Retention of Latent Centromeres in
the Mammalian Genome
G. C. FERRERI*, D. M. LISCINSKY*, J. A. MACK, M. D. B. ELDRIDGE, AND R. J. O'NEILL
*Both authors contributed equally to this work.
From the Department of Molecular and Cell Biology, U-2131, University of Connecticut, Storrs, CT 06269­2131
(Ferreri, Liscinksy, Mack, and O'Neill); and the Department of Biological Sciences, Macquarie University,
New South Wales 2109, Australia (Eldridge).
Address correspondence to R. J. O'Neill at the address above, or email: roneill@uconnvm.uconn.edu.
Abstract
The centromere is a cytologically defined entity that possesses a conserved and restricted function in the cell: it is the site of
kinetochore assembly and spindle attachment. Despite its conserved function, the centromere is a highly mutable portion of
the chromosome, carrying little sequence conservation across taxa. This divergence has made studying the movement of
a centromere, either within a single karyotype or between species, a challenging endeavor. Several hypotheses have been
proposed to explain the permutability of centromere location within a chromosome. This permutability is termed
``centromere repositioning'' when described in an evolutionary context and ``neocentromerization'' when abnormalities
within an individual karyotype are considered. Both are characterized by a shift in location of the functional centromere
within a chromosome without a concomitant change in linear gene order. Evolutionary studies across lineages clearly
indicate that centromere repositioning is not a rare event in karyotypic evolution and must be considered when examining
the evolution of chromosome structure and syntenic order. This paper examines the theories proposed to explain
centromere repositioning in mammals. These theories are interpreted in light of evidence gained in human studies and in
our presented data from the marsupial model species Macropus eugenii, the tammar wallaby.
Introduction
The ectopic emergence of a neocentromere occurs most
frequently to provide mitotic stability on otherwise acentric
chromosome fragments resulting from rearrangement
(Amor and Choo 2002; Warburton 2004). Approximately
70 described cases of neocentromeres have been identified
on 19 human chromosomes (Warburton 2004). Almost 10%
of these cases are meiotically stable and heritable (Amor et al.
2004; Knegt et al. 2003). This has implications for the role
that neocentromeres may play in creating karyotypic diversity
and potential repositioning of a centromere. Three clear hot
spots for neocentromeres have been identified within the
human karyotype (3q26­qter, 13q21­32, and 15q24­26)
(Amor and Choo 2002), implying a nonrandom mechanism
for their appearance. The appearance of centromere repositioning
in several primate taxa, exemplified in the shifts
identified between Old World monkeys and New World
monkeys (Eder et al. 2003)--as well as in cattle (Band et al.
2000), mouse (Armengol et al. 2003), and several marsupial
lineages (Eldridge and Close 1993)--indicates this type of
chromosome rearrangement may be a significant feature of
chromosome evolution in mammals.
Methods
Fluorescence In Situ Hybridization
Bacterial Artificial Chromosome (BAC) fluorescence in situ
hybridization was performed as described previously
(Ferreri et al. 2004) with modifications. Briefly, BAC
DNA was labeled by nick translation, incorporating either
a biotin or digoxigenin labeled nucleotide as per standard
protocols (Invitrogen). Labeled DNA was hybridized to
metaphase chromosomes of M. eugenii in the presence of 6
lg sonicated M. eugenii DNA under stringent conditions,
and posthybridization washes were performed at 458C in
50% formamide, 23 SSC, and 0.13 SSC (prewarmed to
608C). Hybridization to Petrogale assimilis chromosomes was
performed as above without blocking DNA and with a low
stringency wash series of 50% formamide, 23 SSC, and 23
SSC at 458C. After blocking with 5% bovine serum
albumin in 43 SSC, 0.2% Tween 20, detection was
performed using the appropriate secondary antibody
(antibiotin fluorescein, antidig Texas Red) (Vector Labs
and Molecular Probes). Images were captured on an
Olympus AX70 using a CCD cooled camera and Applied
Imaging Cytovision software.
