TOPOGRAPHICAL RELATIONS BETWEEN ELEMENTS
OF CONTROL SYSTEMS IN MAIZE

Barbara McClintock

‘In Year Book 60, parallels were drawn
between gene-control systems in maize
and those in bacteria. In both organisms,
the comparable systems are composed,
basically, of two genetic elements: an
“operator” element, located adjacent to
the structural gene(s) and directly con-
trolling genic activity; and a “regulator”
element that in turn controls the func-

tioning of the operator element. Each
operator responds only to a_ specific
regulator. Therefore, each operator with
its corresponding regulator represents a
gene-control system. In bacteria, the
position of the regulator element on the
chromosome is not the same in all
examined systems: it may be located
either near by or at a distance from the
DEPARTMENT OF GENETICS

operator. Probably the elements of a
control system in maize express similar
topographical relationships; evidence sup-
porting that probability will be presented
here. ;

In maize, the controlling elements of a
system are transposable. Therefore, dif-
ferent genes may come under the control
of the same system, or the same gene
under the control of different systems.
Inception of control of gene action by a
particular system sometimes occurs when
the operator of the system is inserted
near the locus of the gene, the regulator
element being located elsewhere in the
chromosome complement. At other times,
inception of control is associated with the
appearance of the regulator near the locus
of the gene. In each such example that
has been examined adequately, however,
a clearly expressed two-element system
has subsequently arisen, the operator
element residing near the gene locus and
responding to the regulator element, now
located elsewhere in the chromosome
complement.

With advancing knowledge of control
systems in bacteria and of the topograph-
ical relations among individual compo-

nents of a system, the cases in maize in’

which the regulator element of a known
system initially occupies a position close
to the structural gene require close
scrutiny. In maize genetics, the devised
test methods allow ready detection of the
presence of the regulator element of a
system, either when it is located close to
the gene under the control of the system
or when it is located elsewhere. When the
regulator is close to the gene locus, the
concomitant presence of the operator
element is not so readily detected. Never-
theless, as was stated above, a clearly
expressed two-element system of control
of gene action does subsequently appear.
The origin of such a system poses several
critical questions: Are both elements of
the system initially located close to the
structural gene, and does the two-element
system arise by removal of the regulator
only? Or can a regulator element alone

449

directly modify the action of a gene when
located close to it, and does the two-
element system that subsequently ap-
pears result from an alteration of the
regulator element which converts it into
an element that thereafter behaves as an
operator?

Experiments eliciting direct answers to
the above questions may not be conducted
readily with maize. Nevertheless, certain
relevant facts are known. Although it is
possible to consider that insertion of a
known regulator element close to a
structural gene may initiate control of
action of the gene, and also that an
operator element may originate by some
modification of the regulator, no certain
evidence has yet been obtained of the
reverse process, that is, conversion of an
operator into a regulator. Moreover,
examination of the behavior of each of

‘the elements of a two-element system has

shown that transposition of both elements
to new locations in the chromosome
complement may occur coincidently in a
single nucleus, so that the operator is
removed from the locus of the gene and
the regulator is moved to a new location
in the chromosome complement. In other
cells, on the other hand, only one of the
elements is transposed at a given time.

Now, if it is assumed that both an
operator and a regulator are located
adjacent to the structural gene in those
instances where only the presence of the
regulator at that location can be deter-
mined with certainty, the following
possibilities are predictable. The operator
and regulator might undergo simultane-
ous transposition, leaving the structural
gene with neither element adjacent to it.
Other transpositional events might re-
move only one of the elements, the
complementary element of the system
remaining in location. With regard to
control of gene action, the effects pro-
duced by these kinds of transposition
would lead to three distinctly different
consequences. Removal of both elements
would release the structural gene from
control by the system to which the
450

elements belonged, and genetic tests
would reveal the absence of the regulator
element from the vicinity of the gene.
Removal of the operator element only
would likewise release the structural gene
from control by the system to which it
belonged; but the regulator element
would still be located close to the gene
locus. Removal of the regulator alone,
however, would give rise to a clearly
expressed two-element system of control
of gene action—the same system that was
operating before the transposition. More-
over, the location of the regulator element
at some distance from the structural gene,
or in another chromosome of the comple-
ment, could readily be determined.

The Cold Spring Harbor cultures
include five identified instances in which
the initial location of the regulator
element of a known control system close
to a structural gene resulted in control of
the gene’s action by the system to which
the regulator belonged. In the three
instances that have been adequately
examined so far, all three of the expected
consequences of transpositions outlined
above have been confirmed. The evidence
is reviewed below.

Three of the five examples involve the
Ac (Activator) system, and two the Spm
(Suppressor-mutator) system. Insertion
of Ac, the regulator element of the Ac
system, close to the locus of the bronze
(Bz) gene in chromosome 9 initiated
control of action of this gene by the Ac
system. The modified locus was desig-
nated bz”-?, and some discussion of it
appears in Year Books 54 and 55. Two
independently occurring insertions of Ac
close to the locus of the Wx (waxy) gene
in chromosome 9 resulted in control of
action of this gene by the Ac system.
These modified loci were designated
wae"? and wx"-*. Insertions of Spm, the
regulator element of the Spm system,
near the locus of the A, (anthocyanin)
gene in chromosome 3 gave rise to two
modifications designated a,"~? and a,”—*.
Extensive examinations have been made
only of bz™~?, a,"-?, and a.”7*.

