h a l f b a k e r y
Why did I think of that?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
Gene therapy is the mitigation of an undesirable
phenotype by way of modifying the genetic information.
For many years this held enormous promise, however,
much has happened in the clinic.
The reasons for this centre around delivery of the new
genetic information. The best way of getting
cells is by virus. Viruses are really fantastic at this job.
The viruses that scientists tend to pick for this task fall
into two main categories: those which infect and then
express their DNA cargo directly, such as adenoviruses.
those which infect and subsequently incorporate selected
information into the host genome, such as lentivirus,
which is a kind of retro virus.
Each have their own problems, adenovirus expression
tends to be temporary, and repeated doses are far from
cheap/simple solution. Lentiviruses tend to incorporate
themselves into the genome in an indiscriminate
where they may interrupt an important gene... doing
harm than good. Specifically, they may cause cancer.
Mitochondria have their own genome. They used to be
bacteria in the dark history of evolution, before they
captured, and they now live as obligate intra cellular
symbiotes. Most of their genome has been stolen away
and now resides in the dusty library of the nucleus
(carefully robbing them of their independence,
However, a number of genes remain. Almost
these code for the proteins of the electron transport
(ETC). A clever series of proteins that makes eukaryotes
whole lot more efficient. However, there are a number
defects associated with these proteins leading to very
The example I will use is Leber's hereditary optic
neuropathy <link>. This disease is caused by mutations
the mitochondrial genome, all of which have been
which prevent the normal operation of mitochondrial
Now, there are viruses that infect bacteria;
There are even bacteriophage which infect bacteria and
subsequently incorporate themselves into the genome
<link>. This is great, it means that it may be possible to
infect mitochondria with bacteriophage. Then, if you've
been a good little molecular biologist, you can get your
phage to incorporate itself into the mitochondrial
and start expressing a FUNCTIONAL copy of the affected
mitochondrial complex 1 sub unit.
Here's the first problem: The phage will not be able to
infect or enter a eukaryotic cell. It will never reach the
mitochondria. Worse, in a clinical setting it would just
tool around in the bloodstream before getting destroyed
the immune system.
Here's the solution: create an adenovirus (more
an adenovirus associated vector; AAV) which encodes
phage genome. The phage will then be directly
manufactured and constrained within individual human
cells. The phage genomes will fit; AAV inserts go to
phage genomes are rarely bigger than 4.5, and we may
need all of it.
Then, your AAV infects human cells, it's DNA starts being
transcribed and replicated, the Phage then self-
inside the host cell, the functional phage infects the
mitochondria, incorporates into the genome and then
SOME of the complex 1 subunits begin to function
There should be no problems with causing cancer: while
there are MANY mitochondrial proteins associated with
control of cell death etc. They are all encoded for by
nuclear genes. There should be no problems with the
you can give pretty massive doses of AAV which are
replication-deficient. The phage should only ever be
present INSIDE cells, so there should be no immune
1. There may be no phage capable of infecting
mitochondria; it's been a while since they were free-
and so their outer membrane may have become a little
specialized (would make a cool paper to infect mitos w/a
micro-injected phage though).
2. The phage may not self-assemble in a eukaryotic cell:
phage proteins may not be correctly processed and get
stuck in the ER/fold incorrectly etc.
3. There may not be a way of controlling where the
incorperates its DNA. You may fix complex1, but destroy
complex 2. This may be mitigated by the multiple copies
of mitochondrial DNA.... you should end up with a full
complement of functional proteins, even if they're from
diverse DNA molecules.
Leber's hereditary optic neuropathy
[bs0u0155, May 08 2013]
Phage genome incorperation
[bs0u0155, May 08 2013]
Great though this idea is...
