Review
Homologous pairing and chromosome dynamics in meiosis and mitosis
Bruce D. McKee*
Department of Biochemistry and Cellular and Molecular Biology and Genome Sciences and Technology Program,
University of Tennessee, Knoxville, M407 Walters Life Sciences Building, Knoxville, TN 37996-0840, USA
Received 23 September 2003; received in revised form 18 November 2003
Abstract
Pairing of homologous chromosomes is an essential feature of meiosis, acting to promote high levels of recombination and to ensure
segregation of homologs. However, homologous pairing also occurs in somatic cells, most regularly in Dipterans such as Drosophila, but
also to a lesser extent in other organisms, and it is not known how mitotic and meiotic pairing relate to each other. In this article, I summarize
results of recent molecular studies of pairing in both mitosis and meiosis, focusing especially on studies using fluorescent in situ
hybridization (FISH) and GFP-tagging of single loci, which have allowed investigators to assay the pairing status of chromosomes directly.
These approaches have permitted the demonstration that pairing occurs throughout the cell cycle in mitotic cells in Drosophila, and that the
transition from mitotic to meiotic pairing in spermatogenesis is accompanied by a dramatic increase in pairing frequency. Similar approaches
in mammals, plants and fungi have established that with few exceptions, chromosomes enter meiosis unpaired and that chromosome
movements involving the telomeric, and sometimes centromeric, regions often precede the onset of meiotic pairing. The possible roles of
proteins involved in homologous recombination, synapsis and sister chromatid cohesion in homolog pairing are discussed with an emphasis
on those for which mutant phenotypes have permitted an assessment of effects on homolog pairing. Finally, I consider the question of the
distribution and identity of chromosomal pairing sites, using recent data to evaluate possible relationships between pairing sites and other
chromosomal sites, such as centromeres, telomeres, promoters and heterochromatin. I cite evidence that may point to a relationship between
matrix attachment sites and homologous pairing sites.
D 2004 Elsevier B.V. All rights reserved.
Keywords: Chromosome pairing; Recombination; Cohesion; Meiosis; Mitosis; Nuclear matrix
1. Introduction
Pairing of homologous chromosomes is a fundamental
event in meiosis, where it is normally accompanied by high
levels of genetic recombination and results in the segrega-
tion of homologs into separate cells. However, homolog
pairing can also occur in a variety of other contexts. Meiosis
in some organisms proceeds without recombination but the
homologs nevertheless pair and segregate with great regu-
larity. Moreover, pairing also occurs in vegetative, somatic
and germ-line mitotic cells of Dipteran insects. In these
cases, the function of pairing is less clear as homologs do
not segregate; various investigators have suggested a role in
preparing for meiosis, promoting DSB repair, or facilitating
interactions between homologous transcriptional regulatory
sequences. Sister chromatids also associate with one anoth-
er, usually quite closely, and these cohesive associations are
essential for chromosome segregation in both mitosis and
meiosis, as well as for DNA repair.
An important set of unanswered questions concerns
whether these various meiotic and non-meiotic pairing
processes are mechanistically similar, despite their often
diverse morphological appearances. Is there a single under-
lying mechanism for pairing of homologous loci? Does the
presence or absence of recombination in meiosis imply
fundamentally different pairing processes, or is recombina-
tion a separate process that can be associated with pairing in
certain pathways? Are there common mechanisms for link-
ing sister chromatids and chromosomes in various segrega-
tion pathways? These and similar questions have been posed
numerous times, and although there is a voluminous de-
scriptive literature on pairing and related processes, the
underlying mechanisms of homolog pairing have remained
elusive. Progress has been hampered until recently by the
inability to monitor pairing directly, so that most analyses
have of necessity relied on indirect assays of downstream
0167-4781/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.bbaexp.2003.11.017
* Tel.: +1-865-974-5148; fax: +1-865-974-6306.
E-mail address: bdmckee@utk.edu (B.D. McKee).
www.bba-direct.com
Biochimica et Biophysica Acta 1677 (2004) 165­180
events such as homolog segregation, recombination, synap-
tonemal complex (SC) formation, or transvection. While in
principle, abnormalities in these downstream processes
could reflect failure of homologs to pair, in practice it has
not been possible in most situations to determine whether
this is the case. However, the development of molecular
assays for pairing based on hybridization of fluorescently
tagged DNA probes (FISH) and, more recently, GFP tag-
ging of single loci have begun to circumvent this limitation.
Studies based on these approaches, combined with a rapidly
growing arsenal of mutations and other genetic and molec-
ular reagents, have yielded a great deal of new information
about pairing in both mitotic and meiotic contexts and how
it relates to other processes such as recombination, synapsis
and gene expression.
In this article I will review recent molecular studies of
pairing and attempt to integrate the results with data from
classical cytological and genetic approaches. My primary
focus will be on homolog pairing in Drosophila, taking
advantage of the presence in this organism of an unusual
wealth of pairing pathways, including both recombinational
and non-recombinational meiotic pathways in females and
males, respectively, and ``somatic'' pairing in virtually all
cell types. However, I will compare the Drosophila results
with relevant data from other organisms, in an attempt to
achieve as broad an overview as possible. The major
purpose will be to explore similarities and differences in
the dynamics of pairing and the distribution of pairing
ability in different cellular contexts, in the hope of identi-
fying key conserved features of chromosome pairing and
gain clues to the underlying mechanism.
2. Homolog pairing in meiosis
The early stages of meiotic prophase in most organ-
isms involve a gradual alignment of homologous axes and
their subsequent incorporation into an elaborate meiosis-
specific structure known as the synaptonemal complex
(SC), which holds the homologs in close register as the
events of recombination are being completed [1­3]. With
the development of FISH methods for determining the
locations of homologous loci, it has become possible to
track the progress of pairing in early stages of prophase.
In most organisms, homologs are unaligned at leptotene,
when chromosomes first begin condensing, but achieve
high levels of pairing in a relatively short period of time,
coincident with the achievement of full synapsis at
pachytene. However, as described below, there are inter-
esting exceptions in which premeiotic pairing is thought
to be an important prelude to meiotic pairing [4,5]. The
SC is removed at mid-prophase and homologs from that
point until anaphase I are connected only at discrete sites
known as chiasmata, which are thought to be products of
crossing over. There are numerous excellent reviews on
synapsis (the process of forming SC), meiotic recombi-
nation, and chiasmata, to which the reader is referred for
details [2,3,6­10].
Detailed and elegant studies in yeast have elucidated
many of the details of meiotic recombination and synapsis
and have shown those processes to be closely intertwined
both temporally and mechanistically (reviewed in Ref. [7]).
The early events of recombination, especially the formation
and processing of meiotic double-strand breaks, which are
induced by the Spo11 protein, coincide temporally with the
initiation of synapsis, and mutations that disrupt those
events, such as mutations in Spo11, also prevent synapsis
[11­13]. Recent results have indicated that the basic pro-
cesses of meiosis are conserved but have also pointed to
substantial variation among organisms in the relationship
between recombination and synapsis. Mammals and plants
seem to follow the pattern in yeast [14,15], but in Drosoph-
ila and C. elegans synapsis can occur in the absence of
meiotic double-strand breaks, which seems to indicate that
homolog pairing is independent of recombination [16,17].
Indeed, it has long been known that homolog pairing and
segregation can occur independently of recombination and
chiasmata [18]. Drosophila males provide the best-studied
example of ``achiasmatic meiosis'', and the only case in
which crossing over has been demonstrated to be absent
throughout the genome, but it is thought to be universal in
Dipteran males and Lepidopteran females and has been
described in numerous other taxa. Synaptonemal complex
can be present, as in Lepidopteran females where it persists
in modified form until anaphase, providing a chiasma
substitute, but is absent in Drosophila males [1]. Until
recently, little has been known about pairing patterns in
achiasmatic meiosis, but the recent application of molecular
methods has pointed to some unexpected and striking
parallels between recombinational and non-recombinational
meiosis.
3. Mitotic and meiotic homolog pairing in Drosophila
3.1. Mitotic pairing
Mitotic chromosomes in somatic and germ-line cells in
Drosophila and other Dipterans have been known for almost
a century to exhibit high levels of homolog pairing [19].
Homolog pairing is also the rule in polytene chromosomes
which are arrested in an interphase-like state. However, it
was not known until the development of FISH methodology
whether homologs are also paired during interphase in
mitotically cycling and other non-polytene cells. Fluores-
cent hybridization probes for the repetitive euchromatic
histones locus, which contains 100­150 copies of a 5-kb
repeat, were first used to address whether and when mitotic
pairing occurs in cleavage-stage nuclei of Drosophila em-
bryos, using a relatively non-disruptive whole-mount pro-
tocol [20]. Two separate signals, indicative of an unpaired
condition, were seen in virtually all nuclei in the first 11
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180166
embryonic divisions. These divisions last only 10­17 min
each and consist entirely of alternating S and M phases.
However, single signals, indicative of pairing, were seen at
low levels in cycle 12, which is slightly longer than the
previous cycles, and then in the majority of cells in cycle 13,
which is the first division in which there are significant gap
phases. Subsequent FISH studies involving probes from
different chromosomal regions have established that high
levels of pairing occur at interphase throughout the chro-
mosomes in all stages of development after the early
embryonic cleavage stages, and in germ line as well as
somatic cells [21­24].
A GFP tagging assay developed in yeast has recently
been applied in Drosophila, and the first results provide
strong confirmation for the FISH studies. In this method, a
multimeric array of bacterial LacO repeats is inserted, via P-
element transformation, at various chromosomal sites,
which are then tagged by expression of a GFP-LacI chime-
ric protein [25,26]. In the initial study, 14 different euchro-
matic lacO insertions exhibited comparable frequencies of
pairing, around 50%, in spermatogonia [27].
The cell cycle dynamics of mitotic pairing have not been
very thoroughly investigated, but the general picture is that
pairing is seen throughout most of the cell cycle. One FISH
study reported a partial loss of pairing (from 80% to 60%) at
S phase in mitotic larval brain nuclei at two loci. In this
case, maximal pairing levels were not restored until G1 of
the following cycle [22]. In another report, pairing frequen-
cies at the histones locus were observed to be much lower in
early G1 of embryonic cycle 14 than in late G2 of the
preceding cell cycle, suggesting that pairing is disrupted
during mitosis and then reestablished quickly during G1 of
the next cell cycle. Interestingly, no loss of pairing was
observed in prophase or metaphase, indicating that the
disruption of pairing must occur at anaphase [21]. However,
no disruption of pairing was detected in the GFP-tagging
study of pairing in Drosophila spermatogonia [27]. Classi-
cal studies have documented the occurrence of anaphase
pairing in somatic nuclei (e.g., Ref. [28]), but no reliable
estimates of frequency have been available. Clearly, addi-
tional studies of mitotic pairing in the context of the cell
cycle need to be carried out.
Another important unresolved question is whether pair-
ing within a cell cycle (prior to anaphase) is reversible. In
the Fung et al. [21] study, pairing frequencies were tracked
from 2 h after egg deposition to 5 days, and found to
increase throughout this period to frequencies in the 90­
100% range. This was interpreted to indicate that pairing is
irreversible. In support of this interpretation, it was found
that the mean distance between unpaired loci increased over
time, rather than decreasing as would be expected if many
unpaired loci at later time points had recently been paired.
However, other studies have detected maximum pairing
frequencies below 100% at multiple loci, even in mitotically
arrested cells, which would appear to more consistent with a
scenario in which pairing and unpairing are in dynamic
equilibrium [22­24,27]. However, a plausible alternative
might be that pairing can occur only at the beginning of the
cell cycle, and is permanent until anaphase. Resolution of
this question will require tracking of single loci over time in
living cells, which can now be done using the GFP-tagging
system.
Somatic/mitotic pairing in Drosophila has been consid-
ered of significance because of the evidence for a variety of
genetic phenomena that are thought to depend upon homo-
log pairing [29­31]. One of these is ``transvection'' which
refers to situations in which mutant alleles of a gene
complement one another, but this complementation is dis-
rupted by heterozygosity for chromosome rearrangements
with breakpoints on the same chromosome arm as the locus.
