Worm Breeder's Gazette 11(4): 58

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.

A Screen of Patterns of Gene Expression in C. elegans

Ian A. Hope

One way to screen patterns of expression for a large number of genes 
might be to place a reporter, whose expression can be easily monitored,
under the control of the promoter elements of many individual genes.  
Attempts (over the preceding 2 years) to generate such fusions in C.  
elegans through transposition of Tc1 and derivatives had failed.  So 
instead, fusions were created in vitro using conventional molecular 
procedures and patterns of expression assayed after germ-line 
transformation of C.  elegans.A library of potential promoter fusions 
was constructed by ligating 5-10 kb partial Sau3A C.  elegans genomic 
DNA fragments into the BamHI site of pPD22.11 (an Andy Fire vector).  
The lacZ gene of this vector lacks promoter elements so expression of  -
galactosidase should depend on promoter elements within the insert and 
translational fusion with a coding region reading out from the insert. 
Therefore only 1 in about 120 inserts (about 5% coding, 3 frames, 2 
orientations) was expected to drive  -galactosidase expression.  So, 
DNA pools each prepared from 96 independent clones were introduced 
into C.  elegans by syncytial injection and rol-6 selection, which 
gives a high rate of co-transformation (as described by Craig Mello).  
(Pooled DNA at 180  g/ml, pRF4 at 20  g/ml).  Transmitting lines were 
established before staining for  -galactosidase using a modification 
of Andy Fire's procedure (see below).
Patterns of expression observed reproducibly so far (31 pools 
examined): Pool 2.  staining in the nerve ring observed post-
embryonically.  Pool 4.  Nuclear localized staining in adult body wall,
probably a subset of muscle cells.  Pool 9.  Cytoplasmic staining of 
cells forming the uterus, appearing as the gonad develops, L4 onwards. 
Pool 23.  Staining of the tips of head and tail in adult.  Pool 24.  
Nuclear localized staining, subset of cells (maximally <20) in embryos 
at approximately the 50-200 cell stage.  Initially thought expression 
turned off completely in later embryos but  -galactosidase may persist 
until L1 in 2 cells (possibly Z1 and Z4).  Pool 26.  Nuclear localized,
2 cells in rectal region (possibly blast cells F and U) observed post-
embryonically.  Pool 29.  Nuclear localized staining in ventral cord 
observed post-embryonically.  Pool 38.  Staining of the spermathecal 
valve (L4 onwards) and the intestinal-rectal valve (late embryo 
onwards).  Staining is subcellularly localized in an annulus, possibly 
the filamentous structure seen by EM.  Pool 42.  A ring of staining in 
the nerve ring observed post-embryonically.  (Note; pPD22.11 lacZ has 
a nuclear localization signal, though this signal may be overridden 
for some translational fusions).  
14 pools gave expression within the pharynx, many in distinct 
patterns.  This high rate suggests that in these cases  -galactosidase 
is not being expressed to accurately reflect the pattern of expression 
of a gene to which lacZ has been fused.  Many explanations are 
possible, but at least for the moment these pools have been put aside. 
To identify the individual plasmids responsible for the expression 
patterns of pools 2, 4, 9 and 24, further DNA pools were prepared.  
The original clones had been picked into 8x12 multi well plates.  Now, 
DNA pools were prepared for all clones in each row and each column of 
all 4 multiwell plates, generating 12 new pools of 32 clones and 8 new 
pools of 48 clones.  Each new pool was assayed as described above.  
Each pattern observed for the original pools should be observed for 
one row and one column, unambiguously identifying the individual 
plasmid responsible.  This was successful for pools 4, 9 and 24, but 
not for pool 2.  Transformed lines have been established using the 
identified plasmids or pools 4 and 24, (not yet for pool 9), and stain 
with the appropriate pattern.  The locations on the genomic physical 
map of the DNA inserts for the 3 identified plasmids have been 
determined by a combination of probing a 'YAC polytene grid (provided 
by Alan Coulson) and fingerprinting (carried out by Ratna Shownkeen).  
None coincide with a likely genetic locus.  
Comments: No constitutive expression has been observed yet.  Perhaps 
a weak, cryptic enhancer on the vector plasmid causes constitutive or 
incomplete promoters to give  -galactosidase expression only within 
the pharynx, hence the preponderance of pharyngeal patterns.  (Other 
explanations possible).  The non-pharyngeal patterns observed probably 
accurately reflect the expression of the genes to which lacZ has been 
fused because translational fusion should be necessary for expression. 
However any enhancers downstream from the point of fusion will have 
been lost.  Confirmation of expression patterns will require 
independent work with the identified genes.  
This screen is labor intensive, particularly compared to the 
approaches that had been initially intended utilizing Tc1.  However, 
the first 22 pools were screened in one month and interesting patterns 
of expression have already been obtained.  A larger screen has just 
Staining procedure: 3 l aliquots of worms were placed in each well 
of an 8 well multi-well microscope slide, a coverslip was applied and 
the worms frozen on dry ice.  The coverslip was flicked off and the 
slide placed in acetone, 5 min.  -20 C.  Worms were then air dried and 
mounted in 25 l/slide of staining solution under a coverslip sealed on 
with nail varnish.  This procedure is convenient for examining large 
numbers of lines but preservation of structure is variable.