SIA
"
N,
Mr.
2
.
MULAN
UNCLASSIFIED
NA
.
-Y
ORNL
18
654
.
.
..
'.
.
-
:-
!
.
1
isi
)
.
.
.
.
.
.
1
L
!
.
.
1
>
.
..
i
1
1
.
.
.
.
1*;"
21
2.4
.
.
ARA
IN
.
V
ly
YA
S
.
.
.
I.
1.
.
W
.
-
"
.
.
Wi.
(2
OP
121
.
71
21
SEK
SIA
INC
.
7
W
s
.
NI
.
NEL
W XV
TA
ru
.
2.
"
7
12
1
w
.
!
1
.1.-..
.
*
IN .
SIR,
VY
X2
7
7
wa
TER
12
'
.
(.
.
-
!
"
NY
744
.
ORNO-p-6054
0.11-2
-3
NOV 1 3 1964
Chromosome Pairing, Crossing Over and
Disjunction in Drosophila melanogaster*
-
Rhoda F. Grell
Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee
---LEGAL NOTICE
The report mo prepared a an mirom i of Carennual sponsored work. Neldber the United
Non, nor the Comwonlon, nor any person acting on behall of the Councilou:
A. Makes my miranty or repromalama, exprent or implied, su roeperi to the acry.
racy, completenes, or w hom of haformation contained us to report, or what there we
ol way talonulon, manau, mohon or prono deloord la we report way sot talringe
print owned right ar
D. Aonmi w Innleko w roapect to the of, or for damagn reiwure from the
We of war information, aurats, unlod, or proctus dixcloud in Weroport,
Ao out in the aboro, "person why ou bowall of use Coa msalon" Includes way na
ployee or contructor of the Commission, or uplore of such coatraclor, to the erunt that
Such anployna or contractor of the Conamlaslon, or saplosae of much coutractor prepara,
# Atalay, or provides a to, wy taformation purnaal to No toplojacal or contract
will the court calom, or wo enploymnl vid much contractor.

