SPONTANEOUS ALTERATIONS IN CHROMOSOME
SIZE AND FORM IN ZEA MAYS

BARBARA McCLINTOCK

Spontaneous aberrations in maize leading to
changes in size and form of the chromosomes have
not been investigated from the point of view of de-
termining, systematically, the frequency and posi-
tions of breakages and reunions of broken ends of
the chromosomes of the complement, as has been
done in Tradescantia (Giles, 1940), in Allium
(Nichols, 1941) and in other forms (Darlington and
Upcott, 1941). Nevertheless, through studies of
various problems not directed toward this goal,
much has been learned of the process underlying
the origin of changes in size and form of the chromo-
somes of maize which are not conditioned by the
usual methods of inducing aberrations, such as
X-radiation, ultraviolet radiation, high tempera-
tures and aging.

In the early cytological studies of maize, it be-
came clear that spontaneous aberrations were oc-
curring to give rise to various types of altered chro-
mosomes. In many cases, the time of occurrence or
the conditions which gave rise to the alteration
were not known. These aberrations were first ob-
served in various plants of particular strains which
were under cytogenetic investigation. These aberra-
tions included reciprocal translocations, inversions,
deficiencies, ring-chromosomes, a duplication, frag-
ments, and a secondary trisome. Through further
studies, it became apparent that chromosome modifi-
cations were occurring in individual plants under
investigation. A single plant of a culture may show
one of the various types of aberrations mentioned
above. Although, in some cases, it could not be
determined whether all of the cells of the plant
possessed the aberration, in other cases it was deter-
mined that the plant was sectorial for the modifica-
tion. In these latter cases, it was obvious that the
modification occurred during the development of the
individual plant. The factors responsible for these
spontaneous aberrations were not apparent in any
of these cases. However, there are types of chromo-
somal aberrations which are induced by known
factors or are correlated with known conditions.
These will be considered under appropriate head-
ings in the following discussion,

SPONTANEOUS CHROMOSOME ABERRATIONS UNDER
GEnic CONTROL

There are two well investigated cases which indi-
cate that the rate of spontaneous chromosome aber-
ration in maize may be controlled by the genic
composition of the nucleus. The first case is strik-
ingly illustrated by the sticky gene studied by

Beadle (1932, 1937). Sticky is a recessive mutant -

located in chromosome 4 which causes a tre-
mendous increase in the rate of spontaneous chromo-
some aberration in all types of tissues. At the first
meiotic anaphase in homozygous plants, the chromo.
somes appear adhered to one another. This stick-
ing together of the chromosomes of the complement
suggested the designation sticky for this mutant,
As a consequence of this sticking, many of the
chromosomes are ruptured during the meiotic ana-
phase. In the mitotic divisions, numerous types of
chromosomal aberrations were observed in plants
homozygous for the sticky mutant. The continued
production of spontaneous aberrations during devel-
opment causes plants homozygous for sticky to be
stunted in growth and to possess numerous streaks
of tissues with altered phenotypes. The endosperm
tissues are likewise a mosaic of various types of
aberrant cells. The extremely high rate of spon-
taneous chromosome alterations, both in the sporo-
phytic and endosperm tissues, undoubtedly is the
cause of the observed phenotypic alterations of the
cells and tissues, for various grades of chromosomal
unbalance must be present in these cells. It is
likewise of particular interest to note that the
sticky mutant is responsible for a marked increase
in the rate of spontaneous mutation.

The second case of spontaneous chromosome al-
teration under genic control has been studied by
Jones (1937, 1940). In some of his strains of maize,
the endosperm tissues give unmistakable evidence
of a high rate of spontaneous chromosome aberra-
tion. Unlike the sticky mutant, the high rate of
chromosome aberration appears to be confined to
the endosperm tissues alone. The genetic evidence
indicates that reciprocal translocations are occurring
between non-homologous chromosomes and _ that
chromosomes with unstable broken ends are likewise
produced. Other types of chromosomal aberrations
could not be detected genetically. Cytological ob-
servations of the endosperm tissues of these plants
have produced direct evidence of a high rate oi
spontaneous chromosome aberration (Clark and
Copeland, 1940).

It is needless to say that any systematic study of
spontaneous chromosome aberration in maize must
be considered with reference to the genic compo-
sition of the plants under investigation.

CHROMOSOME ALTERATIONS INDUCED By CROSS
ING OVER BETWEEN HoMoLocous SEGMENTS OF
CHROMOSOMES

The normal process of crossing-over may be
responsible for the production of chromosomes with

[72]
CHROMOSOME ALTERATIONS IN ZEA

“ered sizes and forms. When individual plants are
~ terozygous for an inversion, a duplication or some
“arrangement in the linear organization of the
“sromosome, predictable types of chromosomal
“rerations may follow crossing-over between homol-
ous segments of chromosomes. Although struc-
gral heterozygosity greatly increases the rate of
roduction of altered chromosomes following cross-
‘az-over, it will be shown that chromosome aberra-
ions likewise may be induced within a normal com-
“jement following the regular process of crossing-
‘ver, Several examples illustrating the part that
crossing-over plays in the production of chromo-
«mal aberrations will be considered.

73

more deficient than the broken chromatid entering
the sister nucleus. The latter broken chromatid will
possess a duplication besides a deficiency. Several
such inversions have been investigated in maize
(McClintock 1933, 1938b). In several cases, a defi-
cient broken chromosome, presumably derived from
crossing over within an inverted segment, has been
recovered in the following plant generation.

