Re: >H Protein Antifreeze

From: Robert Bradbury (
Date: Fri Jan 07 2000 - 19:01:34 MST

On 7 Jan 2000, Anders Sandberg wrote:

> "John Clark" <> writes:
> > I have no knowledge of the practical application of Cryonics so I'd like
> > to ask those that do if the following is a nutty idea. As I understand it
> > some otherwise promising cryoprotectants are rendered useless
> > because they're too viscous to infuse at body temperature, much
> > less during cool down. But what if you engineered a bacteria to produce
> > the cryoprotectant and then infused the bacteria; wouldn't that make the
> > viscosity of the substance irrelevant and as a bonus be dirt cheap too?
> You have to time it with the immune system. If it is too active it
> will attack the bacteria and you won't get much use from them.

Anders, I think we have to solve this problem anyway for biobots to
work. Two possibilities are temporary immune system suppression
or reengineering the bacteria so they aren't quite so immunogenic.

> On the other hand, infusing them when the patient is cooling might mean
> they have a too short time window to produce cryoprotectants before they
> freeze too.

This is true, but you could engineer the bacteria so they function
optimally slightly above freezing. So you have the body in relative
"stasis" while the bacteria are happy as clams in mud. Once you
take the temperature down to near freezing you have significantly
longer windows in which to do therapeutic interventions.

> In my opinion the best solution would be to add genes for
> cryoprotectants connected to a hormone promotor from some insect
> either prenatally or through gene therapy during adulthood, and then
> activate the cryoprotectant production when it becomes necessary. Why
> involve bacteria when there are trillions of cells?

Oppps, Anders slips a little (such a rare occurence, that I've got
to grab the fleeting meme). The problem with this is the same problem
with Greg Stock's "added chromosome" discussed at Extro4. Any protein
manufactured in your body post-immune-system training is going to be
viewed as "foreign" (i.e. viral in origin) and will cause a killer
cell attack on those cells producing such proteins. Whether this would
be significant in the time frame of cryonics suspension is questionable,
but you will have to suppress that arm of the immune system on cryonic
reanimation until you purge those proteins. In Greg's approach, you
either have to suppress that arm of the immune system entirely (not good
from the perspective of cancer & viral disease prevention) or find a way
to train the immune system to tolerate any "new" proteins. Not impossible,
but at the edge of current science.

It remains an open question whether or not we will easily (or soon)
achieve reliable exact and specific universal cell modifications
in vivo. I've devoted a fair amount of thought to the problem and
while I think it is doable, I think it requires a significant amount
of protein engineering beyond current state-of-the-art. The problem
with Anders' approach is that you might only get 20-80% of the cells
producing the proteins with current/near-future technologies and that
probably doesn't bode well for reliable reanmimation. On the other
hand you can probably distribute bacteria uniformly enough to get
close to 100% treatment. Getting the bacteria across the blood-brain
barrier might be a bit of a problem, but if the proteins are small
enough, you can probably get the normal transport mechanisms to do
the job for you. Might have to be a bit warmer than 0C for that to
work however. (Of course it is worth noting that current gene
therapies cannot cross the blood-brain barrier as well, you have
to administer the delivery vehicles directly to those parts of
the brain you want to treat.)

I'll note in passing that this isn't a "simple" problem since you
probably want to get the antifreeze proteins in both intracellular
and intercellular spaces so you either have to have bacteria or
cell made proteins be transported around or be producing them in
both locations.

However, John, this was a great reference for me and I will keep this
in the bottom of my tool-kit box.


This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:02:07 MDT