Transvecting alleles are typically mutations in sequences
that would normally be expected to be cis-acting, such as
enhancers and promoters, suggesting that these sequences
can interact in trans in certain circumstances, but only when
the loci are well-paired [32]. An apparently related phe-
nomenon is pairing-dependent silencing of transgenic re-
porter genes. In the most common examples, the reporter is
silenced only when the transgene is homozygous. Two
nonhomologous insertions can usually interact only weakly
if at all, and in some cases, heterozygosity for a rearrange-
ment on the same chromosome arm has been shown to
disrupt the silencing interaction [29,31]. However, these
phenomena are seen only in special mutant or rearranged
genotypes and thus do not provide much insight into how
pairing benefits wild-type organisms.
3.2. Premeiotic and meiotic pairing in Drosophila
It has been proposed by several investigators that mitotic
pairing in the germ line of Drosophila and other Dipterans is
a prelude to and important for the establishment of homolog
pairing in meiosis (e.g., Ref. [33]). Both premeiotic oogo-
nial and spermatogonial nuclei of Drosophila were shown
by classical staining methods to exhibit homolog pairing in
mitotic prophase, metaphase and anaphase, leading to the
idea that the chromosomes enter meiosis fully aligned
[33,34]. Mosquitoes also exhibit strong homolog pairing
at mitotic metaphase and anaphase in spermatogonia
[35,36], and this pairing has been reported to continue
uninterrupted from premeiotic cells through pachytene
[37]. However, until the development of molecular tagging
methods, it was impossible to determine the arrangement of
homologous loci in the interphase period immediately
preceding meiotic prophase, or to track the changes that
occur in chromosome organization and pairing during
prophase.
Compelling evidence for premeiotic pairing in Drosoph-
ila was recently obtained from observations of the pairing
behavior of GFP-tagged loci in cysts of spermatogonia and
young spermatocytes [27]. Meiosis in Drosophila males
takes place in cysts of 16 spermatocytes that are
interconnected by cytoplasmic ring canals and that arise
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 167
from a single primary spermatogonium by a succession of
four rapid mitotic divisions. Each primary spermatogonium
in turn arises from an asymmetric division of a stem cell
attached to the inner wall of the testis. Primary spermato-
cytes undergo S phase almost immediately after the com-
pletion of the last gonial mitotic division, then enter a
prolonged G2 phase that lasts more than 4 days during
which time they increase in volume more than 25-fold. At
the end of the growth period, the chromosomes condense
rapidly and enter the meiotic divisions, which last approx-
imately 4 h [38]. Chromosomes in growing spermatocytes
prior to the late condensation stage cannot be visualized by
staining with non-fluorescent dyes because they are too
decondensed. However, chromosomes in growing sperma-
tocytes can be visualized by staining with fluorescent dyes
such as DAPI. In the youngest spermatocytes, the chromatin
appears as a brightly staining but indistinct mass that fills
the interior of the nucleus. As the nucleus grows, the
chromatin remains loosely associated with the expanding
membrane, and a distinct nuclear lumen free of staining
develops. At about the midpoint of G2, the chromatin
separates into three distinct territories, each of which
remains associated with the nuclear membrane as the
nucleus continues to grow. After another 2 days or so of
growth, the chromatin condenses into four compact masses
at the onset of the division phases that are easily visualized
by classical microscopy. Autosomes appear in both EM and
light microscope analyses to be aligned in parallel, with the
most stable connections being medial or distal and the
centromeres clearly separated, and sex chromosomes appear
to be connected at one or more discrete heterochromatic
sites known as collochores [38­41]. The fact that there are
three major territories from mid-G2 onward rather than six
suggests that the homologs are paired in some way through-
out G2, but the poor morphology of the chromosomes has
precluded any detailed description of pairing, and it has not
been possible to rule out the alternative that the two
homologs are colocalized to a common territory by a
mechanism independent of pairing, e.g., by binding to
nearby sites on the nuclear membrane.
In the Vazquez et al. [27] study, 14 different insertions of
LacO repeats at a variety of euchromatic loci on chromo-
somes 2 and 3 exhibited similar (about 50%) levels of
pairing in spermatogonia. No loss of pairing in young
spermatocytes was observed, suggesting that mitotic and
meiotic pairing are continuous. However, the earliest dis-
tinguishable spermatocyte cysts have much higher pairing
frequencies, > 95%, indicative of a dramatic transition in
pairing frequency. Unfortunately, it is difficult to distinguish
the earliest spermatocyte cysts from 16-cell G1 spermato-
gonial cysts in living preparations. Consequently, it is not
clear if this transition occurs before or after premeiotic S
phase, although the authors of the study considered the
former possibility more likely. The very high pairing fre-
quencies of GFP-tagged loci that were observed in young
spermatocytes persisted throughout the early G2 growth
phase. However, a second dramatic change occurred at
mid-G2: pairing of both homologous and sister loci at all
14 LacO insertions disappeared completely, with four spots
becoming visible in most nuclei. This mid-G2 transition
occurred shortly after the separation of the chromatin
masses into distinct territories. Although a brief transition
period was seen when nuclei in the same cyst can show one,
two, three or four spots, the distribution of numbers and
brightnesses indicate that there is no particular order to the
loss of pairing and that homolog pairing and sister chroma-
tid cohesion are lost essentially simultaneously. This un-
paired state persisted through the end of G2. During this
period, homologous and sister loci were confined to the
chromosome territories but were on average no closer to one
another than to nonhomologous loci on the same chromo-
some. At diakinesis/prometaphase, sister and homologous
spots moved closer to one another as the chromosomes
condensed, indicating that both the sister chromatids and
homologs had remained connected somehow, but the spots
never became fused again.
Three important conclusions emerge from this study.
First, as has been proposed for years, Drosophila chromo-
somes apparently enter meiosis already paired; no interrup-
tion between premeiotic and meiotic pairing was observed.
Second, nevertheless, meiotic pairing differs dramatically
from mitotic pairing in frequency; nearly 100% pairing was
observed at all 14 tagged loci in young spermatocytes,
approximately double the frequency in spermatogonia. This
is interesting because it shows that the very high levels of
pairing associated with mid-prophase in meiosis do not
depend upon synapsis or recombination. It also suggests
that the pairing process in Drosophila males may be more
similar to that in recombinational meiosis than had been
suspected previously.
Finally, pairing is lost suddenly and completely at a point
in mid-G2 after the homologs have separated into territories.
Thus, intimate molecular pairing at most loci in Drosophila
male meiosis is confined to, approximately, the first half of
meiotic G2, despite the fact that chromosomes in late
prophase and prometaphase appear in all classical light
and electron microscope preparations to be very tightly
linked. This result is surprising at first glance, but less so
on reconsideration, as there is a similar transition in mid-
prophase in recombinational meiosis. As described above,
the transition from pachytene to diplotene involves removal
of the SC and lateral elements, often during a ``diffuse
stage'' in which axes lose resolution and chromatin becomes
highly decondensed. When the chromosomes recondense at
diplotene, homologous axes are clearly separate except for
the chiasmata, which may be as few as one per chromo-
some; indeed, in several organisms, homologous axes ap-
pear to occupy positions as far from one another as possible,
and are often said to be in ``repulsion'' (reviewed in Ref. [7].
FISH analyses show clearly that homologous spots are
separate at this stage, although of course they are con-
strained by the chiasmata to remain within the confines of
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180168
the bivalent (e.g., Ref. [42]). As in the Drosophila case,
homologous loci never become closely associated again. An
important difference, however, is that sister loci remain
paired in recombinational meiosis for the remainder of
prophase I (two spots are seen, not four) and the sister axes
remain closely associated. This persistent cohesion in the
arms is now thought to be critical for maintenance of the
chiasmata [43]. Presumably since male Drosophila lack
chiasmata, they have no need for close sister chromatid
cohesion after the end of the pairing phase of meiosis.
Nevertheless, the parallel behavior of the homologs in the
two types of meiosis is quite striking, and again suggests the
possibility that the underlying homolog pairing pathways in
recombinational and non-recombinational meiosis may be
quite similar, despite the more obvious differences in
associated structures.
4. Somatic and premeiotic pairing in other organisms
There has been considerable controversy about whether
somatic and/or premeiotic pairing occurs in other organisms
besides Dipterans. Although there have been reports of
pairing of certain chromosomal regions in mammalian cells
[44,45] most studies have been negative. There is strong
evidence that chromosomes occupy discrete territories in
interphase human cells, but homologous territories are
typically separated from one another [46,47]. Moreover,
there is clear evidence that homologs are unpaired in
spermatogonia of humans and rodents and do not exhibit
significant levels of association until late leptotene/zygotene
of meiosis [48­52]. Pairing is preceded by a reorganization
of chromosomes that features clustering of centromeres and
of telomeres on the nuclear membrane, the so-called bou-
quet configuration [50­52].
The situation appears to be similar in C. elegans where
FISH analysis has shown that homologs are unpaired in the
premeiotic germ cells [16]. Significant levels of pairing are
first seen in the transition stage nuclei which correspond to
the leptotene and zygotene stages of meiosis, and are
preceded by a reorganization of the nucleus in which the
chromosomes cluster at one end. This arrangement differs
from the bouquet arrangement in most organisms in that
only one end of each chromosome is in contact with the
nuclear membrane. Pairing frequencies average around 50­
60% at most loci in the transition stage nuclei, then increase
to nearly 100% in pachytene nuclei, which exhibit full
synapsis [16,53,54].
There is evidence for somatic and premeiotic pairing in
cereal species but it has been disputed (see Refs. [4,55]
for reviews). In wheat, homologous chromosomes have
been reported to be closer to each other than to homol-
ogous or heterologous chromosomes both in root tip
metaphases and in early meiotic prophase [56­58]. More
recently, FISH and GISH (genomic in situ hybridization)
analyses have provided additional evidence for homolo-
gous pairing in premeiotic interphase (prior to leptotene),
especially of centromeric regions [59­61]. This is fol-
lowed by chromatin condensation and pairing at other
loci, with telomeric pairing usually occurring prior to
pairing at interstitial sites. However, several studies have
failed to detect any homologous associations in premeiotic
mitoses in cereal plants [62­64], suggesting that the pre-
leptotene centromeric pairing may represent the initial
onset of meiotic pairing, rather than a result of mitotic
homolog alignment.
The situation appears to be somewhat different in
Arabidopsis. Recent chromosome painting studies have
documented preferential association of homologous chro-
mosomes, especially in the proximal regions, in immature
parenchymal (mitotic) cells [65­67]. In these cells, chro-
mosomes occupy distinct territories centered about con-
densed chromocenters consisting mostly of pericentric
heterochromatic, with the euchromatin forming one or
several loops emanating from the chromocenters. Nonho-
mologous chromocenters were not significantly clustered,
but homologous chromocenters were associated at frequen-
cies ranging from 10% to almost 80%. Painting of the
euchromatic region of one chromosome arm also detected
close homolog pairing over a length of 500 kb to 2 Mb,
but only at a frequency of 5­6%. Interestingly, euchro-
matic pairing was always accompanied by pairing of
chromocenters, suggesting that pairing may initiate in
centric regions and sometimes spread to the euchromatin.
GFP-labelling of single loci has recently been achieved in
Arabidopsis, and an initial report indicates significant
levels of homologous pairing in guard cells, which are
diploid [68]. However, the centric associations seen in
somatic cells do not seem to contribute to meiotic pairing.
At leptotene, centromeres are dispersed and telomeres take
the lead in first clustering on the surface of the nucleolus,
then pairing. Telomeric clustering on the nucleolus was
suggested to be a functional substitute for the bouquet
configuration in other organisms in which telomeres cluster
on the nuclear membrane [69]. Thus, although there does
seem to be at least some mitotic pairing in Arabidopsis,
especially in centric regions, there is no evidence at this
point that it contributes to meiotic pairing.
