COMPONENTS OF ACTION OF THE REGULATORS Spm AND Ae
Barbara McClintock

Gene control systems other than those
associated with the regulators Spm
(Suppressor-mutator), Ac (Activator),
and Di (Dotted) are known to be present

in maize. Just as proved true of the Ac
and Spm systems, aspects of which still
need to be elucidated, many types of test
extending over a number of years will be
528

required to identify these additional
systems, their components, and their
modes of operation. Because time will not
be available for such extended studies, it
was decided to concentrate attention
during the past year on some of the un-
resolved aspects of the Ac and Spm
systems.

The Suppressor-mutator control system
is so called because its regulator element,
also designated Spm, has two components
of action. The suppressor (or inhibitor)
component, component-1, directly regu-
lates the expression of a gene that has
come under the control of the Spm
system. Such control, it will be re-
membered, arises through insertion of the
operator element of the system at the
locus of a gene. When component-1 of
Spm is in an active phase, the expression
of the gene is suppressed; when it is
inactive, gene action is expressed. One
exception to this general rule concerns the
modified gene locus ai”°*. This gene is
active when component-1 is active, and
suppressed when it is inactive. Com-
ponent-2 of Spm is the mutator or trans-
position-inducing component. The re-
sponse of the operator element to
component-2 often gives rise to a mutant
expression of the gene that releases it
from control by the Spm system. Some
responses, on the other hand, effect other
modifications that are not associated
with release but instead alter the sub-
sequent types of response of the operator
element to the components of Spm.
These modifications have been called
“changes in state’ of the gene locus.
With some altered states, the operator
element loses its capacity to respond to
component-2, although it retains its abil-
ity to respond to component-1. No further
mutations occur, nor is the gene released
from control by the system. Its action
remains permanently under the control
of component-1 of the Spm regulator
element.

Both components of Spm may undergo
change. Component-1 exhibits alternat-
ing cycles of activity and inactivity,

CARNEGIE INSTITUTION

whose regulation has been outlined in
previous reports. Changes of component-2
resemble mutations in that they arise
from single events, each of which alters
the effectiveness of this component, both
for inducing responses of the operator
element that lead to changes in gene
expression, and for inducing  trans-
positions of Spm. Such alterations in
action of component-2 have been detected
only in those cells of the plant in which
component-1 is active. Some alterations
eliminate all activity of component-2
whereas others effect altered times and
frequencies of occurrence of mutation-
inducing responses of the operator ele-
ment. Each mutant of component-2 may
undergo still further mutation, the fre-
quency of occurrence differing with
different mutants.

The components of Spm and _ their
characteristic modes of action were
detected originally in observations of the
activity of one gene that is under the
control of the Spm system. Although it is
possible to distinguish a change in action
of either component through such obser-
vations, distinctions are greatly facili-
tated when two or more genes, each under
the control of the system, are present in
a plant or kernel. The operator elements
at the different gene loci respond in like
manner to modifications affecting either
of the components. Through such combi-
nations it has been possible to determine
with a considerable degree of accuracy the
types of modification undergone by each
of the components of Spm.

Although direct evidence is lacking, the
mechanism that releases a gene from the
control of the Spm system is believed to
be transposition of the operator element
away from the gene locus. This as-
sumption is based on studies of the Ac
system, which have provided direct
evidence of release of a gene from its
control through transposition of the
operator element away from the gene
locus. Transpositions of the regulator
element Spm, on the other hand, may be
readily detected, and indirect evidence
GENETICS RESEARCH UNIT

had suggested that component-2 regulates
them. Tests designed to obtain direct
evidence were completed during the year,
and are described below.

