Directed Evolution Improves Gene Therapy Vector for… : Neurology Today – LWW Journals

By Richard Robinson December 16, 2021 The Science Explained Article In Brief

A new gene delivery vector and methodology called directed evolution enables more potent delivery of corrective genes to the muscles in mice and primate models, and holds promise for reducing the cost and risk of treatment for muscle disorders.

A new gene delivery vector, building on over 20 years of basic virology research, holds promise for accelerating gene therapy research and reducing the cost and risk of treatment, according to a paper published September 16 in the journal Cell.

We are watching the validation of basic science translated into therapeutic reagents to eventually treat unmet needs for muscle disorders, said R. Jude Samulski, PhD, professor of pharmacology and director of the Gene Therapy Center at the University of North Carolina, who was not involved in the new study.

Hope for delivering therapeutic genes for human diseases has focused in recent years on the handful of naturally occurring adeno-associated viruses (AAV), whose various serotypes differ in their affinity for different tissues. For neuromuscular diseases such as Duchenne muscular dystrophy and spinal muscular atrophy (SMA), AAV9 has been the preferred vector, due to its relative tropism for muscle tissue.

But that tropism is still only partial and is in fact rather limited; autopsies of patients who have received AAV9 for SMA suggest that about 90 percent of the vector is taken up by the liver instead. This is part of the reason that it requires an enormous number of viral particles to treat a patient, explained Mohammadsharif Tabebordbar, PhD, lead author of the new study and until recently a research scientist at the Broad Institute of MIT and Harvard University.

That necessary excess accounts in part for the astronomically high cost of even preclinical gene therapy studies. It also increases the risk of liver toxicity and immune reaction in patients receiving treatment. The risk and expense of high doses of AAV vectors is such that older and heavier patients are largely excluded from clinical trials altogether.

That led Dr. Tabebordbar and colleagues to try to design an AAV capsid that could overcome these limitations. The problem we were trying to solve was whether we could engineer the capsid of the virus in such a way that we could switch it from mainly going into the liver, to mainly going to the muscle, he told Neurology Today.

To do so, the team turned to a technique called directed evolution, whose developer, Frances Arnold, PhD, was one of three recipients of the 2018 Nobel Prize for Chemistry. Like evolution through natural selection, directed evolution is an iterative process, in which a large number of variants are produced and then screened by a selection process. Through multiple rounds, the most fit variant emerges.

In the current study, in the first round, Dr. Tabebordbar generated more than 5 million variants of one protein in the AAV9 capsid, each carrying the gene for its own variant, and then injected the resulting vectors into mice.

He then used a stringent selection process to find the most suitable vectors for muscle gene therapy. Much previous work in the field has identified myotropic vectors by identifying virus genomes from variants that entered the muscle, based on the assumption that more virus genomes in meant more therapeutic protein produced. But that assumption has proved faulty, Dr. Tabebordbar said.

There are many steps involved in effectively transducing a muscle cell, he said. It begins with binding and internalization, then intracellular trafficking and escape from the endosomes, entering the nucleus, release of the genome, and conversion to double-stranded DNA. Only variants that get through this whole process can express the transgene.

So rather than measuring the number of viral genomes per cell, the team looked instead at the level of capsid mRNA. From the original 5 million variants, they selected about 30,000 that expressed the highest levels of mRNA, and used these for another round of variation and selection.

That second round led to an interesting discoverythe top 12 variants all bore a three amino-acid motif (arginine-glycine-aspartic acid, or RGD). Further work with the top candidate, dubbed MyoAAV-1a, showed it to be between 10 and 29 times better than AAV9 at delivering a reporter gene to muscles, including the heart, and reduced delivery to the liver. They observed improved transduction in multiple breeds of mice, and treatment with two different therapeutic genes in models of Duchenne muscular dystrophy and X-linked myotubular myopathy both led to functional improvement.