217
Journal of Heredity 2005:96(3):217­224  2005 The American Genetic Association
doi:10.1093/jhered/esi029
Advance Access publication January 13, 2005
Southern Analysis
Genomic DNA (10 lg) from the representative species was
digested with the named enzyme in the presence of buffer at
378C (New England Biolabs). KERV was labeled with P32dCTP
by random priming and hybridized to DNA (transferred
to a Nylon N membrane) at 558C in 500 mM
Na2HPO4 (pH 7.0), 7% SDS, and 1 mM EDTA. Posthybridization
washes were performed in 23 SSC, 0.1%
SDS, 13 SSC, 0.1% SDS, or 0.53 SSC, 0.1% SDS at either
658C or 558C. Species analyzed include: Aepyprymnus rufescens,
Bettongia penicillata, Dendrolagus goodfellowi, D. lumholtzi, Dorcopsis
luctuosa, Hypsiprymnodon moschatus, Isoodon macrourus, I.
obesulus, Lagorchestes conspicillatus, Macropus agilis, M. antilopinus,
M. dorsalis, M. eugenii, M. giganteus, M. parma, M. parryi, M.
robustus, M. rufogriseus banksianus, M. rufogriseus rufogriseus, M.
rufus, Monodelphis domestica, Notoryctes typhlops, Onychogalea
fraenata, O. unguifera, Petrogale mareeba, P. penicillata, P.
persephone, P. purpureicollis, P. xanthopus, Potorous tridactylus,
Pseudochirulus herbertensis, Pseudochirus peregrinus, Sarcophilus
harrisii, Setonix brachyurus, Thylogale thetis, Vombatus ursinus,
Wallabia bicolor.
Neocentromeres and Centromere Repositioning
Three theories describing the possible mechanism for
neocentromere emergence in the context of centromere
repositioning have been posited (Figure 1). The first proposes
an epigenetic mechanism for repatterning of a segment
of chromatin to perform as a competent site of kinetochore
attachment and assembly. A lack of identifiable, shared
satellite sequence features among cases of dicentric chromosomes
and neocentromeres (Alonso et al. 2003; Barry
et al. 1999; Lo et al. 2001a; Lo et al. 2001b; Sullivan and
Willard 1998) suggests that this mechanism is likely independent
of DNA sequence. Another theory, not necessarily
exclusive of the first, proposes that a latent centromere may
act as a primer for centromere emergence (Choo 1997; du
Sart et al. 1997). Under its initial description, the latent
centromere hypothesis relies on the presence of a centromere-specific
sequence at the site of imminent centromere
formation. Recently, this hypothesis has been modified to
suggest that there may be latent chromatin and/or genomic
structures that act as a mark for centromere formation
(Ventura et al. 2004). The third combines elements of the
first two, proposing that an increase in instability in a locus
may induce repair mechanisms that ultimately can trigger
chromatin repatterning (Ventura et al. 2003).
``Satellite DNA,'' a phrase that originally applied to
satellite bands observed in ultracentrifuge density gradients,
is commonly used to describe any tandemly repetitive
sequence (John 1988).While satellite DNA families can be
species or chromosome specific (Singer 1982), their seemingly
ubiquitous presence at or near centromeric domains
across a wide variety of organisms suggests that they play
a role in centromere function (Eichler 1999; Henikoff et al.
2001; Willard 1990). Earlier studies in simian cell lines
transfected with human alpha satellite sequences showed
that these repetitive regions could form de novo centromeres
on existing chromosomes (Haaf et al. 1992).
However, later studies characterizing neocentromeric
DNA in dicentric chromosomes showed that classical
human alpha satellites are not the essential element dictating
centromere location in humans (Barry et al. 1999; du Sart
et al. 1997; Sullivan and Willard 1998). Concordantly,
Williams et al. (1998) have shown that neocentromeres in
Drosophila do not contain classical repeated DNA sequences.
