Q&A: How Penn's CAR-T manufacturing discoveryhas drastically reduced timelines
https://endpts.com/qa-penns-car-t-discovery-all-started-with-an-innocent-question/
It takes a minimum of two weeks to pull off some kinds of CAR-T cell therapymanufacturing. But now team at Penn says it has discovered a way to do it inmerely hours. A paper published in Nature Biomedical Engineering andauthored by Penn’s Perelman School of Medicine researchers Michael Milone andSaba Ghassemi builds upon research from 2018 that reduced traditional CAR-Tmanufacturing to just three days.
Traditional manufacturing calls for T cells to be activated to replicate andexpand in number. Milone, Ghassemi and their team used lentiviral vectors thatcome from the HIV virus to deliver the CAR gene to T cells. That process wasn’tefficient, though, so scientists worked to find a way to deliver the genesdirectly to non-activated T cells isolated from the blood. This sped up theprocess dramatically.
The two wrote that they found that functional CAR-T cells can be made within 24 hours“from T cells derived from peripheral blood without the need for T-cellactivation or ex vivo expansion, and that the efficiency of viraltransduction in this process is substantially influenced by the formulation ofthe medium and the surface area-to-volume ratio of the culture vessel.”
Milone took some time on his drive home from the lab last week to talk to meabout what his team is doing, and how it all started.
Congrats on this news about shortening the manufacturing time. This ispretty remarkable.
It’s been sort of a long road to try to reduce this to something that what Ithink could be hours, really. We used 24 hours as our time, but it’s really asclose as you’re going to get to in vivo during a gene transfer, but doing it exvivo. You really just have to mix the virus with the cells long enough for thecells to enter. The time afterwards, the viral genome is not integrated at thetime we inject them because that process takes a very long time, so it’sprobably still occurring as we inject the cells.
I sort of thought … the challenge in gene therapy is the major obstacle of gettingthe cell you want into the gene it needs to go to, and not getting it to go toother cells. So that, to some extent, we mitigate that problem by taking thecells out. We can isolate the cells we want to introduce this into, do that exvivo and put the cells back in. That was first done, probably by StevenRosenberg …back in the early 1990s. But in that case, they were using aretroviral vector.
How did this whole project start? What’s the origin story?
This is kind of an interesting story: So Saba[Ghassemi], she’s actually an engineer by training, came to my lab, has a PhDin mechanical engineering, and really had no biology experience at all when shejoined my lab. I kind of acquired her as a post-doctoral fellow, because wewere collaborating with a group at Columbiaon some mechanobiology-related studies. And she was fascinated by this fieldand so she asked if she could come join by lab.
She had to basically switch her whole field of study into this. This projectoriginated from a naive question that she asked, not knowing anything. She waslearning about these vectors, and she asked, ‘Why do we have to stimulate the Tcells?’ and I said ‘Well, we don’t technically, but the efficiency is low.’ Andso then we started talking about that and saying, ‘How can we make this moreefficient?’
Fortunately, we’ve learned a lot about HIV and how it affects cells, and one ofthe other authors of the paper is an HIV scientist. So this is sort of acollaboration between our labs to see if we could increase the efficiency ofthe process. We basically took what we had learned about HIV over the years andtried to make that process more efficient such that, Sabawas able to go from 1% or 2% transfer into gene cells to 50% transfer, due tosome changes that were quite simple.
One of them was that we tend to culture cells with serum, since HIV has toenter the cell and entering into the cell and the receptor that it binds to,serum kind of internalizes that. We found that if we take away the serum, thatincreases the efficiency. Just the way we culture and keep cells alive inhibitsour ability to do the transduction. Another thing was that the studies we haddone to look at the reverse transcription process, was limited by theconcentration of nucleotides. So just by increasing nucleotide concentrationsin the media by a substantial amount, we could increase the efficiency of that.
Do you know what’s coming next and do you have any inclination of whereyou’re going from here, in terms of developing the approach?
We’re kind of internally discussing ways to do this and how to apply it. Ithink it’s applicable to any of the core programs out there that are doing invivo modifications. There might be modifications that we can make to the virus.When these viral vectors first did this … investigators in Italy basicallytook HIV, and to make it safe, they took out most of the genes from the virus.And when we make the viral vectors, we exclude a lot of the genes that areimportant for HIV biology, and they allow the virus to replicate, but they’renot needed to make the lentiviral vector. But some of those genes mightactually be helpful here. And that’s one of the things we’re going back to, tosee if we can make the process even more efficient.
(20min delay)