The original state of a,"-? has at the
a," locus an Spm element with a highly
active component-2. This Spm is des-
ignated Spm*. It induces many germinal
mutations at the ai"? locus, and the
events responsible for them are usually
associated with transposition of the Spm
away from the gene locus. Often the event
is accompanied by release of the gene
from control by the Spm system. The
germinal mutants that have been released
from Spm control may be placed in two
categories: those that resemble the wild-
type gene in their action, and those that
give rise to the diffuse-mottled phenotype,
described in Year Book 61 (pp. 448-460).
In plants carrying the original state of
a,™-?, component-2 of Spm may undergo a
mutation that alters its capacity to induce
mutations at a,"? and to induce similar
responses of the operator element at other
gene loci. One of these mutations alters
component-2 in such a manner that it
induces change in gene action only very
late in the development of a tissue, and in
only a few cells. This mutant is designated
Spm”, In plants carrying Spm” at the
ai”? locus, no germinal mutants are
produced. Also, this Spm is not removed
from the a,”°? locus in those cells of the
plant that contribute to formation of the
gametes. If component-2 back-mutates
to a high level of activity, germinal
mutants appear and Spm in many in-
stances is transposed away from the gene
locus. One Spm” isolate, however, proved
to be very stable; back-mutation of
component-2 was rare. This isolate
was selected for the tests to be de-
scribed.

It was suspected that this Spm” would
undergo early transpositions if an Spm
with a highly active component-2 was
also present in the nucleus. To obtain
. direct evidence and to determine whether
or not the components of Spm” would be
maintained unaltered after such a trans-

529

position, testcrosses were conducted with
plants that commenced development with
the selected Spm” at the locus of ay"? in
chromosome 3 and an Spm: located close
to the pr marker in chromosome 5. The
constitution of the plants was Spm»
a1”? Sho/a; she; Pr/pr Spms; wz/wea. The
presence of Spm induced some germinal
mutations at the a"? locus, most of them
giving rise to the diffuse-mottled pheno-
type in the kernel. Kernels with this
phenotype were detected on ears pro-
duced on the plants by a cross with plants
that were homozygous for a1, she, and pr
and had no active Spm. (In general the
kernels on these ears were similar to those
illustrated in Year Book 63, plate 1B,
following page 601.)

Plants were grown from 30 of the
kernels exhibiting a  diffuse-mottled
phenotype; 27 of the selected kernels had
purple pigment and thus had received the
Pr marker from the heterozygous parent,
and three were red, having received the
pr marker from this parent. Five of the
plants derived from the Pr kernels
contained Spm” but no Spm‘, as shown by
the response given to Spm” by the gene
wa" 8 which had been introduced by the
pollen parent in some of the crosses.
All 30 plants were tested for the presence
of Spm, its type, and its location with
reference to the genetic markers carried
in the plants.

In all, 78 fertile ears were produced by
the 30 plants. The pollen parents in the
crosses were of several types: homozygous
for ay"!, she, pr, and wz™ 8 and having no
active Spm; homozygous for a1, she, pr,
and wz™* and having no active Spm; or
homozygous for ai}, shy, and wz,
having no Spm, and either homozygous or
heterozygous for the Pr marker. One
particular state of a:”1 was utilized in
these crosses because it responds in a very
clear way to activity of the components
of Spm and also allows the type of Spm in
a plant to be registered in both the She
and the shz class of kernels on the ears of
the plant. The presence of wr8 in a
kernel serves as an additional means of
530

scoring the Spm elements it may contain.

The tests revealed the following.
Among the five plants derived from Pr
kernels that were known to contain Spm”,
two had one Spm” element, carried in
chromosome 3 and linked with the diffuse-
mottled locus but removed from it. Two
others had one Spm”, not linked with the
diffuse-mottled locus. The fifth plant
had two Spm elements, neither of which
was linked with that locus. Of the remain-
ing 22 plants derived from Pr kernels, 12
had no Spm, 4 had one Spm” linked to
the diffuse-mottled locus but removed
from it, 3 had one Spm” not linked
with that locus, and 2 had one Spm*
closely linked with the Pr marker. One of
these two plants had in addition an
Spm”, not linked with the diffuse-mottled
locus. The remaining plant of the 22 had
one Spm* not linked with Pr or with the
diffuse-mottled locus. Of the three plants
derived from the pr kernels, one had one
Spm?, one had two Spm’, and one had one
Spm? and also one Spm“, not linked with
the diffuse-mottled locus.