Thus, it is clear that while satellite DNA may be sufficient
for centromere function, it is not required (Csink and
Henikoff 1998; Willard 1990). Such a disparity between
sequence features among active and inactive centromeres
argues against a latent centromere hypothesis for neocentromerization.
However, this position becomes more difficult
to support when considering centromere repositioning in
karyotypic evolution. In this context, cryptic repeats may be
lost due to increased recombination at recently inactivated
centromere locations (Jackson 2003; Ventura et al. 2003), as
well as through genetic drift.
Using fluorescence in situ hybridization of BAC probes
spanning several chromosomes and in silico analyses, Ventura
et al. (2003) identified an ancestral centromere in HSA15q25.
This centromere was inactivated at the time of the fission
event that resulted in HSA14, HSA15, and the emergence of
two new centromeres. This ancestral location coincides with
neocentromere formation in 15q24­26 in at least two human
cases, supporting the latent centromere hypothesis. The
geographic coincidence of these neocentromeres, however,
does not extend to the sequence level as one neocentromere
maps .8 Mb from the ancestral centromere (Ventura et al.
2003).
Figure 1. Schematic diagram of the models proposed to
explain the derivation of a centric shift/centromere
repositioning. The key is shown to the right.
218
Journal of Heredity 2005:96(3)
Neocentromeres have also been identified at HSA3q26qter
(Muller et al. 2000; Ventura et al. 2004), HSA13q21­32
(Ventura et al. 2003), and HSAXq13 (Ventura et al. 2001). In
the case of chromosomes 3 and 13, neocentromeres are
found at higher frequencies than expected under random
models of chromosome aberration (Amor and Choo 2002),
implicating the latent centromere hypothesis in defining
centromere competent locations. These observations are not
taken from sequence analyses or mapping data alone but are
considered in light of the evolution of active and ancestral
centromere locations across primates. Each hot spot for
neocentromere emergence in humans colocalizes to an
ancestral centromere that has been inactivated during
primate evolution. Like the case of the ancestral centromere
in HSA15q25, there is little sequence homology between the
inactive and active centromere sites. The reuse of centromere
locations, often referred to as centromere seeding, implies
that a latent centromeric site may function as both euchromatin
and as a neocentromere.
Thus, while the DNA obviously is not a strict dictator of
centromere competence, features conserved at inactive
centromere locations may retain the capacity to become
active centromeres. These features may exist in the form of
as yet unidentified epigenetic marks or architectural features,
such as segmental duplications, of the surrounding genomic
landscape.
Ventura et al. (2003) have proposed a link between
persistent recombinogenic segmental duplications and neocentromere
potential. In this model, a locus characterized by
a high frequency of segmental duplications, such as that
found in the region surrounding HSA15q24­26 and its
homologous segments in other primate species, may develop
sites for neocentromerization through repair processes
following chromosome rearrangement induced by these
duplications. A coincidental association between an increased
frequency of segmental duplications and chromosome
rearrangement at loci near ancestral centromeres is
supported by evidence of duplication events at 15q25 in
human samples (Gratacos et al. 2001). A similar observation
has been found for the ancestral, inactive centromere located
in HSA6p22 (Eder et al. 2003). This region is also defined as
rich in segmental duplications, and instability associated with
these duplicons has been linked to 21-hydroxylase deficiency
(Tusie-Luna and White 1995). In contrast to these findings,
there is a paucity of segmental duplications in the
ancestral centromere-neocentromere association identified
in HSA3q26 (Ventura et al. 2004).
When examining centromere repositioning and centromere
seeding across active and ancestral centromere locations
within a karyotype, several factors influencing the
sequence structure and content of these regions must be
considered. Duplicons that are the result of pericentric
segmental duplications likely undergo heterochromatic
reformation to convert to euchromatin and become subject
to an increased recombination rate once they are no longer
restricted to centric domains (Jackson 2003). Timing the
duplication event is difficult as it may be the result of
pericentric duplications from the centromere prior to its
inactivation rather than from the newly active centromere
into an euchromatic site. Likewise, instability at the ancestral
centromeric regions may be the result of the decay in
pericentric heterochromatin following epigenetic repatterning
once centromere repositioning has taken place (Jackson
2003).
