Worm Breeder's Gazette 10(2): 69

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.

Genes that can be Mutated to Reveal Hidden Antigenic Determinants in the Cuticle

Samuel M. Politz, Miguel Estevez, Karl J. Chin and Peter J. O'Brien

In our initial attempts to isolate mutants that express stage 
specific surface antigenic determinants at inappropriate stages, we've 
turned up some mutants with a novel, unexpected phenotype.  In 
immunofluorescence tests, these mutants stain uniformly with a rabbit 
antiserum that does not stain the surface of wild-type animals.  This 
antiserum was prepared by treating an anti-wild-type cuticle serum (by 
adsorption with wild-type adults) to remove antibodies capable of 
binding to the wild-type surface.  Thus, antigenic determinants that 
are not available on the wild-type surface are exposed in the mutants. 
We describe here isolation and preliminary genetic characterization 
of the mutants.
Mutants were isolated by screening synchronous populations of F2 
progeny of animals mutagenized with EMS.  Populations were 
synchronized by refeeding F2 dauer larvae, or by gently washing plates 
to remove egglaying F1 adult hermaphrodites while leaving F2 eggs 
stuck in the E.  coli lawn.  In the latter case, eggs were allowed to 
hatch for 1-2 hours and then L1's were washed off and transferred to 
fresh plates for growth.  The egg-hatching and transferring steps were 
usually repeated to obtain a second 'wave' of synchronous F2's.  
Populations were harvested at the L4 stage or younger and screened 
with L4-adsorbed antiadult serum (Genetics 117:467 1987) by indirect 
FITC immunofluorescence.  Samples in 100-200 l PBS were spread in thin 
layers onto Pyrex Petri dishes; up to 10,000 animals per sample were 
screened at a time.  Samples were screened using an epifluorescent 
stereomicroscope (Leitz-Wild).  Immunofluorescent larvae were observed 
at low frequency; these were picked using a mouth tube and a drawn-out 
100 l capillary micropipet.  Clones were established and cloned stocks 
were rescreened by immunofluorescence to eliminate false positives.  
Viable mutants were obtained at a frequency of about 5x10+E-4 per F1.
Of 12 mutants that we have analyzed, 11 fit the pattern described in 
the first paragraph above; the remaining one produces an incompletely 
penetrant 'small adult' phenotype that may be of interest, but this is 
not its story.  Initially, we looked at which stages besides adult 
stained with the L4-adsorbed antibody.  Eight of the 11 mutants 
stained at all stages, while the remaining three stained only at late 
larval stages.  We looked at several of the mutants to see if they 
expressed adult lateral alae at earlier stages, but surprisingly, 
found no larvae that carried alae.  yj10 is the only mutant in the set 
that has a readily apparent visible phenotype; it is somewhat scrawny 
and has a cold-sensitive weak Left Roller phenotype.  These results 
led us to consider that the mutant phenotypes might result from 
cuticle structural defects rather than, or in addition to, precocious 
expression of an adult-specific cuticle type.  The 11 mutants were 
then tested for staining with the same parent antiserum after it had 
been adsorbed with wild-type adults to remove all antibodies capable 
of reacting with the wild-type surface.  All of the mutants were 
antigen-positive; wild-type was antigen-negative (this is the result 
summarized in the first paragraph).  The 8 mutants that light up at 
all stages with this serum are termed 'super-bright' in our lab jargon 
and the other three that exhibit apparent stage-specificity stain much 
more weakly.
We considered the possibility that known apparent cuticle phenotypes 
like Sqt and Rol might share this phenotype with our new mutants.  We 
obtained a set of ten sqt and rol mutants from Bob Edgar and tested 
them with adult-adsorbed antibodies.  None were antigen positive.  
Thus our mutants have a new phenotype that is not shared with the 
classical morphological cuticle mutants we have tested.
We have made some progress in analyzing the 'super-bright' mutants 
genetically.  It has been possible to use the difference between 
wildtype and mutant antigen phenotypes to do linkage, complementation, 
and mapping using methods similar to those described for mapping of 
the srf-1 antigenic polymorphism (Genetics paper again).  All mutants 
were backcrossed to wild-type twice and an antigen-positive (mutant) 
segregant picked after each backcross.  Penetrance of the surface 
antigen phenotype is complete; it is very unusual to find a non-
staining individual in a stained population of one of the homozygous 
mutant stocks.  This is in contrast to the srf-1 phenotype which we 
found to be incompletely penetrant both in wild-type and srf-1 mutant 
strains.  All of the mutant antigen phenotypes are recessive, the best 
evidence for this is that in crosses with unlinked unc markers, 25% or 
less of Unc segregants issuing from an unc +/+ srf parent are antigen-
positive.
Linkage was determined by using unc markers on the autosomes as 
selected markers and staining for antigen-positive Unc segregants of 
the double heterozygote.  Absence of the double homozygous recombinant 
in a small population was taken as evidence for linkage.  For 
complementation testing, srf-a/+ males were mated with srf-b 
hermaphrodites.  F1 male progeny were collected and stained.  Failure 
to complement was indicated by presence of antigen-positive 
heterozygous males.  So far there are two complementation groups.  At 
least 5 non-complementing mutations are linked to unc-13(e51)I, and 
yj10 shows linkage to unc-24(e138)IV.  We are presently designating 
the chromosome I mutations as srf-2 and yj10(IV) as srf-3.  We have 
mapped srf-2(yj262) by 2 and 3 factor crosses; it is on IR in the 
vicinity of lin-11 and unc-75.  A 3 factor cross with these two 
markers is in progress.
Mutant phenotypes were compared serologically by the adsorption 
method.  Antiserum adsorbed with srf-2(yj262) adults binds to yj10 
adults but not to yj262 or any of the other srf-2 mutants.  The same 
parent antiserum adsorbed with yj10 adults doesn't bind yj10 or any of 
the srf-2 mutants.  Thus the two complementation groups correspond to 
two distinct serological phenotypes, and the results fit a model in 
which the antigens exposed in srf-2 mutants are a subset of those 
exposed in srf-3(yj10).To summarize, all mutants in this set share the 
general characteristic of expressing surface antigenic determinants 
that are not on the wild-type surface.  The fact that the antisera 
were raised against a wild-type cuticle immunogen indicates that the 
antigens exposed in the mutants are probably wild-type antigens and 
not novel structures.  The fact that the mutant phenotypes are 
recessive, even though they constitute an apparent gain of 
antigenicity, suggests that the mutant lesions are defects in cuticle 
structure that uncover normally hidden antigenic determinants.  Our 
hope is that these mutants and the genes that they identify will 
contribute to understanding genetic control of the layered 
organization of the cuticle in a way that is complementary to what can 
be learned from morphological mutants.