The evidence obtained from the tests
indicates that an Spm” element that is
unable to induce its own transpositions
early in development will undergo such
transpositions if a potent component-2 is
supplied by an Spm located elsewhere in
the chromosome complement. It also
indicates that the transposition event
does not modify the components of the
Spm*; they remain unaltered after the
event.

The Component of Spm Responsible for
Preset Patterns of Gene Expression

To identify the component of Spm that
is responsible for inducing the preset
patterns of gene expression discussed in
Year Book 63 (pp. 594-599), plants carry-
ing state 7977B of ai"? and no active
Spm were crossed with plants that had
Spm” at the locus of a1”, or Spm” at the
locus of a8, or Spm” linked with Pr in
chromosome 5. Other crosses were made
that introduced into some of the kernels

CARNEGIE INSTITUTION

an Spm with a highly active component-2.
From the ears produced by these crosses,
kernels were selected that had and that
had not received the Spm. Plants were
grown from both types of kernels, and
testcrosses were conducted with their
ears to determine whether or not they
would bear kernels exhibiting types of
anthocyanin distribution similar to those
illustrated in Year Book 63, Plate 2B.
Sixty-one plants derived from kernels
that had received the introduced Spm
were tested. Since most of them produced
two to four fertile ears, different types of
tests could be conducted with many of
them. At least one ear, and usually more
than one, took part in a cross with a plant
that was homozygous for aj, she, and
wa and had no active Spm. On all the
ears there were kernels that exhibited
preset patterns of anthocyanin distribu-
tion: and showed no indication of the
presence of an active Spm. There was no
evidence that would relate the type of
preset pattern to the particular Spm
present in the ear-bearing parent: Spm*
and Spm” were equally effective in this
regard. This observation suggested that
component-1 of Spm is responsible for
induction of preset patterns. Confirma-
tion was provided by the kernel types on
ears of four additional plants, which
carried an Spm whose component-1
remained in an inactive phase throughout
plant development but returned to an
active phase in many cells during develop-
ment of the kernels. No kernels with
preset patterns appeared on any of the
ears produced by these four plants.
Sixty-one plants derived from kernels
that had not received Spm from the
pollen parent were also examined. On 68
of the ears on these plants, produced
either by self-pollination or by crosses
with plants that were homozygous for a1
and had no active Spm, kernels exhibiting
preset patterns did not appear. The
kernels were colorless or nearly so. In
addition, 49 ears produced by 36 of the
plants were utilized in a cross that
introduced an active Spm’. No kernels
GENETICS RESEARCH UNIT

with preset patterns appeared on these
ears. The phenotype expressed by ay"?
(state 7977B) in those kernels that
received Spm was typical: deep-pig-
mented spots in a lighter-pigmented
background. Kernels that did not receive

Spm from the pollen parent were colorless
or nearly so.

Transmission of the Preset Pattern

In the study of preset patterns de-
scribed in Year Book 63, pp. 594-599, it
was found that the patterns usually did
not reappear in the following generation.
Most of the ears of plants derived from
kernels exhibiting preset patterns showed
no evidence of retention of the patterns:
their kernels were colorless or nearly
colorless. On a few ears, however,
produced by plants carrying state 7995 of
a,™"*, several kernels exhibited a pattern
of anthocyanin distribution resembling
that in the kernel from which the plant
was derived. Five such exceptional ker-
nels were present on one ear, four on
another, and two on a third. This year,
plants were grown from the exceptional
kernels to determine whether or not the
pattern would again reappear.

The constitution of the plants derived
from these kernels was a1”? (state 7995)
Sh2/a1 she, wx/wa, and no active Spm was
present in any of them. Nineteen of the
fertile ears on these plants were utilized in
a cross with a plant that was homozygous
for a1, she, and wz or wx™8 and had no
active Spm. If more than two ears were
produced by a plant, one of them was
used in a cross with a plant that was
homozygous for a; and shz and had one
or more Spm elements. This cross,
conducted with one ear on each of five of
the eleven plants, was made to test
whether some modification had occurred
at the ai”? locus that would be revealed
by the locus’s response to an active Spm
element. On all five ears the response to
Spm was normal: deep-pigmented spots
appeared in a lighter-pigmented back-
ground.