The obvious shuffling of chromosomal segments in an
evolutionary context, typified by breaks of synteny (BOS),
implies that centromere repositioning may be a much more
common occurrence than previously appreciated. Intriguingly,
recent genomic analyses have shown that segmental
duplications are strongly correlated with BOS between
human and mouse (Armengol et al. 2003). Therefore, the
contribution of segmental duplications to genomic diversity
must be discussed in the context of centromere emergence.
Segmental Duplications
The assembly of sequences from large contiguous chromosomal
regions from human, mouse, and a few other
mammalian models has made possible high-resolution comparative
mapping of regions of synteny on both large and small
scales. What has become clear is that the organization of the
genomes of these organisms has been heavily influenced by
duplication, followed by tandem insertion or transposition of
regions encompassing single genes or large DNA segments.
Furthermore, these high-resolution comparative maps reveal
that the distribution of these segmental duplications frequently
coincide with unstable genomic regions and disease loci.
Comparisons of the assembled mouse and human
genome sequences have led to some insights into the gross
karyotypic changes between these two distantly related
mammals. Remarkable power to trace the evolutionary
trajectories of duplicated segments (duplicons), however, has
come from concerted sequencing and mapping efforts in
several primate models. It is estimated that approximately
5.2% of the human genome exists as duplicons, many
derived in the last 35 million years (Bailey et al. 2002; Bailey
et al. 2001; Consortium 2001). Interestingly, these duplicated
segments tend to be concentrated in pericentromeric and
subtelomeric regions, showing tenfold enrichment in these
regions compared to euchromatin (Bailey et al. 2001).
Furthermore, duplicons localizing to pericentromeric and
subtelomeric regions are more likely to originate interchromosomally,
while euchromatic duplicons tend to be intrachromosomal
in origin (Bailey et al. 2001; Eichler 2001).
Duplicated segments in euchromatin frequently comprise
low copy repeats (LCRs) that likely arise from and are prone
to nonhomologous recombination. Nonhomologous recombination
serves not only to expand LCRs but can also
lead to micro-deletions and inversions within or encompassing
these segments. Such instability imbued by intrachromosomal
segmental duplications can have dramatic
effects on gene expression and chromosomal integrity.
Analyses of intrachromosomal segmental duplications in
humans have shown them to be associated with 169 known
regions of instability and 24 human genomic disorders
(Bailey et al. 2002; Stankiewicz and Lupski 2002).
219
Ferreri et al.  Latent Centromeres
Tracing the trajectory of both intrachromosomal duplicons,
as highlighted in studies of instability, and interchromosomal
duplicons in an evolutionary context can uncover
events responsible for species-specific genomic alterations.
Interchromosomal duplications have been found in phylogenetic
mapping studies of paralogous regions between
species of great apes. A complex series of rearrangements
was described in analyses of a pericentromeric DNA fragment
found on the long-arm of human chromosome 21
(HSA21). This sequence was found to be the product of
intrachromosomal duplication (Potier et al. 1998), followed
by interchromosomal transposition (to HSA2q, HSA13, and
HSA18) in great apes after the divergence of orangutans
(Golfier et al. 2003).
Another example of the complex history of sequences
that may have contributed to the formation of rearrangements
during primate karyotypic evolution can be found in
the elegant studies of the evolution of the ancestral fusion
site in HSA2q13­2q14.1 (Fan et al. 2002a; Fan et al. 2002b;
Martin et al. 2002; Mefford and Trask 2002). Chromosome 2
in humans derives from the fusion of two chromosomes that
have remained separate in other primate species. Fan et al.
(2002a) traced the history of the sequences present at this
fusion site and found that this region has undergone
a complex series of rearrangements and duplicative exchanges,
including several rounds of intrachromosomal and
interchromosomal duplications, an inversion and subsequently
a fusion between duplicated segments to form
HSA2.