...it's practically identical to the one immediately below it. (Copy-number variation?) [Wrongfellow, May 11 2013]
||Viral evolution is pretty snappy. A free-living
mitochondria-like bacterium would be a useful
development model. Then, just culture it in an
Intracellular-like medium and screen.
||Damn. Not to detract in any way from the idea,
but I had this self same thought a while back, and
it's still on my list of potential "cool shit" projects
to try someday.
||With regard to your three problems:
||(1) finding a phage that can infect mitochondria.
Easy, as long as you can find one that infects them
to even a very limited extent. Just cook purified
mitochondria and phage in a cytosol-like soup;
spin down the mitochondria; extract the phage
DNA; amplify and mutate; repeat for several
cycles. A complication is that you probably want a
phage that can be grown up on an easy bacterial
host, as well as infecting mitochondria; but this is
||(2) self-assembly in eukaryotic cells. Again, in
vitro selection should work; just purify any intact
(packaged) phage from infected cells, mutate and
||(3) This will be fixable; it's just Lego.
//No [beanangel] ?// If this were a [beanangel]
idea, it would go something like:
||"Research reported in WallMart Weekly suggests
that mitochondrial oldening is a key factor at the
ageing process. People could be made happylong
by transmitting improved mitochondrial genome
sequences, with extra sweetness molecules, via
nanotube receivers using phased interference.
Handyphones could be made to transmit the
sequences, if two were used on opposite sides of
the body. it would also be possible to discover
the mitocodes of beatiful women an upload to
less pretty people, possibly crowdsourced."
||Have you just bunned [beanangel]'s version of this idea, [MB]? He wouldn't've put any capitals into the title.
||No, I was bunning [bsetc]'s.
||It is an interesting question as to whether mitochondria suffer from phage. They might be pretty safe in there, all domesticated. Complacent.
||If this nested vector thing works it could work for other genetic engineering projects too. The idea of sidestepping cancer by targeting the mitochondria is super slick.
||Most of the proteins needed in eukaryotes are encoded in the nucleus. Messenger RNA is imported into the organelle where it is translated into the protein.
It seems to me that rather than trying to get DNA in there one might use the host cell machinery:
||1) Introduce DNA to the cell as a plasmid. If memory serves these sometimes and somehow end up in the nucleus and can be transcribed. The plasmid also carries both selection and counter-selection.
2) Plasmid encodes RNA tagged for transport to mitochondria: the mitochondrial genome patch, a reverse transcriptase and a recombinase.
3) All three RNAs are transferred. The reverse transcriptase and recombinase are translated to produce proteins.
4) The genome patch is reverse transcribed (it has a signal sequence to prime transcription, the other mRNAs do not).
5) The DNA patch is integrated into the Mt genome by the recombinase.
6) This process is kept up for long enough to convert most copies of the genome.
7)The cells are grown. The plasmid cannot replicate, so is lost from some cells.
8) These plasmid-free cells are selected by some counter-selection.
9) The cells are grown some more to create clonal lines
10)The cells are screened for successful integrants by PCR.
||The human mitochondrial genome is only around 17 kb. That's smaller than many eukaryotic genes. It might be easier to try for a complete replacement rather than patching. I can imagine other issues with that, though.
||[Loris] I like the idea of using the intrinsic import
machinery to get an RNA version into the
mitochondrion - that's a neat idea. It depends on
a bit more stuff going on in the mitochondrion in
order to produce the DNA, but the simplification
of the import process is a huge potential
||But I'm not sure about your ideas for
selection/ counterselection/ screening. Surely
what we're after is whole-body, in situ
mitochondrial genome relacement (or patching)?
