Worm Breeder's Gazette 7(2): 9

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.

Somatic Mosaics Produced by Loss of Free Duplication

B. Herman

Figure 1

The free duplication mnDp3(X;f) carries unc-3+ and osm-1+.  About 
1/1000 of the Unc-3 self-progeny of unc-3 mnDp3 
hermaphrodites segregate non-Unc-3 self progeny.  Six such Unc-3 
animals were stained by fluorescein isothiocyanate (FITC) according to 
the procedure devised by Ed Hedgecock; FITC stains particular neurons 
of amphids and phasmids of wild-type but not osm-1 animals (Hedgecock 
et al., last Newsletter).  Five animals were completely negative and 
the sixth showed staining of neurons in the left amphid only.  The non-
Unc-3 self progeny of all six animals stained normally.  The FITC 
staining results thus confirm that all six Unc-3 animals were mosaic, 
i.e.  they did not carry mnDp3 in certain somatic cells but retained 
it in their germ lines.  If one assumes that osm-1 is cell autonomous 
with respect to FITC staining, the above results are easily 
interpreted on the basis of the known lineages (Sulston et al., last 
Newsletter): in five cases, duplication loss occurred in the AB 
founder cell and in the sixth case, loss occurred in ABp.  A model in 
which amphidial staining depends on the genotype of the associated 
sheath cell or socket cell (both of which also belong to the AB line) 
does not fit the pattern of staining found in the sixth animal unless 
more than one event of duplication loss occurred in the development of 
that animal.  In any case, the results strongly suggest that the Unc-3 
phenotype is specific to AB cells.  The somatic losses that gave rise 
to Unc-3 animals have all occurred early in the AB lineage.  The 
reason for this could be that duplication loss is much more likely in 
the early divisions or that the full Unc-3 phenotype is expressed only 
when a distantly-related set of AB cells is mutant.
To further test the idea that unc-3 expression is specific to AB, I 
made use of unc-93 III and sup-10 X (Greenwald & Horvitz, 1980); sup-
10 is a recessive suppressor of unc-93, which by itself confers an 
uncoordinated phenotype that is very different from that of unc-3 and 
that has been attributed to abnormal muscle structure (Greenwald & 
Horvitz, 1980).  Because mnDp3 carries the dominant allele sup-10+, 
animals of genotype unc-93; sup-10; 
rally segregate two types of self 
progeny: Unc-93, with the same genotype as the parent, and Unc-3 non-
Unc-93, of genotype unc-93; sup on the assumption 
that sup-10 is specific to muscle cells, a mosaic animal generated 
from the zygote unc-93; sup-10;  
mnDp3 is retained in AB cells but lost from body muscle cells should 
be non-Unc-3 non-Unc-93, i.e.  wild type.  Given the body muscle cell 
lineages, there is only one way a single somatic duplication loss will 
lead to a large majority of muscle cells lacking mnDp3: loss by P1, 
which would also lead to absence of the duplication in the germ line.  
Thus all self progeny of unc-93; sup-10; 
wild type by virtue of being somatic mosaics 
should give only Unc-3 non-Unc-93 offspring.  Wild-type recombinants, 
on the other hand, should give some wildtype self progeny.  I have 
found four mosaic animals satisfying these criteria, and the 
calculated frequency of loss in P1 (~ 1/2000) agrees with the 
frequency of loss in the sister cell, the AB founder cell, observed in 
the earlier experiments; thus it seems likely that the Unc-3 mosaics 
were also produced by single events, as already supposed.
Unc-93 animals are egg-laying deficient, presumably because of 
defective vulval and uterine muscles.  If loss of mnDp3 occurred in an 
unc-93; sup-10; fter 
formation of P1 but prior to formation of the cell that is precursor 
to all vulval and uterine muscles (a sequence of 9 cell divisions), 
the body muscles responsible for movement would be largely Unc-93, but 
the vulval and uterine muscles would be non-Unc-93, which could lead 
to egg-laying proficiency; such animals would be non-Unc-3 and retain 
mnDp3 in their germ lines.  I have found 15 animals meeting these 
predictions, at a relatively high frequency (~ 1/350), which may be a 
consequence of the many divisions during which duplication loss may 
As a control for the experiments involving sup-10, I have looked for 
both wild-type mosaics and Unc-93-moving egg-laying-proficient mosaics,
as defined above, among the progeny of unc-93; )/+;
sup-10 hermaphrodites.  Since mnDp1 is 
translocated to LGV and is not free, I expected to find no mosaics.  
None were found.  This indicates that the exceptional animals 
identified in the mnDp3 experiments were not simply the result of 
occasional incomplete expression of the single sup-10+ allele.  I 
conclude that unc-3 expression is specific to the AB lineage and sup-
10 expression is specific to muscle cells.
The free duplication mnDp26 carries the right end of X, including 
unc-3+, as well as various genes from LGV, including unc-60+, which 
affects muscle structure and confers a paralyzed phenotype (Waterston 
et al., 1980).  An Unc-3 non-Unc-60 segregant from unc-60; 
osm-1; d that was 
completely Osm-1, by FITC staining, and gave some non-Unc-3 non-Osm-1 
self progeny; this mosaic animal thus indicates that the absence of 
unc-60+ function in AB cells does not give the Unc-60 phenotype.
The frequency of duplication loss seems to be a function of 
duplication size.  The small duplication mnDp2 is lost ten times as 
often as mnDp3.  And mnDp26, which seems to be large, is lost one-
tenth, or less, as often as mnDp3.  In screening for the relatively 
rare mosaic animals, I have also recovered new mutants or genetic 
recombinants.  All cases that I have analyzed proved to involve mutant 
or recombinant duplications.  Altered forms of mnDp1, 
n recovered.  In a few cases, the 
duplication has been shown to have a deficiency.
I am working to extend these studies to other genes and other 
duplications.  I also plan to test mutations or treatments that might 
enhance the frequency of duplication loss; all of the work so far has 
involved spontaneous events.  Finally, I think these results enhance 
the attractiveness of Bob Waterston's idea of putting sup-7 (or sup-5) 
on a free duplication, which could then be used to give mosaic 
expression of any gene that is suppressible by one dose of the 

Figure 1