The breakage of a chromosome at a meiotic ana-
phase is the starting point in the production of
chromosomes with various modifications of size and
chromatin constitution. The direct cause of these
modifications is related to the subsequent behavior
of the broken end. When a chromatid is broken at a
6 A

ee Broken end

B_A
= Prophase
BOA
c 8 A Anaphase
t
cB AA Tetophase Af
CB A_A 8 oC
—_— Prophase Cc
cB A A BC
¢ 8B A A A_A BB C Anaphase co 8
t 1
cs
CB A A AA BC Telophase cB 86 £

 

 

Fic. 1. Diagrammatic representation of the method by which a chromosome with a broken end gives rise to chromosomes
with altered sizes and constitutions. The diagram at the top of the figure represents a chromosome with a broken end. The
centromere is represented by the clear oval. The organization of the arm with the broken end is represented by A, B and C,
A being adjacent to the broken end. Following reduplication of the chromosome, the two sister chromatids are fused at the
position of previous breakage (Prophase, second diagram from top). The centromeres pass to opposite poles in the succeeding
anaphase. This produces a bridge configuration (Anaphase, third diagram from top). If breakage of this bridge configuration
occurs at the position of the arrow, a broken chromosome will enter each telophase nucleus (Telophase, right and left, fourth
diazram from top). The broken chromosome to the left possesses a duplicated segment, that to the right is deficient for a
terminal segment. Continuation of this breakage-fusion-bridge cycle in succeeding nuclear divisions may result in the
production of chromosomes with various duplications, deficiencies or duplications plus deficiencies as illustrated in the
diagrams below each of these telophase chromosomes. (From McClintock 1941a, through the courtesy of Genetics.)

Plants heterozygous for an inversion which does
not include the centromere will give rise to altered
chromosomes following a crossover within the in-
verted segment. It is well known that this results
In the production of a dicentric chromatid and an
acentric fragment. Passage of the two centromeres
ul the dicentric chromatid toward opposite poles
‘n the meiotic anaphase spindle produces a chroma-
tin bridge configuration. Rupture of this bridge
vccurs either before or following the formation of
the cell plate. The position of rupture varies. It may
be adjacent to one centromere or at any position
detween the two centromeres. In all cases, a rup-
tured chromatid possesses a deficiency because the
dicentric chromatid itself is deficient for a segment
Carried by the acentric fragment. If the position of
breakage in the bridgé configuration is non-median,
the broken chromatid entering one nucleus will be

meiotic anaphase, fusions occur at the position of
breakage between the two sister halves of this
broken chromatid (Prophase, second line, fig. 1).
As the centromeres of the terminally fused sister
chromatids pass to opposite poles in the following
mitotic anaphase, a chromatin bridge configuration
is produced (Anaphase, third line, fig. 1). As the
centromeres approach the poles of the spindle fig-
ure, tension is exerted on this bridge causing it to
rupture. The rupture may occur at any position
between the two centromeres. Following a non-
median rupture of the bridge configuration (arrow,
upper Anaphase, fig. 1) the broken chromosome
entering each sister telophase nucleus will differ in
chromatin constitution (upper Telophase, right and
left, fig. 1). The behavior of the broken end in each
daughter nucleus is similar to its behavior in the
previous mother nucleus. Following reduplication of
74 BARBARA McCLINTOCK

the chromosomes, fusion occurs between the two sis-
ter chromatids at the position of the last breakage
and a bridge configuration is produced in the suc-
ceeding anaphase which is followed by rupture and
the inclusion of a chromosome with a broken end
in each sister telophase nucleus. If this process
continued, each succeeding mitosis would possess
an anaphase bridge configuration because each pre-
ceding telophase nucleus had received a chromosome
with a broken end. The continuation of this break-
age-fusion-bridge cycle should produce chromosomes
with various deficiencies, duplications and reduplica-
tions of segments following non-median rupture of
the bridge configurations in successive anaphases.
This subsequent behavior is illustrated in Figure 1.
Thus, the production of a dicentric chromatid fol-
lowing crossing-over at a meiotic prophase may ini-
tiate a breakage-fusion-bridge cycle. It has been
demonstrated that this cycle will continue in all
subsequent gametophytic and endosperm mitoses
following its origin at a meiotic anaphase (McClin-
tock 1939, 1941a). However, this cycle will cease
whenever such a broken chromosome is delivered
to the zygote. The broken end heals. This healing
is permanent for no further fusions and breakages
will occur in the sporophytic mitoses or in any tis-
sues of succeeding plant generations. Because this
cycle occurs in the gametophytic divisions (two in
the male, three in the female) preceding the for-
mation of the zygote, a wide range of newly or-
ganized chromosomes with stable broken ends could
be recovered in the sporophytic tissues. These could
possess deficiencies of various lengths, duplica-
tions of various lengths, deficiencies plus duplica-
tions or simple or multiple duplicated segments if
the original broken chromosome possessed at least
a complete complement of genes of the chromosome.

Although plants heterozygous for an inversion
which does not include the centromere produce
chromosomes with broken ends at meiotic anaphases
following crossing-over within the inverted segment,
these inversions usually may not be used to recover
chromosomes with modified constitutions because all
of these broken chromatids are deficient for a seg-
ment of chromatin. In most cases, the deficiency in
the genomic complement of the spore receiving a
broken chromatid is sufficient to hinder the func-
tioning of the gametophyte arising from it. Con-
sequently, other structural modifications have been
used which will produce at meiosis a chromatid
with a broken end but with no deficiency of genes
of this chromosome. A spore nucleus receiving such
a broken chromosome has no deficiency in its ge-
nomic complement. A functional gametophyte could
be produced from such a spore. Two alterations
in the structural composition of chromosome .9 in
maize have been used for this study. One is a mod-
erately complex rearrangement of segments com-
posing the chromosome (fig. 2: a, normal chromo-
some 9; b, rearranged chromosome 9). In the
second case, the chromosome 9 possessed a duplica-

tion of the short arm as shown in Figure 3. In each
of these two cases, a dicentric chromatid is pro.
duced following a crossover between the norma|
chromosome 9 and the modified chromosome 9 (dq,
fig. 2; c, fig. 3). If breakage of the dicentric chroma.
tid occurred at or to the left of the arrow in the

 

4 : 7/6 9 Y a
Gide at 58 0
a
io 8 7 6 oi 3 4 5 8 v
St ore ee ee sree ee

 

eon sot 2 3 4 6 6 ? 8 ¢ © H .