Thus, genome architecture, as defined by nongenic
regions of the genome, can act as a catalyst for chromosome
rearrangements through the action of segmental duplication
and nonallelic homologous recombination. The propensity
toward intrachromosomal rearrangements in human chromosome
evolution (Pevzner and Tesler 2003a; Pevzner and
Tesler 2003b; Postlethwait et al. 2000) has made delineating
the complex history of human chromosomal segments
across other mammalian lineages challenging. Comparisons
of mouse and human genome sequences have shown that
53% of evolutionary breakpoint rearrangements defined at
BOS between these two species associate with segmental
duplications (Armengol et al. 2003). Additionally, Thomas
et al. (2003) have identified recent (;5­7 mya) pericentromeric
segmental duplications on mouse chromosomes 5 and
6 that have been implicated in a chromosome fission event,
resulting in the derivation of two new centromeres, as well as
chimeric transcript formation. While defining the trajectory
of some of these rearrangements and determining whether
the duplications are associative or causative has not been
accomplished, an intriguing correlation between karyotypic
evolution and segmental duplications is emerging.
Marsupial Karyotypic Evolution
Descriptions of chromosome homologies and karyotypes
within Marsupialia are extensive, with over 70% of known
marsupial species karyotyped (Hayman 1977; Hayman 1990).
Chromosome evolution within Macropodidae (kangaroos,
wallabies, and rat kangaroos), a group that radiated into ;77
species over 22 mya (Kirsch and Lapointe 1995; Rens et al.
2003), has been comprehensively studied in terms of Gbanding,
chromosome rearrangements, and homologies.
This group of marsupials is characterized by an extensive
diversification of karyotypes, with diploid numbers ranging
from 2n10, 11 in W. bicolor to 2n32 in A. rufescens, all
derived from an ancestral 2n22 karyotype (Hayman 1990;
Rens et al. 2003; Rofe 1979).
Central to the description of this karyotypic diversity is
the involvement of the centromere, either through its
location on the chromosome or its involvement in fissions,
translocations, inversions, fusions, and shifts. The involvement
of the centromere in the chromosomal rearrangements
giving rise to the karyotypic diversification within Macropodidae
has been elegantly highlighted in a recent study
tracing the phylogenetic distribution of 19 conserved
syntenic segments across several marsupial families (Rens
et al. 2003).
Further evidence for the correlation between centromere
dynamics and karyotypic diversity in marsupials has been
described in detailed analyses of genome rearrangement and
instability found in dysgenic interspecific hybrids within the
Macropus genus. Several hybrids display karyotypic aberrations
almost exclusively associated with centromeric abnormalities,
including translocations and amplifications (O'Neill
et al. 1998; O'Neill et al. 2001). Detailed analyses of
five hybrids from two different species crosses have
shown instabilities linked to kangaroo endogenous retrovirus
(KERV) (see below), attributed to a significant copy-number
increase of this sequence in the centromere (unpublished
data; O'Neill et al. 1998). Recent research has shown that this
centromeric amplification also associates with fusion and
fission events, as well as knob-formation, a potentially
meiotically driven element (Rhoades and Dempsey 1966).
Thus, it appears that the centromere, or at least centromereassociated
sequences, may have played a pivotal role in
chromosome restructuring and centromere repositioning in
macropodines.
Retroviruses and Centromere-specific Sequences
The link between segmental duplications, BOS, and centromere
repositioning becomes more intriguing in light of new
data from the marsupial model species M. eugenii. A
previously identified retroviral element, KERV, has been
localized to all of the active centromeres of this species
(Ferreri et al. 2004). Previously characterized as a centromeric
element within another macropodine (kangaroos and
wallabies) species, this sequence is preserved as a centromeric
repeat as well as a functional retroviral element. BACs
containing this sequence have been mapped onto metaphase
chromosomes of M. eugenii (Ferreri et al. 2004; for a current
map, see Figure 2). Significantly, 72% of these clones (35
total) map to BOS between conserved chromosome
segments as determined through cross-species reciprocal
chromosome painting (Ferreri et al. 2004; Rens et al. 2003).