531

The 19 ears produced by the first-
mentioned cross were examined for
kernels with patterns of anthocyanin
distribution and intensity resembling
those in the parent and grandparent
kernels. On the ears produced by plants
derived from the parent ear that had four
such kernels, no kernels of this type
appeared. All were colorless or nearly so.
On one or more of the ears produced by
the remaining seven plants, some parent-
type kernels did appear, in numbers
ranging from one to five per ear. Their
distribution on an ear was not random:
most were located in the upper third,
several at the base of the ear. When more
than one was present on an ear, they were
not clustered, exhibiting in that respect
the same distribution as on the parent
ears. There is no evidence that con-
tamination contributed to their presence
on the progeny ears, as all the remaining
kernels were colorless or nearly so.

The exceptional kernels had pigment
intensities resembling those of the kernels
from which their respective plants arose.
They did not show the wide range of
intensities represented among kernels
with preset patterns appearing on ears of
plants that have an active Spm. It seems
probable, therefore, that the condition
responsible for a particular expression of a
preset pattern was retained for two plant
generations in the ancestor cells that
produced the exceptional kernels, but was
lost from those ancestor cells that
produced the remaining kernels on the
same ears. At present there is no adequate
explanation for this phenomenon. Some of
its aspects recall a type of gene expression
that is produced by one of the operator
elements of the Ac system. This
operator functions at the loci of c,™-2,
we, and wx, It was noted that some
responses of the operator to Ac did not
immediately result in a stable expression
of the gene; instead, the level of expression
appeared to oscillate. This behavior was
observed in the descendent cells of a cell
in which such a response had occurred.
The results were apparent in kernels that
5382

had received one such ‘excited” gene
locus from the pollen parent. Plants were
grown from some kernels that exhibited
this type of metastability at the wae
locus. In kernels on the ears of the plants
the level of action of the Wz gene was now
uniformly expressed throughout the cells
of the endosperm. The “oscillation” had
ceased. Furthermore, no responses of the
mutant locus to Ac occurred thereafter.
The locus had acquired stability.

Components of Action of Ac

Study of wa’, first reported in Year
Book 63 (pp. 599-601), was extended in
order to examine the activity cycles of
Ac in greater detail. The wx? modi-
fication arose by insertion of Ac at the
Wz locus in chromosome 9. The initial
effect was a marked reduction in activity
of the gene. Transposition of Ac away
from the locus restores a high level of gene
action and releases the gene from control
by the Ac system. Ac is known to
regulate the time and frequency of occur-
rence of self-transposition and of re-
sponses of the operator element at other
gene loci. Initially this Ac at the wx?
locus exhibited the type of dose expression
that characterizes Ac: the higher the
dose, the later the time of occurrence of
such responses. It was observed that this
Ac undergoes cycles of activity that alter
the component responsible for the re-
sponses and for induction of transposi-
tions. When that component is inactive,
no such responses occur nor does the Ac
contribute to dose expressions should an
active Ac also be present elsewhere in the
chromosome complement. It responds,
however, to such an active Ac by under-
going transposition that releases the Wa
gene from further control by the Ac
system, and the time and frequency of
occurrence of transposition reside in the
active Ac. When the Ac at the Wz locus
returns to an active phase, its capacity to
induce responses of the operator element
located elsewhere, to induce its own
transpositions, and to contribute to

CARNEGIE INSTITUTION

dose expressions is restored. Thus, the
activity cycles affect a component of Ac
that is comparable to component-2 of
Spm, but different in that it is also
responsible for dose expressions, which
are not exhibited by component-2 of Spm.
No component of Ac comparable to
component-1 of Spm has yet been
identified.

