T-Cell Gene Therapy Delivery System

T-Cell Primer

Skip this section if you already know how T-Cells work, broadly speaking.

Every cell in your body (it is biology, there are exceptions) turns RNA into proteins. When it does this, most of those proteins go off to do protein things. Some small portion however are chopped up into short fragments (peptides). The cell then takes these fragments and presents them on its surface (in a MHC complexes). Every T-Cell in your body is programmed to look for a randomly chosen specific peptide, that isn't one of your naturally occurring peptides, presented on any cells in your body. When it finds a match it does two things:

1. It binds to the cell and delivers a toxic payload to the cytoplasm.
2. It undergoes clonal expansion so there are more T-Cells looking for that particular peptide.

Cool T-Cells Features

Skip this section if you already think T-Cells are cool

There are several nice properties that we humans spend lots of money trying to reproduce.

1. They are highly specific. Without this specificity, T-Cells would attack healthy cells, causing autoimmune disease.
2. They can target cells based on proteins they produce internally, not only surface markers.
3. They self expand, so you don't need to manufacture a lot to reach a lot of cells.
4. During clonal expansion, they end up everywhere in your body no matter where they found their first hit.
5. They can deliver a payload into the cytoplasm. Lots of effort has gone into things like LNPs for getting stuff into the cell, T-Cells already can do this reliably.

Leveraging T-Cells

To utilize T-Cells to deliver a gene therapy we need to do two pieces of engineering:

1. Engineer the T-Cell to target a peptide of interest to us that occurs naturally in the body, rather than a random non-self peptide.
2. Change the cargo it delivers from a cytotoxic payload to something we want delivered.

The first one is already done by people in the form of CAR-T engineering, so not a huge lift. The second one is more novel, but in theory should be possible. We would need to stop the engineered T-Cells from producing or packaging granzyme, and instead convince them to produce/package some other thing of interest to us. Initially, this likely will be just a protein but perhaps future efforts could try to get mRNA or even DNA produced and packaged.

First Step - GFP

The first obvious research step here is to produce engineered T-Cells that can deliver GFP to cells instead of granzyme. This would allow us to verify that the general technique can work, at least in theory. It would require not only changing the targeting of the engineered T-Cells to a self protein, but it would also require preventing delivery of cytotoxic payloads and introducing the delivery of a novel payload. There may be a step before this where the cytotoxic payload is still included but the delivery is to cytotoxic resistant cells if that turns out to be easier/cheaper, but the first major milestone is being able to deliver GFP in vivo so we can see the whole system working in a live organism.

Second Step - Knockout

The next thing to do after this is an actual gene therapy. Knockouts are easiest because we can do them with a single nucleotide change, which means we can use a protein-only gene editor like a recombinase, zinc finger, TALEN, etc. What is particularly interesting here is that once you knockout the protein of interest, it will no longer be produced by the cell and thus stop presenting the targeting peptide! This means that the T-Cells can undergo clonal expansion and eventually reach every cell in the body, and then stop expanding because there is no more work to do. This feature in particular is the most interesting thing about using T-Cells as a delivery vehicle, because it means we can target every cell in the body about once each. Most existing gene therapy delivery systems will happily deliver your delivery system to the same cell over and over again, which means if you want to reach every cell in the body you need to produce a huge volume of your delivery system if you want to reach those last few cells.

For in vivo experimentation and eventual human translation, we can knock out a protein that we believe is unnecessary late in life or potentially detrimental. A good example of such a protein would be myostatin, a muscle growth inhibitor. Without this protein mammals are leaner and more muscular and the side effects are relatively minimal (perhaps some issues with tendon stiffness that needs further investigation).

Third Step - Knock-in

One issue with this concept is that you can only target cells that are already expressing a protein of interest, and you need to knockout that protein so the T-Cells don't expand forever. However, we can work around this limitation by delivering both a knock-out and a knock-in at the same time to the same cell. So along with knocking out a protein like myostatin, we would also deliver an editor that would add some new gene. This is notably more complex because you now need a guide RNA or a very large recombinase. However, if you can manage to come up with a deliverable editor, we can do a knockout and a knock-in at the same time.

Fourth Step - Repeat Dosing

The above strategy has an obvious problem which is that you cannot redose. Once you knockout myostatin, if you want to knock-in something new you need to find a new target to knockout in parallel. If you want to do a lot of gene therapies, then you'll need to find a lot of genes that you can safely knockout and you may quickly run out. To address this problem, we can add a third payload to our T-Cell delivery package: a gene that produces a protein that is immune compatible with the individual, but also doesn't do anything interesting. We can then use the produced protein as our knockout target for future therapies! Each time you do a therapy you will thus do 4 things:

1. Target a particular protein.
2. Knockout the targeted protein.
3. Deliver your actual therapy of interest.
4. Add a placeholder gene that can be used for targeting in the future rounds of gene therapy.