In which case screening or selection are pretty
much out, unless you wanted to kill off a small
minority of cells that hadn't received the
||Heh, I usually work with bacteria. They're just easier.
||I suppose one could try an RNA virus vector's packaging machinery for a one-shot conversion, or something like that. I'd be a bit wary of putting something into human cells which would be modifying the genome and wouldn't be eliminated afterwards.
||I don't think it would matter if some part of the
system replicated in vivo; in fact you might need
the whole shebang to replicate in order to achieve
high enough infection levels (not my area,
||Whatever you do, it'll be easy enough (compared
to everything else) to put an Achilles' heel into
the carrier virus so you can zap it once its job's
done. Or, obviously, the immune system would
clear it - though hopefully not before it's done
that thang it does.
||There is no way a eukaryotic nucleus is going to transcribe what is on a random plasmid. Bacteria sometimes make with that sort of anonymous bathhouse exchange and sometimes it works for them but eukaryotes have too much at risk.
||//The human mitochondrial genome is only around
17 kb. That's smaller than many eukaryotic genes.
It might be easier to try for a complete
replacement rather than patching. I can imagine
other issues with that//
||I'm not trying to replace the whole mitochondrial
genome, admittedly there are several phage that
are big enough (T4), and there are hundreds of
very large mammalian virus genomes (vaccinia),
however that just steps up the difficulty. Working
with Lambda/AAV is pretty trivial to the right
||I'm looking for a replacement of a single defective
protein (a single subiunit of C1 in my LHON
||incidentally, LHON was a good example, and the
mitochondrial ETC is VERY amenable to screening.
||Once you've made your custom phage, you can simply
microinject them to LHON-HeLa or whatnot, and use
the suite of crazily specific complex inhibitors.... 3NP
for CII for example. If your cells live, your onto a
||//There is no way a eukaryotic nucleus is going to transcribe what is on a random plasmid.//
||Well, yeah. I don't think you understand my proposal. I think we can take it as read that this requires carefully engineered sequence to provide the functionality required, however it's done.
||//I'm not trying to replace the whole mitochondrial genome ... I'm looking for a replacement of a single defective protein (a single subiunit of C1 in my LHON example).//
||The bulk of my comment covered 'patching' the mitochondrial genome. I was just pointing out at the end that if you can perform a full replacement then you have a 'one size fits all' treatment. That is, there'd be no need to match the treatment to the particular mutation, it would just work for everybody.
The potential disadvantages I alluded to are that the larger RNA would be harder to generate reliably, and the larger the replacement sequence the greater the risk of introducing other mutations.
||//requires carefully engineered sequence to provide
the functionality required, however it's done.//
Exactly. Vectors for transient expression in
mammalian cells are well-known. That is not the
||Wrongfellow: Thanks for the HB link. I did a quick
search on this but didn't catch it. I think there is
sufficient novelty here to differentiate the two
ideas. The differences are more genOME therapy vs.
gene therapy. In addition, the nested nature of dual
independent virus vectors is new here. It gets
around the delivery issues associated w/the straight
up phage idea.
||Oooo! Oooo! How about this idea! Take out stem cells and put in whole new mitochondria. Just get them from other cells and plunk them in there. You might not even need to get rid of the defective ones first. Then transplant cells back! The new mitos will plug along shoulder to shoulder with the old, picking up their slack.
||No need for dna jiggery-pokery! Just good old cell microsurgery.
||An anno from the linked idea, by the formerly
||//I'd thought of the problem of getting a virus
that'll get into both eukaryotic and prokaryotic-
like cells. What we need is a Trojan Horse virus.
basically, you make a phage genome but you
package it in vitro with eukaryotic viral coat
proteins. This infects the eukaryotic cell. The
phage genome encodes the packaging proteins
necessary to re-package it as an active phage to
enter the mitochondria. With a nifty bit of coding,
you could (hopefully) ensure that the phage-coat-
encoding genes were inactive in the
mitochondrion (by suitable choice of promotors
perhaps). The phage would also encode, of
course, a fully active mitochondrion.//
||Very nice, I thought that at about the same
moment*. Mitochondria, when injected into cells
do retain some functionality... they move and fuse
as normal, in fact these experiments were among
the first to demonstrate mitochondrial fusion. This
was back in the days when we couldn't switch
fluorophores on and off. (take a cell w/red
mitochondria, remove mitos, put in cell w/green
mitos observe as some mitos become red and
||There's also research showing how stem cells can
actually transfer mitos to other cells (not sure it's
published, saw it at a conference).
