-
-
-
-
--
-
-
--
-*---.2
10F2
70
PL
ORNL
UNCLASSIFIED

hu
S
.



.
**!**:
_-
..
1
.
LEGAL NOTICE
This report was prepared as an account of
Government sponsored work. Neither the
United States, nor the Commission, nor any
person acting on behalf of the Commission:
A. Makes any warranty or representa-
tion, expressed or implied, with respect to
the accuracy, completeness, or usefulness
of the information contained in this report,
or that the use of any information, appa-
ratus, method, or process disclosed in this
report may not infringe privately owned
rights; or
B. Assumes any liabilities with respect
to the use of, or for damages resulting from
the use of any information, apparatus,
method, or process disclosed in this report.
F
.
TA
-
2
.
t
PA
-
-
-
.
L
.
As used in the above, "person acting on
behalf of the Commission” includes any em-
ployee or contractor of the Commission, or
employee of such contractor, to the extent
that such employee or contractor of the
Commission, or employee of such contractor
prepares, disseminates, or provides access
to, any information pursuant to his employ-
ument or contract with the Commission, or
his employment with such contractor.
_**..
---
.
.
*
.
.
-
-
14
it
i
7:
24
**
-*
-
.
.:-
.1
1
*.*..
7
Bar
aut
+
+
Sant
A
-
.:
1
Y
1..
to ..A
-
.
---
.
*
*:
T
!
11.SI
:..
.
DTIE
MICROCARD
ISSUANCE
DATE
8 / 7
1964
-
-
-
*
*
--
-
-
-
-
.
:
-
.
.
..
.
.. to!
-
11,
.
IR
.
..




6 R. W 2-p-90 ne
CONF-613-1
-
.
1981 g Lanr
- LEGAL NOTICE -


Two ropurt m. povered unitat al Covennt opon ord work. Nordhor the United
kalos, bor We Couals.loa, sor my pornoa suing on ball of the Cosweslea:
A. Maui wymruty or represmu tod, riproud or impun, mu roopass to the accu.
ry, completas, or wtuin.. of He istoranuor conduwe roort, or that we
of any laloraaloa, oppuretu, wchod, or proces diklound u wa roport may not talringe
privately owned rudu; or
B. Asme. wy labtine nu respect to the we of, or for denuo reowux
Nu of uy wtor usulon, spauratus, souhod, or proces dioclorodu do ropon.
As used ja ne oboro, "person kung an bowlf al dhe ConwSLO" uciudeo may m.
ploys or connector of Coao!son, or eaploys of such contractor, toe omni hot
such employee or contractor of the Cow aluton, or employs of such contractor prepare.,
disuninavy, or woridoo accm ko, may lasormuoa pumunt home suplojuat or contract
viu the Counoslas, or M. eoployarat med such contractor.

!
MASTER
-
T
..
T
'
I
.::
3
.a
..
y GENETIC AND FUNCTIONAL MOSAICISM IN THE MOUSE"
.
*
1......
.
.
..
.
.
..
.
.
.
...
ule was commence Liene B. Russeck
the undrestine Biology Division, Oak Ridge National Laboratory,
Oak Ridge, Tennessee

i

i
, in
"
IA
Facsimile Price S_ 3.160
Microfilm Price $_lizi
"
,
1"}
-"
Available from the
Office of Technical Services
Department of Commerce
Washington 25, D. C.
NA
* Research sponsored by the U.S. atomic
Research ponsored the lion Corlide exporation
Energy
Commission under
.
1.
For Growth Soc. Symp.
June 16-18, 1964
1.
GENETIC AND FUNCTIONAL MOSAICISM IN THE MOUSE
In a sense, the whole process of differentiation is one of directed functional
mosaicism. It is, however, the accidental or random juxtaposition in the same organism
of cells having actually or effectively different genotypes that is generally thought of
as true mosaicism. The study of this condition constitutes a perfect meeting place
for the fields of genetics and developmental biology, providing, as it does, interrelated
information on, mutabilitycell lineage (including the special problems of cell lineage
of the germ line), and the effect of genotype on part of the organism versus the whole.
The discovery in recent years that most or all of one X chromosome of the normal
-- an event there leari, to
mammalian female becomes randomly inactivated early in development, lesding to
functional mosaicisma provides us with a great potential tool for the study of gene
action.
The present paper will attempt to bring together results of diverse observations and
experiments (many of them as yet unpublished) bearing on both genetic and functional
:::.
mosaicism in the mouse. Although most of the mosaics to be discussed involve coat color
.
.
characte:s, this is only incidental to the circumstance that mosaic animals are detected
-.
.
by external observation; and coat color per se will not be a topic of discussion. Instead,
.
attempts will be made to derive answers to (and, where this is not possible, to outline
. .
the problems involved in) general questions of mutability, cell lineage, and gene action.'
.
:.
'
7
..
*
...
...
....
. -- ...
I. GENETIC MOSAICISM
Three
two main groups of pathways could lead to genetic mosaicism. First, the
"zygote" itself may already be mosaic as a result either of a genetic events)
in one of the gametes, involving only one strand of a double-stranded chromosome;
1 (see I A2).
or as a result of abnormalities of egg maturation or fertilization, Second,
two normal zygotos or voryto ourly ombryos may fuse to form a singlo, chmuric.
animal. The end result of this is in no way distinguishable from partici.
patio. of one of the polar bodies in formation of the embryo and
will be discussed in the same section (880. IA2). Third,
starting with a normal zygote, any of a number of genetic events may occur in
subsequent cell multiplication, so that the finished organism contains cells
1 (see IB).
of more than one genotype, Among such genetic events are some that produce
deficiency,
only one new genotype (somatic mutation, chromosome loos) and others that
produce two reciprocal new ones (somatic crossover, non-disjunction).
A. Mosaic "Zygote"
...
.
1.
"Partial" mutations in gametes
I.
.
-
::;.
A genetic event occurring spontaneiously in only one strand of a gamete
I end result
chromosome cannot be distinguished in its eftseta from one occurring in the
..
a.
course of the first cleavage division: in each case one of the first two blastomeres
..
is the bearer of the new genotype.
The distinction becomes particularly meaningless
.
in the case of maternal chromosomes since no separate female "gamete" stage
exists in the mouse, the second meiotic division not being completed until
after sperm entry.
·
· ·
..
...
It 16, however, legitimate to ask whether genetic changes can be induced
(1964)
in only one strand of spermatozoal chromosomes. W. L. Russell (mae Haguey has
found that while irradiation of spermatozoa causes a relatively high frequency
of total-body mutations, the frequency of mosaice in the progeny 18 no greater
than in controls. These mosaics are presumably the result of spontaneous.
...
... ..
.
...
--•..
see TBI
mutations in cleavage (see below); or, if they should result from events in
spermatozoa, such events are not suaceptible to radiation -- a rather unlikely
situaticn.
2. Abnormalities of egg maturation and fertilization. Fusion or
foles.
of the various abnormalities of e88 maturation and fertilization that
have been described (LR-23, table 1, summary), those involving "Imediate
cleavage" (Braden, 1957) can lead to mosaicism. "Imediate cleavage" 18,
essentially, the participation of one of the polar bodies in the formation of
O resulting
the embryo, and then mosaics-formed can be of the types 10/2n, 2n/2n, 20/3n, etc.
depending on the circumstances of fertilization.
Diploid-triploid mosaics have been fcund in man (Böök, Penrose) and the
cat (Chu). The euploid 2a/2n type, however, cannot be detected unless an
obvious gynandromorph has been formed (for a human case, see Gartler, Waxman,
and Giblett, 1962), or
less special genetic markers are present so that the
mosaic can be shown to contain two meiotic products from at least one of the
parents. · A mosaic of this type was observed by us in a cross involving three
" AY
alleles at one locus as well as closely linked markers (Woodiel and Russell, 1963;
Russell and Woodiel, 1963 ORNL-3498).
The female, which came from a cross of *4. un we 9 x AT e o had yellow
ke at ++ ** kra
and black areas arranged roughly in transverse bands. She transmitted A, a*,
and :&, and was presumed to be a mosaic of follow patches) and (black
patches). The segregation of the linked markers, to be described in detail
* Ayun we a* + +
elsewhere, indicates that her actual constitution was
s kr a + + //
which is compatible with two possible origins involving retention and fertilization
of the first or second polar body, respectively (see Fig. 1). Two other modes
of origin which originally seemed possible can now be eliminated. One was
retention of an unfertilized first polar body, ruled out because of recovery of
It should be noted that if the mosaic femals had been formed by fusion
of two originally independent zygotes or embryos. The results would be
indistinguishable from those obtained:

• 1/xo T/ko.”
a + 2 + + crossovor product. The other was mosaicism for trisomy-5, mado
unlikely by now cytological data obtained soparatoly for yollow and black aroas.
Gynandromorpnisin, which is rare in the mouse, oxcopt, possibly, in certain
Inbred strains (Hollander, Gowen, and Stadler, 1956), can also be explained
L' of.a fertilized polar borlig
by polarebody retention or by fusion of Indopondont zygotes. It would, howevor,
1. This possibility that the gynandromorphs were mosaic for
be nocossary to rule out, pretin sox-chromosomest'only. , 0.8., XY/XO, The
cases reported to dato wore observed beforo adequate cytological techniques
became availablo.
B. Mosaicism Arising After the Zygote Stage
Any of a number of possible genetic changes occurring in the formation
or maintenance of the organism can create at least a temporary state of
mosaicism, but whether or not a mosaic organism actually persists depends on
the survival of the newly formed cell type. företelow for a discusion or
call lothality and Similarly, the finally observed proportion of mutant to
original tissue need not give an accurate indication of the stage at which the
genetic change occurred.
It should also be noted that the first cleavage division presents a
special case in that certain events occurring then can lead to an organism
types. This has been diagramatically outlined for 1088 and nondisjunction of
sex chromosomes (L. B. Russell, 1961, F1.g. 4). As discussed earlier (LR-20
),
we believe that a relatively high proportion of spontaneous X-monosomy in the
mouse is the result of chromosome 1038 at the first cleavage; and radiation
sensitivity of chromosomes has been found to be very nigh in the early pronuclear
stage (LR-822; LR-20; LR-24).
There is no reason to believe that autosomes at these stages behave
differently from the sex chromosomes (LR-24). Autosomal monosomy affecting the
whole organism would very probably be lethal (L. B. Aussell, 1962, pp. 243-244);
but since autosomal trisomy in the mouse 16, at least in some cases, compatible
with viability (Cattanach, 1964), a tricomic/nonosomic mosaic, such as would be
formed from nondisjunction at the first cleavage, could, perhaps, survive --
Dossibly as a pure trisomea.
Gcnctic events affecting single chromatids at the first cleavage would
resemble those at all later divisions in that they potentially produce mosaics
of the mutant and original genotypes. The remainder of th:.8 discussion will
be devoted to mosaics of this type.
1. Spontaneous somatic forward mutations
Ever since the early 1930's there have been several reports of mice with
coat color patterns that could be interpreted in terms of mosaicism resulting .
from somatic forward mutation (Bittner, 1932; Duna, 1934; Feldman, 1935; Morgan
and Holman, 1955). It should be realized that in order for such a mutation to
be obrious from the phenotype, the animal has to be heterozygous at the locus
in question.
In recent years, the opportunity for this type of detection has been very
much increased by the large-scale mutation-rate studies of hots Hussett and
mis-group that use the specific-locus method (refs.).
In this method, wild-type
irradiated and non-irradiated mice are mated to animals homozygous for seven
specific recessive mutations. Therefore, over the past several years, many
thousands of Fj mice have been observed that start life heterozygous for seven
recessives, five of them affecting coat-color (a = non-agouti; b = brown; cºn =
chinchilla; 2 - pink-eyed dilution; d = maltese dilution). Two other features
of this set-up aid in making it a good one for the detection of somatic mutations.
First, of the five coat color recessives in the cro88, three have linked markers
lg for g; g for c; se for a), so that it 18 possible to distinguish between
mutations, deficiencies, or other events lavolving only part of a chromosome,
on the one hand, and whole-chromosome losses on the other. Second, at two of
the loci, the allele to which most mutations appear to occur 18 different from
strert
Of 112 "mottled" and "questionably mottled" offspring in specific-locus
experiments that have been analyzed io date (many experiments only partially analyzed
are excluded) the distribution was as follows at least 26, and probably as many as 40
were found to be somatic and germinal mosaics for one or another of the coat color loci
for which the population is heterozygous; in 13 others there were mutations to sex-linked
genes that cause w "mosaic" phenotype; 4 carried X-autosome translocations involving
one of the markers (sec II C); 2 were heterozygous for new mottling alleles at one of the
specific loci (sec Il B); and 2 hod mutations to dominant spotting. In most of the remaining
51, tests ruled out mutation to a dominant or sex-linked factor, but were otherwise
inadequate to discover a cause for the mottling. Adequate tests thus could have swelled
i
1...
the mosaic category even more. Results on only a small portion of the 51 are of a nature
. .
.
that could indicate somatic mosaicism without germinal involvement.
.
.
.
.
.
Of the 40 certain and probable specific-locus mosaics, 21 had one irradiated
.
parent and 19 came from contemporarycontrol matings. There is thus no evidence against
all of them being of spontaneous origin, the more so since in 12 of the irradiated parents,
the germ cell stage treated was spermatogonia (see also sec IAI).
The 40 can be subdivided into a group of 20 (group 1) in which germinal mosaics
can be diagnosed with certainty because of close linkage or because the mutation is to a
different allele; and another group of 20 (group 2) in which the evidence for germinal
involvement comes from upset ratios only (probable repeat mutations, no close markers,
e.g. mosaics of type 116). In group 2, there were only 6 in which the upset in ratio
was so extreme as to prove mosaicism with certainty (group 2A). The other 14 will be
referred to as group 28. The distribution among loci involved was as follows o, 1 certain;
(rev.)
ionc
tain
b, 3 certain and 2 questionable; c, 9 certain (6 c viable, 3 new viable alleles) and
3 questionable (cch); d, 7 certain (6 d', id viable) and 2 questionabie; p, 6 certain
(1 p lethal, 3 p viable, 2 p untested) and 2 questionable; dse, 4 questionable; cche,
I questionable.
Loss of a chromosome is ruled out in the case of 29 of the mosaics (c,d,p loci).
it is a possibility in 5 of the questionable cases (d se and cche). However, although'
whole-chromosome lossos could possibly survive somatically (see sec TI E on cell lethals)
and even form gonadic mosaics, it is unlikely that they would be transmitted to viable.
progeny, since there is evidence (LR-23) that autosomal monosomy is lethal. Animals
mosaic for such chromosome loss should then appear "partially sterile" since some of their
progeny would die as embryos. One of the presumed d se mosaics in group 2B indeed had
a clearly reduced litter size. The other 3 cases of d se and the presumed och mosaic
were, however, fully fertile. In addition, for most of the few mottled that were clearly
particlly sterile in the unassigned group of 51, somatic chromosome loss could be ruled out.
Before making any further calculations, it was necessary to determine whether there
was any selection against the mutant type, either gametically or in the progeny. If the .
frequency of the gene that gave rise to the mutation (in our cases, the wild-type allelle)
is w, that of the new gene m, and that of the uninvolved homologous marker h, then, if
there is no selection, the ratio (m + w)/(m + w + h) should equal 0.50. Actually, for
individual animals of group 1, it ranged from 0.38 to 0.58, the average being 0.50
(based on 1383 classified progeny).
The progeny ratios can therefore be used with confidence in calculating cell-type
proportions in the gonad. This was done by various different methods (to be described
in detail elsewhere) depending on the individual mutation and the type of test-mating used.
-7-freri)
Proportion of heterozygous mutant cells in a mosaic gonad can also be calculated for
m
another group of animais not yet mentioned. In addition to those somatic
iradidit lathtothobe-somatic mutations that are detected by mosaic
phenotype in the heterozygous F, population, others reveal themselves by the
breeding results of the supposedly homozygous wild-type parent group. An
occasional mouse in this group may produce in the cross with the mw.tiple recessive
test stock a cluster of like mutant progeny. Provided the animal he.s not had a
radiation treatment that could account for this result (namely, depletion and re-
population or the testis encountered in measurements of spermatogonial mutation
rates), one can assume that the supposedly homozygous wild-type mouse was
aneons
1.6.