Maize meiosis appears to follow a similar pattern to that
in yeast and mammals with initiation of pairing during
meiotic prophase following telomere clustering on the
nuclear membrane [70]. Maguire [71] claimed that homo-
logs are associated in premeiotic mitosis in maize, but no
such associations were detected by Palmer [72]. Recent
FISH studies have also provided evidence against premei-
otic homolog pairing in maize [70,73].
Somatic and premeiotic pairing have been reported in
budding yeast, but the evidence is contradictory. Painting of
chromosome V in spread meiotic nuclei demonstrated that
many if not most homologs are unpaired in early prophase,
prior to zygotene, and that close pairing at pachytene is
preceded by parallel alignment at a distance [74]. However,
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 169
several other FISH studies have reported quite high levels of
pairing of a substantial number of loci, at levels from 25­
50%, in premeiotic interphase, i.e., at time 0 after transfer to
sporulation medium [75­78], which was substantially dis-
rupted by passage through premeiotic S phase [75,78].
Comparable levels of pairing were also reported in vegeta-
tive cells [77]. However, other FISH studies have failed to
detect significant pairing, at least of centromere-distal
probes, in vegetative cells [5,79]. Centromere-proximal
sequences were found to be paired or nonrandomly close
to each other at a significant frequency [79], but this was
attributed to general centromere clustering in yeast nuclei
throughout the cell cycle, an interpretation that has been
confirmed by later studies [80,81]. Moreover, GFP tagging
failed to detect any significant chromosome pairing in
premeiotic cells [82]. Indirect assays for mitotic pairing in
yeast have also been equivocal; site-specific recombination
frequencies are somewhat higher between allelic pairs of
insertions of Cre-Lox target sites than between nonhomol-
ogous insertions in mitotic cells, but the difference is
relatively slight when the effects of the prominent clustering
of yeast centromeres and the positions of the insertions
relative to their centromeres are taken into account [83,84].
Homologous mitotic recombination frequencies are also
very similar when the recombining loci are on homologous
versus nonhomologous chromosomes as long as position
relative to centromere is controlled [85,86]. A recently
developed physical assay for proximity of DNA sequences
shows some advantage for homologous loci, but again, the
effect is not terribly dramatic [87]. These data contrast
sharply with the data on site-specific recombination in
Drosophila mitotic nuclei where FLP induced recombina-
tion frequencies can exceed 20% between allelic targets in
the germ line [88], but are less than 1% between targets on
nonhomologous chromosomes [89]. One interpretation of
these data is that pairing in yeast is an artifact of the small
size of the nuclei, the strongly polarized organization of its
chromosomes and structural constraints that might act to
sort chromosomes by size [81]. However, the issue of
somatic/premeiotic pairing in yeast remains unresolved.
The strongest evidence for premeiotic pairing outside the
Diptera is from the fungus, S. pombe, in which meiosis is
normally azygotic, occurring in transient diploids in which
the maternal and paternal chromosome sets do not encounter
each other until after premeiotic S phase. However, diploids
can be induced to undergo mitotic proliferation, and then to
undergo an azygotic meiosis by manipulation of the media.
FISH analysis in such diploids showed that the homologs
occupy joint territories in more than 95% of mitotic nuclei,
an arrangement that was proposed to occur also in azygotic
meiosis as a preliminary to meiotic pairing [90]. It was also
shown that both homologous and nonhomologous centro-
meres but not telomeres are typically clustered, and that
sequences throughout the chromosomes are paired at fre-
quencies ranging from less than 20% for interstitial probes
to 60% for centromeric probes in vegetative nuclei. Induc-
tion of meiosis led to clustering of telomeres (bouquet
formation) and to increases in pairing frequency for all
but the centromeric probes. Interestingly, pairing at most
sites never exceeded 50%, presumably reflecting the lack of
synapsis in this organism. Pairing increased more rapidly
and reached higher levels at telomeric than at interstitial
sites, presumably reflecting the importance of telomeric
clustering in early prophase, but the differences were fairly
modest. Pairing occurred independently at all linked sites
except the telomeres in both vegetative and meiotic cells,
thus indicating that homology throughout the chromosome
is used. Similar pairing frequencies were measured in an
independent FISH study and in two applications of the GFP
tagging method [91,92]. Interestingly, this latter method
demonstrated that the intermediate pairing frequencies in
S. pombe at meiotic prophase reflect dynamically unstable
interactions, with GFP signals coming together and separat-
ing throughout meiotic prophase [91].
Overall, then, most eukaryotes do not exhibit high levels
of homolog pairing in somatic and premeiotic cells, and
apparently pair their chromosomes de novo after premeiotic
S phase, but S. pombe and Dipterans generally provide
exceptions in which chromosomes are paired premeiotically.
5. To what extent does pairing in recombinational
meiosis depend upon the synapsis and/or recombination
pathways?
It has been suggested that chromosome pairing is a
consequence of the homology search process that underlies
recombination, but the genetic data thus far are largely
unsupportive of this idea. Mutations that disrupt processing
of double-strand breaks, such as rad50S, or meiotic strand
transfer, such as null mutations in RAD51 and DMC1, have
only mild effects on homolog pairing frequency despite
severe effects on both recombination and synapsis
[75,76,93,94], although Peoples et al. [84] report a stronger
impact of dmc1 mutations on meiotic site-specific recom-
bination. Mutations in genes required for double-strand
break formation such as SPO11, RAD50, MEI4 and
REC102 tend to have stronger effects on prophase pairing
frequencies, but typically do not eliminate pairing. The
strongest effects have been seen in null mutations in
SPO11 which reduce pairing frequencies to 10­20% of
wild-type levels [75,76]. However, this effect is not due to
the role of Spo11 in induction of meiotic double-strand
breaks; a missense mutation in the active site of Spo11
causes no reduction in pairing at prophase [95]. These
results argue strongly that meiotic pairing is independent
of the formation and repair of DSBs, although a significant
deficit in pairing maintenance is manifested in several of the
mutants.
Similar results have been reported in other organisms. Null
mutations in homologs of SPO11 completely block meiotic
recombination but have no effect on synapsis in either C.
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180170
elegans or Drosophila [16,17]. The spo11 mutant phenotype
manifests only after pachytene, after removal of the SC.
Homolog pairing occurs at normal frequencies in Spo11-
deficient worms in both transition zone and pachytene nuclei
[16]. Loss of Spo11 function in S. pombe reduces pairing
frequencies in prophase but only slightly [91].
It is also clear that pairing precedes and is under
separate genetic controls from synapsis. Homologs remain
fully aligned but unsynapsed, and homolog pairing occurs
at wild type frequencies in a null mutation in the ZIP1
gene, which encodes a component of the transverse
filaments of yeast SCs [93]; meiotic site-specific recom-
bination frequencies are also unaffected by this mutation
[84]. Mutations in other yeast genes required for synapsis
and/or axial element formation, such as HOP1 and RED1,
also do not eliminate pairing, although they may reduce it
[74,75,84,93]. Mutations in the C. elegans syp1 gene,
which encodes a coiled-coil protein that localizes to the
central elements of the SCs, completely block synapsis
but do not prevent pairing. Pairing initiates with normal
timing and at normal levels in the transition stage nuclei,
but subsequently decays, and is present only at low levels
in nuclei that should be in pachytene [54]. However, the
conserved Hop2/Mnd1 complex, which localizes to chro-
mosomes in early meiosis, seems to play an important
role in homolog pairing in both budding and fission yeast
[91,96,97]. Mutations in the yeast homologs strongly
reduce homolog pairing frequencies and meiotic recombi-
nation without affecting formation or processing of dou-
ble-strand breaks and lead to extensive nonhomologous
synapsis. Disruption of the fission yeast Hop2 homolog,
Meu13 causes a significant reduction in pairing and
recombination without effecting telomere clustering or
movements [91]. It will be very interesting to learn what
the functions of these proteins are in meiotic homolog
pairing.
6. Do cohesins play a role in homolog pairing?
Cohesin is a four-member protein complex required for
sister chromatid cohesion both in mitosis and meiosis
[43,98]. It consists of two long coiled-coil proteins, SMC1
and SMC3, and two regulatory subunits, SCC1/RAD21 and
SCC3. Cohesin is loaded onto chromosomes prior to or
during S phase all along chromosomes and plays a crucial
role in the vicinity of centromeres in mitosis and meiosis II,
where it is needed to prevent premature separation of sister
kinetochores. The removal of cohesin by proteolysis of its
SCC1/RAD21 subunit at anaphase is thought to trigger
chromatid separation.
During meiosis, cohesin plays other roles as well, and
cohesin mutants therefore have multiple meiotic pheno-
types. This has been most clearly established for the
meiosis-specific REC8 subunit, a paralog of SCC1, found
in most or all eukaryotes [43]. Yeast Rec8 protein largely
replaces Scc1 and is present both on the chromosome arms
throughout the first division, where it is essential to maintain
chiasmata, and at the centromeres, where it is essential to
prevent premature separation of sister centromeres [43,99].
Interestingly, rec8 mutations also block axial element for-
mation and synapsis, and severely reduce recombination,
but do not affect double-strand break formation, suggesting
that cohesin plays a fundamental role in establishing meiotic
chromosome structure. Parallel results were obtained for
smc3 mutations [99]. rec8 mutants also interfere with axial
element formation and strongly reduce recombination in S.
pombe [92,100,101] as do mutations in rec11 which encodes
a meiosis-specific homolog of Scc3 [102,103]. In other
organisms, recombination has not been assayed directly,
but depletion of C. elegans REC8 by RNAi prevented
synapsis and crossing over but not the initiation of recom-
bination, based on the accumulation of small chromosome
fragments at diakinesis [104].
The role of cohesins in meiotic chromosome pairing is
not yet clear. In rec8 
yeast, both sister chromatids and
homologs are unpaired in approximately 70% of cells in
prophase [99], but it is not clear if this represents a defect in
initiation or maintenance of pairing. In S. pombe, mutations
in rec8 reduce homolog pairing about 2-fold during early to
mid-prophase, but the reduction is limited to the interstitial
regions of the chromosomes, consistent with the localized
defect in recombination [100]. As with budding yeast, this
effect is difficult to separate from the defects in cohesion
and axial element formation [92]. In Arabidopsis, loss of
Rec8 function has been reported to completely suppress
sister chromatid cohesion, homolog pairing and synapsis
[105,106], but another group has reported a milder pheno-
type in which synapsis appears normal, but homologs
exhibit abnormal condensation in leptotene and extensive
fragmentation and mis-segregation at anaphase I [107,108].
In C. elegans, the situation is clearer. The initiation of
homolog pairing in transition zone nuclei was unaffected by
depletion of REC8 by RNAi despite a strong defect in
synapsis, but pairing was not maintained in pachytene
[42,104]; similar results have recently been reported for a
mutation in the single C. elegans SCC3 homolog [109]. An
especially interesting observation is that in C. elegans,
SMC1 and SMC3 localize to meiotic prophase chromo-
somes even in the absence of REC8, and depletion of SMC1
in a genetic background in which REC8 fails to localize to
chromosomes caused a substantial reduction in homolog
pairing in the transition zone [42], suggesting that SMC1
and perhaps SMC3 may have a function in homolog pairing
independent of that of REC8. Recent reports indicate that
REC8 and the SMC subunits are loaded and removed from
chromosomes independently of one another in mammalian
meiosis as well [110­112]. Another indication for a role of
cohesins in chromosome pairing is that mutations in the
Coprinus Rad9 protein, which is required for chromosomal
loading of cohesin, causes an approximately twofold reduc-
tion in homolog pairing [113].
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 171
Taken together, these data indicate that cohesins may
play crucial roles in meiotic chromosome pairing. Howev-
er, a critical and as yet unanswered question is whether
this role in homolog pairing is direct or is derivative of
effects on sister chromatid cohesion. In all cases where
homolog pairing has been disrupted in early meiosis by
absence of a cohesin, that disruption has been accompa-
nied by loss of sister chromatid cohesion as well. These
data could indicate either that sister chromatid cohesion
and homolog pairing share a common mechanism involv-
ing cohesins, or that sister chromatid cohesion is a neces-
sary prerequisite for homolog pairing, perhaps due to
formation of axial elements. If the latter explanation proves
to be the case, it will point to divergent mitotic and
meiotic pairing mechanisms, since mitotic pairing has been
shown to occur at G1, when sister chromatids are absent.
