Bad News for Minicircle

when mad science becomes sad science

his plasmids aren’t working

So in my last post I made a stab at evaluating Minicircle’s follistatin gene therapy for building muscle, and I threw it out to the clever people in the comments to bring up any issues I hadn’t addressed.

Max Berry (“biology guy”) delivered, with this excellent writeup, well worth reading in full, on why Minicircle’s claims are implausible.

I started out cautiously optimistic — but I’m now convinced Minicircle is a lot worse than I thought.

I was aware that plasmids are “worse” than viral vectors for gene therapy, but I wasn’t aware quite how bad they are.

Plasmids are very easy to have manufactured (I wasn’t aware of that either.) They’re practically a commodity — you send a lab a file with the sequence you want and you get plasmids back.

If you could get cells to produce a desired protein by injecting them with any kind of bare-DNA plasmid (including minicircles), that would replace not only other gene therapies but also vaccines and monoclonal antibodies — almost the entire biologics industry. Why bother with the extremely difficult process of growing cells in vats if you can just inject a plasmid and get the patient to grow ‘em?

And empirically the pharma industry is eager to make this trade — witness the success of mRNA vaccines and therapies.

So there’s an economic argument here — if plasmids (or any of their variants or tweaks) worked, the pharma industry would be using them already.

Or, as Berry vividly puts it:

The point is: if Minicircle’s claimed numbers are even in the ballpark of being correct, this would easily replace nearly the entire biopharma industry, not to mention receiving a handful of Nobel prizes. Most of the drugs on this list could easily be replaced by plasmid gene therapy if it worked. And it’d cost a few dollars, not $50,000-200,000 or more a year.

You might think greedy big pharma wants to build billion-dollar facilities full of specialized equipment and pump expensive reagents into them with a firehose, but I’m sure they can still make a buck selling a plasmid instead. The vast majority of biologists didn’t get into biology to make money – and a good thing, too, because the pay is shit.

The joy of not having to deal with Chinese hamster ovary cells deciding to stick slightly different sugars onto the antibody for no goddamn reason…it’d make any biomanufacturing technician ecstatic.

In fact, when a technology came along that let us skip making biological molecules in vats, we jumped for it - mRNA vaccines make patients into producers of a protein, instead of needing to grow it in bioreactors, vastly simplifying manufacturing. What Minicircle is claiming is a breakthrough more unbelievable, and of greater importance, than a perfected mRNA vaccine.

There’s also an argument from pharmacokinetics that I missed. A small protein like follistatin has a half-life 300x shorter than an antibody. That means that the steady-state level in the blood you’ll get from an injection is 300x lower for a short-lived protein like follistatin than for an antibody.

That means that the problem Minicircle has set itself is 300x harder than the problem vaccine manufacturers have, when they want to produce adequate antibody levels to resist disease. If non-viral DNA therapies can’t work in a vaccine context (and that’s indeed where you most often see them in the clinic, where they tend to fail) then they’re much less likely to succeed at boosting follistatin levels as high as they’re claiming (20 ng/dL — well, actually they say 20 ug/dL, which must be a typo because normal blood levels are measured in nanograms per deciliter, not thousands of times more than that).

Berry also goes into more detail about how Minicircle is making their plasmids based on their preprint — the bottom line is that they don’t seem to have any “special sauce”, the mysterious “transformation polymer” is just standard PEI, they use off-the-shelf components throughout, so there’s no particular reason to believe they’ve cracked the code and figured out a way to make plasmid gene therapy ultra-efficient.

(They certainly don’t have impressive results — as I mentioned earlier, they only have an uncontrolled “study” of paying customers, in which they make modest body composition improvements that would be consistent with, say, starting an exercise program around the same time as they got the injections.)

So yeah. Doesn’t look great.

The one bright spot here is that basically nobody (including Bishop) denies that if you could get adequate increases in follistatin levels in skeletal muscle, you’d get muscle growth. Follistatin makes sense.

The real problem is that plasmid gene therapy doesn’t work (at adequate efficiency).

When somebody comes up with a gene therapy that Just Works (TM), raising follistatin levels may well be a cool thing to try.

(For that matter, as one commenter mentioned, why don’t we just…inject follistatin itself? Maybe cost of manufacturing recombinant protein, maybe rapid degradation means you’d need a lot of injections, maybe there are bad immune reactions? Or more problems I haven’t thought of?)

Anyhow, unfortunately, things look pretty bleak for Minicircle. I wish there were more biohacker-y projects that were actually viable, but this doesn’t seem to be one.