||So this could work in two ways: 1. By simply
populating the body with cells derived from the
healthy-mitochondria containing stem cells. 2. By
actively transferring healthy mitos to certain
||Some problems that are yet to be resolved: 1. We
don't know if transplanted mitos won't be
selectively degraded over time... i.e. is there and
intracellular immunity governing self/not self
decisions, and is that somehow based upon the
mitochondrial genome. 2. Without selective
pressure, would the donor mitochondrial DNA just
be so dilute as to be negligible.
||*lies, I thought that as soon as I read about 2
words of your comment ;-)
||You want the mitos to be self? Engineer your own exvivo. You could do it in fibroblast culture. Once you get the one you want, sequence it to make sure it has not acquired weirdness. Then back home it goes.
||Yeah, I phrased that badly. I think it's pretty
interesting as to whether there IS a form of
intracellular immunity, there would have to be for
non-self mitos to be an issue. As it is, mitos are
pretty non-self. In crush/burn/necrosis injuries
where a lot of mitochondrial content ends up in
the circulation... you can end up dead from
immune overreaction to mtDNA and protein
fragments with formylmethionine on them.
||This is either an Achillies' heel or great idea,
depending upon which angle you look at it.
||What could be interesting, is that several diseases
are HUGELY potentiated when present on a
background of certain subtypes of mitochondrial
genome. So, you could ameliorate a few diseases
simply by transplanting in the least-worst subtype.
||Additionally, some mtDNA types are associated
with different ATP generating efficiencies... i.e.
some mitos are tightly coupled (many ATP) some
are loosely coupled (some ATP, some heat). So, if
you're planning some super cool Tour de France
cheating, you could transplant in muscle-specific
stem cells armed with central African mitos.
||Just thinking technical details, you could probably purge the cells of native mito with something like chloramphenicol. The cells would do glycolysis until the new mitos took. The new mitos would give a metabolic advantage (aerobic respiration) and allow succesfully mito transplanted cells to overgrow the non-mito cells.
||That's one way, another would be antimycin A, it's
non-competitive, so you could give a quick antimycin
pulse, wash, then inject the new mitos and the old
ones should be degraded.
||It would be very embarrassing if you purged all the
native mitochondria, and then found that your
replantation didn't work in all cell types. Brains
don't really do glycolysis.
||It would be safer to make the new mitochondria
resistant to a drug which is toxic to normal
mitochondria, so you could wait and see before
burning your bridge over troubled water.
||Incidentally, cells regulate their numbers of
mitochondria fairly tightly, so a few rounds of
replacement would dilute out most of the native
||you're forgetting about the ongoing mitochondrial
dynamics, fusion-fission is going to make chimeric
hybrids out of all of them.
||//fusion-fission is going to make chimeric hybrids
out of all of them.// Ah yes. Well, in that case just
keep dosing with new genomes and dilute out the
||Ha. The Donor genome should code for a restriction
enzyme. The donor genome should not be cut by the
enzyme. The recipient genome should be.
||(obligatory quote every time something like this comes up)
Tyrell: The facts of life. To make an alteration in the evolvment of an organic life system is fatal. A coding sequence cannot be revised once it's been established.
Roy: Why not?
Tyrell: Because by the second day of incubation, any cells that have undergone reversion mutations give rise to revertant colonies like rats leaving a sinking ship. Then the ship sinks.
Roy: What about EMS recombination?
Tyrell: We've already tried it. Ethyl methane sulfonate as an alkylating agent a potent mutagen It created a virus so lethal the subject was dead before he left the table.
Roy: Then a repressive protein that blocks the operating cells -
Tyrell: Wouldn't obstruct replication, but it does give rise to an error in replication so that the newly formed DNA strand carries the mutation and you've got a virus again. But, uh, this-- all of this is academic. You were made as well as we could make you.