>>I recently asked the following question and since I got a bunch of interesting
>answers back (thanks !), I thought I might post some of them.
>>> Does anybody if and how often transcriptional regulatory elements are
>> located DOWNSTREAM of the ATG start codon of C.elegans genes ? (i.e. how
>> "dangerous" is it to fuse GFP to the first exon, thereby potentially
>> missing regulatory sequences within the first introns...).
>> Thanks for examples either way !
>>>Richard Zwaal wrote:
>I don't know if you're going to like this, but I have bad experience with
>fusing lacZ into the first exon.
>I first made a clone of lacZ to the first exon of the gene gpb-1 (G protein
>beta subunit), and found little expression in only a couple of cells, mainly in
>the head region.
>I then made fusions in the 5th or 7th exon and these gave enormous expression
>mainly in neurons (circumpharyngeal nerve ring, ventral and dorsal nerve cord,
>I didn't bother to find out where the regulatory elements were actually
>located, but apparently just using the 5' UTR wasn't enough.
>>Cori Bargmanm wrote:
>we have found enhancers in many locations. for tax-2 we found one promoter
>element and one first intron element, which can work independently of one
>another. for both osm-9 and another gene, ras-gef, we found elements in
>the promoter, more elements in introns (multiple elements in multiple
>introns for the ras-gef), and still more elements in the 3' utr for osm-9.
>>Julie Ahringer wrote:
>I made a vab-7::lacZ fusion containing over 10kb of upstream sequence,
>including the coding region for the entire upstream gene, and including the
>first two exons and first intron of vab-7. I now have antibodies and the
>expression pattern of this construct is correct, but missing the entire
>nervous system component - we are now looking for regulatory regions
>directing this expression in other introns.
>>Ralf Baumeister wrote:
>unc-86 is such an example. There are a couple of others, too, the
>most drastically being lin-12. If I remember correctly, the Greenwald lab
>did not get proper
>lin-12 expression when they fused only 5' segments to the reporter.
>Most important, all of the lin-12 L2-L4 expression was lost! Only a fusion
>with the entire 3' sequences in a smg-1 background proved to be working.
>I would consider it rather risky to just fuse GFP to your gene's first
>exon, if you want to rely on the expression data without having an
>antibody to confirm the results.
Harvard Medical School/Department of Genetics
Mass. General Hospital/Department of Mol. Biol.
Boston, MA 02114
Fax: 1-617-726 6893
Tel: 1-617-726 5983
e-mail: hobert at molbio.mgh.harvard.edu