Biopsy a malignant tumor from the patient, place the tissue in an incubator, infect the tissue with a GM virus (The GM virus was created by removal of a gene (called E1B-19kD) that viruses use to disguise themselves, and which prevents cells from noticing they have been infected.), let them pick up the genetic material from the cancer cell, then inject them into the patient.
-- mhuppertz, Aug 20 2010

You must watch this TED talk. http://www.ted.com/...re_of_medicine.html
MUST! [2 fries shy of a happy meal, Aug 21 2010]

Or read the science-fiction story where the guy dying of cancer is sent out to meet the killer robot fleet, and slips them a tumor when they ask for a tissue sample, "just out of curiosity, why do you ask, human?" (Saberhagen's Berserker series.)

Although giving us a gene number is a lot better than most GM handwaving, I've gotta say this is probably obvious to the gene-tamperers. If it would work, I mean.
-- baconbrain, Aug 20 2010


// obvious to the gene-tamperers //

Hey, [MB], you're up.
-- 8th of 7, Aug 20 2010


This is good apart from making no sense.

Setting aside the obvious radical oversimplifications here, how and why do you want the viruses to pick up DNA from the cancer cell? Viruses don't do this in general (though some can, rarely). And how would DNA from the cancer cell make the virus kill other cancer cells? And since almost all the DNA in a cancer cell is identical to that in normal cells...

...I mean, basically, what are you talking about?
-- MaxwellBuchanan, Aug 20 2010


// This is good apart from making no sense. //

[Marked-for-tagline]
-- 8th of 7, Aug 20 2010


I don't think this globally "doesn't make sense," it's just got so many strong assumptions that it probably wouldn't work, plus a large unexplained detail (vs. another assumption). The assumptions are:

1)Cancerous cells are normally killed by an immune reaction against mutated genetic material or products of the mutated gene. 2) "Successful" cancers succeed by escaping this immune surveillance. 3) If the immune response were reactivated at a much later than normal stage of the cancer it would still be effective

The idea is to extract the mutated gene from cancer cells, and express it in normal cells lacking the ability to evade the immune response (using a virus as a vector)

It's assumed (4) that this would provoke an immune response which would (5) be curative, or, at least, therapeutic.

Of course, if the virus gives the normal cells the ability to evade an immune response, the exercise is futile: that's why the virus must lack E1B-19kD (or whatever else it normally uses for that purpose).

It is further assumed that 6a) expression of any non- cancerous genes transferred to the non-cancer cells will be harmless or 6b) some unspecified technique will be used to select only vectors containing the cancer gene.

The logic seems fine -- the problem, so it seems to me, is with the premises. (Of course deliberately transducing normal cells with a cancer gene seems like an odd way to treat cancer.)
-- mouseposture, Aug 21 2010


Geez, when did this site become "bash the contributor"? It was just a thought. I know that viruses splice their DNA into the DNA of the invaded cell, and that's where I was going with this.
-- mhuppertz, Aug 21 2010


// when did this site become "bash the contributor"? //

It's kind of always been that way; your problem with this idea is that there's more than one HalfBaker with serious in-depth knowledge of this topic.

If you're going to post an idea like this, you'd better be damned sure of your biochemistry; the telegraph wires are crowded from end to end with impatient vultures, just aching for a chance to flap down and peck at your liver while it's still warm.
-- 8th of 7, Aug 21 2010


s/biochemistry/molecular immunology/ Mmmm, warm liver. Yummm.

Sorry, [mhuppertz], it was actually intended as a *defense* of the idea, though I guess it didn't seem that way.
-- mouseposture, Aug 21 2010


I for one liked your discussion of actual assumptions amidst the general snarkiness, mouseposture. Thanks.

> Close enough to be [Marked-for-Deletion] ?

It's only magic if the author doesn't care how the details actually work - here, I think the author is thinking about specific mechanisms, and has some model that the Virus is somehow "trained" to attack a specific kind of cell, or maybe stores (parts of?) that target cell in its own DNA, and we could probably talk in more detail about how this does, or doesn't, work.
-- jutta, Aug 22 2010


//general snarkiness// He's my kinda soldier.
-- MaxwellBuchanan, Aug 22 2010


Well, it's certainly half-baked.

There's so many large gaps in the description that we're filling them in with different models. And none of those even seem plausible.

mhuppertz, could you expand on your description?

Yes, retroviruses (a subset of all viruses) integrate into the host's DNA. But how does that help the cancer patient?
* Are the viruses supposed to target cancer cells after growing on them for a few generations? - That's rather unlikely.
* Is the immune system supposed to target uninfected cancer cells after exposure to infected cells? Again, probably won't happen.

A quick google suggests that the gene you mention knocking out inhibits apoptosis (the cell killing itself in response to being infected). So non-cancerous cells wouldn't propagate the virus, they'd kill themselves instead. So is the idea that the virus would run amok inside the tumours (since cancer cells don't undergo programmed cell death)? That may well happen, but you'd still be producing stackloads of infective virus particles, each capable of killing a healthy cell. Then the immune system would kick in, mop up all virons everywhere and you'd be back to square one. Only sicker.
Also, if that's the plan, you don't need biopsy/preincubation step - you might as well just introduce the virus to the tumour directly.
-- Loris, Aug 22 2010



random, halfbakery