Subcloning: Loosing my favorite gene

Peter Ellis via (by pjie2 from
Tue Feb 13 13:50:35 EST 2007

Allan wrote:
> Hello everyone,
> I'm hoping for some advice.  I have cloned a gene into pUC19 and have
> sequenced.  It is perfect.  I have linearized it at an obscure cut
> site (SgrAI)

This cut site is in the gene insert, not the vector.

> and phosphatase treated the resultant 5' protrusions with
> Antarctic phosphatase.  The linearized product runs at the desired
> size (6.3 KB).

However, any small proportion of uninserted vector (which there will be in 
any plasmid prep due to natural revertants) will not have cut, and will 
remain as viable circular plasmid.  This may be far too small a proportion 
to be visible on a gel.

> Upon ligating GFP with sticky ends complimentary to
> the cut site in the gene  (GFP is cut out from a cloning vector
> construct) I get transformants.

Yes, but these transformants simply contain the whole, nonlinearised, 
noninserted vector.  Your ligation is failing, and thus only the tiny 
residual proportion of uninserted, uncut vector is able to transform.  The 
ligation is the step you need to work on, whether that be different 
conditions, a new batch of enzyme, or whatever.

>  I get no transformants on my negative
> control:  linearized, phosphatased gene of interest.

What do you mean by a linearized gene?  Genes aren't naturally circular.  I 
presume you mean that you've cut the gene with SgrAI.

> Now, upon restriction digestion with KpnI I should get  two products:
> pUC19 (2.7 KB) and my favorite gene+cYFP (4.5 KB) I ONLY GET a single
> 3 KB band.
> After some fighting with this problem I sequenced the
> resultant 3 KB plasmids.  They did not contain my gene.  They
> contained pUC19.

The colonies you call transformants are actually derived from residual 
pUC19, which did not get cut by SgrAI and thus remained as viable, 
uninserted vector.   Your ligation failed, and thus only the uninserted 
vector was carried through the subsequent transformation.

> So it appears that I am getting a massive deletion.  Does this sort of
> thing happen frequently?  Given how easy it was to clone the gene I am
> in disbelief how difficult it has been to introduce the ~800bp GFP.

You should ligate the GFP to your gene before cloning it into the plasmid.

Are you trying to make a fusion protein, or just put the one gene downstream 
of the other?  If the latter, you could continue with the existing strategy, 
but linearise using a different enzyme that cuts the vector rather than your 
gene of interest.  That at least would get rid of the background noise of 
vector-only transformants.  It won't do anything about the fact that your 
GFP ligation currently isn't working, though.


More information about the Methods mailing list