GenScript was honored to host its first Gene and Cell Engineering Summit on July 22, 2021. This virtual event offered three different tracks: Expanding CRISPR Toolbox, Genome Editing in Cell and Gene Therapy, and Enzyme and AAV Engineering. A total of 14 speakers, leaders in gene and cell editing, cell therapeutic development, and protein and viral engineering, presented their most current work.
The program kicked off with opening remarks by Ray Chen, President of GenScript USA Life Science Group, and Shiniu Wei, Chief Financial Officer of GenScript Biotech, who welcomed attendees to the summit. Dr. Alexander Marson’s keynote presentation set the tone for a productive and stimulating afternoon, which included presentations by leaders in the field from Editas Medicine, Leidos Biomedical Research, REGENXBIO, and MERCK, among many others. Lastly, attendees were treated to a closing keynote presentation on T cell and tumor determinants of CAR-T efficacy by Dr. J Joseph Melenhorst.
Genomic Immunology and Engineering Immunity
In his talk, Dr. Marson presented work from his team which aims to understand gene programs controlling the behavior and properties of immune cells, particularly T cells. His lab leverages genetic engineering strategies to better understand immune cell functions and instruct T cell modification for future cell therapies for a broad range of human disease states.
His team has been developing CRISPR/Cas9 editing tools, specifically non-viral approaches to gene editing, to enable the precise and efficient modification of immune cells. They found that electroporating the Cas9 ribonucleoprotein complex alone or combined with DNA payloads into T cells proved to be an efficient strategy to knockout or knockin gene sequences of interest. For example, by enabling replacing the TCR alpha and beta sequence, ultimately re-writing T cell specificity.
To improve non-viral gene knockin efficiency, Dr. Marson’s team engineered modifications to the DNA payloads and Cas9 protein to promote co-localization and delivery or “shuttle” to the nucleus. However, realizing the toxicity associated with high levels of double-stranded DNA payloads, his team has adapted the “shuttle” strategy to single-stranded DNA donor templates. Currently, they are leveraging this approach to developing various T cell therapeutics.
To meet manufacturing or “GMP” standards enabling developing T cell therapeutics through non-viral approaches, Marson’s team is partnering with GenScript to synthesize long single-stranded DNA (e.g., CAR construct). His team has not only found a source for GMP grade long single-stranded DNA but is delighted with the improved knockin efficiency achieved by using GenScript’s synthesized payloads.
In the second part of his talk, Dr. Marson shared progress on genome-wide target discovery in primary T cells to enable the development of better cell therapies. Their studies leveraged a hybrid approach to gene editing by relying first on lentiviral delivery of a library of guide RNAs followed by Cas9 delivery into primary T cells. This approach supported identifying genes critical for improved T cell therapeutic properties (e.g., T cell proliferation and cancer cell killing).
Building More Useful CRISPR-Cas Technologies
The Kleinstiver lab is focused on developing new and improved gene-editing technologies. During his presentation, Dr. Kleinstiver discussed protein engineering strategies applied by his team to improve on-target activity, targeting range, and specificity/safety of CRISPR/Cas9 nuclease.
Targeting range is a limitation imposed by protospacer adjacent motif (PAM) sequences on the access of Cas9 nuclease to edit broad genomic sites. Therefore, Dr. Kleinstiver’s team engineered changes in amino acid residues relevant for PAM recognition. In addition, to expedite the screening of Cas9 nuclease libraries, his team developed a high-throughput PAM determination assay for screening (HT-PAMDA) Cas9 PAM specificities. This approach has enabled the Kleinstiver team to identify various forms of Cas9, including SpG (NGN) and later SpRY (NAN), with relaxed PAM requirements.
Cas9 nuclease’s off-target effects limit the therapeutic application of current gene editing tools due to potential unwanted genomic changes. With newly engineered Cas9s having relaxed PAM requirements, a critical question that the Kleinstiver lab had to address was the potential for increased off-target activities. Using a previously developed assay, GUIDE-seq, they confirmed that SpG and SpRY activities targeted an increased number of off-target sites. Introducing high-fidelity changes in SpG and SpRY sequences allowed the reduction of off-target edits, thus improving their safety.