1. i

Fic. 2. Diagrammatic representation of a rearrangement
in chromosome 9 which may lead to the production of a
broken chromosome with at least a full complement of
genes for this chromosome. e, a normal chromosome 9.
The centromere is represented by the clear oval. The short
arm terminates in a large knob. The linear organization of the
chromosome is represented by the dash line, the heavy line
and the lighter line aided by the numerals. By means of
X-rays, the chromosome was broken at the positions of the
arrows. Union of broken ends gave rise to the rearranged
chromosome 9 as shown in b. The synaptic association of
this chromosome and a normal chromosome 9 (with no ter-
minal knob) is shown in c. A crossover in region A will pro-
duce the dicentric chromosome shown in d. This dicentric
chromosome possesses a full complement of genes of chro-
mosome 9 from the arrow to the right end of the chromo-
some. A bridge configuration results in anaphase I. If a
break occurs at or to the left of the arrow, the broken
chromatid to the right will possess at least a full comple-
ment of genes of this chromosome. (From McClintock
1941a.)

diagrams, the broken chromatid to the right would
possess at least a complete set of genes for this
chromosome. The breakage-fusion-bridge cycle
which occurs in the subsequent gametophytic div!-
sions could produce chromosomes with various cot
stitutions by the method illustrated in Figure !-
Thus, chromosomes 9 with the various modifica:
tions described above could be delivered to thé
zygote. A wide range of structurally modified chro-
mosomes 9 have been recovered in the progeny ©
inviduals heterozygous for these two modification
(McClintock, 1941a).

It may be seen that the breakage-fusion-brids¢
cycle is a particularly favorable means of obtaininé
CHROMOSOME ALTERATIONS IN ZEA 75

 

 

esarertizaase

c .

Fic. 3. The synaptic association of a normal chromosome 9 (with a large terminal knob) and a chromosome 9 with a
‘yplication of the short arm in the inverted order (no knob present). The clear oval represents the centromere. A crossover
as indicated, following the aséociation in a, will produce the dicentric chromosome shown in c. Likewise, a crossover as
indicated, following the association shown in 6, will produce the dicentric chromosome shown in c. This dicentric chromosome
i: equivalent to two chromosomes 9 fused at the ends of their short arms. This dicentric chromosome produces a bridge
configuration at a meiotic anaphase. If the break occurs at or to the /eft of the arrow, the broken chromatid to the right will
possess at least a full complement of genes of this chromosome. (From McClintock 1941a, through the courtesy of Genetics.)

chromosomes with altered sizes and chromatin con-
-titutions. It is not understood why this cycle is con-
jined to the gametophytic and endosperm tissues in
the generation immediately following the meiotic
origin of the broken end nor why it ceases in the
sporophytic tissues and never reappears. The re-
covered broken chromosome is as permanent in its
morphology as any normal chromosome of the com-
plement. The behavior of a chromosome initially
broken in the sporophytic tissues is not the same
and will be discussed later.

In several dissimilar cases the normal process
of crossing-over has been held responsible for alter-
ations in the structural composition of the chromo-
some. Two of these will be mentioned (McClintock,
unpublished). The first case involves the nucleolus
chromosome. The appearance of the nucleolus chro-
mosome at prophase is diagrammed in Figure 4.
The nucleolus organizer, a deep staining body ad-
jacent to the nucleolus, is responsible for the or-
vanization of the nucleolus at telophase (McClin-
tock, 1934). During this process, the segment of
chromatin from the organizer to the end of the short
arm (the satellite) is removed from the main body
of the chromosome by growth of the nucleolus. Al-
though removed some distance from the organizer,
it is attached to it by a thread running through the
nucleolus substance. If at a meiotic prophase, a
chiasma forms between the centromere and the nu-
cleclus organizer and if terminalization of this chi-
asma proceeds toward the end of the arm of the
chromosome, will the terminalizing chiasma stop at
the organizer or will it pass through the nucleolus
substance to reach the end of the arm? It appar-
ently cannot pass through the nucleolus. The termin-
alization process either stops at the nucleolus or-
ganizer, or, if the force is great enough, the chroma-
tids involved are ruptured at the position of attach-
ment of the organizer to the nucleolus. Following
this rupture, fusion occurs at the position of break-
age between the nucleolus organizers of the two
chromatids involved. Consequently, a dicentric
chromosome is formed which results in a bridge con-
figuration at a meiotic anaphase. It will be noted
that the segment between the centromere and the

 

Fics. 4 anp 5

Fic. 4 (above): Diagrammatic illustration of the normal
nucleolus chromosome in maize. The large circle represents
the nucleolus. The small, clear oval (4) represents the
centromere. The large, deep-staining body (3) attached
to the nucleolus represents the nucleolus organizer. The
satellite is represented by 1. Because of the growth of
the nucleolus at telophase, the satellite is removed from
the nucleolus organizer but remains attached to it by a
thread (2) which is in or on the nucleolus itself. This con-
dition is maintained from telophase to the following late
prophase. If, in a normal plant, a chiasma forms between the
nucleolus organizer and the centromere (between 3 and 4)
and if terminalization of this chiasma proceeds toward the
end of the arm, obstruction occurs when the chiasma reaches
the nucleolus. The nucleolus organizers may be ripped from
the nucleolus resulting in breakage of the chromatids at this
position. Fusion 2-by-2 then occurs between the broken
chromatids at the position of rupture. This produces a di-
centric chromatid composed of two chromosomes 6 fused
at the distal part of their nucleolus organizers. Breakage
of this dicentric chromosome at various positions between
the centromeres during the following meiotic anaphases
produces chromosomes with various modifications in size
and chromatin content.