*Research sponsored by the U.S. Atomic Energy Commission under contract
with the Union Carbide Corporation.
Running head:
Send proof to: Dr. Rhoda F. Grell
Biology Division
Oak Ridge National Laboratory
P. 0. Dox y
Oak Ridge, Tennessee 37831
The study of chromosome behavior during me10818 no longer
ocoupies the paramount position that it did in the early days of
genetic investigation. This does not mean that all or even most
or the problems of chromosome transmission have been solved. Among
many organisms, customarily used for genetic studies, such as
Neurospora, maize and the mouse, transmission genetics 18 in a
fairly elementary state. Even Drosophila, where studies have been
most extensively pursued and where the mechanisms were considered
to be fairly well understood, has proven to be a still fruitrul
organism for investigation. Here, some amazing findings have recently
come to light which have led to a radical revision in our concepts of
the meiotic mechanism. It 18 these findings and these changes that
I will talk about today.
Classical Models of Meiosis. We will start out with the pre-
mise that the basic processes that we are dealing with during meiosis
are chromosome pairing, chromosome replication, chromosome exchange
and chromosome disjunction. Our need is to fit these events into a
causal relationship that will be consistent with our fund of genetic,
biochemical, and cytological data. Superficially, this would appear
to be a simple problem because the possibilities are few and because
they must be restricted to accomodate the following facts: First,
that some type of pairing must precede crossing over; secondly, that
crossing over in the four strand stage means chromosome replication
must also precede or be concurrent with exchange; third, that
crossing over must occur before disjunction; and fourth, that
disjunction itself !mplies a previous association of chromosomes.
The scheme that was enthusiastically adopted by most geneticists
in the thirties was Darlington's "precocity theory" of me 10818 (1)
which took as its basic tenet that homologs pair at zygotine because
they enter prophase precociously in an unreplicated state in contrast
to mitotic chromosomes which are visibly double at the beginning of
prophase. The cytological observations of split chromosomes before
zygotine in a number of organisms as well as the recent finding that
DNA replication occurs before zygotine (2, 3, 4) no longer permits
serious consideration of this hypothesis. From the purely genetic
viewpoint, we are concerned with that aspect of Darlington's theory
which holds that the presence of a chiasma in a prerequisite for the
regular segregation of homologs. According to this model, the causal
sequence of events is chromosome pairing, followed by chromosome
exchange leading to regular disjunction. Darlington's scheme assumes
that chiasmata are generally formed in all paired chromosomes at
pachytine and that failure to form a chiasma results in a univalent
at metaphase and some nondisjunction or loss at anaphase (1).
From the time of its inception, it was apparent that this model
was incompatible with certain genetic data. Even if we ignore the
male of Drosophila, where regular segregation occurs in the absence
of exchange, and limit our considerations to the female, we find
that here too, exchange is not a prerequisite for regular disjunction.
Evidence for this comes from the small fourth chromosomes which are
invariably nonorossovers in the displaced female but which segregate
in a highly regular fashion; from the larger x chromosomes whion by
tetrad analysis are found to be noncrossovers ca. f'ive percent of
the time but which nondisjoin with only about one hundredth of this
f'requenoy, or 0.05 percent (5); from those cases where noncrossover
X tetrads occur with exceedingly high frequencies due to the presence
of heterozygous X inversions, yet where nondisjunction remains less
than one percent (6); and from studies with the mutant CIIIG which
practically eliminates exchange but where chromosome assortment shows
a marked departure from randomness (7). As emphasized by a number
of workers (8, 9), these observations indicate that something other
than exchange 18 capable of governing or modifying disjunctive behavior.
As an alternative to Darlington's "chiasma hypothesis of meta-
phase pairing", Dobzhansky, on the basis of the behavior of trans-
location heterozygotes, formulated the hypothesis of competitive
pairing (8). According to this view, crossing over and disjunction
are negatively correlated. The relation, however, is not a direct
causal one in the sense that crossing over inevitably leads to regular
disjunction and failure of crossing over leads to random assortment.
Rather, Dobzhansky assumed that both processes are predetermined by
the intimacy of synopsis between specific homologous loci, prior to
crossing over. Rearrangements provoke a conflict between the attraction
forces, weaken the intimacy, as well as decrease the frequency of
.
.
'
.-
7
.
:
;-
.
..
MO
synops18 between homologs and lead to both decreased crossing over
and increased nondisjunction. If intimacy of pairing is a more
critical requirement for the disjunctive process than 1t 18 for
exchange, this hypothesis could account for the high frequencies
of nond 18 junction in translocation heterozygotes despite almost
normal cross over values. From inversion heterozygotes, however,
where crossing over 18 known to be greatly disturbed, disjunction
1s expected to be even more irregular. Instead we find 1t 18 highly
regular and it becomes clear that the solution to the problem must
be sought elsewhere.
The Phenomenon of Nonhomologous Pairing
This is the way matters stood until the mid-fifties. At this
time a remarkable new phenomenon was discovered in the Drosophila
female, which, as it turned out, held the key to many of the principles
underlying the behavior of chromosomes during oogenesis. It was found
that under the proper conditions nonhomologous chromosomes would
segregate from one another with great regularity, and it has been
assumed here that, as with homologs, regular segregation is preceded
by some type of regular pairing.
That associations between nonhomologs might be occuring was first
postulated by Cooper and his co-workers (10) in 1955 to explain the
increase both in x nondisjunction and in dominant lethality where
heterozygous X and outosomal inversions were simultaneously present
in the genome. Following this, in 1956, Sandler and Novitski (11)
proposed that pairing between nonhomologs could account for the
failure to recover the different classes of progeny from triploids
in the frequencies expected from independent assortment. The
following year we (12) were able to provide unequivocal genetic
evidence for associations between nonhomologs by studying the off-
spring of females carrying on extra y chromosome and a single free
fourth chromosome. It was found that in over eighty percent of the
progeny, if the Y was present, the four was not or conversely if the
four was present, the Y was not. More important, we could establish
that the frequency of eggs carrying the maternal free four and no Y
was equivalent to the frequency of eggs carrying a Y and no maternal
free hour (12, 13). In other words, these two kinds of eggs represente
the two reciprocal products resulting from the pairing and subsequent
segregation of the two nonhomologs, the Y and the fourth chromosomes.
Clearly, this constituted a complete contradiction to the law of
Independent assortment of nonhomologous chromosomes.
Since then, the combined work of a number of geneticists has
disclosed not only that all of the chromosomes of the Drosophila
genome are capable of participating in nonhomologous associations
but that such associations can occur with the utmost regularity
reaching in one case a frequency of 99.98 percent (14). Table 1 pro-
vides an indication of the generality and the frequencies of this
phenomenon.
:
..
-
.
:
i
..
While it was of great interest to discover that chromosomes
could behave in this unorthodox manner, at the same time this
phenomenon presented a serious enigma. All pertinent evidence
indicates that crossing over is an extremely precise process,
calling for exact pairing at the nucleotide level to insure that
the DNA template is correctly conserved. Furthermore, it is well
known that nonhomologs do not undergo exchange. How then can the
prerequisite for completely specific pairing before exchange be
reconciled with high frequencies of nonhomologous pairing at this
time. Even if it were conceded that chromosomes might conceivably
resort to illegitimate pairing in the absence of their homolog, as,
for instance, in haploid strains of plants, the paradox is heightened
in those cases where both homologs are present, although one may
carry a partially inverted sequence, yet a nonhomolog, possessing
an entirely dissimilar nucleotide sequence, 1s apparently preferred.
Sandler and Novitski (11) had proposed, as a possible solution,
that all of the chromosomes of the Drosophila genomi possess homology
in common in their proximal heterochromatic regions. Viewed in this
way, nonhomologous pairing becomes entirely a matter of proximal
heterochromatic pairing. We were able to test this hypothesis by
studying the associations among three chromosomes, two of which
carried the same heterochromatic base but enchromatic segments of
different origin and length while the third chromosome possessed no
known homology for either of the other two (18). If non homologous
pairing is entirely a proximal heterochromatic affair, associations
should occur at least as frequently between the two elements with
identical bases as between either of these and the third chromosome.
Instead, it was found that the two chromosomes with identical bases
assorted randomly and that associations occurred exclusively between
one of these two and the third chromosome. Apparently then, factors
other than proximal heterochromatin, are involved in nonhomologous
a8.sociations. Results, to be described later, have now made it
abundantly clear that proximal heterochromatin, per se, is inconsequen
tial in this process.
Tests of the Single Pairing Model
1. Nonhomologous Pairing
As alternatives to the Sandler-Novitski hypothesis, it may be
assumed that either nonhomologous pairing occurs along the entire
length of the chromosome or that it occurs solely in the eudiromatic
consequences are inevitable. First, chromosomes that have been
associated nonhomologously must be noncrossovers, and secondly,
crossing over between homologs must be reduced to the extent that
nonhomologous associations occur. Both of these predictions may be
tested genetically and such a test was carried out (19).
For this test females were constructed that carried a particular
chromosome for which a choice in pairing existed. It might pair and
perhaps cross over with its homolog or it might pair with a non-
homolog. Preliminary studies had already shown that in such females,
-
'
7-
..
.
1
.
this chromosome paired with the nonhomolog about fifty percent
of the time. Sister females, identical in every respect to the
test females except that they lacked the nonhomolog, were used as
controls. On the single pairing, 1.e., classical, hypothesis,
nonhomologous associations occuring ca, fifty percent of the time
must lead to a reduction in crossing over of ca. fifty percent.
Moreover, those chromosomes engaging in the nonhomologous associations
are expected to be noncrossovers. Analysis of the results revealed
that those chromosomes that segregated from the nonhomolog, were
indeed noncrossovers, but contrary to prediction, crossing over
was the same whether the nonhomolog was absent or present. The
test was repeated using two other genetic situations but the results
were the same in each case. The data tell us, therefore: 1. That
exchange between homologs is not altered by the presence of a non-
homolog; 2. Only those chromosomes that have not participated in
exchange with a homolog are available for associations with a non-
homolog. Thus, it becomes clear that nonhomologous associations do
not produce noncrossovers but rather that noncrossovers associate
nonhomologously after exchange has occurred.
2. Secondary Nondisjunction
A situation closely analogous to nonhomologous pairing is that
which gives rise to secondary nondisjunction. This process has been
extensively studied over the years, beginning with the classical
work of Bridges (20) but it has remained something of an enigma,
-
10
leading Stiertevant and Beakle (21) to remark that "the general
problem of the mechanism of secondary nondisjunction is unsolved."
Secondary nondisjunction may be defined as the increased frequency
of X nondisjunction that occurs in females bearing a Y chromosome.
The consequences of X nondisjunction are recognizable among her
progeny as either females possessing both maternal X's or males
possessing neither and both types are collectively referred to as
secondary exceptions. Since the nondisjoining X's in contrast to
the regularly disjoining X's are noncrossovers, Bridges postulated
that the difference between the two types was initiated as a stage
preceding crossing over and that the exceptions arose from those
cases in which a competition for: pouring partners among the two
exceptions would result when the single x happened to pass to the
same pole as the segregating X.
This scheme introduces the same problem encountered with non-
homologous pairing, namely, if pairing for exchange is based on
homology, and there is every reason to believe it 18, why should
an X find the Y, with which it has only one known locus in common,
more attractive than its own homolog? Fortunately, the Bridges'
model is subject to the same kind of test that was applied to
nonhomologous pairing. Bridges' model likewise assumes a single
pre-exchange pairing, so that the incidence of XY bivalents must
CAL
"
?
i
.
!
!
1
7
1.!
!
decrease in X exchange. Bridges had, in fact, looked for such a
decrease in crossing over and had failed to find it. Since the
genetic backgrounds of the two types of females (XX and XXY) 'that
he used were uncontrolled, his test could not be considered a
critical one.
In our tests, sister females were utilized, half of which were
XX in genotype and the other half XxY, and the frequencies of
X-nondisjunction and X-crossing over were measured for the two
groups. Studies were made on females that carried two isosequential
X's as well as on females that carried a variety of different hetero-
zygous X inversions (5). The results demonstrated that in all cases
the presence of a Y in the mother causes an approximately 50 fold
increase in X nondisjunction and, as 18 well recognized, these
secondary exceptions are noncrossovers. Crossing over studies,
with two normal x's however, showed that a y caused an increase.
rather than the predicted decrease in crossing over. With inversion
heterozygotes the Y caused localized effects on crossing over that
could be traced to the position of the inversion. In some cases
these effects resulted in an overall increase, in others a decrease,
but in the latter case the alteration in crossing over was never
as great as that predicted from the Bridges model.
That the Y was not producing additional noncrossovers was
most clearly shown with normal x's where we could perform a tetrad
analysis to actually determine the number of noncrossover x bivalerts.

