Question on Plasmid Cloning

John Dixon jpcd100 at
Mon Aug 23 11:38:47 EST 1999

> Student <mucineer at> wrote (snipped):
> >> is there anything in the process that prevent different molecule of the
> >> insert from "self-ligating" to form species such as
> >> HindIII-------BamHI--------HindIII or BamHI--------HindIII------BamHI?
> The answer is that NOTHING prevents this happening in the ligation:
> indeed it is bound to happen unless you do some clever phosphatasing*.
> However, to make a recombinant plasmid you must first make trimolecular
> HindIII-------BamHI--------HindIII--------BamHI forms of the insert in
> order for it to ligate into the vector to complete the circle. Note
> that the insert will now have two overlapping palindromic
> (head-to-head) sequences. HOWEVER, despite an earlier respondent's
> claim, you NEVER recover such molecules intact after transformation
> because standard E coli cloning hosts will not tolerate palindromic
> sequences above a certain small size (I am excluding insert sizes of,
> say, < 150 bp). You cannot clone perfect uninterrupted palindromes in E
> coli: this is thoroughly documented. In fact, a cloning system was
> devised precisely to take advantage of the inviability of uninterrupted
> palindromes:
>What you can get, however, is a recombinant in which a mutational event
>has disrupted the palindrome(s) in some way so as to restore viability.
>I have never seen this happen despite giving ample opportunity.


Hi all,

I have been working with some plasmids that contain two HSVtk genes
which are in the opposite orientation to each other (for eventual
negative selection in targeting constructs - derived from lambda KO
(Nehls et al., Biotechniques 1994)). The plasmids dont yield as much
DNA as some but when an insert is placed between the TK cassettes, they
grow as normal. 

I was assuming that the bugs (XL10 - recA- etc) are not entirely happy
growing the "empty" vectors where the TK cassettes are back to back
with only 60-80bps of polylinker between, but can replicate the repeats
no problem when the are interrupted by the insert of >5kb. 

More information about the Methods mailing list