An argument against the latter explanation is that meiotic
pairing and synapsis takes place at normal frequencies in
Coprinus in msh5-22 mutants which fail to undergo
premeiotic DNA replication and therefore lack sister chro-
matids [114,115]. Interestingly, the rad9 mutant still shows
a partial defect in homolog pairing in a msh5-22 back-
ground, indicating that the role of Rad9, and presumably
cohesin, in homolog pairing is not entirely derivative of its
role in sister chromatid cohesion [113].
7. Distribution of pairing sites
7.1. Pairing via general homology
In general, despite the important roles of telomere
sequences and of a few specialized pairing sites in meiosis
discussed below, homolog pairing in most organisms
appears to utilize sequence homology throughout the
chromosomes. Evidence for this includes the fact that ring
chromosomes can pair and recombine despite lacking
termini; that synaptic initiation can occur interstitially in
most organisms, although subtelomeric sites are often the
earliest synaptic sites; that homologous synapsis requiring
interstitial initiation can and often does occur in individ-
uals heterozygous for chromosome rearrangements such as
inversions and translocations; and that homologous synap-
tic switches occur at interstitial sites, often at high fre-
quency, in triploids and higher ploids of many organisms
[1,4,116­118].
Similar observations have been made in Drosophila male
meiosis. In males heterozygous for transpositions of euchro-
matic segments of chromosome 2 into the Y chromosome,
the Y chromosomes were found to pair with a normal 2 at
frequencies proportional to the size of the transposition.
Pairing was monitored in late prophase or prometaphase
by the presence of quadrivalents involving the X-Y and 2-2
pairs, which were seen at frequencies above background for
all transpositions involving euchromatin, including one that
involved only a few salivary bands (probably less than 100
kb). These quadrivalents were effective in orienting the
connected chromosomes to opposite poles, as exhibited in
excess segregation of the Y chromosome from the normal
chromosome 2 [119]. It is not known whether the interme-
diate frequencies of quadrivalents seen with most trans-
positions directly reflect the pairing frequencies in early
prophase. A plausible alternative could be that all loci
become fully paired in prophase, but that the probability of
forming (or maintaining) a stable connection between the
transposed segment and the normal chromosome 2 is depen-
dent on the length of the transposition. It would clearly be
informative to evaluate pairing directly using the GFP-
tagging protocol in these transposition heterozygotes.
This democratic approach is also characteristic of mitotic
pairing in Drosophila. In embryonic and somatic cells,
every locus or chromosomal region that has been tested
by FISH has shown substantial frequencies (50% or higher)
of pairing [20­24]. The FISH probes have included a
variety of cosmid and BAC clones containing unique
sequences as well as both complex and simple sequence
repeats derived from euchromatic and heterochromatic
regions. Moreover, 14 different LacO insertions in the
autosomal euchromatin showed comparable levels of pair-
ing (approximately 50%) in spermatogonia, and these same
loci also paired at comparable levels in young primary
spermatocytes [27].
As in political democracies, some individuals (or loci)
are more equal than others. In the Fung et al. [21] study,
pairing was observed at all 11 tested sites, but at different
frequencies. The histones locus, a repetitive locus in the
proximal euchromatin of 2L [120], was found to be the
strongest mitotic pairing site, achieving high levels of
pairing (61%) as early as embryonic cycle 13 when most
other loci are still paired at less than 10% [21]. Intriguingly,
the histones locus also appears to be especially potent in
meiotic pairing, based on the fact that 2-Y transpositions
that carry this region pair with a normal chromosome 2 at
frequencies disproportionate to their size [119]. Unique-
sequence euchromatic probes from proximal, medial and
distal regions of the chromosome 2 arms were found to pair
at more moderate rates and frequencies in both mitosis and
meiosis [21,119]. Simple sequence satellite repeats also
exhibited more moderate pairing frequencies.
Mitotic as well as meiotic pairing in Drosophila initiate at
multiple sites rather than spreading from a single major site.
One line of evidence for multiple initiations is the lack of
significant inter-locus correlations when probes from the
same chromosome are hybridized simultaneously [21,23].
Especially compelling are examples in which unpaired loci
are flanked by paired loci. A second line of evidence for
multiple independent initiation sites is that pairing of euchro-
matic regions in Drosophila mitotic cells is unaffected or only
partially disrupted by heterozygosity for translocations or
pericentric inversions with breakpoints on the same chromo-
some arm as the monitored locus. Pairing in the vicinity of the
brown locus, which is near the tip of chromosome arm 2R,
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180172
seems to be almost completely independent of chromosome
location, as relatively small transpositions of distal 2R
sequences into the proximal heterochromatin pair with the
native locus at the same frequency (70­80%) as do unrear-
ranged copies in larval brain cells [24]. Pairing at the BX-C
complex, which is located medially in chromosome arm 3R,
is partially disrupted in embryonic cells by heterozygosity for
two different translocations that move the BX-C complex
either to the opposite arm of chromosome 3 or to the X
chromosome. However, pairing still occurs at frequencies of
25­30% despite the fact that both translocations completely
abrogate transvection at the Ubx locus [23,121,122]. Similar
conclusions were reached using an indirect assay for pairing
in which frequencies of FLP-induced recombination between
homozygous target sites on chromosome 3 were measured in
the male germ line. These events occur at high frequency
(15­20%) between allelic insertions on unrearranged chro-
mosomes, but at much lower frequencies between non-allelic
insertions, indicating a dependence on chromosome pairing.
Several inversions and translocations with breakpoints prox-
imal to the target sites were tested and most were found to
reduce recombination frequencies, but never by more than
half, and some had only minor effects on recombination
frequency. The severity of the effects correlated with the size
of rearrangement and the proximity of the breakpoint to the
FLP insertion [88]. These data indicate that euchromatic loci
are capable of finding their homologs even when linked to
different centromeres and/or at different distances from the
centromere, but that these types of rearrangements do inter-
fere with pairing to some degree.
7.2. Roles of telomeres and centromeres in pairing
There is little evidence for a special role of telomeres in
mitotic pairing, but numerous recent observations indicate a
prominent role for telomeres in the early stages of meiotic
pairing. A common scenario is for telomeres to form a cluster
on the nuclear membrane near the centrosome/spindle pole
body in early prophase shortly before or commensurate with
the onset of homologous pairing. The resulting bouquet
configuration and associated chromosome movements have
been postulated to promote the homology search process in
interstitial as well as distal regions [123]. An interesting
variation on this pattern is seen in Arabidopsis in which
telomeres associate with the nucleolus prior to meiotic
prophase, then pair homologously at leptotene, leading to
initiation of synapsis in terminal or subterminal regions [69].
However, as noted above, there is a large body of evidence
that pairing can initiate at interstitial sites independent of
telomeric pairing, so the role of telomeres is likely to be more
kinetic than essential. This is consistent with the observation
that mutations in yeast that disrupt telomere clustering lead to
delays in pairing and reductions in recombination but do not
prevent either [124,125].
Roles for centromeres in pairing have been extensively
discussed and investigated. As described above, in some
organisms centromere clustering is a prominent feature of
chromosome organization in mitotic cells, but it is often
difficult to ascertain the role of homology in centromeric
clusters. A common situation is for clusters to include
variable numbers of homologous as well as heterologous
centromeres [5,81]. However, in mitotic Arabidopsis cells,
centromeric associations appear to be almost exclusively
homologous; moreover, homologous centromeres were
found to be paired whenever euchromatic regions of the
same chromosome were paired, suggesting that pairing
might spread from centric heterochromatin into the euchro-
matin, although other explanations have not been ruled out
[66]. In S. pombe, both mitotic and premeiotic pairing
frequencies are highest in centromere regions [90,92]. In
Drosophila mitotic pairing, however, there is no obvious
bias for or against heterochromatic sequences and no
evidence for spreading, as discussed above. Moreover,
pairing in the X heterochromatin was found to be almost
completely disrupted by heterozygosity for a large inversion
that moves the X heterochromatin to a distal location [126],
suggesting that heterochromatic loci may not have the same
independent ability to find their homologs that euchromatic
loci evidently do.
Overall, there is relatively little evidence for a prominent
role of centromeric heterochromatin in meiotic pairing,
despite numerous suggestions for such a role; indeed, the
bulk of the data suggests that centric heterochromatin is
partially or completely excluded from an active role in
pairing in most organisms, as well as being recombination-
ally inert. There is some evidence in wheat for the initiation
of chromosome pairing in premeiotic interphase, with pair-
ing of centric regions preceding that of more distal regions
[59­61] but this seems to be rather untypical. In Arabidop-
sis, although some pairing of centric regions has been
documented in mitotic cells, centromeres are clearly un-
paired in early meiotic prophase [69,127] and pair later than
distal regions. In yeast, centromere clustering is common in
vegetative cells, but is rapidly lost upon induction of meiosis
[78,81]. In humans and rodents, centromere movements to
the nuclear membrane precede telomere movements in early
prophase, but no evidence for homolog pairing is detected
prior to telomere clustering and bouquet formation [50], and
centric sequences were found to be the last to pair. Even in S.
pombe in which centric heterochromatic regions appear to
enter meiosis paired at high frequencies, early prophase is
dominated by telomere-led chromosome movements [128],
and both telomeric and interstitial regions pair independently
of the centromeres [90,129]. These observations are consis-
tent with numerous classical observations that SCs in het-
erochromatin are shorter than in euchromatin (or even absent
in some cases), that heterochromatic SC is often structurally
abnormal, and that synaptic initiation sites are absent from
heterochromatin [1,4,130].
There is considerable evidence that centric heterochro-
matin is largely excluded from pairing in Drosophila male
meiosis. Centromeres tagged with GFP-Cid (the Drosophila
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 173
homolog of CENP-A, a histone H3 variant specific for
centromeres [131]) in young primary spermatocytes, the
stage at which euchromatic loci exhibit nearly 100% pair-
ing, were observed to form variable numbers of clusters but
to be unpaired in a substantial fraction of cells. This is
consistent with several classical studies analyzing pairing at
prometaphase or metaphase and/or segregation patterns in
heterozygotes for various types of chromosome rearrange-
ments, all of which have failed to detect pairing ability in the
centric heterochromatin of the autosomes (see Ref. [132]).
Of particular relevance to the comparison with mitotic
pairing is the finding that a large transposition of 2R
heterochromatin into the Y chromosome proved unable to
induce formation of quadrivalents despite being much larger
than several euchromatic transpositions that did induce
quadrivalents and despite carrying the Responder locus
which was found to be a strong mitotic pairing site
[21,119]. X-Y pairing clearly does take place in the hetero-
chromatin in Drosophila [40,132], but as described in more
detail below, the pairing site is a specific repeat, and most of
the centromeric X heterochromatin is clearly devoid of
pairing ability [132].
Distributive pairing in Drosophila females provides a
major exception to the exclusion of centric heterochromatin
from pairing. Distributive pairing is a back-up system that
ensures the segregation of non-exchange homologs, espe-
cially the fourth chromosome pair which almost never
undergoes exchange. Genetic evidence indicates that dis-
tributive pairing is based on heterochromatic homology
[133,134]. Moreover, heterochromatic pairing has been
directly documented by FISH in late prophase (after SC
removal) in Drosophila females [135]. However, there is no
evidence either for robust homologous distributive pairing
systems or for post-synaptic pairing in the centric hetero-
chromatin in other organisms; indeed, centric regions of
homologs are usually widely separated in late prophase [7].
Although heterochromatic pairing clearly can exhibit at least
some specificity, the limited ability of the distributive
system to sort chromosomes homologously is shown by
the chaos that ensues when exchange is drastically reduced
in Drosophila females and most or all chromosomes are
thrown into the distributive pool. In general, heterochromat-
ic pairing is probably counterproductive in meiosis because
of the high density of repeated sequences shared with other
chromosomes, and the consequent danger of rearrangements
and/or nonhomologous pairing and segregation.