CARNEGIE INSTITUTION OF WASHINGTON

Origin from ay"—* of a Two-Element
Control System

The original isolate of a1"~* had an Spm
element located close to the A: gene; the
degree of closeness was not made apparent
in the initial tests of 28 plants carrying
a,™-*. In 18 of these plants, more than
two Spm elements were present; 6 had
two Spm elements, one obviously linked
with a,;"—5; and 4 had one Spm element,
linked with ai:7—*. The location of Spm
in the immediate vicinity of the A: gene
was not recognized in the tests conducted
with these last 4 plants, because frequent
transpositions of Spm, occurring in some
sporogenous or presporogenous cells, re-
sulted in the production of a number of
gametes in which Spm occupied a new
location in the chromosome complement.
The intimate proximity of Spm to the A;
locus was made evident, however, in
tests of certain progeny of these plants,
in which transpositions of Spm occurred
go late in development that few or no
gametes carried a transposed Spm ele-
ment.

With respect to A: gene action, trans-
position of Spm away from the locus of
a,"-* leads to one or the other of two
quite different results: either release of
gene action from the control of the Spm
system, or continued control by that
system. Approximately half of the trans-
positional events that effect release from
control of the Spm system result in an A1
gene capable of a high level of action. T he
other half of these events bring about a
much lower level of A: gene action, or
occasionally the absence of such action,
in both plant and kernel.

That the Spm system can continue to
contro) A, gene action after some trans-
positions of Spm away from the a,;"~*
locus was discovered in two ways. One of
these utilized selected kernels on ears of
plants of the constitution a,”~° Sho,‘a1 she
(sho, shrunken endosperm; ai, standard
recessive allele of Ai) that had Spm
located close to a,"-*, produced by 4
cross with plants homozygous for a1 and
DEPARTMENT OF GENETICS

sh. and having no Spm. Plants were
grown from 29 Sh2 kernels that exhibited
uniform anthocyanin pigmentation in the
aleurone layer, intense in some kernels
and pale in others. The plants were tested
for presence or absence of Spm and, if it
was present, for the number of Spm
elements and their relative locations in
the chromosome complement. When Spm
was absent, it was introduced by means
of a cross into the endosperm nuclei of
kernels on the ears produced by these
plants, in order to test the expression in
its presence of the A, gene in the She-
carrying chromosome. The tests indicated
that, in 28 of the 29 plants, action of the
gene A, was no longer under the control
of the Spm system; the same level of
genic expression appeared both in the
presence and in the absence of Spm. In
some of these 28 plants, no Spm was
present. In others, one or more Spm
elements were present but were not
located close to the A, gene. In 2 plants,
however, Spm was found to be located
very close to the gene, even though genic
action had been released from the control
of the Spm system.

The remaining plant of the 29 was

derived from a pale-pigmented kernel. :

Tests for the presence of Spm were
negative. Nevertheless, the action of the
Ay gene in the Sh2-carrying chromosome
remained under the control of the Spm
system, as was shown when Spm was
introduced by a cross of this plant with
one carrying Spm. The response of the
gene to Spm was similar to that given by
the class II states of a:”—' described in
Year Book 57. With this state of ai™—*, a
medium level of Ai gene action is ex-
pressed in plants and kernels that have
no active Spm in their nuclei, but gene
action is suppressed if an active Spm is
present. Here, then, the Spm system
continued to control the action of the Ai
gene although Spm no longer occupied a
position close to it. A typical two-element
system of control of gene action had

evolved from an apparently one-element
system.

451

The second demonstration that not all
transpositions of Spm away from the
locus of a,"~° release the genic action from
the control of the Spm system was
provided by an examination of plants
derived from other selected Sh kernels
from ears produced by the same type of
cross as that producing the 29 kernels
whose constitutions are described above.
Each of these kernels exhibited a marked-
ly altered pattern of pigmented and
nonpigmented areas, as compared with
that of kernels carrying the original state
of a"-*. It was suspected that each of
these kernels had received an a,™—* locus
whose state had been altered in a cell of
the a,"-*-carrying parent plant. To test
this assumption, plants derived from 10
such kernels, each selected from a differ-
ent ear, were examined, and extensive
tests were subsequently conducted with
the progeny of 4 of them. All 10 plants
carried a modified state of a,"—°, and also
an Spm element. In 3 of the plants, the
single Spm had remained in intimate
association with the a:”~* locus; that is,
the event responsible for the alteration of
state had not resulted in its removal to a
more distant location. In the other 7
plants, however, Spm was located else-
where in the chromosome complement,
and in 6 of them it was not linked with
a,"-5; in the seventh, it was located
approximately 30 crossover units from
a,"~>, Thus a typical two-element system
of control of gene action was operating in
each of these 7 plants, and Spm was its
regulator.

The illustrations given above show that
with a,"-° the three anticipated conse-
quences of different types of transposition
of elements of the control system were
observed: (1) Release of Ai gene action
from control by the Spm system, associ-
ated with removal of Spm from the
immediate vicinity of the gene. (2)
Release of such control, not accompanied
by transposition of Spm. (3) Continued
control of Ai gene action by the Spm
system, after removal of the Spm element
from the immediate vicinity of a:”—°.
452

Analysis of a"?