A fascinating aspect of inactive Ac at
wz™7 relates to its control of the level of
action of the Wx gene in the starch-bear-
ing cells of the endosperm. Different
levels of action are induced during kernel
development. To examine this aspect it
was necessary to utilize a technique that
could reveal these levels in individual
cells. The Wz gene is responsible for the
production of amylose starch in the
pollen grain, the embryo sac, and the
starch-bearing cells of the endosperm. of
the kernel. The associated enzyme, which
has. been identified by O. E. Nelson, is
affixed to the gtarch-forming granules
within the cells. When the Wz gene is
acting normally, approximately 25 per
cent of the starch in the endosperm is
amylose, the remainder being amylopec-
tin. If the activity of the gene is reduced,
the amount of amylose formed is also
reduced. If the gene is totally inactive,
all the starch is amylopectin. The two
types of starch stain differentially with a
solution of potassium iodide and iodine:
the amylose stains blue and the amylo-
pectin red-brown. The red-brown stain
may be removed, either by hot water or
by exposure of the cells to the rays of a
lamp. To examine the level of Wa gene
action in different cells of the endosperm
of a kernel, a cut is made to expose
surface of endosperm cells, which are
then stained with an I-KI solution. The
intensity of blue coloring in the starch
granules of different cells may be com-
pared. It was learned many years ag0
that the intensity of this staining in the
granules of a cell reveals the level of Wx
gene action in that cell. It is believed,
therefore, that the observed differences 10!
intensity of blue coloring among the
GENETICS RESEARCH UNIT

endosperm cells of an individual kernel
reflect differences in level of JVx activity
in the cells. Such differences are
illustrated in Plates 1 and 2.

To aid in interpreting the illustrations,
a brief review should be made of endo-
sperm development, some aspects of
which were first revealed during the
course of this study. The primary
endosperm nucleus, produced by fusion of
a sperm nucleus with two haploid nuclei
contributed by the female gametophyte,
divides in two, and then each nucleus
divides again. Each of the resulting four
nuclei gives rise by subsequent divisions
to a column of nuclei, before cell walls are
formed. The columns are arranged around
a central, nonnucleated core. Cell-wall
formation occurs later, but does not
eliminate the central core, often visible in
the mature kernel. The cells in the
columns divide tangentially, and the
outermost cells continue to divide, leaving
behind cells in which endoreduplication of
the chromosomes occurs. In the mature
kernel the innermost cells of the endo-
sperm are highly polyploid and very
large, containing many starch granules.
The degree of polyploidy becomes lower in
cells farther removed from the middle ‘of
the kernel, and the cell size is cor-
respondingly reduced. The chromosomes
in the nuclei of cells toward the periphery
do not undergo endoreduplication, and
these cells are small. The aleurone layer
is the outermost layer of cells of the
endosperm; only in these cells is antho-
eyanin pigment produced. Although a
change in control of action of a gene
contributing to anthocyanin pigment
formation may occur within a cell during
endosperm development, it can be ex-
pressed only in cells of the aleurone
layer that are descended from that cell.

To examine the effects produced by Ac
in its active and inactive phases, it is
necessary to know when it is in one or
the other. This is made possible by the
presence of a gene that is associated with
anthocyanin pigment formation and is
under the control of the Ac system.

533

The modified A, locus ai”? was chosen
for the purpose. The state of a,™3
selected for the experiments produces a
lightly pigmented aleurone layer when
Ac is absent or inactive. When Ac is
active, the responses it induces in the
operator element at the ai” locus give
rise to altered A; gene expressions, often
restoring full or nearly full activity to the
gene. The time of occurrence of these
responses is controlled by the Ac element
that is present in the kernel. An illustra-
tion is given in Plate 14. The responses
to the active Ac in this kernel were
registered in like manner by a,7-? and
we; they occurred late in development
of the kernel.

Kernels that commence development
with a,” and an inactive Ac at the locus
of wx? will show no deeply pigmented
areas in the aleurone layer at maturity if
the Ac remains inactive throughout
endosperm development. If Ac returns to
the active phase in an individual cell
during development, it will induce re-
sponses of the operator element leading to
change in gene action at a,"-* in some of
the progeny of that cell. The event will be
evidenced by an area in the aleurone
layer that exhibits pigmented spots. Ac’s
return to activity will also induce many
reversions of the Wz gene to full expres-
sion. The size of an area having pigmented
spots indicates the time during develop-
ment when the change of phase occurred.