What are some of the benefits of these newly developed Cas9 nucleases? First, Dr. Kleinstiver shared how nucleases with minimal PAM requirement can facilitate base-editing and prime-editing approaches, which require the Cas nuclease to access exact genomic sites. Lastly, he shared how his lab leverages these new enzymes even as molecular tools in vitro, which provide greater flexibility for cloning approaches.
An Engineered AsCas12a Nuclease Facilitates the Rapid Generation of Therapeutic Medicines
Dr. Zuris discussed the rationale and approach for generating a new AsCas12a variant or AsCas12 ULTRA, how the newly developed variant compares in activity and specificity against other nucleases, and its application to T and NK cell gene editing.
Dr. Zuris highlighted how the knockout of endogenous TRAC, B2M, and CIITA genes are edits critical to the safety and success of allogeneic T cell therapies. Additionally, modifications such as CAR knockin would be required to enable desirable therapeutic activity. Therefore, multiplexed gene-editing capabilities are critical to engineering the next generation of cellular therapies due to the number of modifications necessary to achieve both desired activity and safety. To enable T and NK cell therapeutics development, Dr. Zuris’ team focused on AsCas12a, which offered the ability to target genomic regions otherwise unreachable by some available Cas9 nucleases (e.g., SpCas9, SaCas9, and SaCas9 KKH). Although AsCas12a has a more limited genomic targeting range, Dr. Zuris emphasized that this could prove advantageous in reducing potential off-target editing. Combined with AsCas12a’s higher intrinsic specific activity and fidelity, this nuclease represented an attractive candidate for engineering cell therapeutics.
Nevertheless, recognizing that AsCas12a is inefficient in achieving knockout and knockin editing, efforts were directed towards engineering AsCas12a for improved editing efficiency. The resulting variant, AsCas12 ULTRA, showed significantly enhanced editing efficiency in T and HSPC cells while preserving specificity. Overall, his team showed that AsCas12a supports multiplexed T cell editing with high efficiency above 90% when targeting TRAC, B2M, and CIITA for knockout and around 60% knockin efficiency when targeting transgenes into the TRAC and B2M loci. Lastly, Dr. Zuris shared how the new AsCas12a ULTRA variant is highly efficient for multiplexed NK cell editing.
CRISPR in Animal Model Generation - Functionalizing Genome Editing for a Broad Range of Targets
In her presentation, Dr. Pruett-Miller first discussed the main offerings provided by her team to investigators as part of the Center for Advanced Genome Engineering, which include supporting CRISPR/Cas9 gene-editing projects and gene and cell therapy programs.
For transgenic animal model generation, Dr. Pruett-Miller’s team relies on CRISPR/Cas9 knockin approaches, leveraging synthetic single-stranded DNA, such as short <150 nucleotides ssODN and long>150 ssDNA. Long synthetic single-stranded DNA injected or electroporated into embryos results in efficient integration, with over 50% knockin frequency. Dr. Pruett-Miller emphasized that an added advantage of ssDNA donor templates over double-stranded plasmid DNA is their typical shorter homology arms. This feature facilitates verifying insertions and sequences, which they do by targeted next-generation sequencing (NGS) spanning both homology arms to ensure identifying any potential synthesis errors.
Finally, Dr. Pruett-Miller’s team has developed a python-based program ( CRIS.py) for high-throughput analysis of NGS data that facilitates elucidating gene-editing outcomes. Some advantages of CRIS.py for NGS analysis include; efficiently runs on a desktop computer, analyzes over 20 samples, searches more than one donor sequence, generates a searchable and sortable data output. CRIS.py and PrettyCRISP.py, a more recent version, may be freely used and is accessible at https://github.com/jake-steele/prettyCRIS.py.
Targeted DNA Integration without Double-Strand Breaks using CRISPR RNA-guided Transposons
Dr. Sternberg shared methods for targeted DNA integration without double-strand breaks, which rely on CRISPR RNA-guided transposons. The use of RNA-guided transposases/recombinases represents an additional tool that may be leveraged to insert large DNA payloads.