actually a gonadic mosaic. The fact that the animal was very probably also a
in the coat
cometine mosaic, can, of course, not be detected since suck mosaicism is covered
by the wild-type alleles in the homologous chromosome.
Three such animals (group 3) are included in the distribution shown below: two were
and
gonosomic mosaic for c (viable), one for d (prenatal lethal). On the basis of 807, 83,
and 402 progeny tested, the proportions of mutant cells in the gonad were estimated to
be 51.5, 4.8, and 42.8%, respectively.
Many mutations at loci other than those used in specific-locus mutation-rate
experiments have also originated somatically, e.g. a dominant spotting factor, the
sex-linked mutation spf, and 7 of 8 independent sex-linked mottling factors (3 of these
7 first occurred in mosaic males, although they are male-lethal; see sec II E). Because
of problems of viability or penetrance, all of these have been excluded from calculations
of proportion of mutant cells in mosaic gonads.
The distribution of mosaics with respect to percent mutant cells in the gonad is shown
in Figure 2. The average of individual percentages for all 43 animals (of groups 1, 2, and 3)
8(rovat
is 48.3%, based on 3864 total progeny classified. If one restricts calculations to group i
only, the average is 49.5, based on 1815 progeny. It is thus possible to interpret these
results as showing that for the gonosomic mosaics shown in Figure 2, obout half the cells
of the organism are mutant.
Four
Threo types of events might be considered as leading to a half-mutant state (Fig. 3):
(c) mutation in one strand of a double-strunded gamete chromosome; (b) participation in
the embryo of 2 meiotic products of the mother (or corly fusion of embryos, see sec 1 A 2);
(c) forsion of zygotes; and,
1 and (c)
uld) mutation at the first cleavage or in 2-cell stage. Alternative (b),can be ruled out for
un for
the cases here considered because the crosses that yielded the mosaics did not involve
segregating alleles. Alternative (o) seems unlikely, since radiation experiments have
failed to reveal double-strandedness in spermatozoa (soc I A1). Alternative (al is left
as the most acceptable.
Is the first cleavage so much more spontaneously mutable than late cleavages? The
results certainly indicate that this is so. It may be recalled that of 112 "mottled" animals
from specific locus experiments analyzed to date, 40 were among the "half-mutants"
discussed here, 21 had some other genetic explanation, and for 51 no cause of mottling was
discovered (mostly as a result of inadequate tests). Even if all these 51 had been the
result of mutations at later cleavages (a most unlikely assumption) this would constitute a
grave deficit on the hypothesis of equal mutability; for the frequency of mosaics should
increase in proportion to the number of somatic cells available per embryo--ot least to
the limits of visibility, which would not be reached until post-cleavage stages. As will
be show.. iater (sec. IB 3) spontaneous somatic reverse events at the pe-locus also showed
evidence of much higher mutability at the first cleavage than at later ones. In the
pe-case, mutation rate again increased at about 10-day post conception in pigment
10
-9 (revit
precursor cells (gonads not involved). The mutant spots produced by such late ovents are,
however, small enough so that they might be easily missed in specific-locus experiments
where the background color is wild-type, although a few such may be included among the
51 animals for whom no answer was obtained.
The large class of half-mutant animals observed is useful for the drawing of certain
conclusions concerning cell lineage of the gonads. First, it is abundantly clear (as had
already been indicated by some of the early mammalian mosaics) that there is no
such thing as purity of gorm line, or even an early separation of coll
lineages, in mammals; for, 18 teto voro, ono might expoot olthor the
thoro
.
-
8
"
germ line or the coat to be of uniform cell type. Actually, both are a mixture
of cell types, at tcast foot of the animate chanthaltet. If the gonad
primordium is set aside as a random assortment of a cells from a cell population
that 18 half mutant and hall non-mutant, then the distribution of animals w.th
regard to proportion of gonad that 18 mutant should be described by the expension'
of the binomial (.5 mutant + .5 non-mutant)”. A glance at the observed dis-
tribution will indicate that it 18 rather broad, at leastwith regard to animado
" of it were not slinulle
to the very-tut bi-percent-mutant-cell tracketofthe table to biased against
animals at the other extreme of the tatabbution-vince (the chances of detecting
being relatively
a very small gonadic mosaic are low?) This wortt means that, if the gonad primordium
16 indeed set aside as a random assortment of a cells, a must be very small, perhaps
( The theoretical expectation for nosis plotted in Fig 2.)
around 5.1 On the other hand, it could mean that there has been some slight
segregation into cell lineages before the time the gonad primordium 18 set aside
and that the assortment is made from a maroonkamited population that is no longer
half-mutant as is true for the organismi as a whole. It 18 possible that, as
breeding results of more mosaics become available and the observed distribution
Low
can be more accurately determined, a decision between these two bypotheses will
be possible.
2. Induced somatic forward mutations
The last section deaüt in some detail with somatic mutations that occur
very early. Information about later events comes from an experiment in which
somatic mutations were induced by radiation (L. B. Russell and Major, 1957).
Embryos heterozygous for four coat-color genes (b = browa, och s chinchilla,
P - pink-eyed dilution, d - maltese dilution) were produced by mating C57BL
females (a/a, full color) to NB rmales (a/abscb/c he doe/dse). Some of
these were irradiated 10.days postconception and all were observed after
reaching adulthood for spots of altered coat color on the normal black background.
8 12
Parallel irradiated and nonirradiated groups from matings of C57BL females to
C57BL males provided controls for radiation effects on pigment-cell killing or
differentiation and for somatic dominant mutations. By making the proper
allowance for these, it was possible to determine what spots were due to somatic
expression of the recessives at the bed, and g loci.
The mosaics found were not tested for gonadal involvement. For one thing,
it was quite improbable that this could be detected, even if present. Moreover,
it seemed unlikely from embryological facts that any cell of a 104 -day mouse
embryo would be ancestral both.to germinal tissue and tissue involved in coat-
color characteristics. This was borne out by independent results on spontaneous
(sec. IB3),
reverse mutations, which will-band locussed below Since genetic tests were thus i
not available, the mechanisms by which the recessives in the-mosaic spots came
to reveal themselves could not be definitely ascertained. One spot was determined
on the basis of histological analysis of pigment granules to be probably of
genotype + 'p' /home, which would indicate that at least some of the cases were
not the result of chromosome loss or somatic reduction. However, in addition to
gene mutation, there remain the possibilities of deficiency, somatic crossover(sec JBL
free belowly and rearrangements causing position effect.
was considerud.
Since it seemed unlikely the tastoor-partout considerations that
there would be selection against the cells in which the scorable change was
observed, the frequency and size of spots could be used to make various calculations.
One of these was the rate of induced genetic change per cell and per roentgen. Of
more interest to the present paper 18 the calculation concerning modal number of
prospective pigment cells in day-104 mouse embryos and the distribution about
this mode. It was fourid that most embryos are rather narrowly distributed around
a wode of 150 - 200, but that the extremes in the range of spot sizes differed
by a factor oz (20. This may indicate a very rapid cell-division rate at this
10
!
stace (intermitotic interval ca. 4 hours). Alternatively, numbers of prospective
pignent cells could vary widely in embryos at this developmental stage; or, cells
of 104-day embryos which are ancestral to pigment cells could also contribute
cell progeny to other tissue, the relative contribution to these two end products
being quite variable.
The work of Rawles (1947) indicates that by day 10%, some prospective
pigment cells (probably from the neural crest) have elready begun to spread
laterally, from the dorsal midline. Since the induced genetic change, in essence,
gives a label to a cell lineage, the appearance of the mosaic areas provides
<