Doing directed evolution in mice provides you a good research tool, Dr. Tabebordbar said, no mean feat if it can reduce the cost of preclinical evaluation of potential therapies or speed the discovery of new biology critical for developing better treatments. But whether you can actually get a drug out of it is a different question.

The team tested the directed evolution technique in monkeys, and found again after two rounds of selection that the top hits contained the RGD motif, but with different surrounding amino acids, perhaps because of differences in the integrin heterodimer, the surface protein to which the capsid binds, or perhaps because of another part of the story that we don't know yet, Dr. Tabebordbar said. That same RGD motif was independently identified by another group, reported in 2020 in Nature Communications.

Dr. Tabedbordbar is now co-founder and chief scientific officer of Kate Therapeutics in San Diego, a start-up hoping to develop this directed evolution platform for muscle disease therapeutics.

Frustratingly, despite the closer evolutionary similarity to humans, the best vectors in myotube culture were not effective in monkeys. So in vitro and in vivo transduction are two completely different results, he said. It's much more complicated. You can't say that just because it works in myotubes it will work in the human organism.

That presents a challenge for clinical development, but one that may be met by developing better models for further experiments. Using mice with engrafted human muscle, Dongsheng Duan, PhD, professor of neurology at the University of Missouri in Columbia, is pursuing that strategy in an effort to develop a better model for testing new therapies.

This is one step closer, but it will still have to be tested in patients to know if it will work, Dr. Duan said. We hope this new capsid will outperform the previous capsids currently in clinical trials, in order to reduce the dose without compromising gene delivery efficiency. At least for preclinical studies, these results suggest you can use fewer viral particles, which is good, especially for toxicology studies in monkeys.

A single dose of onasemnogene abeparvovec-xioi (Zolgensma), the AAV therapy for SMA, may cost in the low millions of dollars, he pointed out. Cutting that by an order of magnitude or more, which seems possible, should absolutely accelerate the development of gene therapy.

This was an elegant study, said Dr. Samulski. First by testing millions of different possible peptides, then by focusing on the context of the RGD motif that interacts with the heterodimer, they in effect asked the muscle cell which one it liked best. The power of this technique is that you get exactly what you select for.

The role of the integrin heterodimer in capturing and facilitating the entry of AAV emerged from work in Dr. Samulski's lab in 1999.

One of the strengths here is the use of multiple models, from mice to primates to human cells. The field has gone down rabbit holes in focusing almost exclusively on rodent models, Dr. Samulski said, and so the majority of vectors have turned into rodent-specific reagents.

Nonetheless, he cautioned, While it is tempting to think that when you are in the monkey you are closer to the correct species, the reality of it is that until you get into humans, you don't know how close you really are. The animals only give us a guide.

Having seen too many false dawns break over the years, the more seasoned investigators sit back and wait for the clinic, he said.

Dr. Tabebordbar is a co-founder, stakeholder, and full-time employee of Kate Therapeutics

WHAT IT IS

Directed evolution is a technique that rapidly modifies a protein, such as an enzyme or ligand, so that it carries out a reaction faster or fits tighter with a receptor than the original protein.

HOW IT WORKS

Directed evolution is an iterative process, usually involving several rounds of variation and selection. To begin, the gene for the protein is mutated or systematically changed to create many different variants. These gene variants are then expressed in a model system, such as bacteria, that tests the activity of each variant. Those with the strongest activity are chosen for another round of variation and selection, until the desired level of activity is reached or until no further improvement can be detected.

HOW IT IS APPLIED

Directed evolution has been used to develop faster and more efficient enzymes, improving binding ability of therapeutic antibodies, and improving other characteristics of proteins such as stability at high temperatures, for use in industrial production. Additionally, as in the current study, directed evolution can be used to improve the ability of gene therapy vectors to transduce their intended cell targets and minimize off-target transduction.

Continued here:
Directed Evolution Improves Gene Therapy Vector for... : Neurology Today - LWW Journals