.
.
.
.
.
·
12
1.
id
7
7
-
These results are given in Table 2. In series i, the frequency of
Eo's, 1.6., noncrossover X's, 18 the same for the xx and XXX sisters;
in series 2, the Eo's are actually rewer in the presence of the Y,
although probably not significantly so. This is so despite a.fifty
rold increase in x nondisjunction with the y. The overall increase
in X exchange observed with a Y 18 attributable to the increase in
higher rank tetrads, principally Eg's at the expense of Ey's. These
resulta demonstrate that the effect of the Y on X disjunction 18
independent of its effect on X exchange. Furthermore, it is evident
that the role of the Y 18 not to increase X E.'s by pairing com-
petitively with the X's before exchange but instead to pair with
X Eo's after exchange and in this way cause them to nondisjoin.
".
The Distributive Pairing Model of Meiosis
These studies have led to only one harmonious interpretation
of meiotic event in Drosophila females. Nonhomologous associations
do not occur prior to exchange. Rather, they ooour after exchange
and exclusively between chromosomes that have failed to undergo
exchange. On the basis of this evidence, the model shown in Text-
Figure 1 has been proposed. Chromosomes that pos8e88 homology for
another chromosome pair before exchange. This has been called
exchange pairing. If A chromosome has no homology for another,
such as perhaps a Y deleted of its short arm, 1t will not become
involved in exoħange pairing. Following such pairing, exchange
may not ooour. When exchange does ocour, s ohiasma ties the
participating chromosomes together and this tie precludes their
subsequent involvement with other chromosomes so that they remain
associated during the following otages and segregate regularly from
one another. If exchange rails to occur, such chromosomes form a
pool or nonexchange chromosomes called the distributive pool.
Within this pool, a second type of pairing, which we call distributive
pairing, now takes place. Distributive pairing may occur between
two homologs, in which case segregation 18 again regular. Alter-
natively, it may occur between two nonhomologs and when this happens,
the subsequent segregation of the nonhomologs will lead to their
non-randon assortment and part of the time will result in nondisjunction
of homologs and in aneuploid gomites.
A clarification of what is meant by a chromosome in the distri.
butive sense is in order. Just as it is normally defined, a chro-
mosome 18 considered to po88e88 a centromere to which some segment
of chromatin 18 attached. Its form may be metacentric as with the
large autosomes, J-shaped aa with the Y or acrocentric as with the X.
It may be an element deleted of most of its chromatic material such
as a small X-duplication, a duplicated element such as a compound X
which carries two X's attached to one centrome or an element composed
of portions of different chromosomes such as a part of a reciprocal
translocation. All of these types function as a single chromosome.
In the case of the compound chromosome, it has been shown by Ed
Giell (17) that crossing over between its two homologous arms does
not alter its runational rolc as a single chromosome 80 that a
compound X, whether a orossover or a nonorossover within itself, 18
re
5
..
.
14
*
'
-
'-'
.
equally available for distributive pairing.
With this model of meiosis it is now possible to resolve
those problems that appeared insoluble with the single pairing
schemes. First, we see that Darlington's hypothesis that crossing
over leads to regular segregation 18 correct becaus e crossover
bivalents do remain associated, do not become involved with non-
homologs, and segregate regularly at anaphase: Does lack of crossing
over necessarily lead to nondisjunction as Darlington assumed? We
see it need not, at least in Drosophila, ror chromosomes that have
failed to cross over are afforded a second opportunity for associating
with one another at distributive pairing. It is this latter type of
pairing that is responsible for the regular segregation of X-inversion
heterozygotes despite the fact that the majority may be noncrossovers.
Their. segregation is regular, however, only so long as a third
chromosome is not present in the distributive pool. Should a Y,
for instance, be added, association between the X's and Y will result
in high frequencies of X nondisjunction, 1.e., the process described
as secondary nondisjunction. The more efficient the inversion system
in producing nonorossover X-tetrads, the more frequently the X's will
be members of the pool and the more secondary exceptions will occur.
Thus it becomes clear why the Y associates only with noncrossover
X's but does not increase their number. Finally we see that the
requirement for precise preexchange pairing 18 maintained since
associations between nonhomologs occur only after exchange at the
distributive phase.
.
-
-
Characterization of Exchange and Distributive Pairing
Exchange pairing, 18 defined as occuring exclusively between
specific homologous loci which on the molecular level is equivalent
to saying pairing between identical (or complimentary) sequences
of nucleotides. If more than two such loci are present, exchange
pairing 18 competitive since pairing at any particular level 18
assumed to be limited to two chromosomes. Asymmetrical pairing and
exchange, leading to duplicated and deleted chromosome segments, as
described for the zesti and white regions (22,23) are interpreted
to be a consequence of duplicated segments which still possess sequence:
of nucleotides homologous to those of the original segment. This
means, as has been pointed out by Green (24), the term "nonhomologous
pairing" and "nonhomologous exchange" are inappropriate to describe
these processes.
..
-
.
While exchange pairing is characterized as being extremely
precise and limited to homologous regions, distributive pairing,
by contrast, may occur as regularly between nonhomologs as between
homologs. When more than two chromosomes are present in the dis-
tributive pool, it has been shown that pairing is competitive and
that marked preferences are displayed (18). Preference at distributive.
pairing 18 apparently not attributable to homology, since studies
referred to earlier, with three nonrecombinant chromosomes, had
shown that those two that segregated regularly had no known homology
'
1',
'
.