: مه ما 5
lucu
come information on pigment-cell migration and multiplication. The spots, which,
in general, appeared to be distributed randomly over the surface of the animal,
rison
in mai.
...Casos i
were of diffuse outline; and, within these areas there were varying degrees
Eres non
of mixture of mutant and apparently non-mutant hairs. Often, an animal might
have two or more areas which were, however, adjacent, or close to each other,
with a few mutant hairs bridging the gap. A few, but by no means all, of the
spots appeared to start at the dorsal midline; and some gave indication of being
elongated roughly at right angles to the long axis of the animal.
All of these facts are consistent with the interpretations that the cell
lineage from a given pigment precursor cell extends Jatero-ventrally as well
as anteriorly and posteriorly. Cell progeny may never be sent in a dorsal
direction; or, if it 18, it probably does not cross the dorsal midline. Progeny
from a given precursor cell rarely if ever populates the entire distance from
dorsal to ventral midline. Thus, it 18 likely that the ventral and probably
even lateral portions of the fur are populated by the progeay of cells that
have migrated some distance before they reproduce. Finally, there appears to be
a considerable degree of intermingling of the progenies of different jrecursor
cells; and some of the progeny cells may also bave appreciable migratory powers
as evidenced by the occasional observation of two spots from the same mutational event
Do 14
3. Somatic reverse mutations
From the point of view of testing gonadal involvement, somatic reverse
and aguntie forward mutations
mutations, are user at the opposite ends of the scale, from somatte forward
With reverses,
mučationen mall mutant portions of the gonad can be detected; but, on the
other hand, 1t 18 difficult to determine whether the gonad 18 completely
орроя
mutant. Triomis (the hovorocnot to situation thert pertains bith forward repeats,
A number of genetic situations have now been observed in the mouse in .
which a recessive can apparently mutate to wild type with appreciable
spontaneous frequency. One of these (LR-624) has been worked out in some
detail and will be discussed first.
Animals homozygous for the coat-color-diluting mutation, pearl (pe),
occasionally have full-colored patches in their fur. The frequency of such
individuals was 6% in about 800 AV/a pelpe mice observed. (on the AT/A pe/pe
background, even small patches are detectable. Frequency appears less on other,
leos favorable backgrounds.) The frequency was not affected by whether one
cred zrom mosaic or non-mosaic animals, indicating that the degree of instability :
was characteristic of the stock as a whole rather than being transmitted in
sulines. Since garminal tissue was occasionally involved in the somatic event
(see below), it was possible to subject the reverses to genetic tests. In this
manner it could be established that the event had occurred either at the pe-locus
itself or not more than 0.36 crossover units from it; and that it could, at least
tentatively, be considered a reverse mutation of pe to pe* or to some gene
resembling it. This gene proved to be fully viable, both in heterozygous and
in domozygous condition, and completely stable.
of particular interest to the present discussion 16 the distribution of
mosaic animals with respect to spot sizes. Of a total of 61 observed, six had
from 50-100% of their coat full-colored, one had a hill-color spot covering
rent
22
is
about 5% of the surface, and the remainder had much smaller spots (Table ).
In other words, it appeared as though the reverse mutation could occur either
very early or quite late. To estimate the amount or involvement -- If any --
of germinal tissue, one must multiply by two the proportion of pet progeny,
the presumed mosaic gonado being pet/pepe/pe. Since even a completely
mutant gonad would produce only one-half full-colored offspring, it is difficult
to decide whether the gonad 18 ever fully involved. For instance, are the two
100% rull-colored animals the result of reverse mutation in the previous
generation rather than somatic reverses? If so, they should produce 50% pe*
prozeny. Actually they produced 37% (27 out of 73), which is on the borderline
of being significantly less from the (p = 0.03) Cheetot and could indicate that ..
they are mosaics. Annther difficulty lies in deciding whether single pe* animals
in moderately large progenies (1/50, 1/27, 1/34) indicate that a small part of
the gonad is mutant or whether they represent new events.
In spite of these difficulties, the results are fairly clear on the
following points. The spontaneous reverse mutations can occur very early or
later in embryonic development. If they occur early, mutant cells participate
both in the fur and in the gonad. If they occur late, they usually involve the
fur but not the gonad; or -- in the event that the completely full-colored
the result of mutation in the previous generation
animals, observed (see above )were non-masadas -- the gonad but not the fur.
(Animals mosaic in other tissues would, of course, not be detected.) One might
guess, from a comparison with the results of the induced somatic-mutation
experiment discussed above (Sec 18,2), that the late events scored as small spots
in the fur occurred around 10 days post-conception. Presumably, mutations occur-
ring later would result in spot sizes too small to be visible. Animals that were
heavily mottled, with 50__-80% of the fur lovolved, can, by analogy with spontaneous
somatic forward mutations discussed in Sec.IB 1 above, be assumed to have had a
16
27
a.
- as it was
:
over
reverse mutation at the first cieavage or 2-cell stage. The tendency for the
Conad to be either very heavily or very lightly involved 18 here never more
apparent than in the other cases. So far, there 18 only one animal in which
I could have
a mutation apparently occurred at a stage intermediate between days 1 and 10
postconception. This had about 1/20 of the fur and about 1/3 of the gonad
I early.
involved, indicating that mutation occurred to crather torty stuga be is no nepie
that this nuimal, too, traces back to 2-cell stage.
Rough calculations of mutation rate per ceil indicate that the frequency.
may be an order of magnitude higher in the first cleavage (or 2-cell stage) than
it is in 10-day embryos. The difference 18 even greater 1f one includes the
completely full-colored 'animals in the mosaic group. The frequency in
na
intermediate stages 18 probably awtomas or lower than that in 10-day embryos. i
Since the time of the pe pe* reverse mutation study, several other
genes also have been observed to give a relatively high spontaneous frequency
of somatic reversions. Four of these are at the c-locus: two appear identical
to e in phenotypic effect (but one 18 homozygous lethal), and two are intermediate
between c and ce (Russell, unpublished). All of these, either in the homozygous
state or heterozygous with g, give a rather high frequency of animals with
possibly
small dars (, full-colored) patcies in the fur, and one has given a completely
full-colored animal which transmits ct. At the a-locus, also, somatic
"reverse mutations" of the types & + A (Hoecker, upami stiedo ;(Russell, us-
la- Aw (Bhat, 1947)
published) , Diokeys Re# and a → at (Dickey) bave been noted. At least two
Bhre)
one of these (Russell, involved gonadal tissue. Frequent full-colored spots
the dominan
in mice heterozygous for M1**, w, or Va have been interpreted by Schaible
and Gowen (1960) as due to somatic reverse mutation. No breeding results
have been published for these animals.
It should bo notes that the appearance of easily reversible alleles could,
in offoct, roprosont a different genetio situation, 6.8., frequent removal of
a suppressor associated with wild-type, such as the Ds situation in maize.
However, until there is evidence Along these linos, it is easier to speak
of "roverse mutation."
S
ementara
24
17
4. Somatic events that produce two new genotypes
Tr.coretically, some somatic events would produce two new cellular
eenotypes rather than one. These events are: crossing over, nondisjunction,
and reduction to a haploid state. It has not beer definitely established
that any one of these occurs somatically in the mouse, but a number of cases
have been observed that could be explained by one or the other mechanism.
One of these is a mouse, described by
Carter (1952), describecind mouse that had presumably started life as a
.
WV/+* heterozygote (wv, "viable dominant spotting", reduces pigmentation) and
showed an apparent deficiency of the W allele in the coat (full-colored patches)
but a deficient transmission of the ** allele in breeding tests. In his