When the rearrangements these blocks have experienced
220
Journal of Heredity 2005:96(3)
within other marsupials are considered, 100% of these clones
map to latent or active centromere locations within this
species.
Active centromere sites within M. eugenii also contain
another centromeric satellite that experiments show possesses
CENP-B binding capability (K. Bulazel et al., unpublished
data). The colocalization of signals for this satellite and KERV
at all active centromeres within this species suggests that these
elements are tightly associated. Each of the KERV-positive
BACs, therefore, was screened by dot blot hybridization at
high stringency with the complete satellite sequence.
Surprisingly, 100% of these BACs showed positive hybridization
with this probe. Therefore, neither sequence is
restricted to centromeric domains within this species.
Furthermore, 11 interstitial, latent centromere locations thus
far mapped harbor sequences that may indicate centromere
competence.
The appearance of these interstitial sequences is likely the
result of segmental duplications--although replicative transposition
cannot be ruled out, given the direct involvement of
a retroviral sequence. Evidence from hybrid studies indicates
that this element may undergo bursts of activity (O'Neill et al.
1998), enabling mobilization to other locations within the
genome through autonomous replication machinery. However,
the location of interstitial sequences is not randomly
distributed, as might be expected under a mobilization
scenario. While it is evident that KERV has retained activity
(Ferreri et al. 2004; O'Neill et al. 1998), the location of active
KERV sequences has not been defined. The nonrandom
positive correlation between the location of KERV and
centromeres (as well as BOS) indicates that these paralogous
sequences may in fact be the product of segmental duplication
events. Detailed studies of chromosome rearrangements
within humans and homologies between mice and
humans have shown an increased concentration of repeated
DNA, retroelements, and segmental duplications at BOS
(Armengol et al. 2003; Bailey et al. 2004; Bailey et al. 2003;
Dehal et al. 2001), suggesting a mechanistic association
between chromosome rearrangement and these genomic
elements (Armengol et al. 2003; Bailey et al. 2004; Bailey et al.
2003).
The colocalization of the BOS with active centromereassociated
sequences is significant, as each of these sites is
involved in chromosome rearrangements that karyotypically
identify divergent marsupial species. Each of these sites is
also the location of active centromeres in other marsupial
species. The appearance of these sequences may be the result
of past duplication events from active centromere locations;
however, phylogenetic inference from karyotypic studies (see
O'Neill et al. 2004 for a review) indicates that the appearance
of at least some of these sequences predates the centromere
repositioning event. This strongly implies that there are many
latent centromere locations within the marsupial karyotype
that are centromere competent.
Although the order of events over the course of M. eugenii
chromosome evolution in relation to KERV, the functional
satellite, and BOS remains to be determined, examining the
structure and relationship of these sequences as functionally
disparate portions of the genome could lead to major insights
into centromere seeding, competence, and emergence.
Conservation of Breaks of Synteny and Centromere
Sequences
Preliminary analyses of other macropodine species indicate
that the location of at least three KERV and satellite-positive
BACs to BOS is conserved in two other species, M. rufogriseus
(red-necked wallaby) and W. bicolor (swamp wallaby) (Ferreri
et al. 2004). The BAC that maps to the BOS and active
centromere on M. eugenii chromosome 7 was used as a probe
on metaphase chromosomes of P. assimilis (allied rockwallaby),
a genus that diverged early in the macropodine
radiation (Eldridge and Close 1993; O'Neill et al. 2004).
Within this genus, examples of centromere repositioning are
abundant. P. assimilis contains an acrocentric chromosome 7,
the likely result of a centric shift or other unknown
mechanism (Eldridge and Close 1993).This BAC maps to
the interstitial, inactive centromere location, not the derived
acrocentric location (Figure 3). This indicates that centromere
repositioning has occurred on this chromosome and
further reinforces our results from M. eugenii (which indicate
that BOS retain centromere sequences).