.
.
!
.'
"
.
1.
1.5
.
.
!
,
!!
in common, whereas the two with the same basal heterochromatin,
assorted randomly. The question to be answered, of course, is,
what factor or factors, are responsible for preferences in dis-
tributive pairing? In the experiment just referred to, one obvious
possibility was chromosome size, since the two chromosomes that
segregated regularly were large whereas the third chromosome that
assorted randomly was small. The series of investigations that
will now be described examined the role of size in distributive
pairing (14).
The first sets of experiments studied the effect of size on
the frequency of distributive pairing in a noncompetitive situation
between two chromosomes, one of which was kept constant while the
other was varied from much smaller to much larger than the first.
For the variable element, a series of free X-chromosome duplications
were used that had been obtained by Cooper and Krivshenki and measured
by these workers at mitotic metaphase, using the length of the fourth
chromosome as a standard. With respect to chromosome four, arbi-
tranly given a length of one, the duplications vary from <0.3/(Dp1187)
to 3.3 (Dp 1498) times the length of the four as shown in Text Figure
2. The largest duplication, No. 3, which is greater than four times
the length of the four, was obtained by Lindsley and Sandler (25).
"Each X-duplication was introduced into females of identical genotype.
that carried a single free fourth chromosome which served as the
nonvariable element. In this situation, both the free X-duplication
and the single free four are virtually always noncrossovers and are
the only two members of the distributive pool. Asmeasure of the
frequency of pairing between them can be obtained either from the
frequency that they segregate from each other or from the frequency
that they nondisjoin. Since the frequency of nondisjunction 18 a
and segregate, the lower the nondisjunction frequency, the greater
the frequency of pairing. Table 3 gives the frequencies of non-
disjunction between each duplication and the four. The highest non-
disjunction frequency, 4.5 percent, occurs with the smallest dupli-
cation. As the duplications become larger and approach the four in
size, nondisjunction decreases meaning that association increases. .
The lowest nondisjunction, and hence the highest association, occurs
with those duplications closest to the four in length. For Dp 1144
with a length of 1.1, 1t is 0.02% or one nondisjunctional product in
5000 flies. Then, as the duplications become larger than the four,
nondisjunction again increases, levelling off at 2% for duplications
2.5 or larger. It is evident that the amount of pairing between two
nonhomologs, in a noncompetitive situation, 18 a function of their
similarity in size.
Competitive System
Turning now to a study of the competitive situation, the problem was
approached in two ways. The first was to examine the effect of a competitor
noncompetitive conditions. For this it was only necessary to add a third
.
.
...
...
.
.
.
!
!
11'
9.
'
.
Pitiri

*
...
..
ti
.
*-
.
.
.
A
:
.
0
.
."
'
'Y
:
1
'
'
'
' '
i
.'
:
Th
!-
HY
-
.
1
*
L.
.
.
M
.
-
.
14
•
chromosome to the distributive pool. The chromosome that was used, which we
call I, 18 five to six times the length of the four at mitotic metaphase and
has been estimated to be a pool member ca. 96 percent of the time. Again,
nondistinction frequencies between the different X-duplications and chromosome
four were measured. The results are shown in Text-Figure 3, where the upper
curve gives the values when the competitor is present and the lower curve, the
values when the competitor is absent. For duplications close to the four in
length, i.e. between 0.7 and 1.6, the presence of the competitor has little
effect on duplication, four nondisjunction. Por duplications smaller than
0.7 and larger than 1.6, the presence of the competitor has a marked effect
as evidenced by the sharp rise in duplication, pou nondis junction beyond these
points. The effect of the competitor depends, therefore, on the similarity in
Bize of the other two chromosomes.
The second approach permitted a highly critical test of the effect of
size on preference in distributive in distributive pairing. As indicated
above, the starting point for these experiments had been the finding that two
large chromosomes, with no known homology, segregated regularly from each
other, while a small chromosome, with partial homology for one, assorted,'.
randomly with both (18). This situation is shown in Text-Figure 4 by the
rightmost points on curves A, B,' and C. The two percent value for curve A
represents nondisjunction frequency for the two large chromosomes, T, and a
marked Y; the values of close to fifty percent for curves B and C represent
nondin junction frequencies for the small four and the Y and for the four and
T. In other words the large Y and large 7, segregate regularly walle the
small tow assorts randomly with both. ... :
L
.
.
.
:.
.
.
2
..
.
e
If pairing preference is a matter of size, it should be possible, by
...
..
.
.......
preference. Thus, by substituting for the Y, X-duplications of smaller and
smaller size, the point should eventually be reached where the X-duplication
and the four now segregate regularly and the large Ps, assorts randomly. This
point 1sexpected to occur when the X-duplication is very close to that of
the four. By following curves A and B, we see that this is precisely what
happens. As X-duplications of smaller and smaller size are substituted for
the X, nondis junction between the X-duplications and the shown in curve A,
increases. At the same time, nondisjunction between the X-duplication and
the four, shown in curve B, decreases. The point is finally reached, when
the X-duplication (Dp 1144--length 1.1) and four are very similar in size,
that they associate exclusively and the 2 shows fifty percent nondisjunction
(random assortment) with both. Actually, this situation resembles the
original one for in both cases the two chromosomes of like size segregate
regularly and the third chromosome, of unlike size, assorts randomly. These
results demonstrate that preferences, as well as frequencies, in distributive.
pairing are size dependent.
It must be strongly emphasized that pairing behavior at this time 18
neither a property of the heterochromatic nor euchromatic portions of the
chromosome, per se. The X-duplications are predominantly heterochromatic,
the T, 18 predominantly euchromatic while the four has its normal amounts
of both. Yet the pairing behavior. of the three chromosomes correlates
closely with their entire metaphase Je ngth and is independent of the amount
of euchromatin and heterochromatin they contain..