discussion Carter interpretøs three mosaics described in the older literature
as possibly also having been the result of somatic crossing-over or reduction.
- explained
However, all of these can as easily be interpreted by somatic forward mutation.
On the other hand, a cace rather similar to Carter's is being studied in
our laboratory (Russell, unpublished). A female of originally +_+/-chy genotype
was found to have some light patches (possibly of chy/o phenotype?), but, ia ;
mating with cholech, transmitted a highly significant excess of the + +
chromosome (36 ++ /cchin, gechip/e che, 2 tp/cche). It may be possible to
obtain positive proof oi nondisjunction by cytological tests now in progress.
If not, somatic crossing-over proximal to both ch and p must be considered as
a possibility. (Somatic reduction would require that haploid oogonia can go
through meturation to produce viable eggs, which seems, a priori, unlikely.)
Two other mosaic mice, observed by us in the past, couid have been the
result of somatic crossing-over. These were a/acch/c heterozygotes foragishatian
each of which showed spots of white (g/c?) end of near-black (cchyech ?) on the grayish-
tan ground color.
Trey transmitted cca and Hetta in ratios very close to 1:1. No cytological
tests were possible at the time.
If the event responsible had been nondisjunction,
is
25
rather than crossing-over, the dark spots should have been of genotype ech/ech/g
and therefore presumably lighter than near-black. Since, however, no detailed
analysis of piement was made, no firm decision is possible.
Insert, lo
II.
FUNCTIONAL MOSAICISM
Veried phenotypes within an area of uniform differentiation on a given
animal need not always be caused by genetic events occurring in samatic cells. ·
The genetic content may be uniform, and yet certain genes or whole chromosome
portions may function in some cells and not function in others cello of the same
tissue, the distinction being random, with the tissue not showing any other
sims of differentiation within it. This phenomenon will be referred to as
Iw.ctional mosaicism.
A. White Spotting, Probably Not a Case of
Functional Mosaicism
Several mutations in the mouse (e.g., piebald, s; belted, bt; and a
1. It would seem
number of dominants) cause white spots in an otherwise pigmented coat. Beetrico
the sidsspotsheslikti from absence of pigment ils;they-represent-cases-of-
important to determine whether this could represent a case
tamen in two lovimo differentixtriomprather than of functional mosaicism.
White orecis in pied mice.
This absence
Tokie lack of pignent cells, it wisiterereas has been explained in two ways:
either (1) retarded migration of melanoblasts from the neural crest to the skin,
such that the melanocyte population finally formed lacks enough members to cover
the available terrain (Schaible, 1964 dissertation); or (2) en altered quality
of the tissue environment in the white regions which, in some way, inhibits the
establishment oî melanocytes in the hair follicles (Mayer and Maltby, 1964).
Results ou embryonic transplantation experiments do not yet allow a decision
between these two hypotheses.
Insurt ca p. 18
Klein (1963), in his extensive review on genetics of somatic cells,
sucgests anothor possible instance of somatio crossing over in the mouse.
This is the appoarance, in tumors heterozygous for H-2 factors, of stable
variants selectively compatible with one of the parental strains.
Unfortunately, it is roüü yet possible to distinguish betwoon 3: truly
homozygous tumor cells and phonocopies.
76
19
In favor of retarded pigment-cell migration (or multiplication) 18 the
findinc, both in tre guinea pig (Wright, ) and in the mouse (Russell, un-
published), that the presence of white-spotting factors makes for a clearer
separation of colored areas when more than one pigment type 18 present (as,
2.6., in el guinea pigs or in T(X;A) Pemale mice variegated for wild-type and
recessive -- see below). Thus, the very diffuse and irregular outlines of
soraatic mutation spots that one finds in the absence of piebald factors (Sec.IB)
can, perhaps, be thought of as resulting from rapid bursts of migration from
various centers, with an intermingling on the peripheries. (A possible analogy
would be two drops of viscous liquid released from high, with the splashes
mixing; as comared to two drops spreading slowly towerd each other on a surface.)
The effect of embryonic irradiation on white spotting also seems to
support the hypothesis of restricted melanocyte population. Thus, it was found
(L. B. Russell and Major, 1957) that the incidence of small white ventral spots
could be lncreased from 6.4% to 26.8% by 100 r given to offspring of
C57BL ? x CSTBL o matings on day 10 postconception. On the other hand, offspring
of C5TRL 9 x NB o matings had no white spots, regardless of whether or not
they were irradiated as embryos. The interpretation proposed for this was that,
although pigment precursor cells are killed by the irradiation in both types
of embryos, their places in C57BL * NB animals can be filled by a surplus of
surviving cells; whereas the C57BL X C57BL backgrourd normally provides only
barely enough pigment cells to fill up the surface, with the deficiency noted
at the end of the latero-ventrad migration, 1.e., midventrally. The radiation
finding of Russell and Major has recently been confirmed by Schaible (1964)
using the same stage but a different genetic background. It is more in keeping
with known facts to interpret the effect of radiation as killing of pigment cells
than as altering the quality of midventral follicles to render them non-receptive
to melanoblasts.
2720
However, regardless of which explanation of white spotting will be
Pinally proved correct, the phenomenon seems to be one of histological
differentiation, rather than of functional mosaicism as defined at the
beginning of this section.
B. Functional Mosaicism of Autosomal Factors
Ai though little or nothing 18 known yet concerning the mechanisms lavolved,
four instances should be mentioned in which variegating "alleles" at well-known
coat-color loci have been observed. Two of these occurred at the po and two
at the a-locus. (The phenotypes produced by Varitint-waddler, Va, and some
alleles at the dominant spotting, W, locus are often referred to as "variegated"
since they produce white spots, dilute regions, and occasional full-color patches.
Since these last may represent somatic mutations to wild-type (see Sec.1B3] and
since the white spots are considered by Schaible to be of the piebald type
discussed above (Sec.I A), these genotypes will not be further considered here.)
Two p-locus mutants (provisionally named pm and pm2) that arose in
different radiation experiments at our laboratory, produce, in the po/? condition,
a coat in which wild-type areas and areas of typical p/p-phenotype are freely
intermingled. (In cchr/cching the areas are of cht/ccb and cho/che color; .
i.e., the linked c-locus does not appear to be affected.) On the average,
about half the coat 18 of the p/2 color, but individual animals may vary
considerably in this respect. That the condition 18 not caused by a somatically
highly mutable "wild-type" aliele 18 shown by breeding results: D/2 x p/e
crosses yielded at least 50 per cent mottled animals in well over 1000 young
observed, the rest being p/o. If the mottling were the resulü of a somatically
mutable wild-type, alter then mottled (plus, perhaps, fully wild-type) progeny
should not exceed 25 per.cent. It is thus concluded that you 18 present in all
cells but that--at random--it produces full-colored pigment in some and typical.
P/P pigment in others.
....
...
.
-28 21
Could oh be the result of a position effect? Major rearrangements
appcar unlikely in view of the facts that (a) litter size of heterozygotes 16
liigh average and that (b) recombination in the c-o interval is definitely not
rcüuced, at least in the case of pral (ome not yet tested). The karyotypes are
normal (recit. The results cannot yet rule out small rearrangements or other
types of linked variable suppressors of e*. If such are, however, present,
they must be firmly linked with the p-locus since no recombinants were found in
1,354 young.
A situation apparently analogous to one exists at the a-locus in the
"allele" an derived in radiation experiments at this laboratory. Animals of
senotype 2"/a have agouti and non-agouti patches of fur freely intermingled at ,
the edges. As with pas, mottled a"/a animals produce about 50 per cent mottleå
progeny in crosses to a/a, and they are fully fertile. Evidence concerning
possible linked genes or small rearrangements that might act as variable suppressors
is only beginning to be accumulatedx but is, so far, negative.
Another a-locus "allele" with somatically variable effect is AV, viable
