org/10.1101/2022.10.31.514553
3. Huang Q et al. (2023) Targeting AAV vectors to the central nervous system by engineering capsid-receptor interactions that enable crossing of the blood-brain barrier. PLoS Biol. 21, e3002112. doi: https://doi.org/10.1371/journal.pbio.3002112