Transposons are genetic elements containing several genes flanked by conserved inverted repeats. Among different cargo genes, a transposon contains a transposase sequence, which encodes for the enzyme involved in the mobilization of the transposon. By binding to conserved inverted repeats at the end of the transposon, the transposase executes transposon mobilization through a cutting and pasting process from a donor to a target DNA site. Dr. Stemberg highlighted that most transposons exhibit high efficiency for DNA mobilization but lack specificity for DNA targeting. The finding that some transposable elements also encode CRISPR/Cas systems, where the Cas lacks nuclease activity, led them to explore the possible role of CRISPR/Cas on DNA integration.
His lab engineered several plasmids, essentially deconstructing the Vibrio cholera CRISPR-transposon into three individual expression vectors, including mini-transposon, transposase, and CRISPR elements. These constructs, when expressed in E. coli could induce targeted DNA integration guided by RNA. Dr. Sternberg described the elucidated mechanisms involved in the CRISPR RNA-guided integration. This system consists of a complex between Cascade, which binds crRNA and genomic DNA, and other subunits. TniQ, one of the subunits, is critical for subsequent transposase and transposon DNA recruitment. Understanding the structure and basic mechanisms by which the cascade complex operates has enabled Sternberg’s team to develop engineering strategies to improve the efficiency and fidelity of transposition. Lastly, he shared how his lab has implemented these new tools to target large DNA payloads into bacterial genomes.
Therapeutic Gene Editing Enabled by New Delivery Vehicles
In his talk, Dr. Murthy shared his work developing strategies for the efficient and safe in vivo delivery of CRISPR/Cas9 gene-editing tools. He discussed some viral and non-viral methods already developed, such as AAV, microinjection, electroporation, and LNPs. While viral vectors, particularly AAV, provide the most efficient way to deliver CRISPR/Cas9 and genes for insertion in vivo, Dr. Murthy shared some associated disadvantages. Particularly, extended Cas9 expression may lead to adverse immune reactions and unwanted off-target DNA cleavage. In general, non-viral methods (e.g., plasmid, mRNA, RNP) are less efficient in delivering CRISPR/Cas9 editing tools. Dr. Murthy highlighted that the liver is an exception where non-viral delivery methods work very effectively.
To address the current limitations of delivering CRISPR/Cas9 in vivo, Dr. Murthy’s team has developed various strategies. One of these strategies, CRISPR-GOLD, consists of gold nanoparticles conjugated to donor DNA templates, complexed with the Cas9 RNP, and encapsulated with an endosomal disrupting polymer. This approach provides all the components necessary for gene editing in a single particle and is being used to develop a therapeutic for Duchenne muscular dystrophy. Preclinical studies in a Duchenne mouse model (MDX) expressing a mutant form of dystrophin showed that CRISPR-Gold could correct the mutation in vivo following direct injection into the muscle. In addition, because this approach relies on delivering Cas9 as part of an RNP complex, Dr. Murthy’s team found minimal associated off-target DNA damage in mice. Overall, their findings supported the recovery of dystrophin expression and concomitant functional recovery following the use of CRISPR-Gold. Currently, Dr. Murthy is collaborating with GENEDIT for the continued development of this project. In addition, he is also collaborating with GENEDIT in developing CRISPR-GOLD based strategies for gene editing in the brain.
Clinical Scale Gene Editing for Cell and Gene Therapy Applications
MaxCyte leverages its flow electroporation platform to engineer a variety of cell types for research, clinical and commercial applications. The expert® platform delivers high transfection efficiency and viability supporting potent, efficacious, and reproducible cell therapy products. As part of the expert® platform, investigators can access three different instruments to match their preclinical or clinical needs. The ATx instrument is research-focused and enables small scale transfections, STx is scalable and used for transient protein production and other biologics, and the GTx is scalable and clinically validated.
To illustrate how MaxCyte provides gene-editing solutions for research and clinical needs, Rama Shivakumar shared several case studies addressing: optimization of research scale gene-editing in iPSCs, improving the efficacy of TCR therapies, and lastly, manufacturing of cell therapy products.