yellow (Dickie, 1962). Animals carrying this gena are often mottled with agouti
patches ranging from small size to a completely agouti coat. Again, these are
not the result of somatic mutation.
Can these four instances of variable effect be ascribed to alleles producing an
Intermediate levels of gene product, with chance local factors of the environment
determining whether an end product 18 or 18 not made? This seems unlikely,
especially in the case of the p-locus, where allelos producing uniformly
intermediate end-product are known; 1.e.,(e-locus) the end product can be
present at various levels rather than being an all-or-none affair. At the a-locus, .
alleles are known that produce agosti and nonagouti hairs on the same animal, but
the distribution of these hairs follows a definite morphological pattern. The
ده و
moill the apparai altor home espressio
randon mottling that is observed in eme, eta, am, and Amy resembles that force
na autoration one which
22:2.50s here, there 18 heterozygosis of certain X-linked genes or certain
autosomal genes involved in X-autosome translocations. These will be discussed
in the next section.
C. The Mammalian X Chromosome
Functional mosaicism in the mouse has been best documented in the case of the
X chromosome. The "inactive" X or, as I have preferred to call it, single-active-X
chromosome hypothesis is by now so well known that it need not be discussed in detail
-
(see L. B. Russell, 1964, for a recent review). The hypothesis included two main
points: (1) that only one X-chromosome is active in somatic cells of mammals regardless
of the number actually present; and (2) that the choice of which X is the active one is
random, but that once differentiation has occurred it remains fixed in subsequent cell
generations.
Evidence that one X was active (b. B. Russell, 1961) came from the study of
"variegated-type position effects" in the mouse; from cytological and cytochemical evidence
triat the two X chromosomes in a normal female do not act alike; and from the study of
sex-chromatin. That X-differentiation must be random and fixed (Lyon 1961, 1962) was
evident from mosaic phenotypes in females heterozygous for X-linked or X-translocated
genes, and was later borne out by cytological findings. Most of the expected consequences
of the hypothesis--dosage compensation, the coexistence of two cell populations in hetero-
zygous females, and variability of such females--are indeed found.
There have, however, recently been more and more indications that the hypothesis
should not be taken in its simplest form (L. B. Russell, 1964). Evidence from X-autosome
20- 23
translocations in the mouse indicates that inactivation proceeds from a certain point or
region of ine X (L. B. Russell, Bangham and Saylors, 1962; L. B. Russell, 1963). By
studying variegation characteristics of several translocations involving the same two
chromosomes (linkage group I and the X) it was found that degree of inactivation is
dependent on the position of the rearrangement points. A given gene may be inactivated
in a high proportion of cells in the case of one T(X;1), o low proportion in another, and
not at all in a third. This finding, combined with the typical spreading effect that is
observed when two loci are studied together in the same translocation, suggests that
inactivating influences spread in gradients from a certain part of the X, probably not its
centromere region. That this postulated "flow" of the inactivating influence affects
X-chromosomal loci themselves, and not just translocated autosomal loci was shown by a
translocation recently described by Searle (1962) in which the gene Tat had apparently
become separated from the inactivating region of the X and was now invariably active.
The conclusion that inactivation may spread for a certain distance from a certain
point suggests that there may be portions of the X-chromosome that do not normally come
under this influence and are always active. Supporting evidence for this comes from the
finding of a few X-linked genes that do not conform to the simple inactivation hypothesis,
from the lack of complete normality in abnormal sex-chromosome constitutions, and from
cytological data (L. 8. Russell, 1964, review).
D. Mechanisms in functional mosaicism
The problem of how genes are turned on and off in a particular sequence during
development is, of course, the basic problem of differentiation. What is unique about
the mammalian X chromosome--and, perhaps certain autosomal genes also (seci. IlB)--
is the fact that the differentiation is between homologues rather than between different
-ar 24
parts of the genome. (Furthermore, this differentiation occurs in a normal genome,
rather than, as in Drosophila, in one containing rearrangements [Lewis, 1950, chodka
review7 or extra chromatin (Cooper, 19577.)
Although a number of authors have suggested possible mechanisms for X-chromosome
differentiation (L. B. Russell, 1964, review), we are undoubtedly still a long way from
the facts. Only one point seems relatively clear: although the altered state observable
after a certain stage in development is that of the "inactive". X, differentiation must,
instead, somehow single out the X that remains active; for there is always only one of
that type while any number of additional X's can become he teropyk/notic. It is perhaps
not unreasonable to think of X differentiation as being under the control of some part of
the autosomal complement, such that certain autosomal gene products in double dose are
necessary to retain activity of one X. The findings in polyploid cells support this idea.
Elucidation of the mechanisms involved in the induction of one X chromosome as
a continuing active one thus constitutes the first problem. Another set of problems centers
around the ways in which an entire chromosome or, perhaps, its major portion becomes
inactivated. One may have to look for mechanisms that can selectively affect replication
rhyt:ms; for it is easier to think of late replication)
as a cause of inactivity (non-availability of gene products at the proper time), than
vice versa. To postulate a controlling center for replication time of the chromosome as
a whole seems, however, an oversimplification, in view of the increasing evidence
(sec. IIC) that inactivity may not extend to the very ends of an autosomal piece
translocated with the X and, in fact, possibly not even the entire length of the X. If
this is so, and if inactivity is the result of late replication, ihen it may be that the
influence that delays replication "flows" along the chromosome, but only for a certain
distance.
22 25
Dous this influence extend for a fixed distance and then end abruptly, or does
it perer out near its limits? The findings with X-autosome translocations support the
latter alternative. Thus, when the sequence along the translocated autosomal piece is
rearrangement point--c-locus--p-locus, animals are heavily mottled with c areas and
only very slightly, or not at all, with cp-areas, but not with p-areas alone. The
reverse situation exists if the sequence is rearrangement point--p-locus--c-locus. It then
becomes important to determine whether a gene near the limits of the inactivating
iniluence is affected as often but only partially so; or whether, on the other hand, it
has a smaller probability of being affected by complete inactivation. In an attempt to
answer this, the genotype (c p 5 R/cpt has been constructed (R = rearrangement point).
If ct, which is at a considerable distance from R, were only partially affected, mottling
would be with some intermediate color e.g. resembling och/c. However, os neorly as
can be determined, mottling is with white only, i.e., ctis apparently completely
incctivated, only less often so.'
It thus appears that inactivation is all-or-none but may be variable at the extremes
of the range of the inactivating influence. If this should be true also of the non-translocated
complement, i.e., of the normal X chromosome, it would pose new problems of dosage
compensation for certain X-linked genes. Genes located in portions of the X chromosome
that are invariably inactivated are always present in single dose and thus require no
dosage compensation. Those located beyond the range of the inactivating influence are
always present in double dose and possibly come under some system of dosage compensation
similar to that proposed for Drosophila (Stern, 1960, review). But the few that are at the
extremes ci the range could bs present in single dose in some cells of the body of a female
and in double dose in others. Perhaps this would not lead to such violent disorder as
michi, ci sini be anticipated. Obviously, many autosomal genes must be able to function
.
-1.2
.
-
..
-T
a
1
*
1
IT
NWT
A
...
**
: ..
..
.
UNCLASSIFIED
ORNL




