Gene knockout as a method to selectively kill cells?

[mikeyboy]

Anyone know of a human gene that could be knocked out in genetically engineered cells so that those cells and their clones may be killed with a local injection of some chemical? Or, perhaps a gene that can be added that would cause those cells to be sensitive to some chemical?

I am supposed to design a theoretical tissue engineering construct, and my goal is to implant engineered cells that will produce a cytokine which will promote tendon and ligament healing. But a potential problem is that of oncogenesis, and I was thinking of letting the cells recruite the fibroblasts and secrete the necessary proteins, then kill off those cells by injecting a chemical to the wound site. So the chemical should be normally safe to human cells. It would be sort of the opposite of adding antibiotic resistance along with some other gene to select for those cells. One idea I had was a membrane protein which would cause the cell to take up the chemical but other cells would ignore it.

Any ideas?

---

Members don't see this ad.

 

[exmike]

If you killed off the fibroblasts, what cells would maintain the collagen/ecm matrix afterwards?

 

[duck2005]

What chemical are you using to treat the cells? Can you use a protein instead? If so, you can (a) create a transgenic line with a cDNA insert of viral receptors. insert the DNA behind a promoter that is active in the cells you wish to treat, but not all cells. (b) insert the gene seq for the protein treatment into a viral vector (c) infect the cells with the virus.

For example, if someone wanted cells in organ X to be selectively treated with a protein: insert the gene for a viral receptor behind a promotor that is active only in organ X. infect the organism with a virus harboring the genomic DNA (or RNA) for the protein treatment. the virus will only be able to attack those cells that have receptors for it. the only cells expressing the receptor will be those in organ X. so only organ X will be infected by the virus, and only organ X will express the protein of therapeutic interest. of course, its much easier said than done...b/c your DNA constructs will be to be very meticulously made..

if your chemical is not a protein or something of that sort, it may be more difficult..

 

[mikeyboy]

Well, I wasn't planning on killing off the fibroblasts, rather the cells that I used as a bioreactor to produce a growth factor promoting angiogenesis. I don't want the cells to remain there forever, since they are producing excess growth factors (e.g. BMP-12, which promotes collagen formation rather than bone) or other proteins (e.g. a decorin inhibitor which has shown to help healing ligaments). Ideally I would want the healed tissue to be as close to natural as possible. So, the implanted cells would be used primarily as a recruiter and to speed along the tissue formation, but would eventually disappear (through a follow-up treatment), as would the scaffold they were originally seeded on.

Another alternative would be to use mesenchymal stem cells and try to promote their differentiation into tendon or ligaments, but this is probably much more difficult and less efficient. I want to stay away from simple viral delivery because, well, we are supposed to come up with something new, and it has its own issues.

A follow-up viral infection which kills selective cells is interesting, though, and I'll look into that. The reason I wanted to stay away from it is the possibility of the virus mutating and being able to bind to other DNA, but I guess in a theoretical project this risk can be simply noted.

 

[Pinkertinkle]

Is this a BioE final project?

 

[exmike]

Maybe you can insert a gene for a non-native FasL or TNF receptor and selectively initiate apoptosis in those cells when they're no longer needed.

 

[thewebthsp]

This is really out there, but maybe if you knew you had a unique sequence of DNA inserted into the added cells you might selectively kill off the introduced cells with siRNAs?

 

[exmike]

This is really out there, but maybe if you knew you had a unique sequence of DNA inserted into the added cells you might selectively kill off the introduced cells with siRNAs?

ooh, interesting!

 

[jlee9531]

damn so hard to follow but still very interesting to read....

 

[uclacrewdude]

ooh, idea! this is gonna sound really lame compared to all you real researchers out there, but make the response element require more essential amino acids, or preferentially one type of aa. in that way, diet could partially modulate whatever gene you need to be responded to.

 

[SaltySqueegee]

Biochemistry and Molec Cell Bio is da'Bomb! ...Technically speaking... Talk about a cool arsenal of molecular tools!