The analysis of ai"~§, just described,
proceeded rapidly as soon as plants had
been isolated that carried a single Spm
element located close to the Ai gene. The
types of gene action produced by the
stable mutations were not difficult to
interpret; they appeared to express
different levels of standard A. gene
action. The behavior of the modified
states was also readily interpretable.
Those that were associated with the two-
element system of control of gene action
whose origins are described above fol-
lowed the same rules that had previously
been established for the Spm system.

Analysis of a,"-?, on the other hand,
has been complicated. Although the
original state is similar to the original
state of ai7—*, in that Spm resides close
to the A, locus and the Spm system
controls Ai gene action, the types of
expression resulting when the gene is
released from the control of the Spm
system are distinctly different. There are
two classes of mutants. The first has a
phenotype resembling that produced by
the standard A, gene. The other class is
composed of a series of alleles, distin-
guished from the first class by the
distribution and intensity of pigment in
plant and kernel. In the kernel, the
intensity of pigmentation in the aleurone
layer is not uniform, so that kernels
appear somewhat mottled. The different
alleles in this class may be distinguished
from one another by the degree of
intensity of kernel pigmentation, which
ranges from very faint to fairly dark. ‘The
plants also are pigmented, but the color
develops slowly and is markedly affected
by sunlight: the parts of a plant exposed
to direct sunlight become intensely
pigmented, whereas parts not so exposed
remain light in color. Although some
pigment develops in the mid-rib of the
leaf and at its edge, very little or none
develops in the leaf blade. The two classes
of mutants are thus readily distinguished.
The first will be referred to as “Ay”

CARNEGIE INSTITUTION OF WASHINGTON

mutants and the second as “mottled”’
mutants.

Control of gene action at a,"-? by the
Spm system is quite different from the
control exercised by that system when
the Spm element is not located near the
controlled gene. For example, in the
modified loci ai”~! and a,”—!, in the
above-described derivatives of ai”—*, and
in wr 8, gene action is suppressed by an
active Spm element but is expressed in
its absence or when it is present but
inactive. With a"? (original state),
however, the reverse is true: when Spm
is inactive, the action of the gene is
suppressed; when it is active, gene action
is expressed. This fact could be deter-
mined because it was possible to select
gome a,”~2-carrying plants in which Spm
was in an inactive phase of long duration,
and others in which Spm was in an active
phase of long duration. Some tests
conducted with plants having Spm in an
active phase of long duration will be
considered first.

Location of Spm before and after release
of control of gene action at ay"—? by the Spm
system. With the original state of ai”~?,
the location of Spm close to the A, gene
was established by several types of test,
commencing with a cross of plants of the
constitution a:"-? Sh2/a: shy by plants
that had no Spm and were homozygous
for a specially selected state of a,"—? and
also for she Ai gene action in these
last-named plants was under the control
of the Spm system, involving an operator
element located close to the A gene and
an Spm element located elsewhere in the
chromosome complement. With this se-
lected state of a,"~', gene action is
expressed in the absence of Spm (or when
it is present in an inactive phase).
Anthocyanin pigment appears in both
plant and kernel. In the kernel, pigment
of medium intensity is uniformly distrib-
uted over the aleurone layer. When an
active Spm is present somewhere in the
chromosome complement, gene action is
suppressed until there occurs, in some
cells, a response of the operator to Spm
DEPARTMENT OF GENETICS

that effects a release of gene action from
the control of the Spm system. These
releases occur in a relatively few cells late
in the development of plant and kernel,
and most of them lead to an expression
of A, gene action resembling that of the
standard A, gene. Consequently, they
give rise to a distinctive pattern of deeply
pigmented dots in the kernel and small
pigmented streaks in the plant, both
appearing on a nonpigmented back-
ground. In plants and kernels carrying
the original state of a,"~*, on the other
hand, release of control of gene action by
the Spm system may occur in many cells,
both early and late in development. As is
described above, such release may result
either in a high level of Ai gene action, a
lower level that produces the ‘‘mottled”
phenotype, or, rarely, a null expression
of the gene. Thus, kernels carrying the
original state of ay"~? and a fully active
Spm exhibit both large and small pig-
mented areas of various intensities.

In the above-described cross, nearly all
the Sh, kernels on an ear receive from the
heterozygous parent either unmodified
a,"-2 or a modified derivative of it, and
nearly all the sh2 kernels receive the
standard a; allele. This happens because
crossing over between the locus of ay?
and that of Shs is very infrequent, not
exceeding 0.12 per cent. The presence or
absence of active Spm can be detected
readily in the sh, kernels on the ears
produced by the cross, and also in She
kernels that have received a gamete
carrying a stable mottled mutant of
a;"-*. If an active Spm is present in one
such kernel, the distinctive pattern of
deeply pigmented dots produced by the
response of a,"-! to Spm appears in a
mottled background. If Spm is absent,
these dots are absent and the kernels
exhibit only the mottled phenotype.
Table 2 lists the phenotypes of kernels
that appeared on some ears produced by
the cross. The ratios of kernel types were
not the same on all these ears. Neverthe-
less, except on ears of plants whose
numbers are printed in italics, there was