Fic. 5 (below): The nucleolus chromosome in a plant
homozygous for a translocation between chromosome 6 and
chromosome 5. The description of this chromosome is simi-
lar to that given in the legend of Figure 5. The translocation
occurred adjacent to the centromere on the short arm of a
normal chromosome 6 and toward the end of the long arm
of a normal chromosome 5. The centromere of the resulting
nucleolus chromosome is located a considerable distance from
the nucleolus organizer. Chiasma formation between region 3
(the nucleolus organizer) and region 4 (the centromere) is
very frequent in plants homozygous for this translocation.
Consequently, the chromatids are frequently ruptured at the
attachment of the nucleolus organizer to the nucleolus dur-
ing terminalization of these chiasmata. Dicentric chromatids
are produced following 2-by-2 fusions of ruptured nucleolus
organizers.
76 BARBARA McCLINTOCK

nucleolus organizer where an effective chiasma could
be formed is relatively short in the normal nucleo-
lus chromosome. Relatively few bridge configura-
tions following this process appear at meiotic ana-
phases. The true nature of these bridge configura-
tions was clearly revealed during a study of meiosis
in plants homozygous for a translocation which
placed the centromere at a considerable distance
from the nucleolus organizer. This translocation
chromosome is diagrammed in Figure 5. Chiasma
formation is very frequent in the long segment be-
tween the centromere and the nucleolus organizer.
Many bridge configurations arising from fusions of
ruptured nucleolus organizers were observed at me-
jiotic anaphases in these plants. Rupture of the ana-
phase bridge configurations at various positions
between the two centromeres gives rise to chromo-
somes with variously modified constitutions. As ex-
pected, they include various degrees of duplication
or deficiency. Their constitutions may be observed
readily in the prophases of the following spore divi-
sions. Thus, if chiasmata are the result of crossing-
over, as the combined evidence suggests, the normal
process of crossing-over may be a factor in the ori-
gin of modified chromosomes even when no struc-
tural rearrangements are present.

The second case of alteration in the constitution
of chromosomes for which crossing-over is held re-
sponsible is again related to chiasmata. In several
cultures of maize, it has been observed that both the
terminal and the interstitial chiasmata in all of the
chromosomes of the complement are released or un-
raveled only with considerable difficulty at the
first meiotic anaphase. Consequently, as the dis-
joining centromeres of the bivalent chromosomes

‘ pass toward opposite poles, the chromatin between
the centromeres and the chiasma is drawn out into
a very fine thread. Frequently the tension becomes
great enough to rupture these threads before the
chiasma has unraveled and a broken, deficient
chromosome enters the telophase nucleus.

From the evidence reviewed, it may be seen that
the normal process of crossing-over is a means by
which chromosomes with altered constitutions are
produced.

ALTERATIONS IN SIZE, FoRM AND CONSTITUTION OF
CHROMOSOMES FOLLOWING Non-HomoLocous
OR ILLEGITIMATE CROSSING-OVER

Synaptic associations involving non-homologous
parts of chromosomes are regularly present in maize
when the chromosome complement is heterozygous
for some structural rearrangement or when an
unbalanced chromosome complement is present.
The nature of this association has been extensively
investigated (Burnham, 1932; McClintock, 1932,
1933). The synaptic behavior of the univalent
chromosome in monosomic plants or of the extra
chromosome in trisomic plants may be used as an
example. The non-homologous synaptic associations
of the univalent at the meiotic prophase are variable

but one of the most frequent types of associations
is diagrammed in @ Figure 6. This 2-by-2 association
is completely non-homologous. Breakage of the chro.
matin threads and reunions of the broken ends fo].
lowing an illegitimate crossover at « — x’ could
give rise to a chromosome with an inverted segment
or to a deficient rod-shaped chromosome and ap
acentric ring-shaped chromosome depending upon
whether the reunion of the broken ends was diagonal]

y

Prd

3064

Fic. 6. Diagram representing the non-homologous synap-
tic association of a univalent chromosome. The clear oval
represents the centromere. The chromosome is folded upon
itself at the mid-region. The association is completely non-
homologous. If an illegitimate crossover occurs at x-x’ there
are two possible consequences depending upon the resulting
2-by-2 fusions of broken ends. These broken ends are desic-
nated 1; 2, 3 and 4 in b. The fusion of alternate broken ends.
1 with 4 and 2 with 3, will result in a chromosome with
an inverted segment. Fusions of opposite broken ends, 1 with
3 and 2 with 4, will give rise to a deficient rod-shaped
chromosome and an acentric ring-shaped chromosome. If
the breaks occurred at y-y’ in a, fusions of alternate broken
ends will give rise to a chromosome with an inversion. Fu-
sions of opposite broken ends will give rise to a deficient
ring-shaped chromosome possessing the centromere and 4
deficient, acentric rod-shaped chromosome,

or opposite, respectively (, fig. 6). If the illegiti-
mate crossover occurred at y — y’, fusions of diag-
onal broken ends would produce an inversion while
opposite fusions of broken ends would produce 4
deficient rod-shaped chromosome without a cen-
tromere. and a ring-shaped chromosome possessing
the centromere. With respect to opposite fusion.
the deficient rod chromosome could be recovered
from the x — x’ “crossover” whereas a deficient
ring-shaped chromosome could be recovered from
the y — y’ “crossover.” It is interesting to note that
just these types of modified chromosomes have ap
peared in the progeny of trisomic plants. Although
no effort has been made to obtain the frequency ©:
these events, the interpretation of their origin as #
consequence of illegitimate crossing-over betwee?
synapsed non-homologous ‘segments of chrome
somes is strengthened by the types of individual:
which are occasionally produced in the progeny ©
plants heterozygous for reciprocal translocation:
Heterozygous translocations frequently exhibit exte!”
sive non-homologous associations at meiotic pr™
phases. Secondary translocations involving the sam
CHROMOSOME ALTERATIONS IN ZEA 77

two chromosomes have been recovered from such
heterozygous plants. From knowledge of the types
of non-homologous synaptic configurations which
were known to be present in the parent plant, these
secondary translocations may readily be derived on
the hypothesis of illegitimate crossing-over.

