MUTABLE LOCI IN MAIZE

Barspara McCiintock

During the past year the study of mu-
table loci in maize has been continued, in
an effort to determine the mode of origin
of mutable loci from normal loci and to
ascertain the events occurring at a mutable
locus that result in detectable changes in
phenotypic expression. Progress has been
made with respect to both these objectives.

As stated in previous reports, two main
classes of mutable loci have appeared and
are continuing to appear in the maize cul-
tures. One class includes a number of mu-
table loci that undergo changes in action
only when a second locus, the activator
(Ac), is likewise present. Mutable loci of
the second class do not require such an ac-
tivator locus. During the past year, study
has been continued only on the Ac-con-
trolled mutable loci. The decision to con-
fine efforts to these mutable loci was made
because all of them respond to the same
Ac locus, regardless of the diversities of
phenotypic expression they represent. On
the basis of this common response to the
presence of Ac, it could be suspected that
the events leading to a change in pheno-
typic expression are of the same nature in
all the Ac-controlled mutable loci. What
are these events? Also, why do normal,

“wild-type” loci suddenly become unstable
in these cultures?

Previous reports have discussed in detail
the Ac-controlled mutable Ds (dissocia-
tion) locus. It was shown that de may
induce chromatid breaks at the Ds locus
that are followed by fusions of broken
ends, and that these fusions may result in
the formation of a dicentric chromatid and
a U-shaped acentric fragment. It was also
pointed out that each such event is com-
parable, with respect to time and fre-
quency of occurrence, to mutations of
other loci that produce recognizable pheno-
typic changes in genic action. It was con-
cluded that Ac must give rise to a specific
condition in certain cells of the plant that
brings about an alteration in the mode of
reproduction of the Ds locus in these cells
during the mitotic cycle. This alteration
eventuates in the production of breaks in
the sister chromatids at the Ds position, 2§
previously described. By genetical and
cytological test methods, it was possible to
place this Ds locus at a position demark-
ing the proximal third of the short arm of
chromosome g. Continued study, however.
has revealed a type of event involving the
Ds locus that appears to be responsible
DEPARTMENT OF GENETICS

for the origin and subsequent behavior of
all Ae-controlled mutable loci. This event
brings about a transposition of the Ds locus
from one location in the chromosome com-
plement to another. In its new position,
Ds responds to de just as it did in its
previous position. (The position of Ds in
the short arm of chromosome 9, where it
was first detected, has been designated the
“standard position.”) These transpositions
of Ds are not infrequent. In the sporo-
phytic tissues, they usually occur late in
development and in individual cells of the
plant. For transposition to occur, de must
likewise be present. When Ds is trans-
posed from its standard position to another
position in the short arm of chromosome 9,
the new location may be readily de-
termined.

THe MECHANISM OF TRANSPOSITION OF THE
Ds Locus

A number of cases of transposition of Ds
are now under investigation. In some of
these, a gross chromosomal alteration has
accompanied the transposition of Ds.. By
cytological and genetical analyses of the
cases involving gross chromosomal aber-
rations, it has been possible to reconstruct
in considerable detail the events that must
have occurred to bring about a transposi-
tion of the Ds locus. These events are
similar in all analyzed cases, and can be
summarized as follows: During a mitotic
cycle a condition may be produced at the
Ds locus that results in the removal from
one or both chromatids of a submicro-
scopic fragment of chromatin containing
the Ds locus. Both ends of this fragment
are unsaturated; and the mechanism of re-
moval of the fragment may be a tearing
process, since unsaturated ends, capable of
fusion, are produced in each of the chroma-
tids of chromosome g at the position
where the fragment was situated. If, dur-

143

ing the same mitotic cycle, a spontaneous
break occurs elsewhere in the chromo-
some complement, four additional broken
ends may be present in the nucleus. Since
any unsaturated broken end is capable of
fusion with any other unsaturated broken
end, a number of different consequences
of fusion among the twelve broken ends
can arise. If the spontaneous break occurs
in the short arm of chromosome 9g at a
position other than the Ds locus, several
types of altered chromosomes g can be
formed. These may have a deficiency, a
duplication of a segment of the short arm—
either in a normal or in an inverted order
—or an inversion. On the other hand,
fusions of broken ends can bring about a
transposition of the Ds locus without an
accompanying gross chromosomal rear-
rangement. If the spontaneous break oc-
curs in one of the other chromosomes of
the complement, a translocation between
the short arm of chromosome g, at the
position of the Ds locus, and this other
chromosome can be produced. A trans-
position of Ds may likewise accompany
such an event. Examples of these various
kinds of translocation and transposition
have been found. Those involving trans-
positions of Ds within the short arm of
chromosome 9, either accompanied or. un-
accompanied by gross chromosomal rear-
rangements, have been selected for con-
tinued investigation.