7.3. Specialized pairing sites
Other than the rather limited evidence for centromere
dominated pairing in Arabidopsis and S. pombe somatic
cells, there is little evidence that mitotic pairing, as opposed
to nonspecific heterochromatic stickiness, involves special-
ized pairing sites. Indeed as discussed above, there is
substantial evidence that pairing initiates at multiple inter-
stitial sites in Drosophila mitotic cells.
However, there is some genetic evidence for specialized
pairing sites in meiosis. In both Drosophila and C. elegans,
specific chromosomal sites are required in cis for normal
levels of meiotic recombination. In Drosophila the X
chromosome appears to be subdivided into at least three
recombinationally independent intervals by four sites, based
on the fact that heterozygosity for X-4 translocations sup-
presses recombination but only within the interrupted inter-
val [136]. Reciprocal translocations in C. elegans typically
suppress recombination to one side of each breakpoint, and
the breakpoints of these translocations as well as of other
rearrangements have been used to map a single Homolo-
gously Recombining Region (HRR) to one end of each
chromosome (reviewed in Ref. [137]. These ends have been
shown to contact the nuclear membrane at the onset of
meiotic prophase in the transition zone nuclei [16,53,54].
Although both the X chromosomal sites in Drosophila and
the HRRs in C. elegans have been suggested to be pairing
sites, there is no direct evidence for such a role. In fact,
deletions and mutations of HRRs have only minor effects on
recombination when heterozygous, contrary to what would
be expected if they functioned as pairing sites. An alterna-
tive might be that they function as loading sites for synaptic
complexes or other proteins involved in pairing, synapsis or
recombination [138]. Neither of these interesting types of
sites has been molecularly identified as yet, although the
HRRs are known to map in relatively gene-poor regions of
the chromosomes [137].
The most thoroughly characterized specific meiotic pair-
ing site is that of the X-Y pair in male Drosophila, which
has been mapped to a 240-bp repeated sequence in the
intergenic spacers of the rDNA genes. These genes are
localized in two blocks in the heterochromatin of the X
and the short arm of the Y chromosome. Each sex chromo-
some contains some 1000­2000 copies of the 240-bp IGS
repeat, but transgenic studies showed that inserted arrays
containing as few as six or seven repeats are capable of
partially restoring pairing ability to an X chromosome
deficient for the native pairing region [132]. This is clearly
a special case in that the X and Y chromosomes in
Drosophila, as in many other higher organisms, are highly
diverged, and therefore unable to utilize the widespread
homology available to other chromosomes. However, it is
noteworthy that the pairing site nevertheless corresponds to
a region of homology, as does the pseudoautosomal region
of the mammalian X and Y chromosomes.
8. Where does pairing occur?
As we have seen, pairing ability is democratically dis-
tributed in the euchromatin in both mitosis and meiosis, and
is present in the heterochromatin as well, although hetero-
chromatic pairing is usually restricted in meiosis. Although
specialized chromosomal sites have been shown to contrib-
ute to meiotic pairing in some organisms, the data in most
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180174
cases are rather strongly against the idea that pairing
initiates at one or a few sites and spreads to the remainder
of the chromosome. These data indicate that if pairing is
restricted to certain types of sequences, those sequences
must be widely distributed, more on the order of one or a
few per gene than one or a few per chromosome.
The only meiotic pairing site that has been molecularly
identified is the X-Y pairing site in Drosophila male meiosis
(reviewed in Ref. [132]), which as noted above corresponds
to arrays of 240-bp repeats located in the intergenic spacers
between rDNA repeats. These repeats are present in arrays
of 5­12 copies immediately upstream of each pre-rRNA
transcription unit, and have been shown to function as
enhancers of pre-rRNA transcription. Intriguingly, each
240-bp repeat contains a perfect copy of an approximately
80-bp region flanking the rRNA promoter, and these ``spac-
er promoters'' have been shown to be active, albeit weak,
promoters. The most abundant spacer transcripts span the
array from the upstream-most promoter to a termination site
about 300 bp upstream of the pre-rRNA promoter. It has
been shown that arrays of six or more such repeats, when
inserted on an X chromosome deficient for its native pairing
site, can stimulate X-Y pairing in the absence of any other
rDNA-derived sequences. These data are consistent with the
notion that enhancers and/or promoters may function as
pairing sites.
The best-characterized mitotic pairing site is that
encompassing the brown locus located near the tip of
chromosome arm 2R. As described above, even small
transpositions carrying this region pair with unrearranged
homologs as avidly as do normally located copies, sug-
gesting unusual pairing strength [24]. Mapping of the
pairing site by deletion analysis indicated that it mapped
broadly to the upstream regulatory region of the brown
locus, coincident with the sequences that regulate tran-
scription of brown, but could not be mapped more
precisely [139,140]. Thus, the mapping of the mitotic
pairing site in distal 2R agrees broadly with the mapping
of the X-Y pairing site, and is consistent with the idea
that pairing sites may correspond to enhancers and/or
promoters.
Recent studies of pairing-dependent silencing of trans-
genes in Drosophila have also implicated upstream regu-
latory regions of genes in pairing. Transgenic insertions of
constructs that carry a Polycomb Response Element
(PRE) as well as a reporter gene often exhibit partial or
full silencing of the reporter in homozygotes even though
expression may be normal in hemizygotes [31]. Although
trans-silencing interactions are usually limited to allelic
insertions, two particular constructs containing regulatory
regions from the BX-C complex proved to be able to
undergo silencing interactions with copies located at
distant sites in the genome. In each case, the transgene
carries both an enhancer and the PRE, as well as one or
more insulator sequences [141,142]. In one case, long-
distance silencing was seen only when the insulator,
derived from the gypsy element, was added to the
construct. It is not clear whether these types of events
have anything to do with mitotic chromosome pairing, but
the fact that the sites map to complex upstream regulatory
regions is intriguing.
It has been noted previously that the restriction of
meiotic pairing ability to euchromatin in male meiosis is
consistent with the notion of pairing at transcriptional
regulatory sequences or promoters [132]. It is also con-
sistent with the especially strong pairing ability of a
region in proximal 2L encompassing the histones locus,
which contains an unusually high density of strong
promoters. Intriguingly, the single most avid pairing site
in the mitotic pairing studies has also been the histones
locus. It initiates pairing as early as cycle 12, before any
other tested locus or region shows significant pairing, and
achieves higher average pairing levels than any other
locus [21]. This observation is consistent with the idea
that pairing sites generally may correspond to promoters.
However, other data from the mitotic pairing studies
are less easily incorporated into a promoter/enhancer
pairing model. Particularly difficult to accommodate is
the fact that heterochromatic repeats can pair mitotically.
Although simple sequence repeats such as AACAC and
AAGAC were not among the first sites to pair in
Drosophila embryos, they did achieve high levels of
pairing in cells with long cycling times, and apparently
were able to do so independently of flanking sequences.
Moreover, the second most potent pairing site in the Fung
et al. [21] study of chromosome 2 pairing sites was the
Responder (Rsp) locus which is adjacent to the chromo-
some 2 centromere and is a complex heterochromatic
repeat not known to be transcribed or to possess a
promoter. As discussed above, this same site was com-
pletely inactive in the assay for meiotic pairing ability.
Another observation that is difficult to reconcile with
transcription-based pairing models is that mitotic pairing
frequencies seem to be entirely independent of the tran-
scriptional status of the locus. This has been most clearly
demonstrated for the BX-C complex, which pairs at equal
frequencies in cells in which it is silenced as in cells in
which it is actively transcribed [23]. Another pertinent
observation is that mitotic pairing persists essentially
throughout the cell cycle, although there is some evidence
for partial disruption at anaphase and at S phase [21,22].
This suggests that the structure and function of mitotic
pairing sites are not affected by chromatin condensation.
8.1. Could pairing sites be MARs?
From the above discussion, it seems likely that mitotic
homolog pairing occurs at sites that are involved in overall
organization of the chromatin fiber in both interphase and
mitosis, and that are common to euchromatic and hetero-
chromatic regions. This suggests that pairing sites might be
MAR/SARs, short, AT-rich sequences that co-purify with
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 175
preparations of the nuclear matrix in interphase or the
chromosome scaffolds in mitosis [143,144]. They are
associated with DNAase hypersensitive sites, are common-
ly found in upstream regulatory regions where they can
contribute to gene activation, but are present elsewhere as
well, and have been suggested to be especially abundant in
heterochromatin [145­151]. At least some MARs clearly
have boundary element/insulator activity [152­159], and,
conversely, some insulators have been shown to bind
nuclear matrices [160]. Moreover, a recent ChIP study of
some of the complex regulatory regions of the BX-C
revealed that condensin and topoisomerase II, the canon-
ical scaffold proteins, strongly colocalize with Polycomb-
group proteins at PREs [161]. Several of these same
regulatory regions have been shown to contain insulator
activity, and, as noted above, at least two of these regions
can promote long-distance interactions that are suggestive
of pairing [141,142,162,163].
The idea that mitotic pairing sites might coincide with
MAR/SAR sequences could account for the exceptional
pairing strength of the histones locus in both mitosis and
meiosis, as there is a strong MAR/SAR site located in the
largest intergenic spacer of each repeat, which are thus
spaced only 5 kb apart in the histones locus [145]. It is
not clear whether other known pairing sites also contain
MARs. The 240-bp IGS repeat that acts as the major X-Y
pairing site in Drosophila contains regions that are highly
AT rich and similar to consensus MAR sequences.
Unfortunately little is known about the actual functions
of MARs. MARs have been postulated to bind to dispersed
sites on the nuclear matrix during interphase, and to coa-
lesce into linear arrays during mitosis to form the chromo-
some axes, forming the bases of chromatin loops that may
comprise functional domains in interphase nuclei and struc-
tural domains in mitotic and/or meiotic chromosomes on the
order of tens to hundreds of thousands of base pairs. Some
insulators have been shown to form the bases of chromatin
loops as well, and one idea about insulator function is that
they act to separate enhancers and promoters into separate
looped chromatin domains [164,165]. Recent data from
yeast and Drosophila point to preferential localization of
insulators at particular peripheral sites in the nucleus being
essential to their function as insulators [165,166]).
How might MAR elements act as meiotic pairing sites?
The ability to attach to nuclear matrices might contribute to
homology searching but alone would not provide specific-
ity. However, as noted above, MARs are associated with
sites of open chromatin and tend to promote acetylation of
surrounding DNA [150]. Moreover, at least some MARs
have been shown to unwind readily in response to super-
coiling [167]. These properties could be important in the
context of a homology search as they could define sites at
which homology testing could be carried out easily without
a requirement for DNA breaks. They would also provide
regulatory flexibility so that some elements could be active
in pairing in some cells but not others.
References
[1] D. von Wettstein, S.W. Rasmussen, P.B. Holm, The synaptonemal
complex in genetic segregation, Annu. Rev. Genet. 18 (1984)
331­413.
[2] C. Heyting, Synaptonemal complexes: structure and function, Curr.
Opin. Cell Biol. 8 (1996) 389­396.
[3] G.S. Roeder, Meiotic chromosomes: it takes two to tango, Genes
Dev. 11 (1997) 2600­2621.
[4] J. Loidl, The initiation of meiotic chromosome pairing: the cytologi-
cal view, Genome 33 (1990) 759­778.
[5] J. Loidl, Meiosis in budding yeast and in multicellular eukar-
yotes--similarities and differences, Chromosomes Today 13
(2000) 123­137.
[6] N. Kleckner, Meiosis: How could it work? Proc. Natl. Acad. Sci.
U. S. A. 93 (1996) 8167­8174.
[7] D. Zickler, N. Kleckner, Meiotic chromosomes: Integrating structure
and function, Annu. Rev. Genet. 33 (1999) 603­754.
[8] K.N. Smith, A. Nicolas, Recombination at work for meiosis, Curr.
Opin. Genet. Dev. 8 (1998) 200­211.