Haploid plants are characterized by very exten-
sive non-homologous associations. This process may
be initiated in some parts of.the complement by
homologous attractions of unidentified duplicated
segments but much of the observed 2-by-2 synaptic
association is definitely non-homologous. At the
frst meiotic anaphase, several of the chromosomes
may be associated, 2-by-2, by what appears to be a
chiasma. Fragments of various sizes may likewise
be present. Both the chromosome associations and
the fragments may well arise as the consequence
of illegitimate crossing-over between associated
non-homologous segments of chromosomes al-
though legitimate crossing-over between homolo-
gously associated duplicated segments has not been
excluded. As yet, we do not know whether such
duplicated segments are present in the complement
of maize. A more detailed study of the types of
chromosome associations and aberrations at meiosis
in the haploid plants or a study of the chromosome
complements of the progeny of haploid plants could
distinguish between legitimate and illegitimate
crossing-over. The legitimate crossovers would be
expected to give the same chromosomal rearrange-
ment on a number of independent occasions. On the
other hand, illegitimate crossing-over following non-
homologous associations would not be expected to
occur at the same position on a number of inde-
pendent occasions. The progeny of haploid plants
has not been extensively investigated but it is ex-
pected that chromosomes with altered constitutions
would appear.

New Types of CHROMOSOMES ARISING FROM THE
ABERRANT BEHAVIOR OF A TELOCENTRIC
CHROMOSOME

It has been suspected for some time, on good
observational evidence, that true telocentric chromo-
somes—that is, chromosomes with strictly terminal
centromeres—normally are not present in the chro-
mosome complements of organisms. From these ob-
servations, one could conclude that some aberrant
behavior of telocentric chromosomes must result in
their elimination from the complement or that they
become modified in such a way that a true telocen-
tric condition no longer exists. In a recent study
by Rhoades (1940) the behavior of a strictly telo-
centric chromosome has been investigated. The sus-
picion, based on deductive evidence, that telocentric
chromosomes are unstable has been confirmed by
this investigation. This telocentric chromosome in-
vestigated by Rhoades was discovered in a single
plant in the progeny of an individual trisomic for
chromosome 5, It was composed of a complete short
arm of chromosome 5 with the proximal end ter-

minating in the centromere. No chromatin extended
beyond the centromere. Both genetic and cytologi-
cal evidence indicates that the mitotic behavior of
the telocentric chromosome is normal in the major-
ity of mitoses. In some mitoses, however, its be-
havior must be aberrant. Although these aberrant
mitoses have not been observed directly, they may
be inferred from the genetic behavior and the types
of altered chromosomes which are derived from the
telocentric chromosome. Plants were obtained with
two normal chromosomes 5, each carrying the re-
cessive mutant bm (brown mid-rib, located in the
short-arm adjacent to the centromere), and a telo-
centric chromosome carrying the dominant allele,
Bm. Variegation for Bm and bm appeared in some of
these plants. It was concluded that this variegation
was related to aberrant behavior of the telocentric
chromosome which either eliminated the telocentric
chromosome from some nuclei or eliminated the
Bm locus from this chromosome. In several cases, a
bm sector extended into the tassel. This allowed a
cytological determination to be made of the chromo-
some complement of such a sector. In one such
case, the observations showed that the telocentric
chromosome had been completely eliminated from
the nuclei of the sector. In four other such cases,
the original telocentric chromosome had undergone
considerable modification. In two of these cases, the
telocentric chromosome had been modified and re-
duced to a small fragment with a subterminal cen-
tromere. In a third case, a telocentric chromosome
was present but its size was only one-half that of
the parental telocentric chromosome. In the fourth
case, a minute fragment was present composed of
only two or three chromomeres and a terminal
centromere. Although the observational evidence is
insufficient to indicate the methods of origin of these
modifications of the original telocentric chromosome,
it does indicate that telocentric chromosomes are un-
stable. They may be eliminated totally from the nu-
clei or they may produce variants with decidedly
altered constitutions.

One recurring type of modification suggests the
nature of one type of instability of the telocentric
chromosome. When a plant containing this telo-
centric chromosome in addition to the normal com-
plement is crossed by or onto normal plants, three
types of plants are expected in the progeny. These
are (1) normal diploids, (2) plants trisomic for
chromosome 5 and (3) plants carrying the telo-
centric chromosome in addition to the normal chro-
mosome complement. Such plants appear in their
expected proportions when the female parent carried
the telocentric chromosome. Pollen grains carrying
either an extra chromosome 5 or the telocentric
chromosome rarely function in competition with
grains carrying a normal chromosome complement.
However, an unexpected type of plant appeared in
approximately the same relative proportions in the
progeny of these reciprocal crosses. These plants
possessed an extra chromosome. This extra chro-
78 BARBARA McCLINTOCK

mosome was composed of two short arms of chromo-
some 5 joined by a single median centromere—a
true isochromosome. It is known that pollen grains
carrying such an isochromosome in addition to the
normal complement do not function in competition
with normal grains. However, in the case mentioned,
sperm nuclei carrying isochromosomes were deliv-
ered to egg nuclei.