In the analyzed cases of transposition of
Ds, the inserted segment of chromatin
containing the Ds locus is not visible in its
new position with the light microscope.
It is also too small to affect detectably the
percentage of crossing over in adjacent
regions in plants heterozygous for the
transposed Ds locus. Its detection in the
new position is easy, nevertheless, because
it behaves as it did in its former position;
dicentric chromatids and acentric frag-
ments may be produced by subsequent
144

breaks and fusions that now occur at this
new position. Because it behaves in its
new position as it did in its former posi-
tion, transposition from this new position
to still another position may occur sub-
sequently.

The discovery of the transposition of the
Ds locus, and the knowledge gained in
determining the principal events respon-
sible for it, have supplied the information
needed for understanding the origin of
other Ac-controlled mutable loci. It has
also become possible to formulate a more
direct approach for investigation of the
primary effect of Ae on the Ds locus, wher-
ever it may be, and to determine more
fully the various changes that are known
to occur at the Ds locus itself.

THe Oricin of Ac-ControLttep MutTaBLe
Loci

In Year Book No. 47 (1947-1948), the
sudden appearance of an Ac-controlled
mutable ¢ locus was described. It was
found in a single one of the tested male
gametes produced by a plant having one
Ac locus. This plant was also homozygous
for a normal C locus and for Ds in its
standard position. In this gamete, the ac-
tion of the C locus had changed. It be-
haved thereafter like the known recessive
(c) but, unlike this recessive, was capable
of mutating back to a normal C action
when Ac was present.

Study of the ¢”* locus has been of par-
ticular importance in revealing the factors
associated with the origin and subsequent
behavior of Ac-controlled mutable loci. It
is now apparent that the mutable ¢ locus
arose when the Ds locus was transposed
from its standard position to a position
within or close to the normal C locus.
This event occurred late in the develop-
ment of the parent plant, and probably
only in a single cell of this plant. No gross

CARNEGIE INSTITUTION OF WASHINGTON

chromosomal rearrangements accompanied
the transposition. The chromosome 9g car-
rying this transposed Ds locus is morpho-
logically normal. The transposition of Ds
was recognized by the altered position of
the chromatid breaks associated with Ds
behavior and the concomitant disappear-
ance of such events at the standard loca.
tion. Both cytological and genetical test
methods, used to determine the location of
these breaks, were in agreement in placing
the Ds-type activity at the known position
of the normal C locus in the short arm of
chromosome g. In its new position, the Ds
locus presumably inhibits the normal ac-
tion of the C locus. The C locus, although
present, does not appear to function, and
as a consequence no aleurone color is pro-
duced. With respect to pigment forma-
tion, the tissue response is the same as that
given by the known recessive allele, c, or
by a deficiency of the C locus. This in-
hibited C locus, however, can mutate to a
state that re-establishes its former action.
This occurs only when Ac is also present
in the nucleus. The restoration may be
permanent. The restored C locus no longer
shows unstable behavior in the presence
of Ac, and it cannot thereafter be dis-
tinguished from a normal C locus. What
occurs, then, at the inhibited C locus to
restore its normal action?

As stated previously, the studies of a
number of different transpositions of the
Ds locus have shown that Ds may be re-
moved from a chromatid and that the
mechanism of removal involves compound
chromatid breakage at this locus. The re-
moved fragment containing the Ds locus
has unsaturated broken ends, and the ends
formed in the chromatid by its removal
are also unsaturated and capable of fusion.
It is known that Ds activity usually disap-
pears completely at the c”* locus when 4
mutation from ¢ to C occurs. The known
mechanism of removal of Ds from 4
DEPARTMENT OF GENETICS

chromatid, gained from a study of trans-
positions of Ds, suggests an explanation of
these mutations. An event leading to re-
moval of the inserted Ds segment from
the C locus would give rise to two broken
ends in the chromatid. Fusion of these
broken ends would re-establish the former
normal genic order, and remove the in-
hibitory action on the C locus induced by
the inserted segment; and as a conse-
quence a mutation from ¢ to C would be
evident. No further changes at this locus
would occur, for no Ds locus would be
present to produce them. The C locus
would be completely normal again. If
this primary event is responsible for the
c to C mutations, it also explains why a
few of these mutations are accompanied by
detectable transpositions of Ds. Transposi-
tions could take place if a spontaneous
chromosome break, elsewhere in the chro-
mosome complement, occurred in the same
mitosis that removed Ds from the C locus.

The analysis of the events occurring
when Ds is inserted into or close to the

normal C locus has made it possible to

interpret a previously puzzling aspect of
Ds behavior at its standard location. At
this position, two contrasting “states” of
the Ds locus have long been recognized.
When one of these states (state I) is in
effect, the majority of mutational events
occurring at the Ds locus result in the
formation of a dicentric chromatid and a
U-shaped acentric fragment. In the con-
trasting state (state II), there is a markedly
lower frequency at this locus of breaks
and fusions resulting in the formation of
dicentric chromatids or other gross chro-
mosomal rearrangements.

