Chromosoma (2006) 115: 175­194
DOI 10.1007/s00412-006-0055-7
REVIEW
Nancy Kleckner
Chiasma formation: chromatin/axis interplay
and the role(s) of the synaptonemal complex
Received: 1 December 2005 / Revised: 18 January 2006 / Accepted: 20 January 2006 / Published online: 23 March 2006
# Springer-Verlag 2006
Abstract Meiotic recombination proceeds in biochemical
complexes that are physically associated with underlying
chromosome structural axes. In this study, we discuss the
organizational basis for these axes, the timing and nature of
recombinosome/axis organization with respect to the
prophase program of DNA and to structural changes, and
the possible significance of axis organization. Furthermore,
we discuss implications and extensions of our recently
proposed mechanical model for chiasma formation.
Finally, we give a broader consideration to past and present
models for the role of the synaptonemal complex.
Abbreviations L: leptotene . Z: zygotene . P: pachytene .
EP: early pachytene . MP: middle pachytene .
LP: late pachytene
Introduction
During meiosis, chromosomes remain in an organized,
individualized state for an extended period of time. This
same period includes recombination at the DNA level,
which has the unique feature that many of its steps occur
within biochemical complexes that are physically associated
with underlying chromosome structural axes. The
simple linear organization of meiotic prophase chromosomes,
combined with the complex program of global and
local events, provide a unique opportunity to study the
nature and roles of basic chromosome structure and the
functional interplay between that structure and chromatin
status. Most importantly, the DNA events of meiosis
provide a biochemical readout that is not available in
studies of organized mitotic chromosomes. The current
article brings together, and further elaborates upon, a
number of ideas on these subjects previously published by
our laboratory.
Meiotic prophase chromosome axes
Meiotic midprophase chromosomes are well-individualized
linear entities whose organizational structure can be
precisely described (Moens and Pearlman 1988; Zickler
and Kleckner 1999). It will not be surprising if the basic
structure described for these chromosomes is also present
in mitotic prophase.
Co-oriented linear loop arrays
At the pachytene stage of meiosis, homologous chromosomes
are joined via the synaptonemal complex (SC).
Along each side of the proteinaceous SC, the chromatin of
each chromatid is organized into a series of loops. The
bases of these loops comprise a geometric axis, which, as
elaborated by proteins, comprises a structural "axis" or
"core." Furthermore, the axes of sister chromatids are
stacked one above the other, perpendicular to the plane of
the SC ribbon, tightly conjoined into a single morphological
unit and with their chromatin loops emanating outward
from the SC. Thus, within pachytene chromosomes, sister
chromatid linear loop arrays are co-oriented.
Conserved loop density along pachytene axes
Detailed analysis of SC lengths, genome sizes, and/or loop
lengths in spread preparations reveals that the density of
loops along a pachytene axis is quite conserved across a
wide variety of organisms, at 20 loops per micron of axis
length (Zickler and Kleckner 1999; Table 1 of Appendix).
This feature points to the existence of some type of
Communicated by R. Benavente
The synaptonemal complex--50 years
N. Kleckner (*)
Herchel Smith Professor of Molecular Biology,
Harvard University,
Cambridge, MA 02138, USA
e-mail: kleckner@fas.harvard.edu
Tel.: +1-6174954278
Fax: +1-6174958308
organizational "loop base module," whose dimensions or
spacing is evolutionarily conserved. Another implication of
this regularity is that differences in total genome size
among different organisms are accommodated by differences
in loop size or total axis length (see also Moens and
Pearlman 1988), not by variations in basic axial structure.
It is interesting to note that pachytene axis lengths can
vary considerably within a given species, despite the fact
that total genome size/DNA content is essentially constant
(e.g., human male and female SC; reviewed in Kleckner et
al. 2003). If loop densities are the same in all cases,
differences in axis length in such situations should be
accompanied by inversely correlated differences in loop
size. This prediction is strikingly confirmed. Mutant mice
lacking the axis component SMC1-beta exhibit half the SC/
axis length and twice the loop size of wild-type mice
(Revenkova et al. 2004); conversely, human female
exhibits twice the SC/axis length of male and about half
the loop size (Tease and Hulten 2004).
Axial substructure comprises dual loop modules
In a number of organisms, electron microscopic studies of
SC lateral elements reveal regular striations that extend
continuously across the two sister axes (review in Zickler
and Kleckner 1999). This continuity implies that structural
features along the two sister chromatid axes, i.e., their loopbase
modules, must be "in register" as a "dual loop
module." It is not known whether "paired" sister modules
usually or necessarily form on homologous DNA regions
or whether the two loops within a dual loop module
sometimes or often contain different DNA sequences. Put
another way, these considerations suggest that each
homolog axis comprises a linear array of dual loop
modules, with intermodule links both along and between
sisters.
In accord with these two levels of organization, linkages
along or between chromatids can be disrupted differentially:
(1) there is a strong tendency for duality or splitting
of sister axes along their lengths, particularly at certain
stages (Dresser and Moses 1980; Zickler and Kleckner
1999) and (2) coordinate "fracturing" of conjoined axes
also occurs in certain mutants (e.g., van Heemst et al.
1999).
Toward a detailed model for axis organization
The above and other observations can be used to construct
a working model for axis organization.
1. Mitotic late-stage chromatids comprise arrays of
chromatin loops organized along a molecularly definable
axis. At the DNA level, this axis comprises
coalescences of locally AT-rich regions referred to as
an "AT queue" (Saitoh and Laemmli 1994). In yeast,
meiosis-specific axis-associated proteins are bound
preferentially to regularly spaced, locally AT-rich
regions, suggesting that the meiotic prophase axis
analogously comprises an AT queue (Blat et al. 2002).
2. There could be significant amounts of DNA within the
central region of the SC, all along its length (Ortiz et al.
2002), implying that DNA may be protruding "below"
the axis.
3. Two types of cytological studies suggest that axislocalized
protein components occur in multiple
horizontal "layers," with a "supra-axial meshwork"
present above the ultrastructurally defined axis. First,
when pachytene chromosomes are visualized by
electron microscopy, a thin element along each edge
that stains with heavy metals is revealed. This entity is
called the "lateral element" of the SC, and when
observed just before linking of homolog axes via SC
central region components, it is called an "axial
element." The precise nature of the stained entity
varies with the staining method. It is unclear whether
staining delineates epitopes for some particular protein
(e.g., Anderson et al. 2005; Pelttari et al. 2001) or a
clustering of epitopes from multiple components. In
contrast, early immunoelectron microscopic visualizations
reveal that two axis-associated proteins, Topoisomerase
II along mouse SCs and Hop1 along yeast
SCs, are localized "upward/outward" from the lateral
element, rather than along the lateral element itself
(Moens and Earnshaw 1989; Hollingsworth et al.
1990).
Second, immunofluorescence visualizations of
Sordaria chromosomes with the MPM2 antibody,
which recognizes phosphoprotein epitopes perhaps of
TopoII, reveal a thin line of staining in contrast to
antiSpo76/Pds5 and Rec8, which reveal much thicker
lines (van Heemst et al. 1999; D. Zickler, personal
communication). Thus, chromosome axes may comprise
not only the silver staining axis or core but a
much more complex and bulky supra-axial meshwork
of proteins bound to DNA at the bases of the loops
(Zickler and Kleckner 1999).
4. SCs lengths often vary modestly but systematically
over the different stages of pachytene; furthermore,
SCs are prone to twisting (Zickler and Kleckner 1999).
Both features suggest that the SC, while structurally
robust, is also substantially elastic. Elasticity would
presumably derive from DNA components along the
two homolog axes, again suggesting that these axes are
not simply continuous rigid rods but structures like the
proposed complex meshwork assemblies of protein
and DNA components.
5. In some organisms, axis length decreases dramatically,
about twofold, concomitant with SC formation. One
possible explanation is that the "loop base modules"
along the axis remain fixed in number and simply
become more tightly packed. If so, a decrease in length
would be accomplished by "scrunching-up" of axisassociated
DNA.
A model for the organization of meiotic prophase axes
that incorporates the above features is shown in Fig. 1.
176
Axis morphogenesis
Little is known about how chromosome axes assemble.
However, several interesting issues are as follows:
1. How is the positioning of loop-base modules determined?
The most obvious model for axis assembly
would involve binding of loop-base modules to
periodic sites along the chromatin, followed by the
coming together of adjacent modules into a regularly
spaced linear array. Chromatin loop size would thus be
determined by the spacing between bound modules
before assembly. Studies of chromatin loop size further
imply that module positioning has to accommodate
two important constraints. First, loop sizes are
relatively constant and characteristic along the chromosomes
of any given organism. Second, the same
DNA sequence sometimes yields axes having different
chromatin loop sizes according to the type of nucleus
in which it occurs, as seen not only within a species
(above), but also in artificial situations where DNA is
moved into an entirely unrelated species (Zickler and
Kleckner 1999). Neither of these constraints, and most
particularly the second, is explained in a simple way by
supposing that loop-base modules bind to specific
DNA sequences. Thus, we have proposed that chromatin
loop size, and therefore the positioning of loopbase
modules along the DNA/chromatin fiber, is
determined by chromatin fiber stiffness (persistence
length) (Zickler and Kleckner 1999). For example,
installation of one loop-base module might require a
pairwise interaction with another module and, thus, (at
least transient) formation of a chromatin loop; modules
could then be packed in as densely as possible up to the
point where chromatin fiber stiffness precluded further
loop formation.
2. The presence of a sister is not essential for formation of
a regular chromatid axis. In a mutant of Coprinus,
where meiotic prophase occurs without a preceding
round of DNA replication, single chromatids form axes
that are morphologically normal, except that they are
half the thickness of wild-type axes, and which go on
to form SCs (Pukkila et al. 1995). Thus, at least in this
case, axial organization and axis formation are independent
of inter-sister connections.
3. What molecule(s) directly mediate formation of linear
loop arrays? While a number of axis-associated
proteins have been identified by immunocytological
methods (e.g., Page and Hawley 2004), well-organized
chromosomes occur in all of the corresponding
mutants, implying that they do not comprise fundamental
axis building blocks. Included among these
mutants are those lacking the EM-defined axial/lateral
element feature, implying that this entity is not required
per se for development of axial chromosome structure
(e.g., Pelttari et al. 2001; Zickler and Kleckner 1999).
Based on the finding that the cohesin Rec8 localizes
very early along prophase chromosome axes, the
specific suggestion that cohesins comprise the fundamental
axial building block has been made (Eijpe et al.
2003). However, it is difficult to reconcile a fundamental
role for cohesins with the observation that axes
can form in the absence of a sister chromatid (above).
Furthermore, axis development is normal in rec8
mutants of several organisms including mouse,
Arabidopsis and Sordaria (Xu et al. 2005; Bannister
et al. 2004; Bhatt et al. 1999; D.Zickler and A.
