How to do an F3 (precomplementation) screen for roller/cosmid

Erich Schwarz schwarz at cubsps.bio.columbia.edu
Wed Jan 19 15:53:18 EST 2000


Thanks to everybody who replied in less than a day, reminding me why
working in worms is cool.

Here is a summary of the responses, with sources noted:

1. One technical name for this is "precomplementation screen" (Erik
Jorgensen <jorgensen at biology.utah.edu>).

2. There are two references that were cited by one or more people:

    A. Apparently the first instance of this was (Michel Labouesse
<lmichel at titus.u-strasbg.fr>):

    Michael Labouesse and Bob Horvitz (January 1, 1992).  Cosmids as
genetic balancers for the isolation of lethal mutations - A new
let-253 allele.  Worm Breeder's Gazette 12(2): 74.

    The abstract is at: http://elegans.swmed.edu/wli/[wbg12.2p74]/

    B. The next was (Lee Honigberg <lee_honigberg at axyspharm.com>):

    Supriya Shivakumar, Gregg Jongeward, Cynthia Kenyon, Harold
Varmus (June 1, 1994).  wnt homologs in Caenorhabditis elegans.
Worm Breeder's Gazette 13(3): 95.

    N.B.: this method actually used psoralen, not EMS.

    The abstract is at: http://elegans.swmed.edu/wli/[wbg13.3p95]/

    Supriya Shivakumar <sshivakumar at cc.ucsf.edu> describes it as
follows:

    "I managed to isolate 2 Cewnt mutations by reverse genetics
using what we called the roller screen.  We made a transformant line
with about a 30% loss rate (containing the wnt genomic sequence and
rol-6), and mutagenized the line. We then screened for 100% roller
lines. We had to check the roller lines by back crosses for
integration of the array and spontaneous roller mutations.  Also the
presence of dead eggs was another clue that it was working."

    C. Lee Honigberg <lee_honigberg at axyspharm.com> pointed out some
important caveats:

    "The key thing, as I'm sure you've realized, is to quickly
distinguish between lethal mutants and the uninteresting things such
as new Roller mutations or integrations of the array. I remember
that Supriya tested her mutants by crossing in a different rescuing
array with a different coinjection marker and that this was a better
test than trying to look at segration of the Rol in a simple cross
to wild-type. (Integrated Rol arrays can be semi-dominant, variably
penetrant, and generally annoying)."

    D. James Waddle <jwaddle at hamon.swmed.edu> noted:

    "... we and others routinely use EMS (as an alternative to
irradiation) to screen for integrated lines derived from
extrachromosomal arrays.  In our lab, integration events arise at
1/200 to 1/600 clones screened.  However, after 4 - 8 x outcrossing
to N2, we've not yet observed (nor have we screened for) any that
show 100% segregation because they carry a homozygous lethal
balanced by the array; rather, they are simply integration events
without apparent detrimental consequences."


--Erich Schwarz
  schwarz at cubsps.bio.columbia.edu
---







More information about the Celegans mailing list