The above two contrasting states of Ds
may be recognized when it is at the C
locus (c”*). In the original isolate of
c™, a state I Ds locus was present. This
was the same state of Ds that had been
present in the chromosome before its trans-

145

position to the C locus. In kernels having
this state of Ds, only a few mutations giv-
ing a C phenotype appear. This state of
the Ds locus at c™? changes rather fre-
quently, and by a single event, to one that
is comparable to state II of the Ds locus at
its standard position. The event is made
evident by a greatly lowered frequency of
dicentric chromatid formation. The rate of
¢ to C mutations rises to a frequency that
is comparable to the previous rate of di-
centric chromatid formation. It has been
determined that the ¢ to C mutations are
associated with a simultaneous loss of Ds
activity. This relationship indicates that
the change from a c to a C phenotype is
associated with an event involving the Ds
locus itself. A normal chromosome g hav-
ing a fully active C locus but no Ds locus
is the usual consequence. An interpreta-
tion of the event leading to a C phenotype
has been given above. On this interpreta-
tion, the two contrasting states of the Ds
locus reflect the relative frequencies of al-
ternate consequences of the breakage events
occurring at this locus. Both types of con-
sequence are recognized when Ds is at the
ce locus, but only those giving dicentric
chromatids or other gross chromosomal ab-
normalities are detectable when Ds is at
its standard position. At this latter posi-
tion, Ds may inhibit the action of the ad-
jacent loci, but the inhibition may not be
recognized because it results in no obvious
change in a readily detectable phenotypic
character. In this case, neither the inhibi-
tion of genic action brought about by the
insertion of the Ds locus nor the release
from inhibition following its removal
would be evident. Detection of the fre-
quency of breakage events at the Ds locus
would be confined to the fraction that re-
sults in the formation of a dicentric
chromatid and a U-shaped acentric frag-
ment. This fraction may be high or low,
depending on the state of the Ds locus.
146

That the time and frequency of aberrant
events occurring at the Ds locus may be
the same for each of these contrasting
states will be indicated in a later section.
The important difference is in the conse-
quences of the breakage events, not in the
frequencies of the events themselves.
The recognition of different states of the
Ds locus makes it necessary to consider
the factors responsible for the origin of
these states and the conditions present in
each. Two clearly distinguishable states
of Ds have been described above. Other
states of this locus have been recognized.
When Ds is at the C locus (c”*), these
several states are distinguishable, one from
another, by the relative frequencies of the
two main consequences of events occurring
at Ds—that is, dicentric chromatid forma-
tion or ¢ to C mutations. At the standard
position, the comparable states are ‘dis-
tinguished, one from another, by the rela-
tive frequency of only one of these con-
sequences—dicentric chromatid formation.
These states appear to be intermediates be-
tween the extreme state I and the extreme
state II. It has been well demonstrated that
a Ds locus giving a high frequency of di-
centric chromatid formation may change
at a single mitosis to one that gives a low
frequency. A Ds locus giving a low fre-
quency of dicentric chromatids, on the
other hand, does not change to one giving
a high frequency at a single mitosis. This
change from extreme state II to extreme
state I requires several stepwise events,
reflected in the intermediate states. These
observations would suggest that the indi-
vidual states of the Ds locus are indications
of the number of active Ds units that may
be present in a small chromatin segment,
and that the change from one state to
another involves a change in number
and/or distribution of these units within
the segment. Such changes might be ex-
pected to occur as one of the consequences

CARNEGIE INSTITUTION OF WASHINGTON

of thechromatid-breakage-and-fusion mech-
anism associated with the aberrant events
occurring at the Ds locus. On this inter-
pretation, it could be concluded that the
extreme state I] Ds locus has few Ds units
and that the extreme state I Ds locus has
many such units; for the mechanism could
readily reduce the number of units through
losses at a single aberrant mitosis, but
would require a series of such mitoses to
build up a large number of units.

The analysis of the origin and subse-
quent behavior of Ds at the C locus has
served to clarify some other aspects of this
study of mutable loci. Why did new Ac-
controlled mutable loci arise in these
stocks? Why did a normal “wild-type”
locus suddenly behave as a mutable locus?
What event occurred at the locus to bring
about a mutation, that is, a change in
phenotypic expression? The analysis of the
origin and behavior of ¢”1 has made it
possible to approach these questions and
to formulate a concise interpretation of the
origin and behavior of the other Ac-con-
trolled mutable loci. Inhibition of a locus,
either qualitatively or quantitatively, by
insertion of a foreign bit of chromatin can
be followed by release of this inhibition if
the foreign chromatin is removed, trans-
posed, or in some manner altered in posi-
tion with respect to the inhibited locus.
The primary mechanism that allows for
such changes at a locus is associated with
compound chromatid breaks at the locus
and subsequent fusions of the broken ends.
In its initial aspects, it is only necessary to
consider a single locus having the peculiar
faculty of undergoing such breakage
events, at whatever position it may be
located, to account for the origin and be-
havior of many different mutable loci.