Storlazzi, unpublished data). It is true that Rec8 is
required for formation of an axial element in both
budding and fission yeasts (Molnar et al. 2003; Klein et
al. 1999). However, the existence of earlier structural
organization has not been assessed in either case;
moreover, in budding yeast, axial element formation
occurs quite late, essentially concomitant with SC
formation (Padmore et al. 1991), in accord with the
possibility that earlier (basic) axis development is still
normal in a rec8 mutant. It begins to look as though
molecules important for development of mitotic prophase
chromosome axes should be prime candidates
for basic axial building blocks, e.g., Topoisomerase II
(e.g., Saitoh and Laemmli 1994; Gimenez-Abian et al.
1995) and shape-determining architectural proteins
(Strick and Laemmli 1995).
Connections between homolog axes
Cytological studies have detected prominent linkages
between homolog axes at three stages of meiosis.
1. The most prominent, of course, is the synaptonemal
complex (Heyting 2005; Page and Hawley 2004).
2. Before SC formation, many organisms exhibit presynaptic
co-alignment of homolog axes at a distance of
0.4 m. In favorable cases, this configuration is seen
to be mediated by discrete interaxis "bridges" (Albini
and Jones 1987). In some organisms, presynaptic
alignment is a prominent discrete stage that occurs
throughout the chromosomes before initiation of SC
formation anywhere in the genome (e.g., Zickler
Fig. 1 Model for assembly and organization of linear loop axis
array (a) and structure of co-oriented sister arrays characterized by
in-phase "dual-loop modules"
177
1977). In other organisms, bridge formation/alignment
is usually observed in nuclei that, in other regions, have
already begun to form SC, implying greater asynchrony
of different events within individual nuclei. Presynaptic
alignment and/or bridges have been reported
in plants, animals, and fungi; but not, thus far, in the
two best-studied invertebrates, Drosophila melanogaster
and Caenorhabditis elegans. (Further discussion
below).
3. Bridges between homolog axes have also been
observed at a much later stage, after pachytene, at the
sites of chiasmata on diplotene bivalents, e.g., in rat.
These late bridges contain SCP2, SCP3, SMC1, and
SMC3, but not Rec8, in rat (Eijpe et al. 2003).
Crossing-over at the DNA and axis levels via linked
parallel pathways
Three requirements of chiasma formation
Regular segregation of homologs to opposite poles at the
first meiotic division is directed by the presence of a
physical connection between the two segregating units, just
as for segregation of sisters during mitosis. In most
organisms, connections between homologs are provided by
chiasmata, each of which corresponds to a reciprocal
exchange of arms between one sister of each homolog.
Correspondingly, programmed recombination at the DNA
level is a prominent unique aspect of the meiotic program.
In addition, as chromatin is organized along structural
axes from meiotic prophase onward, crossing over with the
DNA/chromatin must be accompanied by an analogous
process of exchange at the axis level, at the same sites as,
and between, the same two chromatids as exchange at the
chromatin/DNA level. The existence of such exchange is
apparent from images of diplotene chromosome in which
the axes are differentially stained (Fig. 2). The need for
specialized axis-related events at sites of chiasmata is also
apparent from the fact that Rec8 is present at diplotene
between sister arms along their entire lengths, except that it
is absent at chiasmata (Eijpe et al. 2003).
In summary, formation of a chiasma involves a coordinate
set of local changes that include not only (1) exchange
at the DNA/chromatin level but also (2) exchange at the
axis level and (3) separation and differentiation of sister
chromatids at both levels, all of which must occur at the
appropriate corresponding positions along the involved
elements (Blat et al. 2002).
Recombinosome-axis association
A particularly striking feature of meiotic recombination is
that the involved biochemical complexes, the "recombinosomes,"
are spatially associated with their underlying
chromosome axes. This fact was first revealed by EM serial
section studies, which identified prominent nodules and
bars associated with the SC at pachytene, and by the
accompanying appreciation that the number and distribution
of these forms matches that of crossovers, as analyzed
genetically, and/or chiasmata, as analyzed cytologically. A
relationship between these "late" or "pachytene" nodules
was further supported by the remarkable finding of
Carpenter that DNA synthesis occurs at the sites of these
so-called "late nodules" (von Wettstein et al. 1984;
Carpenter 1975, 1981). This particular class of nodules is
now known to correspond to immunostaining foci of Mlh1
and Mlh3, which show the same pattern as these late
nodules (Barlow and Hulten 1998; Lipkin et al. 2002).
Other types of recombination nodules and correlated
immunostaining foci occur along SCs earlier in pachytene;
moreover, axis-association of recombinosomes is also
apparent at the presynaptic bridge stage (below).
In principle, there are two ways in which a spatial
juxtaposition of recombining DNA sequences and chromosome
axes could arise: (1) recombination might occur
preferentially between DNA sequences that are located
near the bases of their corresponding chromatin loops and
(2) recombination might occur within sequences that are,
organizationally, far from the bases of the loops, but with
the corresponding recombinosomes becoming "tethered"
to their underlying axes at some point in the process.
Molecular studies have revealed that the second of these
possibilities is in fact the case: recombination occurs within
"tethered loop­axis complexes" (Blat et al. 2002; Fig. 3a).
What is the rationale for recombinosome/axis
association?
Recombination occurs perfectly efficiently in situations
where chromosome axes are absent, i.e., in prokaryotic
organisms and during double-strand break (DSB) repair at
early stages of the eukaryotic mitotic cycle. As a result,
recombinosome­axis association is usually linked to
special requirements of meiosis. Furthermore, most considerations
suggest that axis association exists because it
provides additional useful features to the recombination
process. This feature characterizes many models for the
role of the SC, as well as explanations for the pre-SC roles
of axis components. However, we have proposed a
different idea that recombinosome/axis association exists
Fig. 2 Staining of diplotene bivalent axes reveals exchange at
chiasmata. Circle highlights the fact that chiasmata involve not only
exchange at the chromatin and axis levels but differentiation and
separation of sister chromatids at both levels. (from Blat et al. 2002)
178
due to the requirements of exchange at the structural axis
level. Localization of the recombinosome to the base of its
corresponding loop would target axis exchange such that it
occurs at the appropriate position, specifically along the
axis of the chromatid that is involved in exchange at the
DNA level, and would permit temporal and functional
coordination of relevant events at both levels (Fig. 3b).
It is now clear that recombination at the DNA level
involves a progression of programmed changes that occupy
most of prophase and that the corresponding recombinosomes
are associated with chromosome axes at many or all
of these stages (below). Correspondingly, the progression
of local changes at the DNA level that lead to formation of
a crossover should be accompanied by a correlated
progression of local changes along the axes that lead to
exchange at the axis level. This in turn leads to the idea of a
"linked parallel progression of changes at the DNA and
axis levels," with coordination of events at the two levels
along one particular chromatid on each partner, mediated
by direct physical association between recombination
complexes and their underlying axes (Blat et al. 2002).
DNA events of recombination and their temporal
and functional relationships to basic meiotic prophase
chromosomal stages
Meiotic recombination is initiated via programmed DSBs
mediated by the topoisomerase-like protein Spo11. Formation
of a break is accompanied by specific resection of 5
strand termini to give molecules with 300 nt 3 singlestranded
tails (Keeney 2001; Neale et al. 2005). The
majority of these DSBs appear to go on to interact with the
homologous sequence on a homolog chromatid, rather than
with the sister chromatid (Zickler and Kleckner 1999).
However, only a small subset of these interactions
ultimately yields a crossover. The remainder is resolved
in such a way as to restore two intact DNA molecules but
without accompanying exchange of flanking DNA sequences/chromosome
arms. To put it in another way:
"many are called, but few are chosen." More specifically, at
a particular point in the recombination process, a subset of
interactions is designated for maturation into crossover
products, with exchange of flanking DNA/chromosome
arms, while all remaining interactions are designated for
simple resolution without crossing-over. Designation of
certain recombinational interactions for eventual maturation
into crossovers is a tightly controlled process, which
results in specific spatial patterning of resulting crossovers/
chiasmata (e.g., Kleckner et al. 2004).
It was long thought that crossover/noncrossover differentiation
occurred at the very end of the recombination
process, at the time of (double) Holliday junction resolution.
However, recent studies have shown that this crucial
differentiation step occurs much earlier, before stable,
extensive interaction of DSB "tails" with a homologous
partner DNA sequence (Hunter and Kleckner 2001; Börner
et al. 2004; Bishop and Zickler 2004). The exact nature of
the DNA species upon which crossover control operates
has not been defined.
Once differentiation occurs, crossovers are formed via
two long-lived intermediates: single-end invasions (SEIs)
and double Holliday junctions (dHJs) (Hunter and
Kleckner 2001; Börner et al. 2004; Allers and Lichten
2001). The pathway for formation of noncrossovers has not
been elucidated, but synthesis-dependent strand annealing
is a currently favored possibility (Allers and Lichten 2001).
Two secondary points regarding crossover/noncrossover
differentiation are also noteworthy. First, synthesis-dependent
strand annealing may also involve something like an
SEI intermediate, albeit shorter lived and, thus, not
detected in mutants where the crossover-specific pathway
is abrogated (e.g., Börner et al. 2004). In that case,
crossover/noncrossover differentiation of recombinosomes
at the functional level may occur before appearance of any
difference in the structures of involved DNA intermediates.
Second, in some organisms, a subset of crossovers arise
that do not obey the spatial patterning rules of the majority
"specifically designated" crossovers (e.g., Copenhaver et
al. 2002). These "extra" crossovers have been proposed to
arise either as a minority species along the noncrossover
pathway (N. Hunter in Börner et al. 2004) or via a third
"branch" (Whitby 2005).
The DNA events of recombination occur in a specific,
evolutionarily conserved, temporal relationship to the basic
cytologically defined stages of meiosis in most, possibly
Fig. 3 Recombinosome­axis relationships. a Recombination occurs
in tethered loop­axis complexes (Blat et al. 2002). b Association of
a recombination complex to the base of its corresponding loop
simultaneously marks the relevant position along the chromatin,
along the axes, on the same chromatid at both levels puratively in a
"three-way flaw" (Blat et al. 2002). c Reeling-in model for
formation of interaxis bridges (Tessé et al. 2003; text)
179
all, organisms (Figs. 4a,b and 5).1
DSBs occur at leptotene,
followed by bridge formation and presynaptic co-alignment.
Crossover designation occurs at late leptotene,
followed by nucleation and extension of SC. SEIs and
dHJs occur during pachytene in the context of full-length
SC. Finally, maturation of dHJs into crossovers during late
pachytene. These relationships are known by direct
analysis of DNA events in yeast (Padmore et al. 1991;
Hunter and Kleckner 2001) and, for crossovers, in mouse
(Guillon et al. 2005). Noncrossover products appear either
at the same time as, or before, crossovers depending upon
the exact situation (temperature and/or molecular construct)
(Storlazzi et al. 1995; Börner et al. 2004; Allers and
Lichten 2001). These same relationships also occur in
many other organisms, as shown by analysis of immunostaining
foci corresponding to recombination complexes.
Because the length of the corresponding prophase period
varies from hours to days to weeks according to the
organism (Bennett 1977), this commonality is one of
several indications that the progression of DNA events is
governed by the overall meiotic program rather than by the
rate at which biochemical steps can be completed.
Recombination and axis-related events are not only
temporally correlated, but are directly functionally linked.