The strengths of the expert® platform have supported the effective use of CRISPR/Cas9 editing tools for disease modeling (e.g., Duchene Muscular Dystrophy), development of NY-ESO TCR T cell therapies, and manufacture of modified hCD34+ cells from Sickle Cell Disease and beta‑thalassemia patients for autologous cell therapy.
Genetic Engineering Approaches to Support Clinical Applications of T-cell Receptor Libraries Targeting Oncogenic Mutations
Dr. Cafri started his talk by highlighting the current gap on effective therapeutic strategies for solid tumors, which account for 90% of cancer-related mortality. Therefore, there is a clear need for immunotherapeutic approaches targeting solid tumor antigens. His work focuses on identifying T cells and specifically T cell receptors (TCR) able to recognize oncogenic mutations. An advantage of immunotherapies leveraging TCRs over CART is the ability of these receptors to recognize both cell surface and intracellular processed antigens.
Dr. Cafri has developed a new method to identify T cells reactive to tumor antigens in the blood. In this process, peripheral blood cells from metastatic epithelial cancer patients are sorted to isolate memory T cells to perform in vitro T cell activation assays by-coculturing with neoantigen-loaded dendritic cells. This approach allows for expansion and enrichment of T cells reactive against specific neoantigens. Some of the advantages of this approach include more accessible patient-derived samples and younger T cells. By following this workflow, his team identified HLA-A11 restricted KRASG12V specific TCRs.
Overall, this strategy supports developing personalized T cell therapies and depending on the tumor antigen and HLA restrictions, immunotherapies for a broader patient population. For example, KRAS mutations, such as G12D and G12V, are frequent in pancreatic cancer and could be targeted for immunotherapy to benefit more patients. Armed with a library of TCRs against mutated KRAS, Dr. Cafri’s team is developing a pipeline for autologous T cell immunotherapies leveraging the expression of some of the identified TCRs. Lastly, for T cell gene editing to support clinical applications, his team is considering non-viral approaches, such as non-viral sleeping beauty transposon/integrase system and non-viral CRISPR-Cas9 gene editing.
CRISPR/Cas9-based Genome Editing for Autologous CAR-T Cell Production
As part of the Frederick National Laboratory for Cancer Research, Dr. Vepachedu is involved in the Biomedical Development Program (BDP). This program helps support the development of investigational drug candidates from early stages to manufacturing for phase 1 and 2 clinical trials (e.g., antibody therapeutics, vaccines, viruses). Additionally, BDP offers a gene (i.e., AAV) and cell therapy (i.e., CAR-T) portfolio.
He started by providing an overview of the workflow for CAR-T cell therapy, from collection of cells from patients to re-infusion of the engineered cell product. In their current approach, the delivery of CAR-T constructs is enabled by lentiviral or retroviral vectors. While efficacious for treating B cell lymphomas, CAR-T cell therapies have several challenges, including toxicity, exhaustion, and suboptimal persistence. Beyond improvements to CAR-T design through various CAR-T generations, cell product manufacturing optimization can also help develop more efficient cell therapies. Therefore, new gene engineering technologies that allow knocking out genes to counteract both exhaustion and suboptimal persistence are desirable. Additionally, eliminating random CAR-T integration induced by viruses from the workflow of cell therapy manufacturing would reduce safety risks.
Therefore, Dr. Vepachedu discussed how his team is developing a CRISPR/Cas9 editing platform to engineer CAR-T cell drug products for autologous cell therapies. For this approach, the team is developing methods to target various genes (e.g., TRC alpha, TCRbeta, and PD1) for knockout and knockin the CAR-T construct into the TRAC locus. Additionally, for allogeneic cell therapeutics, they are leveraging CRISPR/Cas9 tools for the knockout of TCR alpha and B2microglobulin to reduce the potential of adverse immune reactions with these cell products. To achieve these modifications, Dr. Vepachedu’s team relies on CRISPR/Cas9 protein in complex with synthetic guide RNA ( sgRNA) and delivery of CAR sequence as a single stranded-HDR template ( ssDNA) containing homology arms for insertion into the TRAC locus.