TRANSPOSITION OF THE Ae Locus

During the past year, an extensive study
of the inheritance behavior of the Ac locus
DEPARTMENT OF GENETICS

was undertaken. This study has estab-
lished that Ac is inherited as a single unit.
It shows typical Mendelian inheritance,
with one important exception. This ex-
ceptional type of inheritance behavior is
the same as that shown by Ds: transposi-
tion of the locus from one position in the
chromosomal complement to another. Two
or three per cent of the gametes of an
Ac Ac plant may be derived from cells in
which a transposition of Ac has taken
place. These transpositions usually occur
relatively late in the development of the
plant. Plants derived from zygotes that
have Ae loci in allelic positions in each of
two homologous chromosomes may give
rise to a few gametes with either (1) two
Ac loci showing no linkage with one an-
other, (2) two Ac loci completely linked or
very closely linked, or (3) no de locus at
all. When an Ae locus is transposed to a
new position, it shows typical Mendelian
inheritance at this new position. Linkage
with known genic markers can be estab-
lished. Here, again, exceptions may arise
as the consequence of a few transpositions
from this new position to still another posi-
tion. The frequency of these transpositions
is not high enough, however, to distort
seriously the statistical data of linkage
studies. It is likely that the mechanism
producing transpositions of A¢ is the same
as or quite similar to that producing trans-
positions of Ds.

Ac itself is a mutable locus. It can be
identified only by its action on Ds. Its
mutations are made evident by changes in
the time and frequency of Ds mutations.
(The events at the Ds locus that result in
either dicentric chromatid formation or a
change in phenotypic expression of a Ds-
inhibited locus will be termed “Ds muta-
tions” in this account.) It is known that
the number of Ac loci in the nucleus con-
trols the time and frequency of Ds muta-
tions. Increased doses of Ac loci (from 1

147

to 3 in the triploid endosperm) result in
an increasingly delayed time of occurrence
of Ds mutations. Similar changes in the
mutational response of Ds will be regis-
tered after a somatic mutation in a single
Ac locus. These responses indicate that
some quantitative change may take place
at the Ac locus when it mutates—probably
an increase or decrease in the number of
subunits at this locus. Thus, superimposed
on those quantitative changes that can be
produced by additions of whole Ae loci
through controlled chromosome combina-
tions in diploid tissues of the plant or in
triploid tissues of the endosperm are those
that can occur at a single de locus. °
There is a ready method of identifying
those kernels on the ears of Ac Ac plants
that are likely to have a transposed Ac
locus. This involves crossing plants having
no Ac locus to plants having a single Ac
locus in which the Ae state is known (de-
termined by its effects on Ds in 1, 2, and
3 doses). The F: plants are selfed and
the F. progeny grown. The F» plants are
then crossed by plants having no de locus
but carrying Ds at its standard location in
each chromosome 9. The ears produced
by the Ac Ac Fz plants are selected, and
an examination is made of the Ds muta-
tion rates in the kernels. If, in the Ae Ac
Fz plants, no mutations have occurred at
the Ac locus and no transpositions have
taken place, all the kernels should show
the same pattern of Ds mutations. In
other words, the control of these Ds muta-
tions should be the same, since all the
kernels should have two de loci in the
endosperm cells and all the Ac loci should
be alike. The majority of the kernels on
such ears do show a remarkable similarity
in the pattern of expression of Ds muta-
tions. A small percentage of the kernels,
however, are markedly different. These
exceptional kernels fall into three classes:
(1) those showing no Ds mutations at all,
148

(2) those showing a few very late-occur-
ring Ds mutations that suggest an increase
in Ac dosage, and (3) those showing a
time and frequency of Ds mutations that
suggest a lowered dosage of Ac. A pre-
liminary test was made in an attempt to
determine the reason for the changed re-
sponses of Ds in the kernels of types (1)
and (2). Twenty-five such kernels were
selected from these ears, and plants were
grown from them. Tests were conducted
to determine (1) the presence or absence
of Ds, (2) the presence or absence of Ac,
and (3) the action of Ac, when present,
in one and two doses. Eleven of the plants
arising from these selected kernels gave no
evidence of Ac at all; the Ac locus was
either absent altogether or completely in-
active. Ten other plants had two inde-
pendent, nonlinked Ac loci. In four plants,
Ac was inherited as a single unit; but this
unit, in a single dose, produced the same
effect on Ds mutations that two units of
the original Ac locus, from which it was
derived, had produced.