w
.
.
.
.
.
.
?
.
.
.
*
-
.
1
.
.
1
1
IK.
.
27
.
..
11
....
.
.. 19
*
-
A
*
'
.
*
A
.
1:
47
1
.1
.
I
.
.
.
.
.
1.'-
-
.
Fourth Chromosome Segregation
With the finding that distributive pairing is size-dependent, it became
possible to elucidate the mechanism responsible for the regular segregation
of the fourth chromosomes (26). According to the distributive pairing model,
the two fours, as noncrossovers, should invariably be members of the distributive
pool and available for nonhomologous associations. When other chromosomes
.
.
.
.
.
.
1.
.
:
:
..:
.
associations with the fours. As one explanation, it might be postulated that
a special mechanism had been evolved to ensure regular fourth chromosome
segregation despite the absence of crossing over and that, as a consequence,
the fours, like crossover bivalents, do not enter the pool. The discovery
that distributive pairing 18 bize dependent suggested, as an alternative
explanation, that the fours, although pool members, are too small to become
Involved with other chromosomes of the genome .
:.To test whether the Powrs are pool members, it was proposed to introduce
the small X-duplications into the pool and to determine whether the fours are
now susceptible to involvement with a nonhomolog. If 80, fourth chromosome
nondis junction should occur.
The results are shown in Text-Figure 5. Nondisjunction of the fours,
which is 0.2 percent with no duplication, steadily increases from a value
Identical to the control, 1.e. 0.2%, for the smallest duplication, to a
maximum of 36.6 percent for a duplication very close to the four in length
(Dp 1244-length 1.1). Since at this point nondisjunction of the duplication
and one of the cows 16 only 26.4 percent, segregation of the nonhomologs
16 vore regular than segregation of the homologo, then as the duplications
1
.
.
become larger than the four, four nondis junction steadily declines, reaching
a value not significantly different from the control (0.3 percent) with the
largest chromosome used, the T20
It is evident from these results that the fourth chromosomes are
always members of the distributive pool and will associate nonhomologously 11
a heterolog of the proper size is present. It follows from this that the
mechanism responsible for regular fourth chromosome segregation, in the absence
of such a heterolog, is distributive pairing. Furthermore, these results demonstrate
again that the critical factor in distributive pairing is size and that, unlike
N
i
pairing for exchange, homology is apparently inconsequential.
tect
!
'.
-
'
-
"
The Time of Exchange and Distributive Pairing. It is appropriate at this
point to examine the relationship between the cytologically characterized
meiotic stages and the events under discussion. This task would be much
simpler 1f the time of crossing over were known. Localization of the exchange
went at two different points in the meiotic cycle has led to two alternative
.
interpretations.
The first, or classical interpretation, assumes that cytologically
visible pairing between homologs at pochytene represents the pairing that
precedes exchange. Since homologous contact presumably occurs along the
entire length of the chromosome, the limiting factor is the exchange event,
whose probability for any given interval 1s considered to be low. With DNA
replication occurring well before pochytene, copy-choice as the sole mechanism
for exchange 18 excluded by tł18 model. If this version is correct,
exchange pairing would occur at the Bygotene stage, exchange during pachytene
and distributive paiting anytime subsequent to exchange and preceding first
anaphase. The speculation (27) that nonexchange univalents might undergo