(1) Formation of interaxis bridges is dependent upon
formation of DSBs and thus appears to be mediated by the
recombination process itself. Indeed, the only molecular
components of these bridges known thus far are recombination
proteins, notably RecA homologs and the singlestrand
binding protein RPA (e.g., Franklin et al. 1999;
Tessé et al. 2003; Tarsounas et al. 1999; Plug et al. 1997;
Moens et al. 2002; de Vries et al. 2005; Fig. 5). It remains
to be determined whether these recombination complexes
comprise the bridges or whether they are layered upon
structural components. Given the existence of tethered
loop­axis complexes (above), it has been proposed that a
DSB present in a recombination complex associated with
one homolog axis would identify the corresponding
sequence in a homolog partner, thus triggering the
"reeling-in" of the partner axis (Tessé et al. 2003;
Fig. 3c). (2) Installation of SC involves specific nucleation
followed by polymerization outwards from nucleation
sites. Nucleation occurs preferentially at the sites of
recombinational interactions that have just been designated
to become crossovers (Keeney 2001). More specifically, it
appears that SC installation is nucleated specifically at sites
of crossover-designated interaxis bridges, concomitant
with their disappearance (for discussion, see Börner et al.
2004 below; Fig. 5).
In both of these phenomena, recombination-related
processes are determinative of structural progression.
Other likely related phenomena were also described.
To cite a few early examples (Alani et al. 1990; Bishop et
al. 1992): in yeast, a rad50 null mutation eliminates DSB
formation and also eliminates formation of axial elements,
as well as SC; a rad50S mutation permits DSB formation,
but eliminates resection at the ends of the breaks and
ensuing interaction with partner DNA permits axial
element formation and only a negligible amount of SC;
and a dmc1 null mutation that permits DSB formation and
resection but blocks immediately following steps permits
axial element formation and substantial (albeit significantly
delayed) SC formation.
It is also clear that, oppositely, axial structure is
important for recombination (below).
Linked parallel pathways of DNA and axis changes
The model proposed above envisions that recombinosome/
axis association targets axis changes to appropriate sites
and permits their functional integration with events at the
DNA level. In accord with the proposed targeting role,
recombinosome/axis association is established very early
in the chiasma formation pathway. It is clearly present by
the time of the "bridge stage" and even earlier, during
bridge formation; for example, in Sordaria, Rad51 foci can
be found on or between progressively juxtaposing homolog
axes (Tessé et al. 2003). Additionally, prominent nodules
can be observed in leptotene along chromosome axes that
are not yet in detectable proximity to their homologs (e.g.,
Zickler and Kleckner 1999). These structures might well
represent DSB-containing recombination complexes that
are associated with their underlying homolog axis even
before the DSB recognizes a homolog partner.
However, it is as yet unclear whether recombination
complexes are axis-associated even before recombination
begins, i.e., before DSB formation, or whether this
association is established immediately after a break has
occurred. Mitotic DSB repair occurs in the context of
chromosome structure proteins, as shown by DNA repair
defects in mutants lacking cohesin or the cohesinassociated
axis component Spo76/Pds5. Recent molecular
studies show that at least one cohesin accumulates in a
large domain around the site of the break (Unal et al. 2004).
It has been suggested, by analogy, that the programmed
DSBs of meiosis occur outside of the context of chromosome
axes and then migrate to the chromosome axes
thereafter (van Heemst and Heyting 2000). Consistent with
this possibility, in Sordaria, where axes are uniquely
discernible from very early leptotene on, Spo11 does not
exhibit obvious spatial association with chromosome axes
(Storlazzi et al. 2003). On the other hand, in yeast, deletion
of any of three known meiosis-specific axis components
(Red1, Hop1, or Rec8) alters the efficiency and/or
1 It has been proposed that the relationship of DNA events to
cytological stages might be different in Drosophila from that
described above as the general case. Specifically, it has been
suggested that DSBs occur after SC formation (McKim et al.
2002). The primary direct evidence for this view is that gammaH2AX
foci, which mark the sites of both mitotic and meiotic
DSBs, can be observed only at pachytene. However, this finding
would imply only that SCs form very rapidly after DSB formation,
e.g., with little or no "bridge stage." Alternatively, gamma-H2AX
foci have recently been reported also to occur during the pachytene
stage in grasshopper (Viera et al. 2004) as a second wave, separate
from the late leptotene wave. This raises the additional possibility
that, for some reason, only the later foci are detected in Drosophila,
thus leading to mistiming of DSB formation.
180
positions of DSBs (e.g., Blat et al. 2002; Glynn et al. 2004;
A. Jordan and N.K. unpublished data), implying some type
of participation of these components in break formation.
Similarly, in S. pombe, mutants lacking Rec8 exhibit severe
DSB defects (Ellermeier and Smith 2005). Furthermore,
Spo11 is also found to associate with chromosome axes in
many organisms at post-DSB stages, implying that there is
the potential for such association before DSB formation.
Fig. 5 Axis association of recombinosomes and temporal relationships
between DNA events and axis changes. Association
established before or immediately after DSB formation, leading to
nodules associated with individual homolog axes. Next stage
involves interaxis bridges that include RecA homologs. Finally,
axes are associated via SC, with associated nodules that represent
both pre-crossover and pre-noncrossover recombinational interactions.
Images are from Albini and Jones 1987 except far left (Stack
and Anderson 1986) and RecA homologs (from Franklin et al. 1999
and Tarsounas et al. 1999)
Fig. 4 Relationship between
DNA events of crossover formation
and chromosome morphogenetic
changes. a. Four
DNA transitions. b. Events relative
to chromosomal stages
depicted by staining of Sordaria
chromosome axes by Spo76GFP
(images courtesy of D.
Zickler)
181
Answering this question is further complicated by the
fact that DSB formation might, in some situations, be
contemporaneous with axis development, thus opening the
way to even more complex functional interrelationships.
Precisely what type of axis changes might be involved in
chiasma formation and when might they occur? It seems
probable that the pathway of local axis changes mirrors the
pathway of DNA events. DSB-mediated bridges appear to
represent essentially all initiated recombinational interactions.
Thus, recombinosome-axis associations before this
transition should be involved in setting up the bridge stage.
Notably, this process will involve concomitant targeting of
appropriate axis sites on the partner axis, as well as on the
initiating (DSB-associated) axis. Crossover designation
then occurs at a specific subset of bridge ensembles,
concomitantly triggering SC nucleation, at the leptotene/
zygotene transition (above). It is therefore expected that the
first crossover-specific change at the axis level should also
occur at this stage. What could this change be? Clearly, for
axis exchange to occur, a "hole" must be created in each of
the two involved axes; perhaps, this is the event that occurs
during crossover designation, with formation of the SC
helping to stabilize axial structure across these holes. This
model could explain why there is a pronounced tendency
for axes to be destabilized in a recombination-dependent
manner in certain mutant conditions, specifically at this
stage (e.g., van Heemst et al. 1999; Storlazzi et al. 2003). It
is also interesting that crossover-designated recombination
complexes are more tightly associated with newly formed
SC than are noncrossover complexes (Sherman et al.
1992), suggesting that recombinosome association may
also provide extra stability around these "sensitized sites."
At the DNA level, finalization of crossing over is wellseparated
functionally and in time from crossover designation.
The same would be true of axis exchange. Given that
crossover designation results in creation of axial holes,
actual exchange of the interrupted axes might then occur
during pachytene, concomitant with late stages of crossover
formation at the DNA level. We have proposed that
this exchange is mediated by twisting of the SC (Börner et
al. 2004; below).
How and why meiotic prophase might have evolved
from the latter stages of the mitotic program
It is becoming increasingly apparent that proteins normally
involved in regulating post-prophase stages of the mitotic
cell cycle are playing key roles in meiotic prophase, e.g.,
polo kinase (Clyne et al. 2003). These molecular findings
are anticipated by appreciation of several cytological
relationships (Kleckner et al. 2004). First, both meiotic
prophase and mitotic prometaphase can involve interaxis
bridges between homologs in the former case (above) and
between sisters in the latter (Gimenez-Abian et al. 1995).
Second, the Sordariaspo76-1 mutation confers analogous
tendencies for sister separation and chromatin diffuseness
at meiotic late-leptotene and at mitotic prometaphase (van
Heemst et al. 1999). Third, meiotic prophase involves
global cycles of histone H3 phosphorylation, a prominent
hallmark of mitotic prometaphase (Kleckner et al. 2004).
Finally, three cycles of global chromatin expansion and
contraction can be identified in meiotic prophase and
suggestively correlated with three corresponding cycles of
the mitotic program that encompass metaphase, preanaphase/anaphase,
and telophase/G1 (Kleckner et al.
2004). Taken together, these findings suggest that meiotic
prophase has evolved directly from the latter stages of the
mitotic program.
Why should this be the case? The key role of meiotic
prophase is to produce chiasmata. Production of chiasmata
involves local changes at the DNA level, and these appear
to have evolved from the mitotic DNA DSB repair process.
However, during meiosis, these DNA changes are
accompanied by corresponding local changes in chromosome
structure, along chromosome axes and between sister
chromatids, and changes in these structural components are
a hallmark of the latter stages of the mitotic cell cycle. To
put it another way, during meiotic prophase, structural
changes that would normally occur globally all along the
chromosomes, e.g., sister separation and axis coiling, are
constrained from occurring by robust axial structures and
instead occur locally, targeted specifically to the sites of
recombination reactions via tethered loop­axis complexes.
Thus, evolution of the meiotic prophase program can be
envisioned, coupling of the local events of DNA repair to
global chromosome structure changes of mitotic prometaphase
and later stages and, concomitantly, the corresponding
cell cycle regulatory mechanisms. The mechanical
model for chiasma formation described below is also in full
accord with this possibility, with changes at the structural
level playing a key role in triggering changes and
progression at the DNA level.
A mechanical model for chiasma formation
Crossover interference implies communication
along chromosomes
In most organisms, the distribution of crossovers/chiasmata
along and between meiotic chromosomes is strikingly
nonrandom. Three features are particularly striking (e.g.,
Kleckner et al. 2004).
First, along a particular chromosome, events occur at
different positions in different meiotic nuclei, implying a
stochastic element to the designation of crossover sites.
Second, every pair of homologs (every "bivalent")
acquires at least one crossover/chiasma, in accord with an
essential role in homolog segregation. Moreover, this
requirement is not met by the occurrence of a very large
number of randomly distributed crossovers sufficient to
ensure that the probability of zero events is very small: the
average number of crossovers per bivalent is small, usually
two or a few, and sometimes one and only one.
Third, when two or more crossovers are present, they
tend not to occur near one another, a phenomenon known
as "interference." The magnitude of this effect is maximal
182
at short intercrossover/chiasma distances and decreases
progressively with increasing inter-event distance.
The entire phenomenon of crossover designation can be
viewed as follows. Initially, a first undifferentiated precursor
interaction is designated for eventual maturation into a
crossover/chiasma. This first event sets up interference,
which disfavors the occurrence of additional crossover
designation events nearby. As a result, any second
crossover designation event will tend to occur away from
the first. More generally, as subsequent designation events
occur, they will tend to fill in the holes between the sites of
prior events. The overall result of these effects is a tendency
for crossovers/chiasmata to be relatively evenly spaced
along the chromosomes.