453

a direct relationship between the per-
centage of kernels in the She class that
received a germinal mutant of a."~? and
the percentage of kernels in the she class
that received Spm. On ears in which all
the Sh» kernels had received unmodified
a,"—? there were no kernels in the shz
class that carried Spm. Among the ears
bearing kernels that expressed germinal
mutations of a1"-? the percentage of such
kernels and the percentage of shz kernels
with Spm were directly related. (This
correlation was exhibited among the
kernels on ears having no detectable
sectors derived from cells in which a
stable mutation of a:*-? had occurred
early in development. Ears with such
sectors are not included in the table.) The
correlation suggested that Spm was
located very close to a,"~? in the hetero-
zygous parents and that its removal from
this location was associated with the
origin of many of the stable mutations.

This possibility was also suggested by
the phenotypes of kernels on ears pro-
duced by testcrosses conducted with other
plants having the constitution a:"~-? Sh2/
a,"~! sho. Ears of 33 plants of this consti-
tution were utilized in crosses with plants
homozygous for a:"—' and sh2 and having
no Spm, and also with plants homozygous
for a, and sh, and having no Spm. Table 3
shows the phenotypes of kernels that
appeared on ears produced by the second
cross. Again, a direct relationship will be
noted between the percentage of kernels
that received a germinal mutant of a:"~?
and the percentage of kernels in the she
class that received Spm.

The cross that produced the kernels
entered in table 2 was conducted after the
above-described correlation had been
recognized. The ear-bearing parent plants
in cultures 79794 and B, 7980A, and
7981A were derived from variegated, Shz
kernels on ears of plants 7799B-1,
7799B-6, and 7800A-5 of table 3. Plants
were grown from kernels in the underlined
classes in table 3 in order to test the
conclusion that Spm resides close to
unmodified a:"-? and that many of the
454

CARNEGIE INSTITUTION OF WASHINGTON

TABLE 2. Phenotypes of Kernels on Ears of Plants of the Constitution a:%~? She/a, she
Produced by a Cross with Plants That Were Homozygous for a,"~! and shz and Had No Spm

Phenotypes of Kernels

 

 

 

 

 

Shz Class* sh, Classt
Plant Germinal Mutations
Number Variegated P. t Dots of Ai P. t-
Mottled for Ai and Genmi ser Pale in Colorless re with
A Mottled Mutatie a! (No Spm) Background eS ~
1 No A; Dots A: Dots Spots uiauem (Spm) pm
(No Spm) (Spm)

T979A-7 0 0 0 57 0 73 0 0
A-8 0 0 0 233 0 189 0 0
B-1 0 0 0 239 0 219 0 0
A-6 0 3 1 69 5.4 59 1 1.6
A-3 1 4 1 71 7.7 64 2 3.0
A-12 3 13 ll 98 21.6 114 12 9.5
A-l 4 25 15 105 29.5 118 20 14.5
B-4 4 36 13 109 32.7 148 29 16.3
A-2 3 33 25 113 35.0 148 25 14.4
A-13 3 35 22 106 36.1 142 17 10.6
A-10 9 46 47 169 37.6 232 51 18.0
A-11 7 61 57 115 52.0 171 53 23.6
A-9 10 26 62 134 42.2 97 157 61.8
B-3 25 44 82 83 64.5 96 137 58.8

7980A-9 ll 32 22 149 30.3 177 31 14.9
A-7 1 5 5 24 31.4 22 6 21.4
A-3 5 20 6 53 36.9 76 13 14.6
A-4 6 38 21 110 37.1 163 30 15.5
A-1 6 62 41 169 39.2 263 64 19.5
A-2 10 55 43 110 49.5 190 49 20.5
A-8 12 33 80 111 §2.9 121 129 51.6

7981A-1 3 49 24 151 33.4 183 22 10.7
A-4 5 51 33 114 43.8 108 Al 27.5
A-5 10 64 33 136 44.0 199 37 16.1
A-7 6 41 66 68 62.4 157 53 25.2
A-8 9 63 66 59 70.0 144 59 29.0

 

* In addition there was one pale, Shz kernel.

+ In addition there were three shz kernels that received a,"7?.

germinal mutations arise when Spm is
transposed to a new location in the
chromosome complement. Among the 26

. plants listed in table 2, 23 had one Spm
element'in the cells that gave rise to the
testcross ear and 3 (whose numbers
appear in italics) had two Spm elements.
There were 4 additional plants, each also
derived from an Sh, kernel whose endo-
sperm was variegated. A mottled pheno-
type was expressed in these 4 plants
rather than the phenotype produced in
plants that commence development with
unmodified a:"-?. The presence of a

mottled mutant in these plants was
confirmed by the kernel types that
appeared on a testcross ear of each (rows
1-4, table 4). Since the endosperm of the
kernel from which each of these plants
arose started development with unaltered
a,"-2, the event that produced the
mottled mutant must have occurred
during development of the female game-
tophyte in the parent plant, or in the
kernel early in development of the
embryo. All 4 plants had one or more
active Spm elements. In 2 of them
(7981A-3 and 7981A-6), one Spm, not
DEPARTMENT OF GENETICS