[9] S. Keeney, Mechanism and control of meiotic recombination initia-
tion, Curr. Top. Dev. Biol. 52 (2001) 1­53.
[10] M. Petronczki, M.F. Siomos, K. Nasmyth, A ménage a quatre: the
molecular biology of chromosome segregation in meiosis, Cell 112
(2002) 423­440.
[11] E. Alani, R. Padmore, N. Kleckner, Analysis of wild-type and rad50
mutants of yeast suggests an intimate relationship between meiotic
chromosome synapsis and recombination, Cell 61 (1990) 419­436.
[12] R. Padmore, L. Cao, N. Kleckner, Temporal comparison of recom-
bination and synaptonemal complex formation during meiosis in S.
cerevisiae, Cell 66 (1991) 1239­1256.
[13] S. Keeney, C.N. Giroux, N. Kleckner, Meiosis-specific DNA dou-
ble-strand breaks are catalyzed by Spo11, a member of a widely
conserved protein family, Cell 88 (1997) 375­384.
[14] P.J. Romanienko, R.D. Camerini-Otero, The mouse spo11 gene is
required for meiotic chromosome synapsis, Mol. Cell 6 (2000)
975­987.
[15] M. Grelon, D. Vezon, G. Gendrot, G. Pelletier, AtSPO11-1 is nece-
ssary for efficient meiotic recombination in plants, EMBO J. 20
(2001) 589­600.
[16] A.F. Dernburg, J. McDonald, G. Moulder, R. Barstead, M. Dresser,
A.M. Villeneuve, Meiotic recombination in C. elegans initiates by a
conserved mechanism and is dispensable for homologous chromo-
some synapsis, Cell 94 (1998) 387­398.
[17] K.S. McKim, A. Hayashi-Hagihara, Mei-W68 in Drosophila encodes
a Spo11 homolog: evidence that the mechanism for initiating meiotic
recombination is conserved, Genes Dev. 12 (1998) 2932­2942.
[18] K.W. Wolf, How meiotic cells deal with non-exchange chromo-
somes, BioEssays 16 (1994) 107­114.
[19] C.W. Metz, Chromosome studies on the Diptera: II. The paired
association of chromosomes in the Diptera and its significance,
J. Exp. Zool. 21 (1916) 213­279.
[20] Y. Hiraoka, A.F. Dernburg, S.J. Parmalee, M.C. Rykowski, D.A.
Agard, J.W. Sedat, The onset of homologous chromosome pairing
during Drosophila melanogaster embryogenesis, J. Cell Biol. 120
(1993) 591­600.
[21] J.C. Fung, W.F. Marshall, A. Dernburg, D.A. Agard, J.W. Sedat,
Homologous chromosome pairing in Drosophila melanogaster pro-
ceeds through multiple independent initiations, J. Cell Biol. 141
1998, pp. 5­20.
[22] A.K. Csink, S. Henikoff, Large-scale chromosomal movements du-
ring interphase progression in Drosophila, J. Cell Biol. 143 (1998)
13­22.
[23] M.J. Gemkow, P.J. Verveer, D.J. Arndt-Jovin, Homologous associa-
tion of the Bithorax complex during embryogenesis: consequences
for transvection in Drosophila melanogaster, Development 125
(1998) 4541­4552.
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180176
[24] G.L. Sass, S. Henikoff, Pairing-dependent mislocalization of a Dro-
sophila brown gene reporter to a heterochromatic environment, Ge-
netics 152 (1999) 595­604.
[25] C.C. Robinett, A.F. Straight, G. Li, C. Willhelm, G. Sudlow, A.W.
Murray, A.S. Belmont, In vivo localization of DNA sequences and
visualization of large-scale chromatin organization using lac oper-
ator/repressor recognition, J. Cell Biol. 135 (1996) 1685­1700.
[26] A.F. Straight, A.S. Belmont, C.C. Robinett, A.W. Murray, GFP tag-
ging of budding yeast chromosomes reveals that protein­protein
interactions can mediate sister chromatid cohesion, Curr. Biol. 6
(1996) 1599­1608.
[27] J. Vazquez, A.S. Belmont, J.W. Sedat, The dynamics of homologous
chromosome pairing during male meiosis in Drosophila, Curr. Biol.
12 (2002) 1473­1483.
[28] E.P. Semionev, A.F. Smirnov, Somatic synapsis of eu- and hetero-
chromatic regions of Drosophila melanogaster chromosomes,
Genetica 65 (1984) 159­166.
[29] S. Henikoff, Nuclear organization and gene expression: homologous
pairing and long-range interactions, Curr. Opin. Cell Biol. 9 (1997)
388­395.
[30] S. Henikoff, L. Comai, Trans-sensing effects: the ups and downs of
being together, Cell 93 (1998) 329­332.
[31] V. Pirrotta, Transvection and chromosomal trans-interaction effects,
Biochim. Biophys. Acta 1424 (1999) M1­M8.
[32] I.W. Duncan, Transvection effects in Drosophila, Annu. Rev. Genet.
36 (2002) 521­556.
[33] C.W. Metz, Observations on spermatogenesis in Drosophila, Z. Zell-
forsch. Mikrosk. Anat. 4 (1926) 1­28.
[34] N.M. Stevens, A study of the germ cells of certain Diptera, with
reference to the heterochromosomes and the phenomena of synapsis,
J. Exp. Zool. 5 (1908) 359­374.
[35] S.M. Grell, Cytological studies in Culex II. Diploid and meiotic
divisions, Genetics 31 (1946) 77­94.
[36] S.R.A.L. Mescher, K.S. Rai, Spermatogenesis in Aedes aegypti,
Mosq. News 26 (1966) 45­51.
[37] A. Wandall, A. Svendsen, Transition from somatic to meiotic pairing
and progressional changes of the synaptonemal complex in sperma-
tocytes of Aedes aegypti, Chromosoma 92 (1985) 254­264.
[38] G. Cenci, S. Bonnacorsi, C. Pisano, F. Verni, M. Gatti, Chromatin
and microtubule organization during premeiotic, meiotic and early
post-meiotic stages of Drosophila melanogaster spermatogenesis,
J. Cell Sci. 107 (1994) 3521­3534.
[39] K.W. Cooper, Normal spermatogenesis in Drosophila, in: M.
Demeric (Ed.), Biology of Drosophila, Wiley, New York, 1950,
pp. 1­61.
[40] K.W. Cooper, Meiotic conjunctive elements not involving chiasma-
ta, Proc. Natl. Acad. Sci. U. S. A. 52 (1964) 1248­1255.
[41] M.T. Fuller, Spermatogenesis, The Development of Drosophila mel-
anogaster, Cold Spring Harbor Laboratory Press, Cold Spring Har-
bor, NY, 1993, pp. 71­147.
[42] R.C. Chan, A. Chan, M. Jeon, T.F. Wu, D. Pasqualone, A. Rougvie,
B. Meyer, Chromosome cohesion is regulated by a clock gene
paralogue TIM-1, Nature 423 (2003) 1002­1009.
[43] K. Nasmyth, Disseminating the genome: joining, resolving and sepa-
rating sister chromatids during mitosis and meiosis, Annu. Rev.
Genet. 35 (2001) 673­745.
[44] J.M. LaSalle, M. Lalande, Homologous association of oppositely
imprinted chromosomal domains, Science 272 (1996) 725­728.
[45] Y. Marahrens, X-inactivation by chromosomal pairing events, Genes
Dev. 13 (1999) 2624­2632.
[46] A.R. Leitch, J.K.M. Brown, W. Mosgoller, T. Schwarzacher, J.S.
Heslop-Harrison, The spatial localization of homologous chromo-
somes in human fibroblasts at mitosis, Hum. Genet. 93 (1994)
275­280.
[47] T. Cremer, C. Cremer, Chromosome territories, nuclear architecture
and gene regulation in mammalian cells, Nat. Rev., Genet. 2 (2001)
292­302.
[48] S.J. Armstrong, A.J. Kirkham, M.A.J. Hulten, XY chromosome
behavior in the male germ line of the human male: a FISH analysis
of spatial orientation, chromatin condensation and pairing, Chromo-
som. Res. 2 (1994) 445­452.
[49] E.Y. Cheng, S.M. Gartler, A fluorescent in situ hybridization analy-
sis of X chromosome pairing in human female meiosis, Hum. Genet.
94 (1994) 389­394.
[50] H. Scherthan, S. Weich, H. Schwegler, C. Heyting, M. Harle, T.
Cremer, Centromere and telomere movements during early meiotic
prophase of mouse and man are associated with the onset of chro-
mosome pairing, J. Cell Biol. 134 (1996) 1109­1125.
[51] H. Scherthan, R. Eils, E. Trelles-Sticken, S. Dietzel, T. Cremer, H.
Walt, A. Jauch, Aspects of three-dimensional chromosome reorga-
nization during the onset of human male meiotic prophase, J. Cell
Sci. 111 (1998) 2337­2351.
[52] C. Pfeifer, P.D. Thomsen, H. Scherthan, Centromere and telomere
redistribution precedes homolog pairing and terminal synapsis initia-
tion during prophase I of cattle spermatogenesis, Cytogenet. Cell
Genet. 93 (2001) 304­314.
[53] A.J. MacQueen, A.M. Villeneuve, (2001) Nuclear reorganization
and homologous chromosome pairing during meiotic prophase re-
quire C. elegans chk-2, Genes Dev. 16 (2001) 2428­2442.
[54] A.J. MacQueen, M.P. Colaiacovo, K. McDonald, K.A.M. Ville-
neuve, Synapsis-dependent and -independent mechanisms stabilize
homolog pairing during meiotic prophase in C. elegans, Genes Dev.
16 (2002) 2428­2442.
[55] T. Schwarzacher, Meiosis, recombination and chromosomes: a re-
view of gene isolation and fluorescent in situ hybridization data in
plants, J. Exp. Bot. 54 (2003) 11­23.
[56] L. Avivi, M. Feldman, Mechanism of non-random chromosome
placement in common wheat, Proc. Int. Wheat Genet. Symp. 4
(1973) 627­633.
[57] M. Feldman, L. Avivi, Non-random arrangement of chromosomes in
common wheat, Chromosomes Today 4 (1973) 187­196.
[58] M. Feldman, L. Avivi, Ordered arrangement of chromosomes in
wheat, Chromosomes Today 8 (1984) 181­189.
[59] T. Schwarzacher, Three stages of meiotic homologous chromosome
pairing in wheat: cognition, alignment and synapsis, Sex. Plant
Reprod. 10 (1997) 324­331.
[60] L. Aragon-Alcaide, S. Reader, A. Beven, P. Shaw, T. Miller, G.
Moore, Association of homologous chromosomes during floral
development, Curr. Biol. 7 (1997) 905­908.
[61] E.I. Mikhailova, T. Naranjo, K. Sheperd, J. Wennekes-van Eden, C.
Heyting, J.H. de Jong, The effect of the wheat Ph1 locus on chro-
matin organization and meiotic chromosome pairing analyzed by
genomic painting, Chromosoma 107 (1998) 339­350.
[62] J.S. Heslop-Harrison, J.B. Smith, M.D. Bennett, The absence of
somatic association of centromeres of homologous chromosomes
in grass mitotic metaphases, Chromosoma 96 (1988) 119­131.
[63] M.D. Bennett, Premeiotic events and meiotic chromosome pairing,
Symp. Soc. Exp. Biol. 38 (1984) 87­121.
[64] T. Schwarzacher, K. Anamthawat-Jensson, G.E. Harrison, Genomic
in situ hybridization to identify alien chromosomes and chromosome
segments in wheat, Theor. Appl. Genet. 84 (1992) 778­786.
[65] M.A. Lysak, P.F. Fransz, H.B.M. Ali, I. Schubert, Chromosome
painting in Arabidopsis thaliana, Plant J. 28 (2001) 689­697.
[66] P. Fransz, J.H. de Jong, M. Lysak, M.R. Castiglione, I. Schubert,
Interphase chromosomes in Arabidopsis are organized as well-de-
fined chromocenters from which euchromatin loops emanate, Proc.