equational division at first anaphase and pair distributively in the
Becondary obcyte cannot be excluded.
The second model, that of Pritchard (28), assumes that two types of
pairing occur during meiosis. The first 18 restricted to short discrite
regions called "erfective pairing" sites and, here, while the probability
of establishing such a site is low, the probability of recombination within
such a site is high. Recombination 18 thought to probably occur by a copy-choice
mechanism at the time of DNA replication. The second type of pairing, that
observed at pachytene, is considered to be a devise concerned sobly with
segregation.
Pritchard's evidence for "effective pairing" sites comes from studies
with Aspergillus (29, 30) where recombination data of two kinds, first
localized negative interference and secondly the increase in exchange
observed In the presence of a triplicated chromosomal segment, support
this notion. When the same lands of data are examined in Drosophila, little
if any aupport for localized negative interference 16 found. Moreover, early
studies of Khoades (31) and Dobzhansky (32) have shown that triplicated
segments, such as Pritchard used, decrease rather than increase recombination
values. Recently, Da Grell (33) completed a series of experiments in which
a triplication (duplication in Probophala terminology) for the base of x
existed either free in the genome or attached at different locations. In
all cases, ; & marked reduction in exchange in the triplicated segment was
observed, but the extent of reduction depended on the proximity of the
triplication to its homologous regions. If the triplication is attached to
the distal tip of X or inserted in chromosome 3, the reduction 16 only
thirty percent; 11 lt 18 free, the reduction is about seventy five percent;
but if it is attached to the base of X; very close to its normal location,
a ninety five percent reduction occurs. Mearly, the availability of an
additional site hinders rather than increases recombination.
Interestingly enough, in Drosophila, the triplication itselt rarely
participates in exchange, whereas in Aspergillus, Pritchard reports that
the triplicated segment 18 as frequently involved in exchange as the two
normal homologs. In Aspergillus, as in Drosophila, the intervals proximal
to the triplicated segemtn, in both the normal chromosomes and in the
nonhomologous chromosomes carrying the triplicated segment, show a marked
reduction in exchange, a result hardly in accord with the "effective pairing"
site potion. In line with this, the occurrence of positive interference
over long intervals in Drosophila, as well as most organisms, becomes very
alfficult to account for with small discreet, peiring sites.
Although Pritchard's model finds little support in Drosophila prepachytene.
exchange remains an attractive hypothesis, An alternative explanation may
be considered. In addition to the Diptera, where homologs are expected
to lle in close proximity at the preméiotic telophase because of somatic
pairing, there exists a variety of observations in the cytological literature
indicating that pairing of homologs in the conventional side by side manner
is initiated at preme.lotic anaphase or telophase (34). Chromosome condensation
followed by parasynapsiis of homologs soon after nuclear fusion has also been
discribed for Neurospora by McClintock (35) and Singleton (36). If this should
turn out to be the general rule, then the paired, condition necessary for
exchange, as well as one which could account for positive interference, has
been established well before moiotio prophase commondes.
Recently, in collaboration with Ann Chandley, who has been visiting
In our laboratory, I have been attempting to obtain more precise information
concerning the time of crossing over in Drosophila. We have used heat .
treatment of 3Tc for a twenty-four hour period so as to increase the
frequency of crossing over. We have brooded daily, thereafter, for fifteen
days In order to determine the number of days that elapse between the
Initiation of the treatment and deposition of those eggs from which progeny
showing an increased frequency of crossing over aribe.
In conjunction with the genetic studies, we have injected females with
tritiated thymidine and prepared daily samples of the ovaries in order to
follow the progression in labeling from the earliest oocyte, where the last
DNA replication occurs to the nature oocyte. In this way it 18 possible to
determine the time interval between the last DNA synthesis and egg laying. .
Our findings are shown in Text-mgure 6. Those eggs laid seven to
eight days after initiation of treatment produce the first progeny to show
a significant increase in exchange. The increase reaches a maximw.of about ·
two and one hall times control value on day nine and then falls reaching
control values again on day . Interestingly, when we study the labeling
sequence, we find label appeare in mature oocytes for the first time on the
eighth day after injection. The coincidence between the times of DNA
replication and the alteration in crossing over by heat means, unless the
heat effect 18 a delayed one, that crossing over 18 occurring concurrently
or very close to the time of DNA replication, i.e., at the earliest phase
of the stage 1 oocyte.
.-
-
.
If oxchange pairing and exchange are proloptoteno evento, associations
commonood at zygotine might correspond to distributiva pairing. A diffioulty
with this intorprotation has been that the excellent correlation between
mitotic metaphase length and begregation behavior suggests distributive ,
pairing occurs lato in prophase or at metaphase when chromosomes are well
condensed rather than at zygotone or pachytene. This difficulty could now
be resolved by the recent findings of Moens (37), who has shown for the
tomato at least, that the traditionally accepted sequence of meiotic prophase
stages has probably been misinterpreted. He finds in the anthers, where an
age gradient exists, that following interphase, chromosomes first become
visible as paired pachytene bivalents, that homologs then begin to separate
from each other corresponding to what had previously been thought to be
synapsis of chromosomes at zygotene, now renamed schizonema, that separation
of homologs continves so that the nucleus becomes filled with a network of
threads, and that following this diffuse stage chromosomes proceed through
diplotene, diakinesis and metaphase.
The events that have been postulated from genetic evidence fit this
altered sequence rather well. Pachytene might then represent a continuation
of exchange pairing, initiated earlier at premeiotic anaphase or telophase,
schizonema a subsequent stage when chromosomes fall apart unless tied
together by a chiasma, and dialdnesis or metaphase the time of distributive
pairing between condensed chromosomes.
Summary
The following points briefly recpaitulate recent findings concerning
Drosophila oogenesis:
1. Monrandom asortment of nonhomologous chromosomes has boon
demonstrated for all of the chromosomes of the Drosophila
genome. The equality of reciprocal classos from such
wysortmento indicate that they are the result of meiotio
segregations in which associations between nonhomologs
precedes segregation of nonhomologs to opposite poles .
Segregation of nonhomologs from one another may be as
regular or more regular than segregation of homologa.
2. Crossing-over studies have shown that chromosomes participating
in nonhomologous associations are always noncrossovers. They
have also shown that the incidence of such events, involving
one of a pair of homologe, does not affect the frequency of
exchange between the two homologs. These two rindings have
led to a new model for me 10818. This model includes two types
of pairing, an early type, concerned with exchange, called
exchange pairing, and a later type, concerned with segregation,
called distributive pairing.
3. Investigations into the nature of distributive pairing have
disclosed that it is a size-dependent process such that both
pairing frequencies and pairing preferences displayed at this
time correlate extremely well with similarity in mitotic
metaphase lengths of the participating chromosomes. In this
respect it differs from exchange pairing where the prime concern
18 homology. The dependence on total lengths excludes the
possibility that associations at the distributive phase between
nouhomologs 19 a consequence of shared homology in proximal
...
.
heterochromatic regions.
4. Distributive pairing has been shown to be the mechanism
responsible for regular segregation of the fourth chromosomes
under normal conditions.
5. The distributive pairing theory provides a satisfactory
explanation for a mimher or heretofore unexplained situations
including regular segregation without exchange, nonhomologous
pairing, the mechanism ox' secondary nondisjunction and
preferential segregation of the fourth chromosomes .
replication is presented. It 18 suggested that parasynapsis
of homologs, prior to premeiotic interphase, followed by
exchange concurrent or close to the time of DNA replication
and distributive pairing at late prophuse or metaphase,
presently provides the most satisfactory correspondence
between genetic events and the cytological picture.
28
LITERATURE CITED
(1). Darlington, O. Dos Recent Advances in Cytology. 559 + 18 pp.
Philadelphia: Blaklston.
(2). Swift, H.: The desoxyribose nucleic acid content of animal nuclei. '
Phys. 2001. 23: 163-198, 1950.
(3). Swift, H. and R. Kleinfeld: DNA in grasshopper spermatogenesis,
oogenesis, and cleavage. Physiol. 2002. 26: 30.1-311, 1953.
(4). Taylor, J. H.: Autoradiographic detection of Incorporation of pe
into chromosomes during meiosis and mitosis. Expt1. Coll Research
4: 164-173, 1953
(5). Grell, R. F.: A new model for secondary nondisjunction: the role of
distributive pairing. Genetics 47: 1737-1754, 1962.
(6). Sturtevant, A. H. and G. W. Beadle: The relations of inversions in
the X chromosome of Drosophila melanogaster to crossing over and
disjunction. Genetics 21: 554-604, 1936.
(7). Gowen, J. W.: Meiosis as a genetic character in Drosophila
... melanogaster. J. Exp. Zool. 65: 83-106, 1933.
(8). Dobzhansky, Th.: The decrease of crossing over observed in
translocations, and its probable explanation. Amer. Nat. 65:
214-232, 1931.
(9). Cooper, K. W.: Normal segregation without chiasmata in female
Drosophila melanogaster. Genetics 30: 472-484, 1945.
(10). Cooper, K. W., 8. Zimmering and J. Krivshenko: Interchromosomal
effects and segregation. Proc. Natl. Acad. Sci. V.8. 42: 922-914,
1955.
...
.
.
.
;
-
.
.
.
-
(21). Sandler, L. and E. Novitold: Evidence for genetic homology between
chromosomes I and IV in Drosophila melanogaster, with a proposed
explanation for the crowding effect in triploids. Genetics 42:
189-193, 1956.
(12). Grell, R. pNonrandom assortment of nonhomologous chromosomes •
Genetics 42: 374, 1957.
(13). Grell, R. 70: Nonrandom assortment of nonhomologous chromosomes in
Drosophlla melanogaster. Genetics 44: 421-435, 1959.
(14). Grell, R. p.: Chromosome size at distributive pairing in Drosophila
melanogaster temales. Genetics, 50: 151-166, 1964.
(15). Oksala, T.: Chromosome pairing, crossing over, and segregation in
meiosis in Drosophila melanogaster females. Symposium on Quantitative
Biology, 23: 197-210, 1958.
(16). Miller, B. and R. F. Grell: Nonrandom assortment of chromosome 3
and e. V. Dros. Inform. Serv. 38: 65-66, 1963.
(17). Grell, E. H.: Distributive pairing of compound chromosomes in females
of Drosophlla melanogaster. Genetics, 48: 1217-1229, 1963.
-
-
-
-
chromosomal elements involved in nonrandom assortment in Drosophila
melanogaster. Proc. Natl. Acad. Sci. U.S. 46: 51-57, 1960.
(19). Grell, R. P.: A new hypothesis on the nature and sequence of meiotic
events in the female of Drosophila melanogaster. Proc. Natl. Acad.
Sci., V.8. 48: 165-172, 1962.
(20). Bridges, C. B.: Non-disjunction as proof of the chromosome theory of
heredity. Genetics, 1: 1-52, 107-163, 1916.
-.
.
.
.
.
vi
(21). Sturtovant, A. H. and Q. W. Beadle: An Introduction to Genetics,
W. B. Saunders Co., Philadelphia and London, 1939.
(22). Green, M. Mo: Non-homologous pairing and crossing over in Drosophila
melanogaster. Genetics 44: 1243-1256, 1959.
(23). Judd, B. H.: Pormation of duplication-deficiency products by
asymmetrical exchange within a complex locus of Drosophila melanogaster.
Proc. Natl. Acad. Sci. U.S. 47: 545-550, 1961.
(24). Green, M. M.; further data on non-homologous pairing and crossing over
in Drosophila melanogaster. Genet:.cs 46: 1555-1560, 1961.
(25). Lindsley, D. L. and L. Sandler: The meiotic behavior of grossly
deleted X chromosomes in Drosophila melanogaster. Genetics, 43: .
547-563, 1958.
(26). Grell, R. P.: Distributive pairing: the size-depenáent mechanism
responsible for the regular segregation of the fourth chromosoñes in
Drosophila melanogaster. Proc. Natl. Acad. Sci. U.S., 52: 227-232,
1964.
(27). Schwartz, Drew: Personal communication.
(28). Pritchard, R. H.: The linear arrangement of a series of alleles in
Aspergillus nidulans. Heredity, 9: 343-371, 1955.
(29). Pritchard, R. H.: Localized negative interference and its bearing
on models of gene recombination. Genet. Res. 1: 1-24, 1959.
(30). Pritchard, R. H.: The bearing of recombination analysis at high
resolution on genetic fine structure in Aspergillus nidulans and the
mechanism of recombination in higher organisms. In "Microbial Genetics"
Tenth Symposium of the Society for General Microbiology held at The
Royal Institution, London, April 1960.
31
(31). Rhoades, M. M.: A new type of translocation in Drosophila melanogaster.
Genetics, 16: 490-499; 1932.
(32). Dobzhansky, Th.: Studies on ohromosome conjugation. III. Behavior
of duplicating fragments. 2. Indukt. Abstamm.-4. Verorblehre 68: 134-162,
1934.
(33). Grell, E. H.; Influence of the location of a chromosome duplication
on crossing over in Drosophila melanogaster. Genetics 50: 251-252, 1964.
(34). Smith, s. Q.; Polarization and Progression in pairing II. Premeiotic
orientation and the initiation of pairing. Canad. J. Research 20:
221-229, 1942
(35). McClintock, B.; Neurospora. I. Preliminary observations of the
chromosomes of Neurospora crassa. Amer. J. Bot. 32: 671-678 (1945).
(36). Singleton, J.: Chromosome morphology and the chromosome cycle in the
ascus of Neurospora crassa. Amer. J. Bot. 40: 124-144, 1953.
(37). Moens, Peter B.: A new interpretation of meiotic prophase in
Lycopersicon esculentum (tomato). Chromosoma 15: 231-242, 1964.
2
.
:-
*
Table 1.
Frequencies of nonrandom assortment of
dirferont pairs of nonhomologo in Drosophila
melanogaster.