455

TABLE 3. Phenotypes of Kernels on Ears of Plants of the Constitution a:*~? Sh2/ay"™ sho
That Had One Active Spm, Produced by a Cross with Plants Homozygous for a, and she and
Having No Spm

Phenotypes of Kernels

 

 

 

 

 

Shz Class shy Class
Plant,
Number Germinal Mutations Hearne Percentage Pale eee Percent-
Germinal A1 age with
Mottled . (No Spm) Background
Ai Mottled Spots Mutations (Spm) Spm
7799B-1 0 0 201 0 0 214 i 0.46

7799B-6* 2 50 176 22.8 1 216 28 11.4

7800A-5 6 60 153 30.1 1 180 21 10.4

7984 -7 6 64 131 34.8 0 172 31 15.2

7984 —4 2 60 109 36.2 0 150 27 15.2

7984 -3 4 62 99 40.0 1 131 30 18.5

TI99A 14 95 119 47.8 2 160 44 21.3

 

* In addition there was one colorless, shz kernel on this ear.

linked to the mutant locus, was present
in the cells that produced the testcross
ear. In the other 2, two Spm were present
in the cells giving rise to the testcross
ear—neither element linked with the
mutant locus in plant 7980A-6, but one
linked with it in plant 7981A~2.

The plant grown from the single she

kernel containing Spm on the ear of plant’

The plant derived from it had no Spm.

7799B-1 (row 1, table 3) proved to be
ay"-! sho/ai shy in constitution and had
two independently located Spm elements
in the cells that produced each of its
tested ears.

Nine plants derived from the mottled
Sh. class of kernels on the ear of plant
7799B-6 (row 2, table 3) were also tested
for Spm constitution. No evidence of its

TABLE 4. Phenotypes of Kernels on Ears of Plants That Were Mottled-Mutant She /ay sho in
Constitution, Produced by a Cross with Plants Homozygous for a1"! and she and Having No Spm

Phenotypes of Kernels

 

 

 

Plant Mottled She Class shz Class

Number No Dots of Deep Dots of Deep Deep-Pigmented

Pigmentation Pigmentation Pale Dots in Colorless

(No Spm) (Spm) (No Spm) Background (Spm)
7980A-6 86 162 73 149
7981A-3 120 107 120 111
7981A-6 82 81 74 82
7981A-2 92 125 111 55
7980B-3 91 85 103 80
7980C-2 60 141 166 45
7980B—-4 30 211 95 127
7981B-1 20 219 254 22
7981B-6 0 274 255 1
7981B-8 80 126 149 71
7981C-3 47 175 51 180

 
456

presence was shown by the kernels on the
ears of 6 of these 9 plants; but it was
present in the cells that gave rise to the
testcross ear in the remaining 3 plants
(7980B and C, table 4). Plant 7980B-3
had one Spm, not linked with Sh2; plant
7980C-2 had one Spm, linked with Shz;
and plant 7980B-4 had two Spm, one
linked with Sh,. Testcrosses conducted
with 8 of the 12 plants derived from
mottled Sh: kernels on the ear of plant
7800A-—5 (row 3, table 3) produced no
evidence of the presence of Spm. It was
present, however, in the remaining 4
plants, as indicated in table 4. Very close
linkage of Spm with the locus of the
mottled mutant was exhibited by plant
7981B-6. Linkage of Spm with the locus
of the mutant was expressed in plants
B-1 and B-8. The ratio of kernel types
on the ear of plant 7981C-3 (355 with
Spm/98 with no Spm) suggests the
presence of at least two Spm elements,
not linked with She, in the cells that gave
rise to this ear.

All together, 42 plants derived from
mottled Sh. kernels on ears of ai"~?
plants that had one Spm, located close to
a,”~?, have been tested for Spm consti-
tution and location. Twenty-six of the
plants showed no evidence of the presence
of Spm. Twelve plants had one Spm: in
2 of them it was situated very close to
the locus of the mutant; in 2 others it was
linked with the mutant locus; and in the
remaining 8 there was no evidence of such
linkage. Three plants had two Spm;
neither was linked with the locus of the
mutant in 2 of the plants but one Spm
was linked with it in the third. The

‘remaining plant of the 42 had three Spm
elements, none of them linked with the
mutant locus. Thus, Spm was present in
only 16 (38 per cent) of the 42 plants
derived from kernels in which a chromo-
some carrying a germinal mutation was
received by both the endosperm and the
zygote nuclei. Among the 1288 mottled
Sho kernels in table 2 that appeared on
ears of plants having one Spm, 552 (43
per cent) carried Spm and 736 had no

CARNEGIE INSTITUTION OF WASHINGTON

Spm. The agreement in distribution of
Spm to the mutant class, demonstrated
by these two types of test, is good.