Natl. Acad. Sci. U. S. A. 99 (2002) 14584­14589.
[67] M.A. Lysak, A. Pecinka, I. Schubert, Recent progress in chromo-
some painting of Arabidopsis and related species, Chromosom. Res.
11 (2003) 195­204.
[68] N. Kato, E. Lam, Chromatin of endoreduplicated pavement cells has
greater range of movement than that of diploid guard cells in Arabi-
dopsis thaliana, J. Cell Sci. 116 (2003) 2195­2201.
[69] S. Armstrong, F.C.H. Franklin, G.H. Jones, Nucleolus-associated
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 177
telomere clustering and pairing precede meiotic chromosome synap-
sis in Arabidopsis thaliana, J. Cell Sci. 114 (2001) 4207­4217.
[70] H.W. Bass, W.F. Marshall, J.W. Sedat, D.A. Agard, W.Z. Cande,
Telomeres cluster de novo before the initiation of synapsis: a three-
dimensional spatial analysis of telomere positions before and during
meiotic prophase, J. Cell Biol. 137 (1997) 5­18.
[71] M.P. Maguire, Chromosome behavior at premeiotic mitosis in maize,
J. Heredity 74 (1983) 93­96.
[72] R.G. Palmer, Cytological studies of ameiotic and normal maize with
reference to premeiotic pairing, Chromosoma 35 (1971) 233­246.
[73] R.K. Dawe, J.W. Sedat, D.A. Agard, W.Z. Cande, Meiotic chromo-
some pairing in maize is associated with a novel chromatin organi-
zation, Cell 76 (1994) 901­912.
[74] H. Scherthan, J. Loidl, T. Schuster, D. Schweizer, Meiotic chromo-
some condensation and pairing in Saccharomyces cerevisiae studied
by chromosome painting, Chromosoma 101 (1992) 590­595.
[75] B.M. Weiner, N. Kleckner, Chromosome pairing via multiple inter-
stitial interactions before and during meiosis in yeast, Cell 77 (1994)
977­991.
[76] J. Loidl, F. Klein, H. Scherthan, Homologous pairing is reduced but
not abolished in asynaptic mutants of yeast, J. Cell Biol. 125 (1994)
1191­1200.
[77] S.M. Burgess, N. Kleckner, B.M. Weiner, Somatic pairing of homo-
logs in budding yeast: existence and modulation, Genes Dev. 13
(1999) 1627­1641.
[78] E. Trelles-Sticken, J. Loidl, H. Scherthan, Bouquet formation in
budding yeast: initiation of recombination is not required for meiotic
telomere clustering, J. Cell Sci. 112 (1999) 651­658.
[79] V. Guacci, E. Hogan, D. Koshland, Chromosome condensation and
sister chromatid pairing in budding yeast, J. Cell Biol. 125 (1994)
517­530.
[80] V. Guacci, E. Hogan, D. Koshland, Centromere position in budding
yeast: Evidence for anaphase A, Mol. Biol. Cell 8 (1997) 957­972.
[81] Q. Jin, E. Trelles-Sticken, H. Scherthan, J. Loidl, Yeast nuclei dis-
play prominent centromere clustering that is reduced in nondividing
cells and in meiotic prophase, J. Cell Biol. 141 (1998) 21­29.
[82] L. Aragon-Alcaide, A.V. Strunnikov, Functional dissection of in
vivo interchromosome association in Saccharomyces cerevisiae,
Nat. Cell Biol. 2 (2000) 812­818.
[83] S.M. Burgess, N. Kleckner, Collisions between yeast chromosomal
loci in vivo are governed by three layers of organization, Genes Dev.
13 (1999) 1871­1883.
[84] T.L. Peoples, E. Dean, O. Gonzalez, L. Lambourne, S.M. Burgess,
Close, stable homolog juxtaposition during meiosis in budding yeast
is dependent on meiotic recombination, occurs independently of
synapsis, and is distinct from DSB-independent pairing contacts,
Genes Dev. 16 (2002) 1682­1695.
[85] S. Jinks-Robertson, T.D. Petes, High-frequency meiotic gene con-
version between repeated genes on nonhomologous chromosomes in
yeast, Proc. Natl. Acad. Sci. U. S. A. 82 (1985) 3350­3354.
[86] J.E. Haber, W.Y. Leung, R.H. Borts, M. Lichten, The frequency of
meiotic recombination in yeast is independent of the number and
position of homologous donor sequences: implications for chromo-
some pairing, Proc. Natl. Acad. Sci. U. S. A. 88 (1991) 1120­1124.
[87] J. Dekker, K. Rippe, M. Dekker, N. Kleckner, Capturing chromo-
some conformation, Science 295 (2003) 1306­1311.
[88] M.M. Golic, K.G. Golic, A quantitative measure of the mitotic pair-
ing of alleles in Drosophila melanogaster and the influence of struc-
tural heterozygosity, Genetics 143 (1996) 385­400.
[89] M.M. Golic, K.G. Golic, Engineering the Drosophila genome:
Chromosome rearrangements by design, Genetics 144 (1996)
1693­1711.
[90] H. Scherthan, J. Bahler, J. Kohli, Dynamics of chromosome organi-
zation and pairing during meiotic prophase in fission yeast, J. Cell
Biol. 127 (1994) 273­285.
[91] K. Nabeshima, Y. Kakihara, Y. Hiraoka, H. Nojima, A novel meio-
sis-specific protein of fission yeast, Meu13p, promotes homologous
pairing independently of homologous recombination, EMBO J. 20
(2001) 3871­3881.
[92] M. Molnar, E. Doll, A. Yamamoto, Y. Hiraoka, J. Kohli, Linear
element formation and their role in meiotic sister chromatid cohesion
and chromosome pairing, J. Cell Sci. 116 (2003) 1719­1731.
[93] D.K. Nag, H. Scherthan, B. Rockmill, J. Bhargava, G.S. Roeder,
Heteroduplex DNA formation and homolog pairing in yeast meiotic
mutants, Genetics 141 (1995) 75­86.
[94] B. Rockmill, M. Sym, H. Scherthan, G.S. Roeder, Roles for two
RecA homologs in promoting meiotic chromosome synapsis, Genes
Dev. 9 (1995) 2684­2695.
[95] R.S. Cha, B.M. Weiner, S. Keeney, J. Dekker, N. Kleckner, Progres-
sion of meiotic DNA replication is modulated by interchromosomal
interaction proteins, negatively by Spo11p and positively by Rec8p,
Genes Dev. 14 (2000) 493­503.
[96] J.Y. Leu, P.R. Chua, G.S. Roeder, The meiosis-specific Hop2 protein
of S. cerevisiae ensures synapsis between homologous chromo-
somes, Cell 94 (1998) 375­386.
[97] H. Tsubouchi, G.S. Roeder, The Mnd1 protein forms a complex with
Hop2 to promote homologous chromosome pairing and meiotic
double-strand break repair, Mol. Cell. Biol. 22 (2002) 3078­3088.
[98] J.Y. Lee, T.L. Orr-Weaver, The molecular basis of sister-chromatid
cohesion, Annu. Rev. Cell Dev. Biol. 17 (2001) 753­777.
[99] F. Klein, P. Mahr, M. Galova, S.B.C. Buonomo, C. Michaelis, K.
Nairz, K. Nasmyth, A central role for cohesins in sister chromatid
cohesion, formation of axial elements, and recombination during
yeast meiosis, Cell 98 (1999) 91­103.
[100] M. Molnar, J. Bahler, M. Sipiczki, J. Kohli, The rec8 gene of Schizo-
saccharomyces pombe is involved in linear element formation, chro-
mosome pairing, and sister chromatid cohesion during meiosis,
Genetics 141 (1995) 61­73.
[101] S. Parisi, M.J. McKay, M. Molnar, M.A. Thompson, P.J. van der Spek,
E. van Drunen-Schoenmaker, R. Kanaar, E. Lehmann, J.H.J.
Hoeijmakers, J. Kohli, Rec8p, a meiotic recombination and sister
chromatid cohesion phosphoprotein of the Rad21p family con-
served from fission yeast to humans, Mol. Cell. Biol. 19 (1999)
3515­3528.
[102] Y. Li, F. Numata, W.P. Wahls, G.R. Smith, Region-specific meiotic
recombination in Schizosaccharomyces pombe: the rec11 gene, Mol.
Microbiol. 23 (1997) 869­878.
[103] T.S. Kitajima, S. Yokobayashi, M. Yamamoto, Y. Watanabe, Distinct
cohesin complexes organize meiotic chromosome domains, Science
300 (2003) 1152­1155.
[104] P. Pasierbek, M. Jantsch, M. Melcher, A. Schleiffer, D. Schweizer, J.
Loidl, A Caenorhabditis elegans cohesion protein with functions in
meiotic chromosome pairing and disjunction, Genes Dev. 15 (2000)
1349­1360.
[105] X. Bai, B.N. Peirson, F. Dong, C. Xue, C.A. Makaroff, Isolation and
characterization of SYN1, a RAD21-like gene essential for meiosis
in Arabidopsis, Plant Cell 11 (1999) 417­430.
[106] X. Cai, F. Dong, R.E. Edelmann, C.A. Makaroff, The Arabidopsis
SYN1 cohesin protein is required for sister chromatid arm cohesion
and homologous chromosome pairing, J. Cell Sci. 116 (2003)
2999­3007.
[107] A.M. Bhatt, C. Lister, T. Page, P. Fransz, K. Findlay, G.H. Jones,
H.G. Dickinson, C. Dean, The DIF1 gene of Arabidopsis is required
for meiotic chromosome segregation and belongs to the REC8/
RAD21 cohesin gene family, Plant J. 19 (1999) 463­472.
[108] A.P. Caryl, G.H. Jones, F.C.H. Franklin, Dissecting plant meiosis
using Arabidopsis thaliana mutants, J. Exp. Bot. 54 (2003) 25­38.
[109] F. Wang, J. Yoder, I. Antoshechkin, M. Han, Caenorhabditis elegans
EVL-14/PDS-5 and SCC-3 are essential for sister chromatid cohe-
sion in mitosis and meiosis, Mol. Cell. Biol. 23 (2003) 7698­7707.
[110] R.D. James, J.A. Schmiesing, A.H.F.M. Peters, K. Yokomori, C.M.
Disteche, Differential association of SMC1a and SMC3 proteins
with meiotic chromosomes in wild-type and SPO11-deficient male
mice, Chromosom. Res. 10 (2002) 549­560.
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180178
[111] M. Eijpe, H. Offenberg, R. Jessberger, E. Revenkova, C. Heyting,
Meiotic cohesin REC8 marks the axial elements of rat synaptonemal
complexes before cohesins SMC1b and SMC3, J. Cell Biol. 160
(2003) 657­670.
[112] J. Lee, T. Iwai, T. Yokota, M. Yamashita, Temporally and spatially
selective loss of Rec8 protein from meiotic chromosomes during
mammalian meiosis, J. Cell Sci. 116 (2003) 2781­2790.
[113] W.J. Cummings, S.T. Merino, K.G. Young, L. Li, C.W. Johnson,
E.A. Sierra, M.E. Zolan, The Coprinus cinereus adherin Rad9 func-
tions in Mre11-dependent DNA repair, meiotic sister-chromatid co-
hesion and meiotic homolog pairing, Proc. Natl. Acad. Sci. U. S. A.
99 (2002) 14958­14963.
[114] P.J. Pukkila, K.B. Shannon, C. Skrzynia, Independent synaptic be-
havior of sister chromatids in Coprinus cinereus, Can. J. Bot. 73
(Suppl. 1) (1995) S215­S220.
[115] S.T. Merino, W.J. Cummings, S.N. Acharya, M.E. Zolan, Replica-
tion-dependent early meiotic requirement for Spo11 and Rad50,
Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 10477­10482.