 

[Gleevec]

Have you considered using a Cre-Lox expression system under the control of a tetracycline promoter?

Basically, surround some vital gene in the cells youre injecting (like DNA polymerase or topoisomerase or something vital to replication) with Lox sites. Place cre under the control of a tetracycline promoter. When you administer tet, it will cause cre to cut out the vital gene (which is surrounded by lox sites) so the cells you injected cant reproduce anymore. The beauty of this is that it will only kill off the transgenic cells, as lox is a bacterial sequence not present endogenously in mammals, while leaving the majority of other cells perfectly intact (more or less, since tet exposure might have some minor effect on surrounding tissue).

I know this works in cell culture, but I have no idea whatsoever if this will work in vivo. Sounds nice in theory at the bench, but as usual, might have some big problems at the bedside.

Anyone know of a human gene that could be knocked out in genetically engineered cells so that those cells and their clones may be killed with a local injection of some chemical? Or, perhaps a gene that can be added that would cause those cells to be sensitive to some chemical?

I am supposed to design a theoretical tissue engineering construct, and my goal is to implant engineered cells that will produce a cytokine which will promote tendon and ligament healing. But a potential problem is that of oncogenesis, and I was thinking of letting the cells recruite the fibroblasts and secrete the necessary proteins, then kill off those cells by injecting a chemical to the wound site. So the chemical should be normally safe to human cells. It would be sort of the opposite of adding antibiotic resistance along with some other gene to select for those cells. One idea I had was a membrane protein which would cause the cell to take up the chemical but other cells would ignore it.

Any ideas?

 

[Treg]

One major caveat to your experimental design is that if you "kill" off these cells in situ, you will create a pool of debris that will recruit APCs to the site, leading to inflammation (i.e. secretion of proinflammatory cytokines like IL-1B and TNFa that will confound your expected outcome).

If I were you, I would use the tet system to do this. I would take a two step approach where the cells would be growth regulatable (i.e. they express SV40 T antigen in the absence of tetracycline, and when you add the drug they stop proliferating). Then I would make them transgenic for my cytokine, again under the control of the tet response element, so that when I added the drug to stop growth I would also stop the cytokine secretion. This would also allow you to titrate the dose of the cytokine by varying the dose of the drug.

A few major considerations-

1. If you want relatively short term effects, use the tet-on system, so you would only have growth and cytokine secretion in the presence of tetracycline (no long term drug usage required)

2. There are side effects associated with tet usage in vivo, so most people use doxycycline, a less toxic and more potent version of tet

3. Make sure your cells are MHC matched

Good luck!

Treg

 

[stoic]

I think a major problem here is the MHC matching. It might be most practical to culture the appropriate cells from your "patient" instead of using a commercially available cell line.

It would be important to choose cells that could be localized to the injury site (otherwise you would lose the localization of desired effect). In fact, it would be really important to place the BMP12 secreting cells appropriately... otherwise you'd have a hard time getting the desired effect. Maybe you could induce mild inflamation at the injured site (if it isn't present already), harvest some monocytes, transfect them, and then reintroduce into the site.

I think your best bet would be to take your cultured cells and transfect them with the cDNA for BMP-12 under the control of a Rapamycin responsive promotor. I'd also tranfect with a constiutively active neomycin resistance to use for postive transfection selection. You could then reintroduce your cultured cells, administer rapamycin to the patient, observe the effects, and discontinue when needed. Hmm.... the only problem with that might be getting the rapamycin to the injured tissue... ligaments and tendons don't tend to be highly vascular. Using this system you'd have to check up on the physiological concentrations needed to activate expression and make sure they were safe for the patient.

Here is a link to a review paper that does a decent job describing several different post transfection expression control methods. http://www.scienceinchina.com/ky/0010/ky0865.pdf

Good luck let us know what you decide to do!

Dave