One type of event, the transposition of
Ac, will account for these results. If, in
these Ac Ac F2 plants, transposition of one
of the Ae loci occurred in a meiotic or pre-
meiotic mitosis, two Ac loci would still be
present in the nucleus, but they would no
longer be allelic with respect to position in
the chromosomal complement. If the
transposed Ac locus were inserted into a
nonhomologous chromosome, meiotic seg-
regations could give rise to gametes with
either (1) one Ac locus, in its original
position or its new position, (2) two Ac
loci, one in each of two nonhomologous
chromosomes, or (3) no Ac locus. Trans-
position within the same chromosome (or
homologue), or insertion of the Ac locus
of one chromatid adjacent to the Ac locus
of the sister chromatid, would give com-
parable meiotic segregations with respect

CARNEGIE INSTITUTION OF WASHINGTON

to the production of gametes with two 4c
loci or with no Ae locus.

In the given cross, the kernels arising
from the megaspores having no Ac locus
would show no Ds activity; for no D;
mutations occur without Ac. Tests for Ac
in the plants arising from these kernels
would be negative, because no Ac locus
would be present. Kernels developing
from megaspores receiving a single Ac
locus, either in its original position or
transposed but unmodified in its action,
would show the characteristic effect on Ds
mutations produced by the Ac locus when
two are present in the endosperm. (lt
should be recalled that the female parent
contributes two nuclei to the triploid endo-
sperm tissue, and the male parent onc.)
Those developing from megaspores with
two Ae loci, either linked or situated in
different chromosomes, would give rise to
endosperms with four instead of two Ac
loci. It is known that increases in the dose
of Ac will delay the time of appearance of
Ds mutations, and that this effect is pro-
portional to dosage—the higher the dose,
the more effective the delay. With four
doses of Ae instead of the usual two, the
delay may be so effective that either no Ds
mutations will occur during the develop-
ment of the tissue or only a few will occur
very late in the development of the endo-
sperm. In either case, the kernels having
such increased doses of Ac will be strik-
ingly different in appearance from the
majority of kernels, that is, those with two
Ac loci in their endosperm cells. It was
this striking difference in appearance of a
few kernels on these ears that allowed the
selection to be made. The analysis of the
Ac composition of the kernels has led to
the conclusion that they develop from an-
cestor cells in which a transposition of Ac
has occurred.

For comparison, plants were grown from
some of the kernels on these F2 ears that
DEPARTMENT OF GENETICS

showed the characteristic type of Ds muta-
tional response known to be associated
with the presence in the endosperm of two
Ac loci. Tests of the Ac constitution of
these plants gave the expected results. One
Ac locus was present in each of the tested
plants, and its control of the time and
frequency of Ds mutations, in one or two
doses, was similar to that in the parent
plant.

These studies have been expanded dur-
ing the summer of 1949; but the results of
the preliminary tests are sufficient to in-
dicate the factors responsible for apparent
exceptions to the expected Mendelian in-
heritance of Ac. They have also made
possible an interpretation of one of the
several kinds of event that occur during
the development of the plant or of the
endosperm to bring about pronounced
changes in the action of Ac on Ds. These
changes are registered by the appearance
of precise sectors showing altered Ds re-
sponses. Tests are now being conducted
to distinguish between changes in state of
the Ac locus—that is, between changes in
quantitative action of an Ac locus that is
inherited as a single unit, and changes
that are caused by an increase in numbers
of such loci after transposition of Ac, as
outlined above. The phenotypic effects of
these two types of change overlap, but the
causative series of events, although related,
are nevertheless separable.

The mechanism responsible for trans-
position of the Ac locus has not been
analyzed. It is thought likely to be the
same as or similar to that producing trans-
positions of Ds. If so, some of the trans-
positions of Ac should be associated with
chromosomal rearrangements. A chromo-
somal translocation was recognized in one
of the cases cited, but it has not yet received
adequate analysis.

149

Tue Action oF de on THE Mutaste Loci
Ir ConTROLS

It has been emphasized repeatedly that
Ac controls the occurrence of Ds muta-
tions and that its quantitative levels con-
trol the time and frequency of these muta-
tions. In this report; it has been shown
that the mutable c™* locus is merely a
transposed Ds locus situated at or close
to the C locus. The analysis of this c™™
locus and of its origin from a transposition
of Ds has suggested that all Ac-controlled
mutable loci arise from transpositions in-
volving, originally, only one Ds locus. Ac-
cording to this interpretation, Ac does not
control the mutability of many different
loci, but only the mutability of a single
locus—the Ds locus—wherever it may be
situated in the chromosomal complement.
Mutations of Ds in these various positions
may result in changes in phenotypic ex-
pression that are strikingly different. The
change in phenotype, in any one case,
depends on the kind of locus that has been
inhibited by the insertion of Ds. In their
normal action, these various Ds-inhibited
loci must control quite different chemical
processes. The events at the Ds locus that
result in a return to partial or complete
action of the inhibited locus must therefore
involve a different series of changes in
chemical processes in each case. Without
an integrative understanding of the events
that occur at such mutable loci, it would
be difficult to understand why de should
control the mutability of loci concerned
with such unrelated processes, and why
each such locus should respond to a par-
ticular Ac locus and dosage in an exactly
comparable manner. There is no difficulty,
on the basis of the given interpretation, in
appreciating the apparent nonselectivity of
control of mutable loci by Ac and the
similarity in response of these mutable loci
to changes in Ac state and dosage.
150