7
..'''.'
P
70
20F2
.
-
.
-
-
ve
Pui-.


..
..
htm .
6

DTIE
micROCARD
ISSUANCE
DATE
.
PORY
iii
2.
8 / 7



E
.
::
.
1964
_
Co
m p
l
ementaren aurrean eta
-23
26
successfully in either single or double dose; for large segments of autosome, when involved
in translocations with the X, become functionally monosomic.
E. Functional mosaicism as a tool in studying gene action
While the mechanisms that bring about functional mosaicism may themselves shed
light on the nature of gene action (when they are elucidated), the very existence of
functional mosaicism provides a tcol for studying certain properties of genes and of
chromosomal portions.
Since the X-chromosome influence appears to cause total inactivation within a
certain range, the hornologous genes are expressed in an effectively hemizygous state.
It thus becomes possible to use females heterozygous for X-autosome translocations as a
specific
test system to determine whether certain alleles introduced on the intact autosome are
Both types have been foundê
amorphs or hypomorphs (b. B. Russell, Bangham, and Montgomery, 1964). In addition,
the system can be used to find out if certain autosomul mutations (possibly deficiences)
known to be homozygous lethal also act as cell lethals. Twelve independent brown-lethals
(general symbol ) and a pink-eyed lethal, pl, have already yielded variegated
b? females, indicating that these mutations are not cell-lethal. It was,
of course, already apparent that some X-linked mutations (Mo-series, Str, etc.), known to
be hemizygous lethal, are not cell-lethal; for these mutations are, in fact, detected in
R(+)/
functionally mosaic heterozygous females. Moreover many of them have arisen by.
somatic mutation in males where they are present in genetically (not functionally)
hemizygous condition in part of the body (sec IBI). .
In the course of the experiments designed to determine whether the
-lethals
were cell-lethal, it was found that female translocation heterozygotes were less viable
if their intact autosome carried the lethal than if it carried the viable recessive.
26 27
. For instance, while in the cross R(+)/6 xb/ 40.9% of 672 classified females
were brown-variegated (Russeli and Bangham, 1961), the cross of R(+)/6 x +' guve
only 12.4% brown-variegated females in 218 classified, the expectation for equal viability
0.004).
being 20.5% (p = 0.01-chockint=2.85). The le type similarly appears in reduced
numbers. Certain conclusions can be drawn from this concerning the time of X-chromosome
1.6.
inactivation. For the particular pe mutation used, we determined in independent experiments
that p/elembryos call forth an implantation reaction but die by day 6-1/2 or possibly
before, but not earlier than day 2-1/2 (no observations, to dote, between days 2-1/2
and 5). If the time of pl-lethal action were to precede the time of onset of X inactivation,
R(4)/{ should be of normal viability, for pl in each cell would then, at the critical time,
still be "covered" by an active pt. The reduced viability of R(+)/pl thus indicates that
X-chromosome inactivation begins before day 6-1/2.
The fact that surviving R(+)/pl are variegated, i.e., pl is expressed in a hemizygous
state, could indicate one of two things. (1) X-differentiation is already complete at the
time of pf-letha! action, i.e., all cells are already effectively +/O or oie', with those
embryos succumbing in which the latter cell-type by chance exceeds a certain threshold ·
proportion. If this is so, one must suppose that death is caused by some organismic effect,
for if it were the result of removal of /p' cells, the surviving embryos would consist of
+/0 cells only and not be variegated. (2) Alternatively, X-inactivation may have occurred
in only some cells by the end of the critical stage, i.e., the embryo may consist of
effectively +/0.0/' and +/pl cells. Again, the proportion of Ole' cells would
determine whether or not the embryo succumbs, but regardless of the mechanism of death
(0/2-cell removal, or organismic), survivors would still become variegated because the
cells that had been +/ed by the end of the critical period would later differentiate into
+/O and O/pl.
290 28
It should be noted that, in the case of alternative (1), the period of X-differentiation
is confined to an interval beginning and ending before day 6w1/2; this interval could be
very short. In the case of alternative (2), X-aifferentiation begins before day 6-1/2 but
extends for some time pust that stage.
The X-autosome translocations may perhaps serve as tools in the solution of an
entirely different problem, the requirements for male fertility. In all of six different
translocations studied by us and one studied by Searle (ref.), the males, although copulating
freely, are completely sterile as a result of disturbances in spermatogenesis. One proposed
explanation of this sterility (LR-25) is that there may be an upset in the rhythms of
chromosome condensation and movement during meiosis. Thus, one translocated chromosome i
may have "divided loyalty" in its affinity for the Y on the one hand and for an autosome
on the other (the X-Y bivalent and autosomes being normally governed by different
rhythms). Juan Valencia (private communication) has suggested a differeni mechanism.
The "sex vesicle" which appears in prophase of the first meiotic division, is normally
formed by the X and Y chromosomes. In translocation malos, however, it would .
consist of Y and presumably only part of an X, namely that part conta ined .
in the translocated chromosome that pairs with the Y. If it can be
supposed that the sex vesicle has some positive function for the
completion of normal meiosis, sterility could be explained by a deficient
sex vesicle. It should be noted that a translocation studied by Cattanach
(1961), which probably consists of insertion of a piece of linkage-group I 1