It is necessary to explain first the origin of this
isochromosome and secondly, how a pollen grain
may deliver such a chromosome to the egg nucleus.
If one assumed that normal reduplication of the
chromatin of the telocentric chromosome occurred
in some of the mitoses in the plant carrying the telo-
centric chromosome, which was accompanied by
some form of misdivision of the centromere of this
chromosome either in the prophase or in the subse-
quent spindle figure, an isochromosome in addition
to the normal complement could enter one telo-
phase nucleus. Under these circumstances, only the
normal chromosome complement could enter the
sister telophase nucleus. If this occurred during the
division of the microspore nucleus, a generative
nucleus carrying an isochromosome and a tube nu-
cleus carrying only the normal complement could
be produced. It is assumed that the functioning of
a pollen grain is controlled by the constitution of
its tube nucleus. Thus, a pollen grain could deliver
an isochromosome to the egg nucleus if its tube
nucleus possessed a normal chromosome comple-
ment and its sperm nuclei carried, in addition, an
isochromosome. Such misdivision of the centromere
of the telocentric chromosome may be one of the
factors responsible for the Bm—bm variegation
mentioned above.

The evidence reviewed indicates that the telo-
centric condition is another factor leading to the
production of a wide range of spontaneous chromo-
some alterations.

Tur RELATION OF CHROMOSOME ForM To CON-
STANCY OF CHROMATIN CONSTITUTION

In the previous sections it was pointed out that
permanency of the constitution of a chromosome will
not be maintained through successive nuclear cy-
cles if the chromosome possesses an unstable broken
end or if it possesses a strictly terminal centro-
mere, Extensive and varied modification in the
size and genic content of chromosomes arise from
these two conditions. There is a third condition
which leads to extensive modification of the compo-
sition of a chromosome. If a chromosome has the
form of a ring rather than a rod, it does not main-
tain itself unaltered through successive nuclear
cycles. Although its form does not change, its chro-
matin composition is continuously subject to altera-
tion. The ring chromosomes may become enlarged
by duplication and reduplication of segments com-
posing the ring or they may decrease in size by
deletions of segments from the ring. Alteration after
alteration will occur if the ring form of the chromo-

some is maintained. A plant possessing a ring-shaped
chromosome may be a complete mosaic of altered
ring-shaped chromosomes. In some cells and tis-
sues the ring chromosome may be deficient for seg-

-ments of various lengths. In other cells and tissues,

the ring chromosome may possess duplications or
reduplications of segments. In still other cells and
tissues, the ring chromosome may possess both de-
ficient and duplicated segments. The size of the ring
chromosome is no indication of its genic content.
Observations of somatic mitoses have indicated
the method by which alterations in the chromatin
constitution of the ring-shaped chromosome arise
(McClintock, 1932b, 1938a, 1941b). It is related
to the mitotic cycle. At some mitotic prophases, the
two sister halves of a divided ring-shape chromo-
some form a continuous, double-sized, dicentric ring
chromosome instead of two freely separating, mono-
centric ring chromosomes (Prophase, fig. 7). This
condition could arise subsequent to reduplication
of the chromonema of the ring chromosome if a
somatic crossover occurred between the two sister
chromatids, or the reduplication process itself could
lead to this condition. At early anaphase, the two
centromeres of the double-sized, dicentric ring chro-
mosome move toward opposite poles of the spindle
figure (Anaphase, fig. 7). Tension on the chromatin
strands between the two centromeres in late ana-
phase or early telophase causes them to rupture.
The position of rupture is variable. Three such
possible positions are indicated by the dash lines
a, b, and c, respectively (Anaphase, fig. 7). The sub-
sequent behavior of the broken strands is illustrated
in the bracketed figures (lower row, fig. 7) for each
of these breakages. Segments of the broken double-
sized ring chromosome enter each telophase nucleus
but the chromatin composition of the segment in
the sister telophase nuclei may differ considerably.
In each telophase nucleus, fusion occurs between the
broken ends of the segment thus reestablishing the
ring form of the chromosome but not necessarily
its original chromatin and genic composition (Telo-
phases, fig. 7). It may be seen that ring chromo-
somes with duplicated segments of ring chromo-
somes with deficient segments may be produced by
this process. Repetition of this process in a later
mitosis could give rise to ring chromosomes with
reduplicated or multiple segments of the original
ring chromosome, to ring chromosomes with still
greater deficiencies or to ring chromosomes with
deficiencies plus duplicated segments. The fre-
quency of occurrence of these aberrant mitoses de-
pends on the length of the chromonema of the ring
chromosome—the longer the chromonema the more
frequent the aberrant mitoses. The aberrant mitotic
configurations may occur in twenty percent of all
mitoses if the chromonema composing the ring is a5
long as the longest chromosome of the normal com-
plement. If the ring chromosome is only one-tenth
of this size, an aberrant mitosis may occur in only
one percent of the mitoses. If the ring chromosome
CHROMOSOME ALTERATIONS IN ZEA 79

i; only one twenty-fifth of the size of the longest
chromosome, an aberrant mitosis may occur in only
0.2 percent of the mitoses. In all other nuclear divi-
sions, the behavior of the ring chromosomes is nor-
mal; the two sister halves of the ring chromosome
separate freely at anaphase along with the rod
chromosomes of the complement.

 

 

segments of the dicentric ring chromosome might
likewise heal and become stable.

