Methicillin revelation

Phillip San Miguel pmiguel at
Fri Jun 25 17:32:37 EST 1999

David Micklem wrote:

> In article <3772755D.2BC2DBBE at>, Phillip San Miguel
> <pmiguel at> wrote:
> <snipped>
> Given that oxacillin and nafcillin both appear to still be useful against
> multi-drug-resistant staphylococcus infections (including
> vancomycin-resistant strains), do you think its really a good idea to
> encourage their widespread use in laboratories? Presumably there could be a
> risk of accidentally selecting in the lab for resistant staph strains.

    I didn't advocate anything, I was reporting that methicillin, oxacillin and
nafcillin do not inhibit the growth of untransformed (non-plasmid bearing)
DH5alpha.  To me this was unexpected given that methicillin, at least, does
appear to potentiate the effectiveness of ampicillin.
    But don't let me sound weaselly here.  If oxacillin works like methicillin
I officially declare here my advocacy for its use in the laboratory.  I am
concerned about MRSA strains in hospitals--as we all should be.  But the
trivial amount of oxacillin I'd be using in the lab is so monumentally
outweighed by the amounts used in medicine as to be inconsequential.  Plus I'm
not injecting any into an animal where creation of oxacillin resistant
pathogenic bacteria is conceivable.  It's just going into media that will be
autoclaved after E coli has been grown on it.

> Use of carbenicillin, or plating a larger number of plates at lower density
> are usually sufficient to prevent the growth of satellite colonies... I
> didn't have a problem getting lawns using the gamma-delta transposon system
> as described by Strathman, PNAS 88, 1247,1991, although he does (IIRC)
> mention that it can be a problem if the cells are plated at too high
> density.
> [...]

    Okay, I didn't go into this, but I was using gamma-delta sequencing for
large plasmids that did not appear to be as easily transferred as the
cointegrate with F.  Thus I needed to plate at higher densities to get
transconjugants in sufficient number to cover the area that was cloned.  I
admit I never tried carbenicillin.  I probably would have but the
methicillin/ampicillin plates work so well I didn't bother.  Plus carbenicillin
was more expensive.
    Again, I don't believe the tiny amounts of methicillin used by me and my
methicillin-using laboratory brethren throughout the world contributed to the
creation of MRSA in hospitals.  But if you've got any evidence to the contrary
I'll certainly take it into consideration next time I'm designing a cloning

Phillip San Miguel
Purdue Genomics Center

More information about the Methods mailing list