Further confirmation that Spm was
located close to a:”~?, and that its
removal from that location was related to
the origin of the stable mutants, was
provided by tests of the progeny of an
ai"? She/a, she plant produced by cross-
ing this plant with one that was homozy-
gous for a,;"—! and sh, and had no Spm.
A very large sector, present in the a,"—?—
carrying plant, was derived from a cell in
which a mutation to a stable mottled
allele had occurred. On the ear of the
described testcross there were 306 mot-
tled, She kernels, of which 150 carried
Spm and 156 had no Spm. Only 18 Sh.
kernels on this ear had received unmodi-
fied a,”~?. Among the 315 sh, kernels on
this ear, 177 were uniformly pigmented
(no Spm) and 138 had dots of deep
pigmentation in a colorless background
(Spm present). Ten plants grown from
mottled Sh, kernels carrying Spm, and 10
plants from She kernels that had received
unmodified a,7—?, were tested for Spm
number and location. On testcross ears
produced by the 10 plants derived from
the mottled kernels the ratio of kernel
types indicated that 9 of them had one
Spm, not linked with She, and that two
Spm elements, not linked with Sho, were
present in the cells that produced the ear
on the tenth plant. On testcross ears of
the 10 plants derived from kernels having
unmodified a,"—-?, the ratios of kernel
types were similar to those entered in
table 3: one Spm was present in each
plant, and it was located close to a,”~°.
It may be concluded, then, that the
a,”~-*-carrying parent of these 20 plants
commenced development with a single
Spm, located close to a,"~?. Early in
development of that plant, transposition
of Spm to a new location, occurring in one
cell, led to the origin of the stable mottled
mutant that was present in all descend-
ants of the cell.

From the above-described series of
tests, it is evident that the origin of many
DEPARTMENT OF GENETICS

of the stable mutants of a:"~? is associated
with transposition of Spm to a new
location in the chromosome complement.
In some of the stable a,”-? mutants, on
the other hand, Spm continues to occupy
a position close to the locus of the
modified A; gene. Thus, in these respects,
a,"—? and a," are comparable.

Origin of a two-element system of control
of gene action at ay". Although the tests
aimed at identifying events in which a
clearly expressed two-element system of
control of gene action arises from ay"?
have not yet been completed, one example
may have been found. A kernel with a
distinctive phenotype appeared on an ear
of a plant that was a1"? Sho/ai she in
constitution and had one Spm located
close to a:"-?, after it was crossed with a
plant of similar constitution. The selected
kernel was weakly and irregularly pig-
mented, with no spots of deep pigmenta-
tion in its aleurone layer. The phenotype
of the plant grown from this kernel was
similar to that expressed by many of the
stable mottled mutants of a:”~?, since
anthocyanin pigment appeared in the
same regions of the plant. Testcrosses
conducted with this plant gave no evi-
dence of the presence of Spm. Its consti-
tution, however, proved to be ay? Sho/
a, she. The kernels on one of its ears were
produced by a cross with a plant homozy-
gous for a, and sh, and having no Spm.
‘All the Sh kernels on this ear exhibited
the same phenotype as that shown by the
kernel that gave rise to the plant. The
kernels on a second ear of the plant were
produced by a cross with a plant that was
homozygous for a; and sh, and carried
one Spm closely linked with the Pr marker
in chromosome 5 (Pr, purple aleurone;
pr, recessive allele, red aleurone). Of the
217 She kernels on this ear, 102 were
variegated for pigmented areas of differ-
ent intensities, in a pattern resembling
that produced by unmodified a:"~*. The
phenotype of the remaining 115 She
kernels was similar to that on the first
ear, just described. From the close linkage
of the variegated phenotype with Pr and

457

the nonvariegated phenotype with pr, it
is concluded that gene action at this
modified a,”-? locus is under the control
of the Spm system, although Spm no
longer resides close to the locus.

The modified a,"~? locus present in the
plant just described could have arisen
from removal of only the Spm element
from the vicinity of the a,"~? locus, the
operator element remaining in location.
That inactivation of the Spm element
was responsible for the modification is.
not probable, as it is well established that
such inactivation results in suppression of
the action of the A: gene (see below);
much pigment appeared in the plant
having this modified a:*~* locus, and
some pigment appeared in the aleurone
layer of the kernels.

Inactive Spm at the locus of ai:"~*.
Testcrosses conducted with plants in
which Spm was in an inactive phase,
changing to an active phase only in a few
cells very late in development, also
served to place Spm close to the locus of
unmodified a:"-?. Two types of testcross
were performed. When plants with in-
active Spm, ai:"-? Sh2/ai she in constitu-
tion, were crossed with plants homozy-
gous for a: and she and having no Spm,
all the kernels on some ears were colorless.
On other ears, however, a few kernels in
the Sh, class had a sector containing
pigment of light intensity, and some of
these sectors, in turn, also displayed small
dots of deep pigmentation. Occasionally,
the entire aleurone layer of an She kernel
exhibited such dots on a lightly pigmented
background. Also, an occasional Shs
kernel was variegated throughout its
aleurone layer, with large as well as small
pigmented areas of various intensities.
Tests conducted with plants derived from
such variegated kernels indicated that
the phenotype of the kernels was pro-
duced by a change in phase of activity of
Spm, from inactive to active, occurring
in a cell late in the development of the
ai™-2-carrying plant.