The phenomenon of interference is of particular interest
because it implies the existence of communication along
the chromosomes. The basis for this communication is
unknown. However, it is important that the patterning
exhibited by meiotic crossovers/chiasmata is also observed
for a number of basic chromosomal features, e.g., replication
origin firings, nucleosomes, and chromosome coils
(Kleckner et al. 2004). Thus, analysis of meiotic crossover/
chiasma distributions could potentially shed light on many
types of fundamental chromosomal phenomena.
A mechanical model for crossover designation
and interference
We have proposed that crossover designation and accompanying
interference occur by a mechanism that involves
imposition and relief of mechanical forces (stresses) along
the chromosome axes (Kleckner et al. 2004). Involvement
of mechanical forces is an attractive explanation for several
reasons, as discussed in detail below. Most importantly,
such a mechanism can automatically explain interference
because, in a physical system, communication can occur
automatically via the redistribution of mechanical stress or
stress relief.
As mentioned above, it appears that crossover designation
and accompanying patterning of crossover sites occur
at the "bridge stage." A specific physical model for the
basic events of this process could be as follows (Fig. 6):
mechanical compression (pushing) stress arises all along
the length of the axes. Interaxis bridges, being discontinuities,
comprise weak points or "flaws" along the axes.
Thus, as the level of pushing force rises, one end of some
bridge will eventually give way, with concomitant buckling
of the axis at the corresponding point. This effect, in turn,
will place the affected bridge ensemble in a new configuration.
And by hypothesis, that configuration would
commit the corresponding recombinational interaction to
the crossover fate.
This model readily accommodates the basic features of
crossover control as described above. Crossover designation
events will occur at different positions along a given
chromosome in different meiotic nuclei according to which
particular sites are affected by stress forces. Occurrence of
at least one crossover along each homolog is easily
achieved by ensuring that the imposed stress reaches a
suitably high level. And most importantly, this model
automatically explains crossover interference as follows.
(1) Buckling of a bridge will result in local relaxation
(relief of compression stress) at the site of the buckle. (2)
Disassembly of the bridge will then result in a corresponding
buckling and local relaxation on the partner axis. (3)
Given the elastic nature of chromosomes, local relief of
stress will automatically tend to redistribute, spreading
outward along the two axes from the two nucleation sites.
(4) The result of this spreading relief of stress is a
corresponding decrease to the probability that additional
(stress promoted) crossover designation events will occur
within the affected region.
A common example of stress redistribution occurs when
a rubber band is pulled and then cut. The rubber band
relaxes not only at the site of the cut, but all along its
length. In our model, the imposed stress is tensile
("pulling") rather than compressive ("pushing"), but the
principle is the same. It is also important in our case that
relaxation and relief of stress will not extend along the
entire length of the chromosome but instead will tend to be
absorbed by other ("stressed") components as it spreads,
Fig. 6 Mechanical model for
crossover designation and interference
via local and spreading
axis relaxation. SC nucleation
and spreading (not shown)
would occur as secondary
effects. (Börner et al. 2004)
183
thus decreasing in intensity with distance. The effect is a
graded spatial domain of stress relief whose dimensions
define the magnitude and extent of crossover interference.
One prediction of this model is that interference would
not involve polymerization of the synaptonemal complex.
This prediction has recently been verified (below). This
model further predicts that the axes of the homologs will
behave asymmetrically, because one axis is affected before
the other. Exactly such asymmetry has been observed in
Sordaria humana. At pachytene, axial bulges are observed
specifically in the vicinity of crossover-correlated recombination
nodules and specifically on only one homolog axis
(Zickler and Sage 1981). Finally, this model can even
accommodate the requirements of a mathematical model,
which proposes that adjacent crossover interactions are
separated by a fixed, integral number of noncrossover
interactions (Lande and Stahl 1993). This model is
interpreted to mean that interference mechanism "counts"
recombinational interactions; the basic point can be
restated more formally to say that the distance along the
chromosome over which interference is "felt" is influenced
or determined by the number of "undifferentiated" ensembles
encountered by the spreading interference signal.
Our stress model for interference readily explains such an
effect: as stress relief spreads along the chromosome axes,
there will be a tendency for it to be absorbed differentially
at the positions of encountered bridge ensembles because
these are especially stressed positions. Correspondingly,
the effective stress relief distance could be influenced, or
even specifically defined, by the number of undifferentiated
bridge ensembles encountered.
Source of the force?
How would compression stress arise along chromosome
axes? We have proposed that stresses are created by
chromatin expansion. It is normally envisioned that
chromatin decompaction is a passive disassembly process.
We suggest instead that when chromatin decompacts into a
large volume that expansion occurs in such a way as to
exert a force; that expansion force would push on any
constraining components, which would push back. These
diverse compression stresses would then perform chromosomal
work (Kleckner et al. 2004). For the particular case
of late leptotene chromosome axes, compression stress
would be generated because the presence of the axis would
constrain free expansion of the chromatin loops of the cooriented
linear loop arrays. Adjacent chromatin loops
would push against one another, initially causing extension
and distension along the axis. But as the axis reaches its
maximum degree of extension, chromatin will tend to
expand differentially in the "tops" of the loops where there
is more "room" as compared to along the axis. As a result,
the axis will tend to bend, with outward forces along the
tops of the loops (generating tension) and opposing inward
pushing forces along the axes (generating compression)
(Fig. 7). (In fact, in addition to compression stresses,
chromatin expansion forces are predicted to result in a
tendency for axis twisting, which will place adjacent loops
out of phase (Fig. 7). Both types of forces could be focused
coordinately, specifically at the flaws created by bridges,
thus promoting bridge buckling.) Buckling of the axis
would alleviate axial compression stress, and this alleviation
would then spread outward for some distance along
the axis, disfavoring the occurrence of additional stresspromoted
crossovers nearby, to a decreasing extent with
increasing distance.
In accord with this model, with respect to meiotic
crossover control, it has been shown that there is a
tendency for global chromatin expansion specifically at
late leptotene, exactly at the time of crossover designation
and interference, followed by a tendency for global
chromatin compaction during zygotene (Fig. 8).
Fig. 7 Predicted effects of
chromatin expansion forces
arising within a linear loop axis
array (from Kleckner et al.
2004)
184
Generalization of the model
In the model described above, global chromatin expansion
creates stresses along the chromosome axes; these stresses
are focused specifically at the sites of axis-associated
recombination complexes because they comprise the
weakest points along the axes. When a weak point (flaw)
gives way, it does so in a programmed way so as to produce
the desired effects. An analogous sequence of events could,
of course, occur at any step of the recombination reaction.
An even more general description of chiasma formation
can be formulated by considering components other than
chromosome axes. Chromatin expansion could promote
local changes at all three of the levels required for chiasma
formation (see above, Kleckner et al. 2004). (1) Chromatin
expansion could also create compression stresses along the
DNA/chromatin fiber that could promote buckling or local
denaturation of the DNA duplex and, thus, important
transitions at the DNA level. (2) Compression stress along
chromosome axes can promote a variety of different types
of axis changes beyond the buckling described above. (3)
Expanding sister chromatin masses would tend to push one
another apart, with a resulting tendency for separation both
within the chromatin and along the two underlying
chromatid axes. Furthermore, all three types of stresses
should be focused specifically to the same positions along
the corresponding components via axis-associated recombination
complexes: in this context, these ensembles
comprise "three-way flaws," which will target stresspromoted
changes to the appropriate corresponding
positions at all three levels (Fig. 3b). Moreover, as stresspromoted
changes begin to occur, coordination of changes
among components involved at the three different levels
can occur via direct mechanical linkage.
Such a scenario could be involved in any or all steps
along the chiasma formation pathway. In each case, global
chromatin expansion would impose stresses within the
chromatin, along chromosome axes and between sister
chromatids at both levels, and those stresses would be
targeted and transduced into appropriate local changes at
all three levels via axis-associated recombinosomes.
Moreover, redistribution of stress relief along the chromosomes
could occur not only along chromosome axes, but
also along the chromatin or along inter-sister boundaries,
and not only at the leptotene/zygotene transition, but also at
earlier stages, e.g., during DSB-formation.
In accord with these ideas, meiotic prophase involves
several different periods of global chromatin expansion;
each such period is accompanied by tendencies for sister
separation, as seen at the chromatin or axis levels, and by
tendencies for chromosome destabilization (e.g., Fig. 8).
Each of the important and known transitions of recombiFig.
8 Events of the leptotene/
zygotene transition. a Chromatin
volume increases and decreases
with concomitant individualization
and reconjunction of sister
chromatid chromatin (Dawe
et al. 1994). b Progressive increase
and decrease in histone
H3 phosphorylation in Sordaria
(D. Zickler in Kleckner et al.
2004). c Disappearance of
bridges with concomitant nucleation
of SC (Albini and Jones
1987)
185
nation at the DNA level occurs during one of these
expansion periods (Kleckner et al. 2004; Zickler and
Kleckner 1999; Dawe et al. 1994; van Heemst et al. 1999).
The crossover control transition modeled above occurs at
leptotene/zygotene. A model for a later stress-mediated
transition at pachytene is described below. It would not be
difficult to construct a model for stress-mediated DSB
formation, with or without axis-association of recombination
complexes.
Complexities of the leptotene/zygotene transition
viewed through the lens of the stress hypothesis
Crossover designation and interference are explained "in
principle" by imposition of stress plus local relief of stress
and redistribution of that stress relief along the chromosome
axes. But this is just the tip of the iceberg with regard
to the multiple requirements and events of this period, and
new ways of looking at these events emerge by considering
additional specific predictions of the stress hypothesis.
Chiasma designation at the axis level Commitment to
crossing over at the DNA level should be accompanied by
commitment to crossing over at the axis level, which, as
suggested above, might involve creation of axial holes at
the appropriate positions along the two involved chromatids.
It is easy to envision how mechanically promoted
axis buckling and ensuing elimination of interaxis bridges
might promote creation of such holes, i.e., at the "bridge
abutments."
Axis extension/distension and relaxation Chromatin expansion
should first cause the homolog axes to become
more extended (distended) as a simple way of alleviating
pushing forces between adjacent loops; then, once the axis
reaches its extension limit, compression stress begins to
arise (Fig. 7). If axis relaxation occurs locally (as the result
of crossover designation) or globally by redistribution of
stress relief (i.e., the "interference signal"), expansion
stresses will be alleviated and axes will become shorter
and floppier. Axis extension and relaxation could play
crucial roles in several aspects of the crossover control
transition beyond crossover designation and interference
in the general sense, as described in the next two sections.