Targeted Integration in Stem Cells: Generating Genetically Engineered Cellular Drugs
Dr. Porteus started his talk with an introduction to “Living Drugs,” or cell and gene therapeutics, which he envisions as the future of medicine much akin to the role of small drug molecules. His group is interested in leveraging precise targeted integration to engineer stem cells, such a human hematopoietic stem cells, to treat human disease. To this end, his lab relies on CRISPR/Cas9 and homologous recombination to precisely target genes for repair or to insert full gene sequences.
Stem cell gene editing is performed ex vivo using a CRISPR/Cas9 ribonucleotide complex and an AAV6 non-integrating virus for gene delivery. The combined use of CRISPR/Cas9 RNP and AAV6 in the Porteus lab is an efficient gene-editing system that has allowed them to modify various human primary cells successfully, including hematopoietic stem and progenitor cells (HSPCs), T-cells, and mesenchymal stem cells (MSCs).
Dr. Porteus shared how this system has allowed his team to introduce a variety of modifications into HSPCs, such as single-nucleotide changes (e.g., sickle cell disease), functional gene correction (e.g., SCIDX1), safe-harbor gene addition (e.g., Mucopolysaccharidosis type I), and transgene insertion (e.g., -thalassemia). He discussed how preclinical studies in his lab have translated to clinical studies to correct various human disease conditions. Overall, by modifying patient-derived stem cells, gene editing approaches allow autologous cell therapies and bypass immune mismatch complications associated with allogeneic cell transplantation. Therefore this approach represents a significant advancement in the treatment of human disease.
Library-selected AAV variants can effectively translate to non-human primates in the spinal cord and cochlea
Dr. Hanlon discussed his work on the development and characterization of AAV variants with enhanced CNS transduction properties. To discover improved AAV variants, Dr. Hanlon and colleagues created an AAV capsid library by insertion of short random oligonucleotides. Library screening was enabled by iTransduced, a newly developed platform for capsid selection. In iTransduced, a CRE recombinase sequence is inserted together with the random oligonucleotides into the AAV vector, enabling AAV capsid library screening in the context of tdTomato expression in Floxed-STOP tdTomato mice. Identification and selection of brain cells expressing the tdTomato reporter allowed Dr. Hanlon and colleagues to recover by PCR the AAV9 capsid sequence with random oligonucleotide insertions for re-incorporation into the library and a new round of screening.
Following this iterative cycle of transduction and screening, two AAV capsid variants, AAV-S and AAV-F, were identified that enabled efficient GFP expression in the mouse CNS. AAV-F induced robust GFP expression in the spinal cord and various brain regions, such as the cortex, hippocampus, and cerebellum. In contrast, AAV-S was efficient in transducing peripheral tissues such as the heart, skeletal muscle, and the inner ear, particularly hair cells in the cochlea. Lastly, Dr. Hanlon highlighted the potential for preclinical and clinical applications of these two vectors as both variants, AAV-F and AAV-S, successfully transduced similar tissues in a non-human primate model.
Improved chemistry by combining enzyme engineering, enzyme immobilization, and flow chemistry
Dr. Camacho provided an overview of protein engineering at MSD that aims at discovering and developing active pharmaceutical ingredients (APIs) by inventing and improving protein functions. Her talk focused on the work of the enzyme engineering group at MSD, which aims to enhance biocatalyst function through evolution. Protein engineering and specifically directed evolution allow the MSD group to address and improve upon biocatalysts’ properties, such as substrate scope and tolerance to environmental conditions. To achieve this goal, the MSD group follows a “Design-Make-Test-Learn” workflow based on the design and development of mutant libraries, high-throughput mutant expression, and testing for selection. Iterative “Design-Make-Test-Learn” cycles enable the group to identify enzymes with improved properties.