CONCLUSIONS

Permanency of chromosome form and constitu-
tion through successive nuclear cycles is.a basic pos-
tulate of genetic theory. This postulate is well

 

 

 

 

 

Anaphase
1 iN
2 ?
Resting Prophase
t 8 4 a 3 7
¢
2 7 2 ? a 5
a }. a
3 s a 4 s] “\I«
4 ‘5 4 5 Ps
6 ae
b anon b
7 2
Not
Late anaphase Telophase Late anaphase Late anaphase Telophase
i) 6 4 s u / 8 ,
2 7 z
2 ’ 2 7 2 ,
3 6 3
3 8 3 6 3 ‘
a s a
a s 4 $s a s ‘
4
5 4 I“
8 3
* ‘ s
‘ 4a
3 6 3
4 3 ‘ 6
7 2 7 2 7 2
T 2 | C) , 2
8 1 : , 8 a A 6 e ‘
\ _—__/ \ —/ X —a/

 

24

Y_ v

Fic. 7. Diagram illustrating a method by which a ring chromosome becomes altered in chromatin constitution. Upper left:
A ring chromosome in a resting nucleus. The clear oval represents the centromere. The individual parts of the ring chromosome
are designated by the numerals. Upper middle: A prophase configuration following a “crossover” between the two sister
chromatids of the divided ring chromosome. A dicentric, double-sized ring chromosome is produced. Upper right : Appearance
of the dicentric ring chromosome in the following anaphase. Breakage of the chromatin strands between the centromeres
may-occur at any position. Three possible positions a, 6 and c, respectively, are indicated by the dash lines. The resulting
broken strands at late anaphase and the new ring chromosomes formed at telophase by fusions of broken ends of these
strands are diagrammed below in the bracketed figures for the breaks a, b and c, respectively. (From McClintock 1941b,

through the courtesy of Genetics.)

It should be emphasized that fusion of broken
ends apparently always follows the breakage of a
double-sized, dicentric ring chromosome during an
aberrant mitosis in the sporophytic tissues. Although
extensively looked for, no cases have been found
where the broken ends had failed to unite. Since
it has been proved (McClintock, 1941a) that a
single broken end which is unstable in the gameto-
phytic tissues may heal and become permanently
stable in the following zygote or early sporophyte,
\t is expected that under certain conditions which
at present are not known, the broken ends of the

founded on extensive observational evidence in a
wide range of organisms. Knowledge of conditions
which will produce changes in this constancy has
been of utmost importance in recent years. X-rays,
ultraviolet radiation, heat, aging, etc., have been
the usual agents producing these desired conditions.
It has been known for a long time that other con-
ditions may lead to changes in the form and con-
stitution of chromosomes. When the previous his-
tory is not known, the observed changes would nat-
urally fall under the heading of spontaneous aber-
rations because the conditions responsible for their
80 BARBARA McCLINTOCK

occurrence were not apparent. In this discussion,
I have attempted to indicate the extent of our knowl-
edge in maize of the conditions which are responsible
for such “spontaneous” aberrations. We do know
that they may occur (1) under genic control, (2)
following legitimate crossing-over, (3) following il-
legitimate crossing-over, (4) during or following
reduplication of the chromonema of the chromo-
some, (5) as the consequence of the instability of
broken ends of chromosomes, (6) during terminal-
ization of chiasmata and (7) as the result of the
aberrant behavior of a strictly terminal centro-
mere.

Studies of spontaneous aberrations in maize have
contributed and should continue to contribute to
our knowledge of the behavior of chromosomes in
general. A few such contributions may be sum-
marized. We know from the study of ring-shaped
chromosomes that the reduplication process of a
chromosome usually occurs along a single plane.
The possibility that some sister strand crossover
chromatids may be present at meiosis is likewise
suggested by these studies for the double-sized di-
centric ring chromosomes represent some form of
sister chromatid exchange. Although the method of
origin of these exchanges is not known, they may
be the result of some process which is shared by
all chromosomes. If so, the frequency of sister-
strand crossover chromatids at meiosis should be di-
rectly proportional to the length of the chromosome.
However, the process which gives rise to these
strands need not be related to the normal process of
crossing-over. We are considerably better informed
about the stability of broken ends of chromosomes
through studies of the types of modified chromo-
somes which arise following mechanical rupture of
chromosomes at meiosis, and following mechanical
rupture of dicentric ring-shaped chromosomes at a
somatic mitosis. With regard to this, we know that
following mechanical rupture of two adjacent chro-
matids at a late meiotic prophase, fusion may occur
between these two chromatids at the position of
breakage. However, if these two adjacent chroma-
tids are broken at the first meiotic anaphase, fu-
sions will now occur at the position of breakage
between the two sister halves of each of these chro-
matids. Likewise, if a single chromatid is ruptured
at a meiotic anaphase, fusion will occur at the posi-
tion of breakage between the two sister halves of
this chromatid and thus initiate the breakage-fusion-
bridge cycle which characterizes the behavior of this
chromosome in subsequent gametophytic and en-
dosperm tissues. We know also that fusions may
occur between two broken ends of a single chromo-
some if this single chromosome has suffered me-
chanical rupture at two points during anaphase of a
mitotic division in the sporophytic tissues. From
these studies we have determined that a recently
broken end of a chromosome is unstable in certain
tissues and under certain conditions but may be-
come completely and permanently stable under

other conditions. Because of these observations,
we know why ring-shaped chromosomes and telo-
centric chromosomes cannot maintain themselves
in nature and thus why they are not frequently en.
countered. Through the study of the composition
of altered chromosomes arising from haploids we
may be able to detect the presence of possible dup.
licated segments in the normal complement of maize
which at present are undetected but suspected from
genetic evidence. Even the concepts of chiasmata
formation and terminalization may be illuminated
through the studies of specific chromosome altera-
tions.

It is obvious that the changes in size, form and
constitution of chromosomes have made it possible
to detect and study some of the processes underly-
ing chromosome behavior in general. .