[116] M.P. Maguire, The mechanism of meiotic homologue pairing,
J. Theor. Biol. 106 (1984) 605­615.
[117] S.M. Albini, G.H. Jones, Synaptonemal complex spreading in Allium
cepa and A. fistulosum: I. The initiation and sequence of pairing,
Chromosoma 95 (1987) 324­338.
[118] L.K. Anderson, S.M. Stack, Nodules associated with axial cores and
synaptonemal complexes during zygotene in Psilotum nudum, Chro-
mosoma 97 (1988) 96­100.
[119] B.D. McKee, S.E. Lumsden, S. Das, The distribution of male mei-
otic pairing sites on chromosome 2 of Drosophila melanogaster:
meiotic pairing and segregation of 2-Y transpositions, Chromosoma
102 (1993) 180­194.
[120] R.P. Lifton, M.L. Goldberg, R.W. Karp, D.S. Hogness, The organi-
zation of the histone genes in Drosophila melanogaster: function
and evolutionary implications, Cold Spring Harbor Symp. Quant.
Biol. 42 (1977) 1047­1051.
[121] J.E. Castelli-Gair, J.L. Micol, A. Garcia-Bellido, Transvection in
the Drosophila Ultrabithorax gene: a Cbx1 mutant allele induces
ectopic expression of a normal allele in trans, Genetics 126
(1990) 177­184.
[122] J.L. Micol, A. Garcia-Bellido, Genetic analysis of transvection
effects involving Contrabithorax mutations in Drosophila mela-
nogaster, Proc. Natl. Acad. Sci. U. S. A. 85 (1988) 1146­1150.
[123] H. Scherthan, A bouquet makes ends meet, Nat. Rev., Mol. Cell
Biol. 2 (2001) 621­627.
[124] P.R. Chua, G.S. Roeder, Tam1, a telomere-associated meiotic pro-
tein, functions in chromosome synapsis and crossover interference,
Genes Dev. 11 (1997) 1786­1800.
[125] M.N. Conrad, A.M. Dominguez, M.E. Dresser, Ndj1, a meiotic
telomere protein required for normal chromosome synapsis and seg-
regation in yeast, Science 276 (1997) 1252­1255.
[126] E. Lifschytz, D. Hareven, Heterochromatin markers: Arrangement of
obligatory heterochromatin, histone genes and multisite gene fami-
lies in the interphase nucleus of D. melanogaster, Chromosoma 86
(1982) 443­455.
[127] L. Comai, A.P. Tyagi, M.A. Lysak, FISH analysis of meiosis in
Arabidopsis polyploids, Chromosom. Res. 11 (2003) 217­226.
[128] Y. Chikagashe, D.Q. Ding, H. Funabiki, T. Haraguchi, S. Yanagida,
Y. Hiraoka, Telomere-led premeiotic chromosome movement in fis-
sion yeast, Science 264 (1994) 270­273.
[129] O. Niwa, M. Shimanuki, F. Miki, Telomere-led bouquet forma-
tion facilitates homologous chromosome pairing and restricts ec-
topic interaction in fission yeast meiosis, EMBO J. 19 (2000)
3831­3840.
[130] S.M. Stack, Heterochromatin, the synaptonemal complex and cross-
ing over, J. Cell. Sci. 71 (1984) 159­176.
[131] S. Henikoff, K. Ahmad, J.S. Platero, B. van Steensel, Heterochro-
matic deposition of centromeric histone H3-like proteins, Proc. Natl.
Acad. Sci. U. S. A. 97 (2000) 716­721.
[132] B.D. McKee, The license to pair: identification of meiotic pairing
sites in Drosophila, Chromosoma 105 (1996) 135­141.
[133] R.S. Hawley, H. Irick, A.E. Zitron, D.A. Haddox, A. Lohe, C. New,
M.D. Whitley, T. Arbel, J. Jang, K. McKim, G. Childs, There are
two mechanisms of achiasmate segregation in Drosophila females,
one of which requires heterochromatic homology, Dev. Genet. 13
(1993) 440­467.
[134] G.H. Karpen, M.H. Le, H. Le, Centric heterochromatin and the
efficiency of achiasmate disjunction in Drosophila female meiosis,
Science 273 (1996) 118­122.
[135] A.F. Dernburg, J.W. Sedat, R.S. Hawley, Direct evidence of a role
for heterochromatin in meiotic chromosome segregation, Cell 86
(1996) 135­146.
[136] R.S. Hawley, Chromosomal sites necessary for normal levels of
meiotic recombination in Drosophila melanogaster I. Evidence for
and mapping of the sites, Genetics 94 (1979) 625­646.
[137] D.G. Albertson, A.M. Rose, A.M. Villeneuve, Chromosome organi-
zation, mitosis, and meiosis, in: D.L. Riddle (Ed.), C. elegans II,
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY,
1997, pp. 47­78.
[138] A.M. Villeneuve, A cis-acting locus that promotes crossing over
between X chromosomes in Caenorhabditis elegans, Genetics 136
(1994) 887­902.
[139] T.D. Dreesen, S. Henikoff, K. Loughney, A pairing-sensitive ele-
ment that mediates trans-inactivation is associated with the Droso-
phila brown gene, Genes Dev. 5 (1991) 331­340.
[140] L.E. Martin-Morris, K. Loughney, E.O. Kershisnik, G. Poortinga, S.
Henikoff, Characterization of sequences responsible for trans-inac-
tivation of the Drosophila brown gene, Cold Spring Harbor Symp.
Quant. Biol. 58 (1993) 577­584.
[141] C.J.A. Sigrist, V. Pirrotta, Chromatin insulator elements block the
silencing of a target gene by the Drosophila Polycomb Response
Element (PRE) but allow trans interactions between PREs on diffe-
rent chromosomes, Genetics 147 (1997) 209­221.
[142] M. Muller, K. Hagstrom, H. Gyurkovics, V. Pirrotta, P. Schedl, The
Mcp element from the Drosophila melanogaster Bithorax complex
mediates long-distance regulatory interactions, Genetics 153 (1999)
1333­1356.
[143] J. Mirkovitch, M.E. Mirault, U.K. Laemmli, Organization of the
higher-order chromatin loop: specific DNA attachment sites on nu-
clear scaffold, Cell 39 (1984) 223­232.
[144] S.M. Gasser, U.K. Laemmli, The organization of chromatin loops:
characterization of a scaffold attachment site, EMBO J. 5 (1986)
511­518.
[145] S.M. Gasser, U.K. Laemmli, Cohabitation of scaffold binding
regions with upstream enhancer elements of three developmentally
regulated genes of Drosophila, Cell 46 (1986) 521­530.
[146] W.C. Forrester, C. Vangenderen, T. Jenuwein, R. Grosschedl, De-
pendence of enhancer-mediated transcription of the immunoglobu-
lin-u gene on nuclear matrix attachment regions, Science 265 (1994)
1221­1225.
[147] B.J. Aronow, C.A. Ebert, M.T. Valerius, S.S. Potter, D.A. Wiginton,
D.P. Witte, J.J. Hutton, Dissecting a locus control region: Facilitation
of enhancer function by extended enhancer-flanking sequences, Mol.
Cell. Biol. 15 (1995) 1123­1135.
[148] W.C. Forrester, L.A. Fernandez, R. Grosschedl, Nuclear matrix at-
tachment regions antagonize methylation-dependent repression of
long-range enhancer­promoter interactions, Genes Dev. 13 (1999)
3003­3014.
[149] L.A. Fernandez, M. Winkler, R. Grosschedl, Matrix attachment re-
gion-dependent function of the immunoglobulin u enhancer involves
histone acetylation at a distance without changes in enhancer occu-
pancy, Mol. Cell. Biol. 21 (2001) 196­208.
[150] Y.L. Chua, L.A. Watson, J.C. Gray, The transcriptional enhancer of
the pea plastocyanin gene associates with the nuclear matrix and
regulates gene expression through histone acetylation, Plant Cell
15 (2003) 1468­1479.
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180 179
[151] L.D. Strausbaugh, S.M. Williams, High density of an SAR-associated
motif differentiates heterochromatin from euchromatin, J. Theor. Biol.
183 (1996) 159­167.
[152] R.A. McKnight, A. Shamay, L. Sankaran, R.J. Wall, L.
Hennighausen, Matrix-attachment regions can impart position-in-
dependent regulation of a tissue-specific gene in transgenic mice,
Proc. Natl. Acad. Sci. U. S. A. 89 (1992) 6943­6947.
[153] G.C. Allen, G.E. Hall, L.C. Childs, A.K. Weissinger, S. Spiker,
W.F. Thompson, Scaffold attachment regions increase reporter gene
expression in stably transformed plant cells, Plant Cell 5 (1993)
603­613.
[154] G.C. Allen, G. Hall, S. Michalowski, W. Newman, S. Spiker, A.K.
Weissinger, W.F. Thompson, High-level transgene expression in
plant cells: effects of a strong scaffold attachment region from to-
bacco, Plant Cell 8 (1996) 899­913.
[155] L. Mlynarova, A. Loonen, J. Heldens, R.C. Jansen, P. Keizer, W.J.
Stiekema, J.P. Nap, Reduced position effect in mature transgenic
plants conferred by the chicken lysozyme matrix-associated re-
gion, Plant Cell 6 (1994) 417­426.
[156] L. Mlynarova, R.C. Jansen, A.J. Conner, W.J. Stiekema, J.P. Nap,
The MAR-mediated reduction in position effect can be uncoupled
from copy-number dependent expression in transgenic plants, Plant
Cell 7 (1995) 599­609.
[157] V.C. Blasquez, M. Xu, S.C. Moses, W.T. Garrard, Immunoglobulin-
kappa-gene expression after stable integration: I. Role of the intronic
MAR and enhancer in plasmacytoma cells, J. Biol. Chem. 264
(1989) 21183­21189.
[158] A. Stief, D.M. Winter, W.H. Stratling, A.E. Sippel, A nuclear DNA
attachment element mediates elevated and position-independent
gene activity, Nature 341 (1989) 343­345.
[159] S.J. Namciu, K.B. Blocklinger, R.E.K. Fournier, Human matrix at-
tachment regions insulate transgene expression from chromosomal
position effects in Drosophila melanogaster, Mol. Cell. Biol. 18
(1998) 2382­2391.
[160] S. Nabirochkin, M. Ossokina, T. Heidmann, A nuclear matrix/scaf-
fold attachment region co-localizes with the gypsy retrotransposon
insulator sequence, J. Biol. Chem. 273 (1998) 2473­2479.
[161] R. Lupo, A. Breiling, M.E. Bianchi, V. Orlando, Drosophila chro-
mosome condensation proteins topoisomerase II and Barren coloc-
alize with Polycomb and maintain Fab-7 PRE silencing, Mol. Cell 7
(2001) 127­136.
[162] J. Mihaly, I. Hogga, J. Gausz, H. Gyurkovics, F. Karch, In situ
dissection of the Fab-7 region of the bithorax complex into a chro-
matin domain boundary and a Polycomb-response element, Deve-
lopment 124 (1997) 1809­1820.
[163] J. Zhou, H. Ashe, C. Burks, M. Levine, Characterization of the
transvection-mediating region of the Abdominal-B locus in Dro-
sophila, Development 126 (1999) 3057­3065.
[164] T.I. Gerasimova, V.G. Corces, Chromatin insulators and boundaries:
effects on transcription and nuclear organization, Annu. Rev. Genet.
35 (2001) 193­208.
[165] T.I. Gerasimova, K. Byrd, V.G. Corces, A chromatin insulator
determines the nuclear localization of DNA, Mol. Cell 6 (2000)
1025­1035.
[166] K. Ishii, G. Arib, C. Lin, G. van Houwe, U. Laemmli, Chromatin
boundaries in budding yeast: the nuclear pore connection, Cell 109
(2002) 551­562.
[167] J. Bode, Y. Kohwi, L. Dickinson, T. Joh, D. Klehr, C. Mielke, T.
Kohwi-Shigematsu, Biological significance of unwinding capability
of nuclear matrix-associating DNAs, Science 255 (1992) 195­197.
B.D. McKee / Biochimica et Biophysica Acta 1677 (2004) 165­180180