Worm Breeder's Gazette 11(2): 22

These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.

On Enhancers and Promoters

Andrew Fire and Susan Harrison

Figure 1

With the eventual goal of setting up a general enhancer trap assay 
similar to that being used in Drosophila, we have been identifying 
segments of the genome which have tissue specific enhancer function.  
In a previous newsletter we reported the identification five random 
genomic segments which could enhance a myo-2-lacZ fusion outside of 
pharyngeal muscle.  In each case the induced expression was in other 
muscle cells, leading to the suggestion that response of the myo-2 
promoter to different enhancers was limited to muscle.
We have used several of these random segments plus some of the 
identified myosin enhancer segments with different promoters to 
determine patterns of expression.  In general these experiments have 
been done by staining the F1 after injection.
[See Figure 1]
Although it is difficult to explain this all with one simple model, 
several conclusions can be drawn.  First the muscle promoters seem 
indeed to have some underlying muscle specificity, since they respond 
specifically in muscle in the vicinity of enhancers which can act in 
other tissues.  Second, the internal enhancer present in unc-54 is 
capable of specifying proper tissue specificity even in the absence of 
a muscle specific promoter, since that enhancer can specifically 
enhance glp-1- -gal fusions (and unc-6- -gal fusions) specifically in 
body wall muscle.  The P element and HSP70 promoter elements fail to 
respond to the unc-54 internal enhancer, and it is worth noting that 
these are both standard 'TATA' box containing promoters.  An evident 
conclusion is that different promoters and enhancers can interact to 
generate a variety of specificities.  In thinking about this it is 
worth keeping in mind that the 'enhancer' segments from random genomic 
DNA could each contain several different enhancer segments, and that 
the promoter segments too could include elements fortuitously 
responsive to different signals.
The failure of the P element promoter to respond to the unc-54 
enhancer is actually a blessing in that these constructs can be co-
transformed with the unc-54 enhancer containing antisense 'twitcher 
plasmids' without spurious induction in body wall muscle.  Most but 
not all of the twitching lines derived with an antisense plasmid 
containing the unc-54 enhancer mixed with glp-1 or myosin derived 
enhancer test constructs give strong body wall muscle expression of  -
gal.The constructs that show limited numbers of larval gut cells 
staining have reproducible patterns, indicating that some feature 
distinguishes posterior, anterior and middle gut cells in young larvae.

Current view of myosin 
regulation
We've previously described sequences both upstream of and inside the 
unc-54 gene which are independently capable of specifying body wall 
muscle specific expression.  We have now narrowed down these sequences 

Internal: An 82 bp minimal region sufficient for body wall muscIe 
specific enhancer function has been defined.  Within this region are 
several short direct and inverted repeats.  At the 5' end is a 
palindrome TCAATTGA.  A deletion of a single A residue (TCATTGA) 
completely abolishes activity.  The palindrome is not sufficient for 
enhancer activity, with sequences at least 24bp downstream also 
required for enhancer function.
Upstream: Sequences between -169 and +12 are sufficient to promote 
body wall muscle specific expression, while sequences from -138 to +12 
are insufficient.  The region upstream of -138 can be inverted with 
retention of activity.  This region is functionally interchangable (
although weaker) than the internal enhancer, and probably represents a 
second tissue specific albeit weak enhancer.  At the 5' boundry of the 
upstream element is a different palindrome from that above: 
'CTACGCGTAG'.  A clustered point mutation eliminating this palindrome 
eliminates specific expression.  Consistent with the suggestion that 
the two palindromes might interact with different factors, the 
relative activities of the two elements in different body muscle 
subclasses appears to be different, with the upstream region having 
relatively poorer activity in vulval muscles.
Promoter sequences downstream of -138, although not sufficient for 
any expression, seem to control the start point of the RNA based on 5' 
mapping with and without upstream and internal enhancer sequences.
Mapping of the myo-2 promoter has revealed an upstream element which 
stimulates expression but does not act as a tissue specific enhancer, 
and a segment near the start site (minimal size now 254 bp) which acts 
both as a weak pharyngeal specific promoter and a pharyngeal enhancer (
when assayed on the myo-3 gene).

Figure 1