Our lab has three research focus areas that are highly complementary and closely intertwined. First, we develop optogenetic tools based on photosensory proteins from nature. We invented a unique technology based on engineered, light-switchable anti-CRISPR proteins that allows controlling CRISPR-Cas genome editing and gene targeting. This technology facilitates spatiotemporally precise (epi)genome perturbations in mammalian cells, which we employ to study human genome regulation and function. Second, we create cell type-specific Adeno-associated virus (AAV) vectors for the efficient and safe delivery of CRISPR components into patients suffering from genetic disorders. Third, we combine artificial intelligence and high-throughput design to accelerate and eventually rationalize the engineering of proteins with novel properties.
Engineering anti-CRISPR proteins for conditional activation of CRISPR-Cas
Anti-CRISPR proteins are potent inhibitors of CRISPR effectors. We engineer light-switchable derivatives of natural anti-CRISPRs by embedding photosensor domains, such as the light-oxygen-voltage 2 (LOV2) domain from Avena sativa, into Acr structures (Bubeck & Hoffmann et al., Nat. Methods, 2018; Hoffmann and Mathony, bioRxiv, 2019). The resulting, chimeric inhibitors block Cas9 in the dark, but release its activity upon blue light irradiation.
This approach which we named CASANOVA, for CRISPR-Cas activity switching by a novel, optogenetic variant of anti-CRISPR proteins, facilitates light-dependent genome editing and epigenome editing and is used in our lab to study genome regulatory processes. While CASANOVA was initially limited to the S. pyogenes Cas9, we are currently extending this original strategy to other Cas9 orthologues, Cas12 and Cas13, the latter of which is an RNA-targeting CRISPR effector. We are also developing designer Acrs that outperform their natural counterparts with respect to inhibition potency (Mathony, Harteveld and Schmelas et al., Nat. Chem. Biol., 2019).
Recently, we started to extend our protein engineering efforts towards a variety of other protein classes of interest for basic research and biotechnology. To render these proteins switchable upon light or chemical induction, we combine computational approaches with mutational scanning, a method for saturating mutagenesis and high-throughput functional analysis of protein candidates.
Developing safe and efficient AAV-CRISPR vectors
One of our central aims is to support the clinical translation of the CRISPR technologies by developing safe and efficient AAV-CRISPR vectors (Senis et al., Biotechnology J, 2014). Towards this end, we use synthetic biology approaches to create switches and circuits that confine Cas9 activity to selected cell types, e.g. by coupling Cas9 activity to the abundance of cell type-specific microRNAs (Hoffmann et al., Nucleic Acids Research, 2019). We also combine wet lab experiments with mathematical modeling to fine-tune Cas9 activity to desired levels, thereby enabling the kinetic insulation of ON- and OFF-target editing events (Aschenbrenner & Kallenberger et al., Science Advances, 2020).
Machine learning-guided protein design
By training neural networks on thousands of protein sequences and corresponding 3D protein structures, we aim at creating algorithms that can inform protein design (Upmeier zu Belzen et al., Nature Machine Intelligence, 2019). Applications range from the prediction of engineering hotspots to facilitate the development of switchable proteins to the re-design of enzymes and – in the future – the creation of immune “stealth” protein therapeutics.