Antisense for gene ablation in mice

Stacy Ferguson sef at
Sat Oct 21 13:08:20 EST 1995

In article <4694dt$p1l at>,
 <mcatania at> wrote:
>I am wondering if anyone knows if creating a transgenic mouse
>expressing an antisense RNA would be feasible in leiu of a transgenic
>knockout.  I have searched the literature, but can find no instances
>of using this method for in vivo gene function analysis beyond the
>embryo stage.  
>I appreciate any comments anyone may have.
>       Michael

I guess the real question is, why would you want to do this? First,
the block probably is never 100% (although functionally it may be)
since you are relying on hybridization kinetics. A knockout is a 
knockout, so there can't be any gene product made. Second, it's 
possible that by using an anti-sense approach, you'd not only block 
the product you want to block but related ones as well (for example,
do you really want to block ALL transcription factors of a given family?
Blocking an entire family of genes is likely to be lethal. 

Here's an example. Let's say you've got a family of 5 related genes,
named "a" through "e". You make an anti-sense construct intended to block
gene "a". Your mouse is now polka-dotted :) Does that mean that gene 
"a" is important in coat color? Not unless you go back and prove that 
the products of genes "b" through "e" are unaffected. After all, if the
genes share homology with your gene, then so do their transcripts. 

Let's say you make your antisense transgenic and nothing happens. Does
that mean your gene is unimportant? Not necessarily. What if the block 
is incomplete and only a tiny amount of your protein is required for 
the cell to function normally? What if you don't have an antibody to 
your protein to find that out? What if you have an antibody and 
Western blotting techiniques aren't sensitive enough to detect those
tiny levels of protein?

You're then stuck with trying to balance things out. You want enough
of your antisense transcripts made to block your gene. However, since
you'll never completely block it without using saturating levels of
antisense production, you'll have to hit levels where less homologous
genes will start being blocked to (because kinetically, the more 
"stuff" you have, the more likely it is to bind to things it won't
bind to efficiently at lower levels).

If you knock out a gene, you CAN'T make the protein the gene encodes.
You can't get more complete than that and since you're going to go through
a heck of a lot of work to make either kind of mouse, you might as well
go with the cleaner system. You're not likely to inhibit the function 
of genes similar to the one you're studying so you don't have to 
worry about them. 

The following is a true story, not involving anti-sense transgenics, 
but similar in principle. 

I've been working on a knockout mouse. It has a screwy immune system (there
is poor antibody production). Another lab has a transgenic mouse that 
encodes soluble protein that's capable of blocking the function of the
protein that's knocked out in our mice. Our knockout and their transgenic
have very similar phenotypes. The difference is that ours is more 
severe. Well, our mouse is missing the gene. Their mouse has plenty of
that gene product but the product is being blocked. Is that blockage 
partial or complete? It's likely to be partial, as the symptoms are
less severe and there's really no way to prove that you've completely
inhibited the function of every single molecule unless you've gotten
rid of it entirely. In fact, it's kinetically unlikely, since saturating
levels of the competitor would be difficult to reach in vivo. It's not that
their mouse wasn't informative regarding the function of the gene, but 
ours is moreso. 

Hope this explanation makes some sense. 


Stacy Ferguson
sef at

More information about the Methods mailing list