To illustrate their work, Dr. Camacho provided an overview of one of their programs, MK-1026, focused on improving a biocatalyst (i.e., a transaminase) to support the synthesis of a small molecule inhibitor of Bruton’s tyrosine kinase (BTK) used in the treatment of CLL and Non-Hodgkin’s Lymphoma. Overall, the team at MDS leveraged site saturation mutagenesis and combinatorial libraries to improve upon several properties of their selected transaminase biocatalyst, such as activity, selectivity, and thermostability.
Discovering the Next Generation AAV Vector Through Capsid Engineering and Expression Cassette Optimization
At REGENXBIO, a clinical-stage gene therapy company, Dr. Liu’s team works on optimizing AAV vectors for gene therapy through capsid engineering strategies and improved expression cassettes. First, Dr. Liu provided an overview of the timeline for AAV discovery and initial use as a vector for gene therapy. Additionally, he introduced critical features of the AAV genome, structure, and transduction pathway, making this viral vector useful for engineering gene therapies.
Dr. Liu and colleagues work to improve the AAV capsid and transgene cassette to maximize the full potential of AAV gene therapy. Mutations are introduced targeting the capsid’s variable loops to engineer capsid variants. These mutations may encode as many as ten amino acids, resulting in the expression of new peptides on the capsid’s surface. Capsid engineering approaches aim at improving manufacturability and enable the development of AAV variants with new affinities for specific tissues. For transgene cassette optimization, the promoter, transgene, intron sequences, and inverted terminal repeats may be modified to improve the expression of genes of interest.
Dr. Liu’s team relies on structure-based rational design approaches to introduce mutations on the AAV’s surface variable loops for capsid engineering. Libraries containing millions of AAV variants may be developed through random peptide insertions or error-prone PCR. The workflow for AAV variant selection is performed in vivo through AAV library injections into animal models, tissue harvesting, and viral recovery for additional rounds of selection. This process, referred to as directed evolution, takes advantage of natural selection to identify AAV variants with desirable properties. Lastly, Dr. Liu shared how these strategies have been implemented in developing an AAV8 gene therapy, RGX-202, for Duchene muscular dystrophy, currently in the precilinal stage.
Response to Second Generation CAR-T Cell Therapy: It Takes (at least) Two to Tango
After introducing the different generations of Chimeric Antigen Receptors (i.e., 1st-3rd CARs generations), Dr. Melenhorst explained the therapeutic workflow, which involves collecting T cells from patients, CAR gene insertion, T cell expansion, and ultimately infusion into the patients.
CLL is one of the most prevalent leukemia types in the western hemisphere affecting predominantly the elderly. Current treatment for CLL is based on small molecule inhibitors (e.g., Idelalisib and Venetoclax), which are not curative; in contrast, CAR-T cells provide a curative approach.
Clinical trials for Chronic lymphocytic leukemia (CLL) with CAR-T therapy were initiated in 2010 and lead to dramatic and durable remission. Dr. Melenhorst shared that one of the lessons learned from this initial clinical trial was the positive correlation between CAR-T expansion and clinical efficacy. Additionally, CAR-T cells recovered from patients showed persistent polyfunctionality even after three years post-infusion (e.g., IFN and IL2 production). Moreover, single cell analysis of CAR-Ts recovered after over nine years post-infusion demonstrated that long-term persistent CAR-T cells were highly proliferative, functional, and metabolically active.
Because both CAR-T clonal expansion and persistence correlated with the patient’s response, Dr. Melenhorst hypothesized that the intrinsic properties of T cells were vital for the efficacy of the therapeutic product. Single cell expression analysis confirmed transcriptomic differences between CAR-T cells infused in responder and non-responder patients. These studies allowed Dr. Melenhorst’s team to identify several genes differentially expressed between responder and non-responders. Significantly, non-exhaustion and early memory T cell signatures were enriched in responders. Therefore, his team has been focusing on the intrinsic properties of T cells before CAR-T cell manufacturing, which could determine therapeutic success. Their work has allowed them to identify T cell biomarkers (CD27+CD45RO-) associated with CAR-T therapeutic efficacy in CLL, identifying a CD8+ T cell population or early memory T cells. Lastly, Dr. Melenhorst also discussed how his team worked to understand extrinsic or tumor factors influencing CAR-T cell efficacy in CLL.