The second type of test utilized the
same plants as those described above, as
458

ear-bearing parents in crosses with plants
that were homozygous for a,"~1 and shy
and had no Spm. On some of the resulting
ears, all kernels in both the She and she
classes were uniformly pigmented and
showed no evidence of the presence of
Spm. On other ears, a few of the kernels
in the Sh, class exhibited sectors of much
lower pigment intensity, and some of
these, in turn, had small deeply pigmented
spots. Occasionally, the whole aleurone
layer of an She. kernel exhibited this
small-spotted phenotype, or one with
large as well as small areas of different
grades of pigment intensity. Kernels
having these last two phenotypes would
be expected to appear if, in the a)"-?-
carrying plant, the inactive Spm under-
went a change to the active phase in some
cells, late during development of the
ovule or in the female gametophyte.

When plants of the constitution a:"-?
Sh:/a: sh, that carried an inactive Spm
were crossed with plants homozygous for
a, and sh, that carried one or more active
Spm elements, all or nearly all kernels
that received a,”—? from the heterozygous
parent and no Spm from the homozygous
parent were colorless. In contrast, all
those that received active Spm from the
homozygous parent exhibited many mu-
tant areas, in a pattern resembling that
produced by a,"~? when the Spm adjacent
to it is in an active phase.

Conclusions derived from the study of
a,"-*, Results of the described tests with
plants having unmodified a,"~?, and with
others having modified derivatives of the
locus, indicate that certainly two and
probably all three of the predicted
consequences of transpositional events,
outlined early in this report, have been
observed. .

In this report, some aspects of the
analysis of a:"—? have been considered in
detail, not only in order to develop the
thesis stated earlier but also to indicate
the nature of the evidence that makes it
possible to relate the mode of control of
the Spm system to that controlling the
alternate action of the duplicate genes,

CARNEGIE INSTITUTION OF WASHINGTON

H, and Hz2, associated with flagella
antigen formation in the bacterium
Salmonella. In maize plants that are
a:"-?/a,"—! in constitution, with an Spm
element located close to a,;"—?, which of
the alleles will be active and which
inactive is determined by the phase of
activity of the Spm element. In Salmo-
nella, which of the two duplicate genes
will be active and which inactive is deter-
mined by the phase of activity of the
controlling element Vh, located close to
the H» gene.

The Derivatives of bz"?

Early studies of bz"-? were reported in
Year Books &4 and 54. It was shown that
Ac, the regulator of the Ae control
system, resides close to the locus of the
bronze gene in chromosome 9, and also
that the Ac system controls the action of
this gene. The behavior of unmodified
bz"? was examined, initially, in 172
plants carrying bz”™-? in one chromosome
9 and the standard stable recessive, bz, in
the homologue, as well as in 13 plants
homozygous for unmodified bz™-?. Sub-
sequent studies were conducted with
plants carrying modified derivatives of
be—*. Since the information obtained is
both diverse and extensive, the present
report will be confined to summary
statements pertinent to the topic in hand.

In a cross of bz”-*-carrying plants to
plants homozygous for standard 62z,
kernels that had received a modified
derivative of bz™-? appeared on some
ears. Most of these kernels exhibited
either a null level or a high level of gene
action at the bronze locus. It was
suspected that in them the action of the
bronze gene, derived from bz"~?, had been
released from the control of the Ac
system. To test this conjecture, selections
were made of 35 independently occurring
examples of change of bz”? to an
apparently stable null-expression allele,
and of 14 independently occurring
changes to an allele expressing a high
level of gene action. It could be deter-
mined readily that, in 33 of the 35
DEPARTMENT OF GENETICS

selected examples, release of gene action
from control by the Ac system was
associated with the origin of a stable null
expression of the bronze gene. In 19 of
these 33, Ac was absent from chromosome
9 in the original plant carrying the
modified bz? locus, although it was
present elsewhere in the chromosome
complement in 6 of the 19 plants. In the
rest of the 33 plants (14), 4c was present
in chromosome 9. It was located close to
the bronze gene in 2 of them, and at
positions away from the locus in 3 others;
but its exact location was not determined
in the remaining 9 plants, 1 of which had
two Ac elements, one in chromosome 9
and one elsewhere.

The 2 remaining kernels of the 35 that
were selected for a stable, null expression
of the bronze gene produced plants in
which no Ac was present; both plants
were totally bronze in phenotype. When
they were crossed with plants carrying
Ac, it was learned that the bronze gene,
derived from bz”—?, was under the control
of the Ac system. The manner of its
response to Ac was similar to that
observed in the many other examples of
two-element control systems in which Ac
is the regulator element. In both these
plants, and in a third plant derived from
a kernel selected in a different manner, a
two-element system of control of gene
action had arisen from bz™-?, Although
Ac was no longer located close to the
bronze gene, it continued to be the
regulator of the system controlling its
action.