Regulated interplay between progression and fate Crossover/noncrossover
differentiation has several requirements
not discussed above. First, regulated differentiation of prechiasma
interactions into pre-crossover and pre-noncrossover
types requires that the chiasma formation process be
arrested specifically at the decision point so that differentiation
can be imposed. Arrest presumably involves some
feature of the bridge ensemble while progression is
presumably licensed by bridge disassembly. Second, by
the scenario, an interaction is specified for maturation into
a noncrossover simply by the "absence of crossover
designation." Taken together, these requirements suggest
that bridge ensembles may all arrest in a pre-noncrossover
mode. Specific programmed events then designate some
ensembles to be crossovers/chiasmata, concomitantly
triggering bridge disassembly, and thus licensing progression
toward that fate. To ensure that "leftover" ensembles
proceed to the noncrossover fate, it would be necessary
only to release the progression block, i.e., to eliminate the
corresponding bridges. How are bridge ensembles arrested
and what promotes the disassembly of bridges that are not
crossover/chiasma designated? We suggest that extension/
distension of axes is required to stabilize and arrest bridge
ensembles at the decision point. If so, local stress relief
will destabilize bridges at sites of crossover designation
while the spreading stress relief signal that mediates
interference will concomitantly trigger destabilization of
encountered nondesignated bridges, thus sending their
interactions toward the noncrossover fate. Furthermore,
any bridges that are too far away from a crossover site to
be affected by the interference signal will be disassembled
automatically during global chromatin contraction and
accompanying global relief of stress at zygotene (Fig. 8).
SC formation Crossover designation is accompanied by
nucleation and polymerization of SC. These processes
clearly require elimination of bridges, which hold homolog
axes apart at a too far distance. We further propose that
SC installation requires relaxation of homolog axes, e.g.,
to permit stabilizing contacts between adjacent central
region transverse filament or central element subunits. By
the combination of both effects, SC would initially be
installed specifically at crossover designated sites and then
would spread outward, following and responding to the
interference signal, with installation in any leftover regions
during global relaxation at zygotene. Involvement of axis
relaxation in SC formation may also help to explain
certain other phenomena. For example, SC is installed
promiscuously at late pachytene in several organisms
(Zickler and Kleckner 1999) and this is a period of global
chromosome contraction (Kleckner et al. 2004). Moreover,
no late leptotene bridges are present at this stage. Thus, the
only requirement for SC formation should be close
juxtaposition of two axes, either by chance or by folding
back of a single axis upon itself, as is commonly observed.
In support of such a possibility, homolog axes are more
"floppy" at the end of pachytene (Kleckner et al. 2004).
Also, in several mutants where SC is installed very late in
an aberrantly prolonged prophase, the resulting chromosomes
have a marked "wiggly" appearance (e.g., Börner et
al. 2004).
The above considerations highlight the fact that, for
organisms that use the "canonical" program, initiation of
recombination (plus, by hypothesis, appropriate global
chromatin status) sets up conditions that prevent SC
formation. This makes it possible for additional features
to ensure that the structure is installed specifically at the
sites of pre-chiasmata, with appropriate timing.
Given this situation, it might be expected that SC would
form promiscuously in the absence of recombination
initiation, analogously to promiscuous SC formation at
late pachytene in other cases. Indeed, this is exactly the
186
case in Drosophila and C. elegans; however, in most
organisms, abrogation of DSB formation severely compromises
SC formation (for discussion, see Page and
Hawley 2004; Pawlowski and Cande 2005). The latter
phenotype may suggest that, in most cases, recombinationrelated
features play essential positive roles in SC formation,
as well as negative roles. For example, SC installation
involves functional interplay between close local juxtaposition
between homolog axes and relaxation/contraction
along homolog axes; in Drosophila and C. elegans, axis
relaxation may be strong enough that fortuitous contacts
between homolog axes suffice to nucleate SC. In other
organisms, in contrast, features that are related to
programmed bridge disassembly might be required. In
certain cases, homolog axes appear to come together and
literally touch one another at sites of SC nucleation (Albini
and Jones 1987; Zickler and Kleckner 1999), perhaps
because of axis buckling and/or via contraction of RecA
homolog/DNA filaments (Franklin et al. 1999). In any
case, the unique behavior of Drosophila and C. elegans is
often interpreted to mean that the two organisms have
significantly different programs during wild-type meiosis.
The above considerations suggest that this need not be the
case. Subtle differences in the relative contributions of
different components to SC formation and resulting
differences in the effects of a DSB-minus mutation could
be sufficient.
Implications of the stress hypothesis for pachytene
events, recombination nodules and synaptonemal
complexes
Our proposal suggests new ways of looking at the
pachytene stage and the roles of the SC and (late)
recombination nodules (RN). Perhaps these structures are
so prominent because they primarily play structural roles.
In the context of our proposed model, the SC might act as a
transducer of mechanical stress forces, while late recombination
nodules might target those effects to the appropriate
sites, as discussed above. Additionally, the structural
prominence of recombination nodules might reflect the fact
that a robust physical entity is required to appropriately
constrain and transduce global stress forces.
We have proposed that global chromatin expansion
during the pachytene stage exerts its effects along
chromosome axes by causing twisting of the SC (Börner
et al. 2004). Twisting will be specifically promoted because
it places the loops along each axis "out of phase" with one
another, giving them more room to expand and thus
alleviating chromatin expansional stress forces (Fig. 7).
Moreover, twists are a common feature of pachytene
chromosomes. Twisting forces, as well as other predicted
effects such as tendencies for bending or buckling, will
naturally be focused to the sites of late recombination
nodules, thereby promoting the appropriate local effects at
those sites.
Exactly what local effects are appropriate to this stage?
At the DNA level, the promoted step(s) could be the SEIto-dHJ
transition and/or, perhaps more likely, the resolution
of dHJs to crossovers. At the structural level, axis exchange
must be finalized, via disruption of old linkages and
formation of new ones. In this context, the SC could serve
two key roles. First, SC twisting is precisely the motion
required to promote axis exchange; thus, the SC might
specifically direct this process (see Börner et al. 2004).
Second, the SC might provide an external "clamp,"
stabilizing homolog associations via links in regions
flanking the exchange point while promoted changes are
in progress.
Domanial chromatin/axis interplay in mitotic
and meiotic cells
The above discussion has focused primarily on local effects
of chromatin expansion forces. There are two sets of
observations that point to functional interactions between
chromatin status and axis status, which could be explained
by global effects of chromatin expansion/contraction
status. Mammalian chromosomes are composed of locally
GC-rich and AT-rich isochore domains whose differential
staining leads to their designation as R-bands and G-bands,
respectively; yeast chromosomes are similarly organized
(for discussion, see Blat et al. 2002). In both cases,
chromatin in GC-rich regions is more expanded relative to
chromatin in AT-rich regions (Gilbert et al. 2004; J.
Dekker, personal communication). In yeast meiotic
chromosomes, the axis-associated protein Red1 is more
abundant along axes in R bands than in G bands (Blat et al.
2002), and in mammalian chromosomes, chromosome axes
are straight in R bands and coiled in G bands (Saitoh and
Laemmli 1994). These features both point to some type of
communication between domanial chromatin structure and
axis status in which chromatin expansion status influences
axis status. In the context of our stress hypothesis,
increased Red1 loading might be explained by a tendency
for wider (expansion promoted) separation between sister
axes in R-band regions. Axis coiling is predicted to be
promoted by chromatin expansion, the opposite of the
correlation observed. On the other hand, effects along and
between sister axes will be mechanically linked. Thus, if
chromatin expansion forces differentially promote sister
separation in R-band regions, stress forces might be
differentially alleviated in those domains. With the result,
there would be a lesser tendency for (stress promoted)
coiling in C. elegans and Drosophila but less so in budding
yeast or Coprinus.
What are the roles of the SC?
This issue can be subdivided into two related questions.
First, what does the SC actually do for the meiotic program
as it currently exists? Second, what was the evolutionary
driving force that has resulted in the appearance of this
structure?
187
Historical background
From the time of its discovery, the SC was known to
mediate the intimate synapsis of homolog axes at pachytene.
The discovery of SC-associated recombination nodules
then triggered considerations of possible roles of the
SC for crossover formation.
One idea of this type suggested that the SC might bring
chromosomes together to make recombination possible.
However, one early (and prescient) challenge to this idea
was provided by was Maguire, whose studies of maize
inversion heterozygotes led her to the conclusion that
nucleation of SC formation is a consequence of, or very
tightly linked in parallel with, what would now be called
crossover site designation (Maguire 1966; Maguire 1972).
A second challenge was raised by von Wettstein and Stern,
whose preselection hypothesis considered that, because
only a small fraction of DNA is located along/within the
SC, there had to be some determinant that ensured that
"corresponding regions on the homologs" were present
there (Stern et al. 1975). The general underlying concern
was the existence of a "chicken and egg" problem: if the
SC mediated synapsis to permit recombination, how was it
determined that the SC would form specifically between
homologs? In any case, subsequent studies have shown that
formation of SC is not the primary mechanism for bringing
homologs together. Rather, it appears to be the very last in a
series of steps (above). In fact, the functional relationship
between recombination and SC formation turns out to be
just the opposite from that originally proposed: in most
situations, recombination is needed to bring homologs
together so as to permit SC formation, rather than the other
way around.
A second type of early model proposed that the SC is
used for physical support of the recombination machinery
(e.g., Moses 1969; von Wettstein et al. 1984; Egel 1995).
This possibility is still viable. However, there remains the
question of why such support would be necessary for
meiotic recombination when it is not required for
recombinational repair of double-strand breaks in mitotically
dividing cells or for recombination in prokaryotes.
One might invoke the possibility that chromosome movements
during meiosis impose special requirements for
physical stabilization; on the other hand, some of the most
violent motions known occur in S.pombe (e.g., Yamamoto
et al. 2001), where no SC is formed. Thus, it seems as if
there must be more going on than physical stabilization.
A third set of proposed models invoked a role for the SC
in crossover interference. The first such model emerged
from Maguire's findings (above) and implied that SC
polymerization might be used to mediate interference (Egel
1995; Maguire 1988). This model also regained prominence
more recently for other reasons (Sym and Roeder
1994). A second early model assumed that crossover
designation occurred during pachytene and envisioned the
use of SC as static structure, i.e., a railroad track along
which the interference signal would travel (King and
Mortimer 1990; Stack and Anderson 1986; Carpenter
1987). Recent studies suggest, however, that neither of
these models is correct. Instead, crossover interference
occurs before and is independent of the SC, as seen in both
in yeast and Arabidopsis (Börner et al. 2004; Heyting
2005; Bishop and Zickler 2004; Fung et al. 2004; Higgins
et al. 2005), leading (e.g.) to the model described above.
Roles for SC in recombination
Mutants specifically lacking an SC central region protein
have now been described in several organisms. Formation
of crossovers is defective in all cases. However, the
magnitude of the observed defects can vary widely among
different situations. At one extreme, severe reductions in
crossovers are observed, e.g., in C. elegans syp1 mutants
(MacQueen et al. 2002) and in yeast zip1 mutants at high
temperature and/or the BR strain background (Börner et al.
2004). At the other extreme, an Arabidopsiszyp1 mutant
exhibits 80% of the wild-type level of crossovers (as well
as normal interference) (Higgins et al. 2005). These defects
likely reflect roles for SC components at the time of SC
nucleation at leptotene/zygotene (before SC polymerization).