into the X (Ohno and Cattanach, 1962) is not completely male-sterile. In
kooping with Valencia 's idea, mice with this translocation would not form
a deficient sex vesicle. However, the finding that translocation males with
two normal autosomes (1.6., with the insertion acting as a duplication) are,
if they survive at all, mora fertile than are balances translocation males
(1.e., with one normal and one deficient autosome) cannot be explained on
the basis of Valencia is suggestion. It may fit better with the idea of
divided affinity.
29.
SUMMARY
The considerable body of information thut has become available in recent
years on both genetic and functional mosaicism in the mouso provides somo answers
to problems of mutability, cell lineage, and gone action.
A. Genetic moscicism
1. Among spontaneous genetic mosaics, there has been one chimera that can
be expico...d on the basis of either retention of a fertilized polor body or fusion of
to genetic events occurring after the zygote stage.
2. Spontaneous mutability appears to be relatively very high at first cleavage
or in the 2-cell stage, becomes low in subsequent cleavages, and then rises again
a
somewhat around 10 days postconception (nothing is known concerning later stages
because the mutant spots would become too small for analysis). This is true both for
forward mutations (or deficiencies) observed in animals heterozygous for standard
coat-color markers, and for reverse mutation of certain genes. Somatic loss of
Gutosomes is rare or may be selected against.
3. In over 40 gonosomic mosaics whose breeding results were analyzed there
was no evidence for selection against mutant cells in the gonad or in the progeny.
The distribution of these mosaics with respect to proportion of germinal tissue consisting
48.3%
ci mutant cells centered about a mean of 426 and was rather broad. If it is assumed
that the gonad primordium is set aside as a random assortment of n cells from a cell
population that is 50% mutant, the observed distribution fits an n value of 5. The re-
sults can alternatively be interpreted on the basis of some slight segregation into cell
lineages before the time the gonad primordium is set aside.
4. A number of animals have been observed that can be interpreted on the
basis of somatic events producing two new genotypes, i.e., somatic crossover,
rondisjunction, or reduction, with the first one the most likeiy. No definite proof
for any of these events has, however, been obtained to date.
5. Instances of induced somatic forward "mutations" and of certain spontaneous
reverses, by providing labels to sht lineages, have been used to make interpretations
concerning migration and multiplication of pigment-coll precursors. Observations on
white sporiing may also be useful for this.
B. Functional mosaicism
A funciional differentiation of genes or whole chromosome portions which is
random and not related to histological or morphological differentiation leads to a
state referred to as functional mosaicism.
1. Four mottling-alleles at autosomal loci have been found. They are not
somatically mutable alleles but, rather, genes that have all-or-none expression at
random. No major rearrangements seem to be involved.
2. Functional mosaicism in the mouse is best documented in the case of the
X chromosome. The single-active-X hypothesis states that only one X is active,
that the choice of this X is random, and that once differentiation has occurred, it
remcins fixed in subsequent cell generations. Recent evidence suggests a modification
of this hypothesis to the effect that inactivation may not involve the whole chromosome
but may spread for a certain distance from a certain point.
3. Mechanisms that need elucidation are (a) how one X is singled out as the
active one, and (b) how major portions of a chromosome can become inactivated.
Inactivity may be the result of late replication. An influence that delays replication
rray "flow" a certain distance from a controlling center in the chromosome. Results
indicate that genes near the limits of this influence are still completely inactivated
out less often so.
4. Functional mosaicism has been used as a tool in de termining whether certain
aileles of autosomal loci are amorphs or hypomorphs. It has also boon used to show
inot a considerable number of mutations known to be lethal do not act as cell lethals.
5. The study of an X-autosome trunslocation, in conjunction with on autosomai
allele whose time of lethal action is known, indicates that X-chromosome inactivation
begins before day 6-1/2 postconception.
in
Diri.winnion
TABLE I
of mice having presumed somatic reveries
of pearl , pe.

Estimaiod %
of coat involved
Number
observed
Number
tested
Number with germinal tissue:
very slightly involved heavily
involved (or.new events) . involves
0.1 -1%
39 + iga
18+2"
19
(40)
0
5%
50 - 90%
100%
o
to
2 (Yaz, Yay)
0
2(474x, '754)
2/19/6, 747
2
NOS is
* Anima's with questionable spots. Diagnosis difficult
either because o unfavorable background color
(Al pe/pe, a/a pe/pe) or because of very small size of
presumed spot
Fractions in parentheses show ratio of pet to total
Size
.
classified
roo
progeny
и
у
Or
꾸십
​9 +
PRI- 45:0T!C DIVISION - SECOND MEIOTIC DIVISION,
SPERM ENTRY .

(At or a

4
+
-
*
*
.
.
.
.
. .
.
-
.

.
: 24 ::
.
.
.
.
.
. .
-
*
i
4
. .
:
om...............
...
...
.
mosaics

-
-
-
gonosomic
--
-
-
...
of
..
•
•
•
Percent
.....
.
:..
-.
L
-
...
........
.
..
-
-
:
:
of
10
20
30
40
50
60
70
80
Percent
gonadal cells
heteroay goue
mutrition
cosa comosoom ...
10. ONOMIIIT
Tis 2.0

Abnormality
Parents
Gametes
Zygote
2-Cell
Stage
+
.
.
.
.
.
-' :
Mutation in
one strand of
double-stranded
+97
pp
+/a
gamete
Polar-body
retention
ja
+/a

tla
Fusion of two
independent
zygotes .
a/a
It'

Mutation at
first cleavage
ata
.
..
;
EL
.
??
.
-.
(END