In order to obtain more specific informa-
tion about the nature of the action of Ac
(other than its known effects in producing
chromatid breaks at the Ds locus and con-
trolling the time and frequency of these
breaks), combinations of Ds loci at various
positions in the short arm of chromosome
g have been made. These combinations
were made in an attempt to answer the
following question: Does Ac produce a
cellular or nuclear condition in a certain
cell, at a certain time in development, to
which all Ds loci will respond? An in-
structive example for this purpose is a
combination of ¢”? (Ds at or close to the
C locus) with Ds at its standard location.
If a plant carrying c”* and wx in its
chromosomes 9 is crossed by a plant carry-
ing ¢* (stable c, nonmutable with 4c),
Wx, and Ds (standard location, to the
right of Wx), kernels will be produced
that are 6” 7 wx/c"" wx/ci Wx Ds. This
combination should show whether or not
mutations in the several Ds loci will occur
at the same time in the same cell, and
whether this response will be of the same
order with one and with more doses of
Ac. Simultaneous mutations would be re-
vealed in these kernels provided an ex-
treme state II Ds locus were present at
ec" (mutation from ¢ to C and few if
any dicentric chromatid formations), and
an extreme state I Ds locus were present in
the c>’ Wx Ds chromosome (high rate of
dicentric chromatid formation). If all Ds
loci respond to some particular develop-
mental change that is brought into being
by the presence of Ac, then when this
changed condition arises in a cell, a muta-
tion of Ds at the ce”? locus should give a
C phenotype in the descendent cells. A
mutation at the Ds locus in the c* Wx Ds
chromosome should also occur. A wx
phenotype would then appear in the de-
scendent cells, because a Ds mutation in
the c° Wx Ds would produce a dicentric

CARNEGIE INSTITUTION OF WASHINGTON

chromatid and a U-shaped acentric frag-
ment; this acentric fragment would carry
the Wx locus, and consequently Wx would
be lost from the nuclei during a mitosis.
The effects produced by such simultaneous
mutations of the several Ds loci should be
visible in the mature kernel. Colored areas
(the ¢ to C mutations) should appear, and
the underlying starch should be wx. Also,
the borders of the sectors having both of
these altered phenotypes should correspond
exactly. In the examined kernels having
these given constitutions, a high per-
centage of the C areas had underlying wx
starch, and the borders of the sectors did
exactly correspond. Exceptions were ex-
pected, and a number were observed.
Some examples were: C areas with under-
lying Wx starch, wx areas with overlying
colorless aleurone, C areas with only half
of the underlying sector composed of wx
starch, or wx areas with only half of the
overlying aleurone layer showing a C
phenotype. It is hoped that an extended
analysis of the various classes of excep-
tional areas in these kernels will reveal the
more unusual consequences of the events
that occur at the Ds loci in these mutation-
producing mitoses, and the resultant or-
ganization in the two affected sister
chromatids.

Tests have also been constructed to de-
termine the relation between the mutations
of Ac and those of Ds. Although the anal-
yses of these tests are incomplete, it seems
apparent that Ac tends to mutate in the
same cell in which a Ds mutation is 0¢-
curring, or in an immediate ancestor cell.
The combined evidence suggests that some
condition, under the control of the de
locus and depending on its state and dos-
age, must develop in specific cells at specific
times, to produce a mutational response
(chromatid breaks) at Ds loci as well as
at the Ac locus itself. The consequences 0!
such mutation are the observed changes 10
DEPARTMENT OF GENETICS

genic action, transpositions or losses of
Ds or Ae, and production of gross chromo-
somal rearrangements with or without ac-
companying transpositions of Ds or Ac.

Murasie Loci ¢"? anp wx?

The Ac-controlled mutable loci ec”? and
wx were described in Year Book No. 47.
A few salient facts and conclusions based
on the continued study of these loci are
as follows: Both loci express their muta-
tions quantitatively. A series of alleles de-
rived from such mutations, which show
gradations of quantitative expression, have
been selected for study. When de is ab-
sent, a particular expression of an allele
can be held constant, for no somatic muta-

tions of these alleles occur. When Ac is ©

present, the alleles may continue to mutate
to either higher or lower levels of quanti-
tative expression. For a study of the action
of any one allele, therefore, it is important
that no Ac locus be present.