The 14 original kernels selected for the
presence of a derivative of bz"? that
expressed a high level of gene action gave
rise to 8 plants having Ac and 6 having
no Ac in their nuclei. Seven of the 8
Ac-earrying plants had one Ac element.
In 4 plants it was not linked with markers
in the short arm of chromosome 9; in the
other 3 it was linked with such markers,
being located close to the Bz-expressing
gene in 2 of them and proximal to the
locus of Wx in the third. The eighth
Ac-carrying plant had two Ac, one located

459

close to the Bz gene and one not linked
with markers in the short arm of chromo-
some 9. Tests of all 14 plants and their
progenies indicated that the presence of
an Ac element in the nucleus did not
effect a modification in action of the
Bz-expressing gene derived from bz™~*.
Action of this gene appeared to have been
released from the control of the Ac
system. :

A modified state of bz"? was recog-
nized early in the study of that locus.
The alteration at the bronze locus that
produced this state did not remove Ac,
which remained close to the locus of the
gene, and the Ac system continued to
control gene action. In contrast to the
original state of bz"—?, this altered state
is characterized by a high level of bronze
gene action. Some of its Ac-controlled
modifications result in recognizable
changes in level of gene action. Others
result in release of the gene from control
by the Ac system, and such release is
often associated with maintenance of a
high level of gene action. Still other
modifications give rise to further altered
states. One of these resembles the initial
state of bz™-2, and another has proved to

. be instructive for the thesis of this report,

for it allows ready selection of kernels
produced from cells in which Ac no
longer occupies a position close to the
bronze gene. This state was recognized,
initially, in a single kernel on an ear. The
aleurone layer of the kernel exhibited
many deeply pigmented spots in a lightly
pigmented background.

Tests of the plant derived from this
kernel, and of its progeny, showed that
Ac occupied a position close to the locus
of the modified bronze gene and that the
observed changes in action of the gene
were expressions of control by the Ac
system. On ears produced by a cross of
plants carrying this state with plants that
were homozygous for the standard reces-
sive bz and had no Ac, kernels that
received the state exhibited deeply pig-
mented areas in a lightly pigmented
background. A few kernels on some ears,
460

however, showed only the light back-
ground pigmentation, with no deep-
colored spots. Plants derived from 5 such
kernels were examined. Tests were made
for the presence or absence of Ac in them,
and for the expression of the weak bronze
allele in their progeny produced by a
cross with plants that were homozygous
for standard bz and had either no Ac or
one or more Ac. It was learned that no
Ac was present in these plants. In the
absence of Ac, the expression of the
bronze gene is constant, and it behaves
as a weak allele of Bz. When Ac is present,
however, deeply pigmented spots appear
in a lightly pigmented background. It
could be determined readily that these
spots arise through the response of an
operator element at the locus of the weak
Bz allele to the presence of Ac. Thus, a
two-element system of control of gene
action was expressed in each of these 5
selected examples, and Ac was the
regulator of the system.

The above-described sequence of events
affecting bz”-? may be interpreted in the

CARNEGIE INSTITUTION OF WASHINGTON

following manner. Insertion of the oper-
ator and regulator elements of the Ac
system close to the locus of the standard
Bz gene gave rise to bz”-?, which exhibits
a null base level of gene action. Thus, the
original bz”-? locus may be symbolized as
bz(op) Ac. Since only the Ac element at
the locus of the gene has been determined,
the symbol for the operator element (op)
is shown in parentheses. Removal of both
the operator and the regulator, Ac, from
the locus of the gene, or removal of the
operator alone, releases the gene from
control by the Ac system. Removal of Ac
alone allows the presence of the operator
to be recognized, and the locus can be
given the symbol bz-op. A change at
bz™-?, assumed to be induced by the
operator in response to the regulator Ac,
produces a new state characterized by a
high level of gene action, which is
symbolized Bz*(op)Ac. A subsequent
modification, again assumed to be induced
by the operator, gives rise to another
state characterized by an intermediate
level of gene action, which is symbolized

TABLE 5. Response of the Bronze Gene to Ac in Selected Derivatives of bz(op)Ac, the Original
State of bz”-? (Part I), and in Two of Its Modified States, Bz'(op) Ac (Part IT) and Bz*(op) Ac

 

 

(Part ITT)
Expression of Bronze Gene Response of Bronze Symbol for No. Cases
in Selected Kernel Gene to Ac Bronze Gene Examined
Part I

Null expression; stable Negative bz 22
Negative bz-Ac 2

Negative bz; Ac in chromosome

9; position not
determined 9
Positive bz-op 3
High level of gene action; stable Negative Bz 11
. Negative Bz-Ac 3
High level of gene action; unstable Positive Bz*(op)Ac 1
: Part II
High level of gene action; stable Negative Bz 5
Negative Bz-Ac 7
Return to bz™~? expression Positive bz(op) Ac 10
Weak expression of gene; unstable Positive Bz*(op)Ac 1
Part III ,
High level of gene action; stable Negative Bz 3
Weak expression of gene; stable . Positive Bz*-op 5
Return to high level of gene action;

unstable Positive Bz*(op) Ac 1

 
DEPARTMENT OF GENETICS

Bz“(op) Ac. Removal of only Ac allows
the presence of the operator to be recog-
nized, and in this event the gene locus is
given the symbol Bz"-op.

Table 5 summarizes the evidence for
the considerations outlined in this section.
The symbols in the table are the same as
those just described. A symbol that does
not include op or Ac indicates that no
evidence has been obtained of the
presence of either element at the locus of
the gene.

The findings presented in this report
are sufficiently extensive to leave no

461

doubt that a two-element system of
control of gene action, composed of an
operator element at the locus of the gene
and a regulator element located else-
where, may arise at a gene locus that
initially carried the regulator of the
system. Although in the examples studied
the origin of the operator element has not
been determined directly, it is neverthe-
less evident that the predicted conse-
quences of removal of either or both
elements from the locus of the gene, on
the assumption that both were present
initially, have been confirmed.