Defects at this stage are apparent in SC central region
defective mutants of Saccharomyces cerevisiae, C. elegans,
mouse, and Arabidopsis as evidenced by (1) delayed
turnover of immunostaining foci corresponding to DSBs
(of RecA homologs, RPA and/or Msh4/5) and (2) in yeast,
at the DNA level, defective progression out of the DSB
stage (de Vries et al. 2005; other references above). The
severity of the recombinational progression block at this
stage varies considerably from organism to organism and in
yeast, according to both temperature and strain background,
approximately in relation to the severity in the
reduction of crossovers, suggesting that the block at this
stage is primarily responsible for the observed mutant
phenotypes.
Given these early defects, it is not possible to assess roles
for the SC at later stages. However, because late steps in
formation of crossovers occur during pachytene and
prominent crossover-correlated recombination complexes
occur along pachytene SCs, a role for the SC in these
events seems probable (e.g., above).
Mutants specifically lacking SC central regions are also
known in Drosophila; in this case, crossover formation is
defective but the reasons as to why and at what stage(s) are
not yet clear (e.g, Anderson et al. 2005).
Additional interesting possibilities for roles of the SC in
recombination are raised by the properties of an Arabidopsis
mutant lacking SC central region components, This
mutant exhibits only very modest reduction in crossovers
(which exhibit interference) (Higgins et al. 2005). In
contrast, the major effect of the mutant is the occurrence of
high levels of ectopic recombination. Although such an
effect has not been reported in other organisms, the very
high levels of homoeologous sequences in Arabidopsis
may make it an exceptionally sensitive system for detecting
such interactions, which are observable by basic cytological
characterization of chromosomes.
188
This phenotype could reflect a role of the SC either in
precluding the identification of ectopic partners at the DSB
stage or in rejecting nascent associations before they are
finalized. The latter seems more probable because (1) DSB/
partner associations arise before SC formation (above); (2)
global rejection of homoeologous associations is known to
occur in allohexaploid wheat immediately after formation
of SC between various pairs of appropriate and inappropriate
partners (reviewed in Zickler and Kleckner 1999); and
(3) the DNA mismatch repair machinery is known to
specifically preclude conversion of homoeologous interactions
into crossovers via a mechanism that requires
formation of significant heteroduplex DNA (Hunter et al.
1996), and such DNA normally arises during meiosis at the
leptotene/zygotene transition, concomitant with SC nucleation
at the corresponding sites (Hunter and Kleckner 2001).
Thus, one specific interpretation of the Arabidopsis
findings is that SC central region components (e.g., patches
of SC) are required for biochemical rejection of homoeologous
interactions after the onset of stable strand
invasion.
Roles for the SC beyond DNA recombination?
Axis exchange? As described above, we have suggested
that the SC might be needed to promote axis exchange
and/or or to stabilize homolog associations while that
process is in progress (Börner et al. 2004). On the other
hand, if the SC has such a role, how does one explain the
case of S. pombe and Aspergillus, where large numbers of
unregulated crossovers occur and homologs successfully
disjoin without benefit of an SC (Molnar et al. 2003;
Egel-Mitani et al. 1982)? The obvious response is that, in
these cases, some other mechanisms come into play.
Sister cohesion? Maguire proposed that the SC per se was
important for maintaining sister chromatid connections
(Maguire 1995). This model was based on the finding that
certain mutants make SC (albeit sometimes slightly
abnormal in structure), but then after dissolution of the
SC, exhibit defects in the number of chiasmata plus
apparent exchange of sister chromatid pairing partners as
indicated by the positions of homolog-specific heterochromatic
knobs. In support of this possibility, (1) a yeast
zip1 mutant does show low levels of sister segregation at
meiosis I in some conditions (N. Hunter and G.V. Börner,
personal communication), as does a Drosophila mutant
defective in SC assembly (Manheim and McKim 2003),
and (2) cytological observations point to a tendency for
sister axis separation along SCs at midpachtyene (Zickler
and Kleckner 1999).
Sensing of interlocks? It has been proposed that SC
polymerization might be important in sensing interchromosomal
entanglements and triggering an appropriate
response (Kleckner et al. 1991), e.g., resolution (Rasmussen
1986) or, more probably, arrest of the meiotic program
(Bishop et al. 1992). Two findings that might fit with such a
role are: (1) occurrence of high levels of multivalents in
Arabidopsis mutants lacking SC (Higgins et al. 2005) and
(2) occurrence of frequent ring bivalent interlocks in
grasshopper in correlation with absence of SC in these
bivalents might be attributable to absence of such a role. On
the other hand, G.H. Jones (personal communication)
points out that a number of organisms exhibit highly
localized SC formation around crossover sites on bivalents
that have only a single chiasma, a situation in which sensing
of interlocking at zygotene/pachytene is clearly not an issue.
Silencing? A recent report shows that chromatin is
silenced specifically at the junctions of translocation
heterozygotes where SC does not form, suggesting a
possible role for the SC in that process (Turner et al. 2005;
Schimenti 2005). However, such effects could also be
explained as yet another consequence of the failure of
DSBs to progress via a normal interaction with a partner.
Evolutionary rationale for existence of the SC
The above considerations still leave unanswered the basic
question of why the SC has evolved as a very common
feature of the meiotic program.
The fact that two fungi, S. pombe and Aspergillus, lack
both SCs and regulated crossover formation (above)
remains intriguing. The association of these two phenotypes
has long been attributed to the fact that the SC is
required for interference, but this explanation is no longer
tenable. An alternative possibility is exactly the opposite:
the SC might be required only in organisms that have
acquired regulated crossover designation, e.g., for "neatness"
of chromosomes that are only connected by a few
recombination-related linkages during pachytene (Kleckner
1996). Such a relationship might be favored by the fact that
crossover designation occurs earlier in prophase than SC
formation, as if the first feature had promoted evolution of
the second. Of course, we would then be left to explain the
evolutionary value of regulated crossover positioning, but
this remarkable feature requires an explanation in any case.
Here is one idea: perhaps interference is valuable because it
ensures that large chunks of chromosomes tend to remain
associated even when recombination occurs, thus mitigating
the tendency for crossing over to eliminate "good
combinations" of genes.
Suffice to say that, like so many other aspects of meiosis,
these issues provide fertile ground for future thought and
experimentation.
Acknowledgements This work was supported by National
Institutes of Health (USA) grants GM25326 and GM44794. The
author is grateful for ideas and comments from Denise Zickler,
Gareth H. Jones and members of the Kleckner lab, and for
manuscript help from Jim Henle.
189
Appendix
Table 1 Calculation of loop densities along pachytene SCs, Denise Zickler, Ruth Padmore and Nancy Kleckner
A. organism B. loop
length in
m
C. BDNA/
m
loop
D. haploid
DNA in Mb
E. haploid
B-DNA m
F. loops/
haploid
genome
G. SC length in
m
H. loop density J. references
Cols. B,D,G
S. cerevisiae 1 7 14 4760 680 28 24 1, 2, 3
0.8-1.2 5.6-8.4 14.9 5066 724 25.5-31.7 28-23; 32-25; 33-25;
34-20,3
1, 4, 5
14 4760 680 22.4-33.1 27-23; 30-20.5; 3123.4;
32-19
- -,2, 6
15 5100 729 21.8-29 mean: 25.9-24.3 - -. 5, 1
21.2-35.7 - -,- -,7
S. pombe 1 7 14 4760 680 34 20 8, 2, 8
0.5-0.6 3.5-4.4 13.8 4692 1082-1117 33.3-38 28-32 8, 9, 8
N. crassa 1.6 11 46 15640 1422 47 30 10, 2, 10
1.5-1.6 10.5-11.2 43 14620 1392-1554 29-45 48-31; 31.6-29.6
(1392)
10, 9, 10
48 16320 1329-1484 44-47 51-33; 33.8-31.7
(1489)
- -, 10, 11
29.5-34;5 23 - -,- -, 12
34 11500 1051 - -, 2- Coprinus
1 7 36 12240 1749 70 25 13, 14, 13
1-1.5 7-10.5 37 12580 1797 36-45 49.9-40 15, 16, 17
1166-1198 32.4-25.9
Zea mays 8-10 300 5000 1700 000 5667 353 16 18, 19, 18
240-300 4053-4200 1 394 000 5800-3560 110-443 mean
353
353: 13.2-16.4 (ref
20)- - 8.1-10.1 (ref 19)
- -, 20, 21
2291-2735 854 400 4647-2848 332-591 - -, 19, 18
219-633 mean
420
- -, - -, 22
7500 2 550 000 10625-8500 420: 11-13.8 (ref 20)
- - 6.7-8.5 (ref 19)
- -, 5, - 20.2-25.3-
-24-30 (ref
5)
Lycopersicum
aesculatum
6 180 2000 680 000 3778 207 18 23, 19, 23
5.25-6.75 158-202 869-1061 295 460 1927-2125 153-296 mean
207
207: 9.1-10.3 4, 20, 4
907-1000 308 380 1713-1888 194-279 mean
234
234: 7.3-9.1 - -, 19, 20
340 000 extremes: 6.5-13.8
Bombyx mori 0.83 25 500 170 000 6800 260 26 24, 2, 25
male 0.6 18 9444 L=180; Z=198;
P=258; LP=347
L=52; Z=47.6; P=36.6;
LP=27.2
26, - -, 26
female 203-220 46.5-43 - -, - -, 27
Mus musculus 5.57-7 165-210 3232 1 098 880 6660-6182 156 36.3-42.7; 41.1-48.5 1, 28, 29
6 180 3000 1 020 000 6105-5667 138-199 48-33.4 30, 5, 30
144-192 41-28.5 - -, - -, 31
190
A. organism B. loop
length in
m
C. BDNA/
m
loop
D. haploid
DNA in Mb
E. haploid
B-DNA m
F. loops/
haploid
genome
G. SC length in
m
H. loop density J. references
Cols. B,D,G
Home sapiens 9.6 288 3000 1 020 000 3542 32, 5, 32
male 4.7 Z=321; EP=207;
P=231;
15.3; 17; 15.3 - -, - -, 34
female 3.1 EP=533;
MP=458;
LP=519
6.6; 7.7; 6.8 - -, - -, 34
1. Moens PB, Pearlman RE (1988) Chromatin organization at meiosis. Bioessays 9:151­3
2. King JS, Mortimer RK (1990) A polymerization model of chiasma interference and corresponding computer simulation. Genetics
126:1127­1138
3. Loidl J, Nairz K, Klein F (1991) Meiotic chromosome synapsis in haploid yeast. Chromosoma 100:221­228
4. Peterson DG, Price HJ, Johnston JS, Stack SM. (1996) DNA content of heterochromatin and euchromatin in tomato (Lycopersicon
esculentum) pachytene chromosomes. Genome 39:77­82
5. Loidl J (1994) Cytological aspects of meiotic recombination. Experientia 50:285­94
6. Byers B, Goetsch L (1975) Electron microscopic observations on the meiotic karyotype of diploid and tetraploid Saccharomyces
cerevisiae. Proc Natl Acad Sci USA 72:5056­60
7. Dresser ME, Giroux CN (1988) Meiotic Chromosome behavior in spread preparations of yeast. J. Cell Biol 106:567­73
8. Bahler J, Wyler T, Loidl J, Kohli J. (1993) Unusual nuclear structures in meiotic prophase of fission yeast: a cytological analysis. J Cell
Biol 121:241­256
9. Orbach MJ, Schneider WP, Yanofsky C (1998) Cloning of methylated transforming DNA from Neurospora crassa in Escherichia coli.