Progeny were obtained from a number
of plants that developed from kernels
having a," and an inactive Ac, for
examination of the cycles of activity
that Ac would subsequently undergo.
Only that aspect of the study relating to
control of expression of the Wx gene will
be reported here. In a number of progeny
kernels, the action of a.”* gave no
indication of a change of phase of Ac
from inactivity to activity. The level of
Wz gene expression in the starch-bearing
cells of these kernels was not uniform, as
shown in Plate 1(C, D, and #). Al-
though most such kernels exhibit a basic
pattern of Wx gene action, produced by
534

a low level of this action in the cells of the
upper mid-region and at the base of the
endosperm, many changes in level may
occur during endosperm development. It
was noted that the initial level of Wa
gene expression imposed by the inactive
Ac at the time of fertilization, regardless
of whether it was contributed by the male
or by the female gametophyte, had a
marked effect on the levels of Wx gene
expression appearing in the kernel. Some
kernels commenced development with a
low level of Wz gene action (Plate 1C),
others with a much higher level (Plate
1D and E). If such kernels had started
development with one active Ac, located
elsewhere than at w27, both a1? and
wz™7 would have responded to it. Early
and late changes in Ai gene expression
and changes to full Wx gene expression
would have occurred, in the manner
exhibited by the kernel in Plate 17. In
that kernel the pattern of Wx gene
expression produced by the presence of
active Ac is superimposed on the pattern
produced by inactive Ac.

The kernel shown in Plate 1B and all
three kernels shown in Plate 2 illustrate
the responses of ai”? and wz™? to a
change in phase of Ac from inactive to
active during endosperm development.
The legends give the constitutions of the
kernels and describe the effects produced
by the changes in phase. It can be noted
particularly in Plate 1B and Plate 2B
and C that the progeny cells of a cell
in which Ac underwent activation are

CARNEGIE INSTITUTION

readily distinguished because they express
either a low or a high level of Wx gene
action. In cells where Ac remained in an
inactive phase throughout endosperm
development, on the other hand, a wide
range of levels may be expressed, re-
sembling the range exhibited by kernels
in which Ac remains inactive throughout
development.

In conclusion, it may again be stated
that the resemblance between the regula-
tors Ac and Spm resides in a component
of each that initiates responses of the
respective operator element and of the
regulator itself which effect transposi-
tions. If the component is inactive, such
responses do not occur. It is this com-
ponent of Ac that is comparable to
component-2 of Spm. In Ac, but not in
Spm, the component is also responsible
for dose effects. The operator element of
the Ac system has not yet given evidence
of differential control of action of the gene
in response to activity cycles of Ac, in the
manner exhibited by the operator element
of the Spm system. Thus, Ac has no
recognizable component that corresponds
to component-1 of Spm.

It is possible that the different levels of
Wx expression produced by inactive
Ac at wx7, the “preset” patterns within
the Spm system, and the “‘oscillations” in
gene expression produced by responses
to active Ac of one Ac operator may all
reflect a common type of event occurring
at the locus of a gene to initiate tem-
porary metastability of its action.

BIBLIOGRAPHY

Burgi, E., see Hershey, A. D.

Cowie, D. B., and A. D. Hershey, Multiple sites
of interaction with host-cell DNA in the DNA
of phage », Proc. Natl. Acad. Sct. U.S., 58,
58-62, 1965.

Davern, C. I., see Hershey, A. D.

Hershey, A. D., and E. Burgi, Complementary
structure of interacting sites at the ends of
lambda DNA molecules, Proc. Natl. Acad. Sci.
U.S., 68, 325-328, 1965.

Hershey, A. D., E. Burgi, and C. I. Davern,
Preparative density-gradient centrifugation of
the molecular halves of lambda DNA, Biochem.
Biophys. Res. Commun., 18, 675-678, 1965.

Hershey, A. D., see also Cowie, D. B.

Ledinko, N., Occurrence of 5-methyldeoxycyti-
dylate in the DNA of phage lambda, J. Mol.
Biol., 9, 834-835, 1964.