Conservation of KERV has been defined in other
marsupial groups in an effort to extend our analyses beyond
the family Macropodidae. KERV sequences have been
identified through Southern analyses in all marsupial species
Figure 2. Ideogram indicating the 19 conserved chromosome segments (Rens et al. 2003) and the location of
KERV/satellite sequences as determined through primed in situ hybridization (Ferreri et al. 2004) and BAC mapping (black
dots). Arrows indicate the location of non-BOS associated BACs at sites that correspond to centromere locations in other
marsupial species.
221
Ferreri et al.  Latent Centromeres
examined thus far, a data set that includes 36 species and six
families (Figure 4A). Comparative analyses for another
organism for which a genome sequence will be available
afford opportunities to examine the potential involvement of
segmental duplications in defining BOS and latent centromere
retention. To this end, a portion of the KERV
sequence containing portions of the gag-pro-pol genes was
used to screen genomic DNA from the South American
marsupial Monodelphis domestica. Southern analyses under low
stringency conditions show several distinct hybridizing fragments
within the genome of this species (Figure 4B). The
data has been confirmed by preliminary BLAST sequence
alignment between the M. rufogriseus KERV sequences and
the initial release of BAC end sequence for M. domestica (data
not shown). Likewise, Southern analyses indicate that the
functional satellite associated with KERV is also conserved
in this species (Figure 4B). However, the retention of
functional domains (i.e., CENPB DNA binding domains)
has not been determined. Conservation of this sequence
across ;180 million years of divergence is remarkable given
the apparent lack of conservation of satellites within other
eukaryotic systems. The extent of sequence conservation,
retention of functional retroviral genes, and correlation with
centromeres or BOS must be determined.
It is evident from studies of neocentromere hot spots in
the human karyotype and ancestral centromere locations
in primates that the latent centromere hypothesis, either
defined by sequence or the involvement of segmental
duplications, may explain centromere emergence in primate
lineages. The data presented herein, in contrast to that from
primates, clearly indicates that latent centromere sequences
are stringently conserved within marsupials. Our data,
showing an involvement of segmental duplications with
BOS within macropodines and retention of latent centromeres
following centromere repositioning, supports the
latent centromere hypothesis in defining centromere locations
within this group of mammals. The remarkable retention
of latent centromeres throughout marsupials, a large
and karyotypically diverse group of mammals, may have
facilitated the chromosomal radiations that several marsupial
families have experienced.
The retention of latent centromeres, either as a sequencespecific
or functional domain, within the karyotypes of
a group of mammals may provide a means of introducing
new karyotypic variability or of restricting gene flow. This
Figure 4. (A) Phylogeny (Amrine-Madsen et al. 2003) of
marsupial families and subfamilies. Superimposed onto the
branches are the terminal lineages in which KERV sequences
have been identified (bold) and the inferred ancestral
conservation of KERV (bold). Lineages in which analyses
were performed are indicated in italics. Note that only these
six families have been studied; an absence of KERV cannot
be inferred in any lineage from this date set. (B) Southern
analyses of M. eugenii KERV (left panel) and the functional
centromeric satellite (right) in M. domestica. A size reference is
shown to the left. Genomic DNA was digested with (a) BglII,
(b) EcoRI, (c) MspI, and (d) PstI.
Figure 3. Centromere shift on chromosome 7 between
M. eugenii and P. assimilis. (A) A schematic representation of
the shift in correlation to the conserved chromosome
segments, indicated on the left. (B) Fluorescence in situ
hybridization of the same BAC (green) to metaphase
chromosome 7 (blue) of M. eugenii (left) and P. assimilis
(right). The centromere for each chromosome is indicated
with a red arrow.