Mol Cell Biol 8:2211­2213
10. Lu B (1993) Spreading the synaptonemal complex of Neurospora crassa. Chromosoma 102:464­472
11. Gillies CB (1979) The relationship between synaptonemal complexes, recombination nodules and crossing over in Neurospora crassa
bivalents and translocation quadrivalents. Genetics 91:1­17
12. Bojko M (1988) Presence of abnormal synaptonemal complexes in heterothallic species of Neurospora. Genome 30:697­709
13. Pukkila PJ, Lu BC (1985) Silver staining of meiotic chromosomes in the fungus, Corprinus cinereus. Chromosoma 91:108­112
14. Zolan ME, Crittenden JR, Heyler NK, Seitz LC (1992) Efficient isolation and mapping of rad genes of the fungus Coprinus cinereus
using chromosome-specific libraries. Nucl Acids Res 20:3993-3999
15. Pukkila PJ, Skrzynia C, Lu BC (1992) The rad3-1 mutant is defective in defective in axial core assembly and homologous chromosome
pairing during meiosis in the basidiomycete Coprinus cinereus. Dev Genet 13:403­410
16. Dutta Sk, Penn SR, Knight AR, Ojha M (1972) Characterization of DNAs from coprinus lagopos and mucor azygospora.Experientia
28:582­4
17. Holm PB, Rasmussen SW, Zickler D, Lu BC, Sage J (1981) Chromosome pairing, recombination nodules and chiasma formation in the
basidiomycete Coprinus cinereus. Carlsberg Res Commun 46:305­346
18. Gillies CB (1983) Ultrastructural studies of the association of homologous and non-homologous parts of chromosomes in the midprophase
of meiosis in Zea mays. Maydica 28:265­287
19. Gillies CB (1983) in Kew Chromosome Conference II. Branham PE, Bennett MD, Eds. George Allen and Unwin, London, page 115
20. Arumuganathan K, Earle ED (1991) Nuclear DNA content of some important plant species. Plant Mol Biol Rep 9:208­218
21. Anderson LK, Stack SM, Fox MH, Chuanshan Z (1985) The relationship between genome size and synaptonemal complex length ih
higher plants. Exp. Cell Res 156:367­378
22. Mogensen HL (1977) Ultrastructural analysis of female pachynema and the relationship between synaptonemal complex lenth and
crossing over in Zea mays. Carlsberg Res. Commun. 42:475­498
23. Anderson LK, Stack SM, Sherman JD (1988) Spreading synaptonemal complexes from Zea mays. Chromosoma 96:295­305
24. Sherman JD, Stack SM (1992) Two-dimensional spreads of synaptonemal complexes from solanaceous plants. V. Tomato (Lycopersicon
esculentum) karyotype and idiogram. Genome 35:354­359
25. Rattner JB, Goldsmith MR, Hamkalo BA (1980) Chromatin organization during meiotic prophase of Bombyx mori. Chromosoma
79:215­224
26. Rattner JB, Goldsmith MR, Hamkalo BA (1981) Chromosome organization during male meiosis in Bombyx mori. Chromosoma
82:341­351
27. Rasmussen Sw (1986) Initiation of synapsis and interlocking of chromosomes during zygotene in Bombyx spermatocytes. Carlsberg Res
Commun 51:401­432
28. Rasmussen SW (1976) The meotic prophase in Bombyx mori females analyzed by three dimensional reconstructions of synaptonemal
complexex. Chromosoma 54:245­293
29. Peterson DG, Stack SM, Heally JL, Donohoe BS, Anderson LK (1994) The relationship between synaptonemal complex length and
genome size in four vertebrate classes (Osteicthyes, Reptila, Aves, Mammalia). Chromosome Res. 2:153­162
30. Goetz P, Chandley AC, Speed RM (1984) Morphological and temporal sequence of meiotic prophase development at puberty in the
male mouse. J Cell SC 65:249­263
31. Moses MJ, Dresser ME, Poorman PA (1984) Composition and role of the synaptonemal complex. In: CW Evans, HG Dickinson (eds)
Controlling events in meiosis, vol 38. Symp Soc Exper Biol, pp 245­270
32. Glamann J. (1986) Crossing over in the male mouse as analyzed by recombination nodules and bars. Calsberg Res Commun 51:143­162
33. Holm PB, Rasmussen SW (1983) Human meiosis. V. Substages of pachytene in human spermatogenesis. Carlsberg Res Commun
48:351­383
34. Tease C, Hulten MA (2004) Inter-sex variation in synaptonemal complex lengths largely determine the different recombination rates in
male and female germ cells. Cytogenet Genome Res 107:208­215
35. Wallace BM, Hulten MA (1985) Meiotic chromosome pairing in the normal human female. Ann Hum Genet 49:215­226
Table 1 (continued)
191
References
Alani E, Padmore R, Kleckner N (1990) Analysis of wild type and
rad50 mutants of yeast suggests an intimate relationship
between meiotic chromosome synapsis and recombination.
Cell 61:419­436
Albini SM, Jones GH (1987) Synaptonemal complex spreading in
Allium cepa and A. fistulosum. I. The initiation and sequence
of pairing. Chromosoma 95:324­338
Allers T, Lichten M (2001) Differential timing and control of
noncrossover and crossover recombination during meiosis. Cell
106:47­57
Anderson LK, Royer SM, Page SL, McKim KS, Lai A, Lilly MA,
Hawley RS (2005) Juxtaposition of C(2)M and the transverse
filament protein C(3)G within the central region of Drosophila
synaptonemal complex. Proc Natl Acad Sci U S A 102:
4482­4487
Bannister LA, Reinholdt LG, Munroe RJ, Schimenti JC (2004)
Positional cloning and characterization of mouse mei8, a
disrupted allelle of the meiotic cohesin Rec8. Genesis 40:
184­194
Barlow AL, Hulten MA (1998) Crossing over analysis at pachytene
in man. Eur J Hum Genet 6:350­358
Bennett MD (1977) The time and duration of meiosis. Philos Trans
R Soc Lond B 277:201­226
Bhatt AM, Lister C, Page T, Fransz P, Findlay K, Jones GH,
Dickinson HG, Dean C (1999) The DIF1 gene of Arabidopsis is
required for meiotic chromosome segregation and belongs to
the REC8/RAD21 cohesin gene family. Plant J 19:463­472
Bishop DK, Zickler D (2004) Early decision; meiotic crossover
interference prior to stable strand exchange and synapsis. Cell
117:9­15
Bishop DK, Park D, Xu L, Kleckner N (1992) DMC1: a meiosisspecific
yeast homolog of bacterial recA required for meiotic
recombination, synaptonemal complex formation and cell cycle
progression. Cell 69:439­456
Blat Y, Protacio R, Hunter N, Kleckner N (2002) Physical and
functional interactions among basic chromosome organizational
features govern early steps of meiotic chiasma formation.
Cell 111:791­802
Börner GV, Kleckner N, Hunter N (2004) Crossover/noncrossover
differentiation, synaptonemal complex formation and regulatory
surveillance at the leptotene/zygotene transition of meiosis.
Cell 117:29­45
Carpenter AT (1975) Electron microscopy of meiosis in Drosophila
melanogaster females: II. The recombination nodule--a recombination-associated
structure at pachytene? Proc Natl Acad
Sci U S A 72:3186­3189
Carpenter AT (1981) EM autoradiographic evidence that DNA
synthesis occurs at recombination nodules during meiosis in
Drosophila melanogaster females. Chromosoma 83:59­80
Carpenter ATC (1987) Gene conversion, recombination nodules,
and the initiation of meiotic synapsis. BioEssays 6:232­236
Clyne RK, Katis VL, Jessop L, Benjamin KR, Herskowitz I, Lichten
M, Nasmyth K (2003) Polo-like kinase Cdc5 promotes
chiasmata formation and cosegregation of sister centromeres
at meiosis I. Nat Cell Biol 5:480­485
Copenhaver GP, Housworth EA, Stahl FW (2002) Crossover
interference in Arabidopsis. Genetics 160:1631­1639
Dawe RK, Sedat JW, Agard DA, Cande WZ (1994) Meiotic
chromosome pairing in maize is associated with a novel
chromatin organization. Cell 76:901­912
de Vries FA, de Boer E, van den Bosch M, Baarends WM, Ooms M,
Yuan L, Liu JG, van Zeeland AA, Heyting C, Pastink A (2005)
Mouse Sycp1 functions in synaptonemal complex assembly,
meiotic recombination, and XY body formation. Genes Dev
19:1376­1389
Dresser ME, Moses MJ (1980) Synaptonemal complex karyotyping
in spermatocytes of the Chinese hamster (Cricetulus griseus).
IV. Light and electron microscopy of synapsis and nucleolar
development by silver staining. Chromosoma 76:1­22
Egel R (1995) The synaptonemal complex and the distribution of
meiotic recombination events. Trends Genet 11:206­208
Egel-Mitani M, Olson LW, Egel R (1982) Meiosis in Aspergillus
nidulans: another example for lacking synaptonemal complexes
in the absence of crossover interference. Hereditas 97:179­187
Eijpe M, Offenberg H, Jessberger R, Revenkova E, Heyting C
(2003) Meiotic cohesin REC8 marks the axial elements of rat
synaptonemal complexes before cohesins SMC1beta and
SMC3. J Cell Biol 160:657­670
Ellermeier C, Smith GR (2005) Cohesins are required for meiotic
DNA breakage and recombination in Schizosaccharomyces
pombe. Proc Natl Acad Sci U S A 102:10952­10957
Franklin AE, McElver J, Sunjevaric I, Rothstein R, Bowen B, Cande
WZ (1999) Three-dimensional microscopy of the Rad51
recombination protein during meiotic prophase. Plant Cell
11:809­824
Fung JC, Rockmill B, Odell M, Roeder GS (2004) Imposition of
crossover interference through the nonrandom distribution of
synapsis initiation complexes. Cell 116:795­802
Gilbert N, Boyle S, Fiegler H, Woodfine K, Carter NP, Bickmore
WA (2004) Chromatin architecture of the human genome:
gene-rich domains are enriched in open chromatin fibers. Cell
118:555­566
Gimenez-Abian JF, Clarke DJ, Mullinger AM, Downes CS, Johnson
RT (1995) A postprophase topoisomerase II-dependent chromatid
core separation step in the formation of metaphase
chromosomes. J Cell Biol 131:7­17
Glynn EF, Megee PC, Yu HG, Mistrot C, Unal E, Koshland DE,
DeRisi JL, Gerton JL (2004) Genome-wide mapping of the
cohesin complex in the yeast Saccharomyces cerevisiae. PLoS
Biol 2:E259
Guillon H, Baudat F, Grey C, Liskay RM, de Massy B (2005)
Crossover and noncrossover pathways in mouse meiosis. Mol
Cell 20:563­573
Heyting C (2005) Meiotic transverse filament proteins: essential for
crossing over. Transgenic Res 14:547­550
Higgins JD, Sanchez-Moran E, Armstrong SJ, Jones GH, Franklin
FC (2005) The Arabidopsis synaptonemal complex protein
ZYP1 is required for chromosome synapsis and normal fidelity
of crossing over. Genes Dev 19:2488­2500
Hollingsworth NM, Goetsch L, Byers B (1990) The HOP1 gene
encodes a meiosis-specific component of yeast chromosomes.