It has been determined that chromatid
breaks may occur at these two mutable
loci; in this respect, they are similar to
c™ 1, Both ce? and wx™* were isolated
from stocks known to have a Ds and an
Ac locus. Unlike c”*, they were not de-
tected at the time of their origin. It is
therefore impossible to reconstruct the par-
ticular events associated with their origin
from a normal C locus and a normal Wx
locus. The presence of Ds-type behavior
at these mutable loci points to a mechanism
similar to the one associated with the
origin of ¢™*.

In the case of c™*, the position of in-
sertion of Ds into or adjacent to the C
locus may differ from its position of in-
sertion in c™7; for two qualitatively dif-
ferent types of phenotypic expression of
the C locus result from mutations of ¢””*,
whereas only one type regularly follows
mutations of ¢"*. Both types of qualita-

151

tively distinguishable mutations at ¢™”*
result in pigment formation in the aleurone
layer. Within each of the two qualitative
types there occurs a series of mutants
showing various degrees of quantitative
expression. The color intensities produced
by the different mutants of both types
range from a faint pink to a deep red (in
pr pr constitutions). The two series of
mutants are distinguished from each other
mainly by the fact that a different dif-
fusible substance (or substances) is pro-
duced by the members of each. Both sub-
stances are concerned with pigment forma-
tion. The diffusible substance produced
by type 1 mutants may be utilized by a
cell having a normal C locus, or by a cell
having a type 2 mutant, to intensify the
color of the cell pigment. The normal C
locus and the type 2 mutants, on the other
hand, both produce a diffusible substance
that can be used by type 1 mutants to in-
tensify pigment color. Thus, the type 2
mutants and the normal C locus are much
alike; they both produce a diffusible sub-
stance that type 1 mutants can use, and
they both can use a diffusible substance
produced by type 1. This relationship sug-
gests that a normal C locus is probably
responsible for the production of at least
two diffusible substances, both of which
are required for pigment formation. It also
suggests that the dosage responses noted
for the normal C locus may be the con-
sequence of a limited production of one of
these substances by a single C locus: the
more C loci were present, the more of this
substance would be produced and the
deeper would be the pigment color. The
quantitative grades of expression of the
alleles within the two types of mutations
arising from c”? may reflect the relative
quantities of the two substances produced
by individual members of a type—limita-
tions in the production of one of these
substances conditioning the amount of pig-
152

ment that can be formed, and thus the
depth of color that can appear.

The conclusions derived from study of
c™* regarding the action of the normal C
locus are noteworthy, in that they consider
a double function of a single unit in in-
heritance. This unit, concerned with pig-
ment production in the aleurone layer of
the endosperm, appears to be composed of
at least two qualitatively different subunits,
both of which determine the production
of substances required for pigment forma-
tion. It is possible that this C locus behaves
as a unit in inheritance not only because
all the subunits are needed for the produc-
tion of pigment, but also because a par-
ticular spatial relation of the units at the
locus is required to assure a definite se-
quence of reactions.

Mutations of the wx” locus have been
similarly instructive in considering the ac-
tion of the normal Wx locus, but for
reasons other than those just discussed for
c™*. Here, alleles showing various quanti-
tative levels of expression are produced by
mutations of wx"". The levels are ex-
pressed by the percentage of amylose in
the starch component of the endosperm
cells. When only the recessive, wx, is
present, no recognizable amylose starch is
produced. The selected alleles derived
from mutations of wx” form a series in
which a single dose (Wx allele, wx, wx)
produces quantities of amylose ranging
from very little (less than 1 per cent) to
as much as the normal Wx locus produces
in three doses, Chemical analyses of the
percentages of amylose starch produced by
several of these alleles have been conducted
by Miss Ruth Sager and Dr. Charles O.
Beckmann, of Columbia University. These
analyses have shown that the type of color
reaction produced by staining with iodine
is a relatively reliable indication of the
approximate percentage of amylose present.

Interest in this case centers not so much

CARNEGIE INSTITUTION OF WASHINGTON

in the appearance of alleles having lower
activity than the normal Wx locus as in
those having higher activity than the
normal locus. Is the normal Wx locus
partially inhibited, or do the Wx alleles
showing greater than normal activity arise
from duplications of the locus? In Ac-
carrying plants, the chromatid-breakage-
and-fusion mechanism associated with mu-
tations at the wx™* locus or its inter-
mediate alleles should give rise, in some
cases, to duplications or multiplications of
units of the Wx locus. It is hoped that a
study of the different amounts of amylose
produced by sister chromatids after muta-
tion of wx", or one of the intermediate
alleles, will furnish some information with
reference to this question.