Genotype
Chromosomes
Involved
Segregation
Reference
Y; T(3;4)86D
T(384)860
YA4
92
Grell, R.F.(13)
Y; Ins(2L+2R) Cy
T(2; 3)rn
90.1
It' s
Ye3
Bu
Y; Ins (LR)
T(2;3)A
Oksala, T. (15)
Miller, Band
R. F. Grell (16)
Grell, R. F.(14)
XDP; T(3;4)86D
TR0)86D
X DPA
4
99.9
XX2
Grell, E.H.(17)
Y; 14
Y
44
98.8
Grell, E.H.(17)
Grell, E. H.(17)
95
T(2;3)A
XX; 14.
XXki14
9
Grell, E. H.(17)
-
-
.
4
.
TABLE 2
Tetrad analysis of crossing over in normal X chromosome 8*
Experiment
b
e
by
by
Distance
tal
Measured
% Total
Crossing Over
67.24 € 0.95
70.09 € 1.00
Sxx
4.73
4.98
57.62
51.77
36.00
41.34
1.64
1.91
Distal tip (BC)
to centromere
xx
LXXY
5.23
3.67
62.91
60.73
30.94
34.61
0.91
0.98
Distal tip (y)
to car
63.76 + 1.14
66.44 $ 0.95
*E.
E
Eng
Ex
No-exchange tetrad;
Single-exchange tetrad;
Double-exchange tetrad;
Triple-exchange tetrad.
TABLE 3
Noncompetitive System
Nondisjunction frequencies between chromosome four
and X-duplications of variable lengths
Duplication
Mitotic
Totals
no.
length
Dp, 4 Nondisjunction
among X-regulars*
(%)
1187
1162
5.0.3
0.5
0.7
816
1204
0.9
1193
1.0
1339
1.2
1144
1.1
1.4
1337
1186
1346
12,113
8,000
9,324
7,391
5,371
7,126
8,476
5,635
5,991
8,012
2,789
2,665
3,343
3,732
2,767
344
4.49 0.19
0.39 + 0.07
0.22 + 0.05
0.08 + 0.03
0.13 10.05
0.08 $ 0.03
0.02 + 0.02
0.38 $ 0.08
0.39 + 0.08
0.42 0.07
0.43 $ 0.12
1.94 $ 0.27
2.20 + 0.25
1.90 6.0.24
1.63.4 0.24
2.00 $ 0.75
1.6
2.0
1328
2.1
1488
2.5
2.6
1343
856
1498
3.0
3.3
4
" X nondisjunction less than 2% in all cases .
.
.
.
W
:

..
.
.
Text-Figure 2
....
12.741
..
...
NATURE OF X-DUPLICATIONS
.
.
-
-
CHROMOSOME
ORIGIN
_
_X-LOCI PRESENT
yt oc* sc* su-wºt dort pnt su-p+ 00*
-
...
-
-
-
-
-.
LENGTH.
AT MITOTIC
METAPHASE
1.0
< 0.3
0.5
0.7
0.9
SC8
SC 8
C-S
T-
SC 8
$C8
Chromosome 4
X-00 # 1187
X-Op # 1162
X-Op # 816
X-Dp # 1204
X-Dp # 1193
X-Op # 1339
X-Dp # 1144
X-Op # 1337
X-Op # 1186
X-Op # 1346
X-Op # 1328
X-Dp # 1488
X-Op # 1343
X-Dp # 856
X-Dp # 1498
X-Op # 3
C-S
C-S
C-S
anno
1.4
+
++
++
++
+
+ + + + + + + + + + + + + + + +
+ + + + + + + 1 + 1 +11+!!
+
+
+
+
+
+
+
+
II 1 + 11 + + + i + 1
+!!!!!!!!!!!!!!!
SC8
1.6
+ + + + + + IIII!!!!!
2.1
C-S
C-s
C-S onunda
onomo
C-S ann
onimo
C-S
C-S
CS anumanmunaa
2.5
2.6
-
-
-
t't
. .
3.0
3.3
.
.
- .
-
-
.
..
-
-
-
-
(Reprinted groni Genelics 50:151-166, 1964)
-

Text - Figure 3. Comparison of De, 4 nondisjunction in competitive
and noncompetitive system's.
.
12,142
1
is
.

..
YUVALU
COMPETITIVE
% Dp, 4 NONDISJUNCTION
NONCOMPETITIVE
J
.-.-.-
0.2
0.3
0.5 0.7 0.9 1.4 1.4 1.6 2.0 2.5 3.0 4.0 4+
DUPLICATION LENGTH
.
NIN
.
... Hanggang
.
(Reprinted from Genetics 50; 151 - 166, 1964)
-.-
.
.-.....
...
.
.
.....
..
.
..
www

Text- Figure 4. Relation of nondisjunction frequencies to duplication tength.
.
.
.
A.
12,745

.
. .
.
.
Tool
... .
...
RANDOM ASSORTMENT
% NONDISJUNCTION
& Dp.4
0.5
07 09 1.1 1.4
2.0 2.5 3.0 4.0
60
PASTA
DUPLICATION LENGTH
(Reprinted from Genetics 50; 151-166, 1964) Eropa
owner op prwt von
.
.
SEMILOGARITHMIC
358.61
KEUFFEL & ESSER CO. MADE IN 5.3.A.
2 CYCLES X 70 DIVISIONS
to) lenglia
Text - Figure 5. Fourth Chromoscnie nondisjunction plotted as a function of duplication and
(A control value of 0.2% was obtained with no duplications

1
clann ITSOFemenu
.
.
i
I.
L
I
.
"
S
+
-
III
rd
-
--
1
en i
xC
II
,ve19-
-
-
-
-
n.
-
---
I----
S-
---
!
CK
--- -
--
---
-.
_
------
1
-
-
11
-
-
-
--
6.9 10 11
12
14
INN
3. 6 33
Tila
2.5
Sa
6.6
Size
0,3
X-duplications, 4 and 74
-
..
Sy
.
.
t
.
.
.
.
DATE FILMED
12/ 30 /64






Yr
.
..
- LEGAL NOTICE

This report was prepared as an account of Government sponsored work. Neither the United
States, nor the Commission, nor any person acting on behalf of the Commission:
A. Makes any warranty or representation, expressed or implied, with respect to the accu-
racy, completeness, or usefulness of the information contained in this report, or that the use
of any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rights; or
B. Assumes any liabilities with respect to the use of, or for damages resulting from the
use of any information, apparatus, method, or process disclosed in this report.
As used in the above, "person acting on behalf of the Commission" includes any em-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such employee or contractor of the Commission, or employee of such contractor prepares,
disseminates, or provides access to, any information pursuant to his employment or contract
with the Commission, or his employment with such contractor.
MA
?
.
.
.
.
SA
.
2
P
are
NX

EUS
RAIN
SA
.
END


.
.
Mr