Cell 61:73­84
Hunter N, Kleckner N (2001) The single-end invasion: an
asymmetric intermediate at the double-strand break to doubleHolliday
junction transition of meiotic recombination. Cell
106:59­70
Hunter N, Chambers SR, Louis EJ, Borts RH (1996) The mismatch
repair system contributes to meiotic sterility in an interspecific
yeast hybrid. EMBO J 15:1726­1733
Keeney S (2001) Mechanism and control of meiotic recombination
initiation. Curr Top Dev Biol 52:1­53
King JS, Mortimer RK (1990) A polymerization model of chiasma
interference and corresponding computer simulation. Genetics
126:1127­1138
Kleckner N (1996) Meiosis: how could it work? Proc Natl Acad Sci
U S A 93:8167­8174
Kleckner N, Padmore R, Bishop DK (1991) Meiotic chromosome
metabolism: one view. Cold Spring Harb Symp Quant Biol
56:729­743
Kleckner N, Storlazzi A, Zickler D (2003) Coordinate variation in
meiotic pachytene SC length and total crossover/chiasma
frequency under conditions of constant DNA length. Trends
Genet 19:623­628
Kleckner N, Zickler D, Jones, GH, Henle J, Dekker J, Hutchinson J
(2004) A mechanical basis for chromosome function. Proc Natl
Acad Sci U S A 101:12592­12597
Klein F, Mahr P, Galova M, Buonomo SB, Michaelis C, Nairz K,
Nasmyth K (1999) A central role for cohesins in sister
chromatid cohesion, formation of axial elements, and recombination
during yeast meiosis. Cell 98:91­103
192
Lande R, Stahl FW (1993) Chiasma interference and the distribution
of exchanges in Drosophila melanogaster. Cold Spring Harb
Symp Quant Biol 58:543­552
Lipkin SM, Moens PB, Wang V, Lenzi M, Shanmugarajah D,
Gilgeous A, Thomas J, Cheng J, Touchman JW, Green ED,
Schwartzberg P, Collins FS, Cohen PE (2002) Meiotic arrest and
aneuploidy in MLH3-deficient mice. Nat Genet 31:385­390
MacQueen AJ, Colaiacovo MP, McDonald K, Villeneuve AM
(2002) Synapsis-dependent and -independent mechanisms
stabilize homolog pairing during meiotic prophase in C.
elegans. Genes Dev 16:2428­2442
Maguire MP (1966) The relationship of crossing over to chromosome
synapsis in a short paracentric inversion. Genetics
53:1071­1077
Maguire MP (1972) The temporal sequence of synaptic initiation,
crossing over and synaptic completion. Genetics 70:353­370
Maguire MP (1988) Crossover site determination and interference.
J Theor Biol 134:565­750
Maguire MP (1995) Is the synaptonemal complex a disjunction
machine? J Hered 86:330­340
Manheim EA, McKim KS (2003) The synaptonemal complex
component C(2)M regulates meiotic crossing over in Drosophila.
Curr Biol 13:276­285
McKim KS, Jang JK, Manheim EA (2002) Meiotic recombination
and chromosome segregation in Drosophila females. Annu Rev
Genet 36:205­232
Moens PB, Earnshaw WC (1989) Anti-topoisomerase II recognizes
meiotic chromosome cores. Chromosoma 98:317­322
Moens PB, Pearlman RE (1988) Chromatin organization at meiosis.
BioEssays 9:151­153 Mol Cell Biol 21:5667­5677
Moens PB, Kolas NK, Tarsounas M, Marcon E, Cohen PE,
Spyropoulos B (2002) The time course and chromosomal
localization of recombination-related proteins at meiosis in the
mouse are compatible with models that can resolve the early
DNA­DNA interactions without reciprocal recombination.
J Cell Sci 115:1611­1622
Molnar M, Doll E, Yamamoto A, Hiraoka Y, Kohli J (2003) Linear
element formation and their role in meiotic sister chromatid
cohesion and chromosome pairing. J Cell Sci 116:1719­1731
Moses MJ (1969) Structure and function of the synaptonemal
complex. Genetics 61(Suppl):41­51
Neale MJ, Pan J, Keeney S (2005) Endonucleolytic processing of
covalent protein-linked DNA double-strand breaks. Nature
436:1053­1057
Ortiz R, Echeverria OM, Ubaldo E, Carlos A, Scassellati C,
Vazquez-Nin GH (2002) Cytochemical study of the distribution
of RNA and DNA in the synaptonemal complex of guinea pig
and rat spermatocytes. Eur J Histochem 46:133­142
Padmore R, Cao L, Kleckner N (1991) Temporal comparison of
recombination and synaptonemal complex formation during
meiosis in S. cerevisiae. Cell 66:1239­1256
Page SL, Hawley RS (2004) The genetics and molecular biology of the
synaptonemal complex. Annu Rev Cell Dev Biol 20:525­558
Pawlowski WP, Cande WZ (2005) Coordinating the events of the
meiotic prophase. Trends Cell Biol 15(12):674­681
Pelttari J, Hoja MR, Yuan L, Liu JG, Brundell E, Moens P, SantucciDarmanin
S, Jessberger R, Barbero JL, Heyting C, Hoog C
(2001) A meiotic chromosomal core consisting of cohesin
complex proteins recruits DNA recombination proteins and
promotes synapsis in the absence of an axial element in
mammalian meiotic cells. Mol Cell Biol 21:5667­5677
Plug AW, Peters AH, Xu Y, Keegan KS, Hoekstra MF, Baltimore D,
de Boer P, Ashley T (1997) ATM and RPA in meiotic
chromosome synapsis and recombination. Nat Genet 17:457­461
Pukkila PJ, Shannon KB, Skrzynia C (1995) Independent synaptic
behavior of sister chromatids in Coprinus cinereus. Can J Bot
73(Suppl 1):S215­S220
Rasmussen SW (1986) Initiation of synapsis and interlocking of
chromosomes during zygotene in Bombyx spermatocytes.
Carlsberg Res Commun 51:401­432
Revenkova E, Eijpe M, Heyting C, Hodges CA, Hunt PA, Liebe B,
Scherthan H, Jessberger R (2004) Cohesin SMC1 beta is
required for meiotic chromosome dynamics, sister chromatid
cohesion and DNA recombination. Nat Cell Biol 6:555­562
Saitoh Y, Laemmli UK (1994) Metaphase chromosome structure:
bands arise from a differential folding path of the highly ATrich
scaffold. Cell 76:609­622
Schimenti J (2005) Synapsis or silence. Nat Genet 37:11­13
Sherman JD, Herickhoff LA, Stack SM (1992) Silver staining two
types of meiotic nodules. Genome 35:907­915
Stack SM, Anderson LK (1986) Two dimensional spreads of
synaptonemal complexes from solanaceous plants. III. Recombination
nodules and crossing over in Lycopersicon esculentum
(tomato). Chromosoma 94:253­258
Stern H, Westergaard M, Von Wettstein D (1975) Presynaptic events
in meiocytes of Lilium longiflorum and their relation to
crossing-over: a preselection hypothesis. Proc Natl Acad Sci U
S A 72:961­965
Storlazzi A, Xu L, Cao L, Kleckner N (1995) Crossover and
noncrossover recombination during meiosis: timing and pathway
relationships. Proc Natl Acad Sci U S A 92:8512­8516
Storlazzi A, Tesse S, Gargano S, James F, Kleckner N, Zickler D
(2003) Meiotic double-strand breaks at the interface of
chromosome movement, chromosome remodeling, and reductional
division. Genes Dev 17:2675­2687
Strick R, Laemmli UK (1995) SARs are cis DNA elements of
chromosome dynamics: synthesis of a SAR repressor protein.
Cell 83:1137­1148
Sym M, Roeder GS (1994) Crossover interference is abolished in
the absence of a synaptonemal complex protein. Cell 79:
283­292
Tarsounas M, Morita T, Pearlman RE, Moens PB (1999) RAD51
and DMC1 form mixed complexes associated with mouse
meiotic chromosome cores and synaptonemal complexes. J Cell
Biol 147:207­220
Tease C, Hulten MA (2004) Inter-sex variation in synaptonemal
complex lengths largely determine the different recombination
rates in male and female germ cells. Cytogenet Genome Res
107:208­215
Tessé S, Storlazzi A, Kleckner N, Gargano S, Zickler D (2003)
Localization and roles of Ski8p in Sordaria macrospora
meiosis and delineation of three mechanistically distinct steps
of meiotic homolog juxtaposition. Proc Natl Acad Sci U S A
100:12865­12870
Turner JM, Mahadevaiah SK, Fernandez-Capetillo O, Nussenzweig
A, Xu X, Deng CX, Burgoyne PS (2005) Silencing of
unsynapsed meiotic chromosomes in the mouse. Nat Genet
37:41­47
Unal E, Arbel-Eden A, Sattler U, Shroff R, Lichten M, Haber JE,
Koshland D (2004) DNA damage response pathway uses
histone modification to assemble a double-strand break-specific
cohesin domain. Mol Cell 16:991­1002
van Heemst D, Heyting C (2000) Sister chromatid cohesion and
recombination in meiosis. Chromosoma 109:10­26
van Heemst D, James F, Poggeler S, Berteaux-Lecellier V, Zickler D
(1999) Spo76p is a conserved chromosome morphogenesis
protein that links the mitotic and meiotic programs. Cell
98:261­271
Viera A, Santos JL, Page J, Parra MT, Calvente A, Cifuentes M,
Gomez R, Lira R, Suja JA, Rufas JS (2004) DNA double-strand
breaks, recombination and synapsis: the timing of meiosis
differs in grasshoppers and flies. EMBO Rep 5:385­391
von Wettstein D, Rasmussen SW, Holm PB (1984) The synaptonemal
complex in genetic segregation. Annu Rev Genet
18:331­413
Whitby MC (2005) Making crossovers during meiosis. Biochem
Soc Trans 33:1451­1455
Xu H, Beasley MD, Warren WD, van der Horst GT, McKay MJ
(2005) Absence of mouse REC8 cohesin promotes synapsis of
sister chromatids in meiosis. Dev Cell 8:949­961
193
Yamamoto A, Tsutsumi C, Kojima H, Oiwa K, Hiraoka Y (2001)
Dynamic behavior of microtubules during dynein-dependent
nuclear migrations of meiotic prophase in fission yeast. Mol
Biol Cell 12:3933­3946
Zickler D (1977) Development of the synaptonemal complex and
the "recombination nodules" during meiotic prophase in the
seven bivalents of the fungus Sordaria macrospora Auersw.
Chromosoma 61:289­316
Zickler D, Kleckner N (1999) Meiotic chromosomes: integrating
structure and function. Annu Rev Genet 33:603­754
Zickler D, Sage J (1981) Synaptonemal complexes with modified
lateral elements in Sordaria humana: development of and
relationship to the "recombination nodules." Chromosoma
84:305­318
194