REFERENCES

Beapre, G. W., 1932, Z. i. A. V. 63 2195-217,
1937, Cytologia, Fujii Jubilee Vol :43-56.

Burwuam, C. R., 1932, Proc. 6th Intern Congr. Genetics 2:
19-20.

Cuark, F. J., and Corezanp, F. C., 1940, Amer. J. Bot.
27 :247-251.

DaruincTon, C. D., and Urcotr, M. B., 1941, J. Genet. 41:
297-338.

Gitxs, N., 1940, Genetics 25 :69-87.

Jones, D. F., 1937, Genetics 22 :484-522.
1940, Amer. J. Bot. 27:149-155.

McCuintock, B., 1932, Proc. Nat. Acad. Sci. 18 :677-681.
1933, Z. Zellf. 19:191-237.
1934, Z. Zellf. 21:294-328.
1938a, Genetics 23 :315-376.
1938b, Missouri Agri. Exp. Sta. Bull. 290:1-48.
1939, Proc. Nat. Acad. Sci. 25 :405-416.
1941a, Genetics 26:234-282.
1941b, Genetics 26 :542-571.

Nicos, C., 1941, Genetics 26: 89-100.

Ruoapes, M. M., 1940, Genetics 25 :483-521.

DISCUSSION

Mutter: What is the rate of apparent “gene mu-
tation” in the sticky stock?

McCiintock: It is very high, although the homo-
zygotes are very low in fertility. The progeny may
also possess chromosomal rearrangements.

Gatrs: How high is the pollen sterility?

McCurntock: Very high. Many plants shed no
polien at all.

BocueE: Would incomplete division be an alterna-
tive to the refusion of broken ends?

McCuintock: We cannot distinguish by ou!
method between incomplete division or fusion fol-
lowing division. All we know is that the two siste?
chromatids are fused at the position of previous
breakage.

Mickey: Do bridge chromosomes break with
greater frequency at one position more often than
at any other position? /

McCuintock: Yes, they tend to break at the pos!
tion of previous fusions.
CHROMOSOME ALTERATIONS IN ZEA 81

Husxins: If this point were near the middle,
the effect could be mechanical.

McCuintock: Yes, but the break would not be
expected to be exactly in the middle, although ap-
proaching it.

Fano: Is crossing over involved in the behavior
of the ring chromosomes?

McCuintock: I think somatic crossing over be-
tween sister chromatids may be responsible for the
double-sized, dicentric ring chromosomes.

Mutter: If a plant is heterozygous for the ring,
can such crossing over occur within the ring?

McCLINTOCK: Yes.

ScHuLtTz: Possibly evidence might be obtained
from this for sister-strand crossing over at meiosis
in maize, which would probably then be more fre-
quent than in Drosophila.

McCuintock: The behavior of ring chromosomes
suggests some form of sister-strand exchange and
also suggests that the frequency with which it oc-
curs would depend on the length of the chromonema
composing the chromosome. The longer the chro-
monema, the more frequent the expected occurrence.

NEBEL: Changes in size might also be due to a
twist in the plane of reduplication within the ring.

McCuintock: Ves, it is possible. However, the
evidence suggests that this must be infrequent.

Mickey: Will a gene such as sticky increase the
frequency of double sized rings?

McCuintock: There is no evidence for this in our
experiments, but no actual measurements of the
frequency in different strains have been made.

SaNsoME: Your ring chromosomes came from
X-ray treatment, didn’t they?

McCuintock: Not in all cases. Ring chromo-
somes may arise following X-radiation but also
spontaneously. ‘

DeELBRicK: Do sister-strands heal following
breakage of a dicentric ring chromosome?

McCurntock: I have not found such a case. Such
healing would give rise to a rod-shaped fragment.
They have been looked for but have not been found.
Since broken ends are known to heal under certain
as yet unknown circumstances such rod-shaped frag-
ments arising from broken ring chromosomes even-
tually may be found.

SPARROW: If we assume that the ring chromosome
is longitudinally bipartite and that it breaks at ana-
phase, sister-strand crossing over as shown in the
diagram is not necessary to explain double or inter-
locked rings.

McCuintocx: The diagram represents the sim-
plest interpretation for this presentation.

MuLieER and Grass (simultaneously): Do you
get interlocked rings?

McCurntock: Because of the small size of the
mitotic chromosomes in maize, interlocking of two
sister ring chromatids would be difficult to detect.
At anaphase they would look like a twisted double
sized ring; this configuration is frequent but one
cannot be sure that it represents interlocked ring
chromatids. Interlocked ring chromosomes occur
following another type of behavior of ring.

CarLson: When there is a break followed by refu-
sion between the nucleolus organizer regions, where
does the nucleolus form in the daughter cells with
reference to the nucleolus organizer?

McCuintock: Development of the nucleolus
takes place at the position of maximum activity of
the two fused organizers, which is at the position
of fusion of the two organizers.

WarmxkeE: Concerning the mechanism of bridge
formation involving the nucleolus chromosome, if
terminalization occurs late, after the nucleolus has
disappeared, then does the nucleolus organizer itself
inhibit terminalization and thus cause the break?

McCrrntock: Probably not. The breakage is as-
sumed to occur before the nucleolus disappears. The
frequency of bridges is low in proportion to the
chiasmata frequency or crossing over within this
region.

WarMKE: In Datura, crossing over apparently
regularly occurs between the centromere and nu-
cleolar organizer, and terminalization takes place
with the formation of what have been called humps;
chiasmata form without breaks in this case.

McCurock: In maize, when the distance be-
tween the nucleolus organizer and centromere is
short, there are few bridges. As this distance in-
creases, the frequency of the bridges increases. An
increase in the distance is obtained from homo-
zygous translocations.