KNIT Genome Editing: A New Designed Tool for Safer Gene Therapy!

Genetic therapy promises to solve many diseases and revolutionize even more fields!

But we’ve started to see the limitations of current genome editors. For classic CRISPR-Cas, indels and off-targets can be too much. Are there alternatives?

Today’s paper is one of the most important of the year. The authors created KNIT, a new editing system with improved efficiency and safety!

Let’s see how they did it.

There are over 10,000 genetic diseases.

And around 5,000 of them are caused by a mutation in a single gene, with the rest coming from combinations of gene variants. Altogether, they affect around 1 in 21 people: not that rare after all!

The treatment options vary, because genetic diseases vary widely too. But there’s one idea that promises to solve many of them: genetic therapy.

At its core, genetic therapy aims to treat or prevent diseases by fixing, replacing, or turning off faulty genes. For many diseases, simply adding a healthy copy of the gene back into the cell could be a real treatment!

And this knock-in strategy is real. But it has two big problems, especially when it comes to replacing kilobase-scale DNA sequences (the so-called DNA donor):

Not ideal! This has been a major barrier for gene insertion and other advanced applications like T cell engineering.

There are alternatives. The most obvious one is single DNA nicks, which are less mutagenic. But they’re also less efficient, especially for homology-based DNA insertion.

So, can we get the efficiency of double-strand breaks without the damage?

Today’s paper says yes!

The authors introduce KNIT (kilobase-scale nickase targeting) editing, a double-strand-break-free, non-viral cell-engineering platform. KNIT editing (KE) uses a DNA donor recruitment system and nCas9, a version of Cas9 that creates single nicks to improve editing safety and efficiency.

And it works!

Across different loci, cell types, and targets, KE inserted DNA fragments from ~0.7 kb to over 10 kb, with efficiencies reaching 89% (!). All while reducing indels, off-target events, and cell mortality.

This is one of the most exciting papers of the year; let’s take a closer look!

The concept behind KE is simple:

The idea is that recruiting DNA close to the nick helps insert it into the genome. It’s so cool: they created a machine that physically manipulates DNA at the nanoscale for genome editing!

And the team didn’t stop here.

Once they made KE work, they designed an improved second version. KE2 improves the DNA donor recruitment to increase the editing efficiency. Cas9 is fused to 24 GCN4 peptides; mSA is fused to anti-GCN4 antibodies.

Now, up to 24 mSA-antibodies can bind to a single Cas9 → in theory, you can have 24 copies of donor DNA recruited at the cutting site! And that made a real difference: in one test, KE2 raised eGFP insertion from 55.8% with KE1 to 81.9%!

Okay, it sounds cool, but where are the results?

Glad you asked. The authors tested the system in many ways! In general, they compared KE1/KE2 vs nickase and traditional, double-strand break Cas9.

For eGFP insertion at the ACTB locus in HEK293T cells:

But the real difference is in the indel rate. KE1 shows a 0.46% indel rate, closer to the baseline single-nickase, and >200x lower than the indel rate for Cas9 (which is set as 100%)!

KE1 has high knock-in efficiency while preserving the genome → what the field has been waiting for!

And it works for large DNA fragments too. The authors inserted 5.7 kb and 10.5 kb fragments into the ACTB locus, and KE1 outperformed Cas9 on both insertion and indel rates!

Accuracy was important for the team. They tested it using the PEM-seq method, showing that KE1:

Amazing!

The team also showed KE1 working at the RAB11A and H2B loci. It’s not limited to one site! And it’s not limited to HEK293T cells either; they also showed successful editing in K562 and NIH3T3 cells.

The authors then moved from eGFP to real therapeutic targets.

They inserted a 1,500 bp IL2RG gene into the AAVS1 locus. IL2RG is mutated in X-linked severe combined immunodeficiency, so it’s a real therapeutic target. In HEK293T cells, KE1 increased IL2RG expression 10x vs the control and achieved 89.1% insertion in single-cell genotyping, compared with 24.7% for Cas9!

The team then targeted lysosomal acid lipase deficiency (LAL-D). LAL-D is caused by mutations in the LIPA gene, and the most common mutation is at the splice site of exon 8 in LIPA, leading to a truncated protein.

So, the team inserted wild-type LIPA DNA into either:

Both worked! With KE1 restoring about 39% of wild-type LIPA mRNA when inserted into the native locus.

Another promising use for genome editors is T cell engineering.

The aim here is to reprogram a patient’s own immune cells to target and destroy diseased cells, mostly in cancer or autoimmune disorders. Sounds cool, but it’s now incredibly hard!

The authors first tested an eGFP insertion in primary human T lymphocytes. KE1 here has a lower efficiency than Cas9, but the cells treated with KE1 have a much higher viability:

And the KE1-treated cells also proliferate much better and maintain long-term expression. This is important because viability is often a limiting factor in genome editing. It’s an expensive process, and every dead cell means lower odds of curing patients.

CAR-T cells are engineered with chimeric antigen receptors (CAR) to target and destroy cancer cells. An amazing new tech getting into clinical trials! But CAR-T cells are often produced using double-strand break-based methods, which can create genome instability and oncogenic risks.

Here, the team inserted a CD19 CAR cassette into 2 loci, getting up to 40% CAR integration with KE2. These are great numbers, right in the range of clinically relevant CAR-T manufacturing!

Their CAR-T cells:

KNIT seems great for non-viral CAR-T manufacturing!

KNIT is a double-strand break-free, non-viral, programmable DNA insertion platform. And it’s simple, using just a nickase and a streptavidin recruitment module! This is a machine that creates a nick in the genome, grabs DNA, and inserts it. So cool!

KNIT combines:

But it’s not perfect, and the authors are the first to recognize it:

But this is a big step for gene editing, and even bigger for CAR-T cell engineering. Manufacturing is a big bottleneck, and even small increases might make or break a treatment.

So, go here and read the paper! It’s worth it.

If you made it this far, thank you! What do you think of genetic therapy? Do you think it’s the future of medicine or just hype? Reply and let me know!