CoNcCLUSIONS

The purpose of the foregoing sections
has been to indicate the progress made dur-
ing the past year in attacking fundamental
aspects of the origin and behavior of Ac-
controlled mutable loci. It was concluded
that only two loci are involved in all these
cases: the Ds locus and the Ac locus. The
origin and subsequent behavior of newly
arising mutable loci depends on the trans-
position of a Ds locus and its insertion into
(or adjacent to) a normal locus, and on the
constitution of this inserted Ds locus. The
genic action of a normal locus may be
inhibited by such an insertion. Subsequent
events at this new position may remove
the inserted segment and its inhibitors
action altogether; or changes in the const
tution or position of the Ds locus may I
sult in changes in the degree of inhibition
of the affected locus. It was also concluded
that the events occurring at Ds during 4
mutation-producing mitotic cycle result in
compound chromatid breaks at this locus.
and that the observed consequences depend
on subsequent fusions of the broken ends.
DEPARTMENT OF GENETICS

The fusion phenomenon, of utmost im-
portance in these cases, calls for no new
interpretations, since the fusion of newly
broken (unsaturated) chromosome ends
has been well investigated and could be
anticipated.

Both Ac and Ds are mutable loci, for
their mode of action changes as the con-
sequence of events occurring at these loci
in certain cells of the plant. Like Ds, de
also undergoes transposition from one loca-
tion in the chromosomal complement to
another. The mechanism of transposition,
although not directly analyzed, is possibly
similar to that associated with the trans-
positions of Ds. The evidence also indi-
cates that changes in de as well as Ds are
associated with chromatid breakage and
fusion. It is necessary to determine, then,
the nature of the events occurring at these
two mutable loci, during a particular
mitotic cycle, that will result in the ob-
served breakage-and-fusion phenomena.
Unquestionably, these events are primarily
responsible for all the observed changes

at these mutable loci. It is suspected that .

they are associated with some aberration in
the mode of reproduction of a particular
type of molecule in the chromosome dur-
ing a mutation-producing mitotic cycle.
Both Ac and Ds are assumed to have such
molecules. If the aberration involves a
chemical bonding of the newly formed
molecule with the original molecule, which
holds at least until after the forced sepa-
ration of sister chromatids during the
prophase period, a rupture of the chro-
matid could occur at the affected- locus
during this separation period. It is known
that the bonds holding the molecules to-
gether in a linear order in the chromosome
may be ruptured by mechanical pull, and
that the broken ends so produced are un-
saturated and capable of fusion with other
unsaturated broken ends. It is therefore
necessary to assume, in this interpretation,

153

that the bond connecting the newly formed
molecule with the original molecule is
stronger than the bond holding the mole-
cules together in linear order.

The study of transpositions of the Ds
locus has shown that a rupturing mecha-
nism of this type, or at least one that leads
to similar consequences, must be involved.
It has been established that the transposi-
tion phenomenon is associated with chro-
matid breakage; the Ds locus is inserted
into a position where a spontaneous break
has occurred. The transposition phenome-
non is readily explained if it is assumed
that break-producing events at the Ds
locus may sometimes result in the tearing-
out of a minute fragment containing Ds
and having two unsaturated broken ends.
The insertion of Ds into a new position
would result merely from fusion of un-
saturated ends. If the broken ends arising
from the spontaneous break are labeled 1
and 2 and those of the fragment 3 and 4,
the fusion of 1 with 3 and 2 with 4 would
accomplish the transposition. The above-
described process of mechanical rupture of
the chromatid at the Ds position could re-
sult in just such a torn-out fragment. The
consequence of any one rupture would de-
pend on the type of fusion of broken ends
that followed. Not only could transposi-
tions occur, but the Ds locus could be lost
altogether, or two Ds loci could enter one
chromatid, leaving none in the sister chro-
matid. Such duplications (altered states
of Ds) could, in turn, initiate a series of
new consequences when the aberrant type
of event, leading to chromatid rupture,
again occurred in a descendant of this
chromatid.

The analysis discussed in this report of
the factors associated with the origin and
behavior of Ac-controlled mutable loci in
maize has led to a relatively simple inter-
pretation of the nature of the events re-
sponsible for changes in action of the genes
154

involved. The types of phenotypic change
that follow mutations of non-dc-controlled
mutable loci are similar to those shown by
the Ac-controlled mutable loci. It is quite
possible that the same or similar events
are primarily responsible for these changed
phenotypes also.

Mutable loci have been described in a
number of organisms. Many of them
show changes in phenotypic expression
similar to those now being observed in
maize. The events responsible for changes
in expression of genic action may be simi-

CARNEGIE INSTITUTION OF WASHINGTON

lar in these organisms to those occurring
in maize. The investigations described in
this report cast doubt on interpretations
that postulate a “true gene mutation,” that
is, a chemical change in a gene molecule,
resulting in a changed specificity of its
active product. Phenotypic change may
well be related to inhibition of the action
of a normal gene followed by partial or
total release of this inhibition, together
with such duplications or deficiencies of the
locus as could be produced by the mecha-
nism outlined above.