222
Journal of Heredity 2005:96(3)
retention does not necessarily constitute a selective advantage
per se, but it implicates latent centromeres in the process
of karyotypic diversification. This may be achieved randomly
through the acquisition of a fertility barrier in the form of
a heterozygote for centromere locations on homologous
chromosomes. Such a mechanism may foster sympatric or
parapatric speciation. While heterozygotes may present as
the result of a random event, as has been observed in
neocentromere cases in humans, they may be found in the
form of intra- and interspecific hybrids. As is evident from
our own analyses of centromere rearrangements in hybrids,
this portion of the genome may be destabilized and rapidly
altered. Such alterations, perhaps through the use of latent
centromeres, may result in the rescue of otherwise
deleterious heterozygosity and/or may allow for introgression
of novel genomic domains. Furthermore, fixation of
new karyotypes within a population may proceed rapidly due
to other forces, such as meiotic drive.
Acknowledgments
We thank M. J. O'Neill for editorial comments and helpful discussion
regarding this manuscript, K. Bulazel for sharing unpublished data, and
C. Metcalfe for chromosome templates used in included figures. This work is
supported by a National Science Foundation grant to R.J.O. (NSF-0093250).
This paper is based on a presentation given at the 2004 Annual Meeting of
the American Genetic Association, ``Genomes and Evolution 2004,''
Pennsylvania State University, State College, PA, June 17­20, 2004.
References
Alonso A, Mahmood R, Li S, Cheung F, Yoda K, and Warburton PE, 2003.
Genomic microarray analysis reveals distinct locations for the CENP-A
binding domains in three human chromosome 13q32 neocentromeres.
Hum Mol Genet 12:2711­2721.
Amor DJ, Bentley K, Ryan J, Perry J, Wong L, Slater H, and Choo KH,
2004. Human centromere repositioning ``in progress''. Proc Natl Acad Sci
USA 101:6542­6547.
Amor DJ and Choo KH, 2002. Neocentromeres: role in human disease,
evolution, and centromere study. Am J Hum Genet 71:695­714.
Amrine-Madsen H, Scally M, Westerman M, Stanhope MJ, Krajewski C,
and Springer MS, 2003. Nuclear gene sequences provide evidence for
the monophyly of australidelphian marsupials. Mol Phylogenet Evol
28:186­196.
Armengol L, Pujana MA, Cheung J, Scherer SW, and Estivill X, 2003.
Enrichment of segmental duplications in regions of breaks of synteny
between the human and mouse genomes suggest their involvement in
evolutionary rearrangements. Hum Mol Genet 12:2201­2208.
Bailey JA, Baertsch R, Kent WJ, Haussler D, and Eichler EE, 2004.
Hotspots of mammalian chromosomal evolution. Genome Biol 5:R23.
Bailey JA, Gu Z, Clark RA, Reinert K, Samonte RV, Schwartz S, Adams
MD, Myers EW, Li PW, and Eichler EE, 2002. Recent segmental
duplications in the human genome. Science 297:1003­1007.
Bailey JA, Liu G, and Eichler EE, 2003. An Alu transposition model for the
origin and expansion of human segmental duplications. Am J Hum Genet
73:823­834.
Bailey JA, Yavor AM, Massa HF, Trask BJ, and Eichler EE, 2001.
Segmental duplications: organization and impact within the current human
genome project assembly. Genome Res 11:1005­1017.
Band MR, Larson JH, Rebeiz M, Green CA, Heyen DW, Donovan J,
Windish R, Steining C, Mahyuddin P, Womack JE, and Lewin HA, 2000.
An ordered comparative map of the cattle and human genomes. Genome
Res 10:1359­1368.
Barry AE, Howman EV, Cancilla MR, Saffery R, and Choo KH, 1999.
Sequence analysis of an 80 kb human neocentromere. Hum Mol Genet
8:217­227.
Choo KH, 1997. Centromere DNA dynamics: latent centromeres and
neocentromere formation. Am J Hum Genet 61:1225­1233.
Consortium IHGS, 2001. Initial sequencing and analysis of the human
genome. Nature 409:860.
Csink AK and Henikoff S, 1998. Something from nothing: the evolution
and utility of satellite repeats. Trends Genet 14:200­204.
Dehal P, Predki P, Olsen AS, Kobayashi A, Folta P, Lucas S, Land M, Terry
A, Ecale Zhou CL, Rash S, Zhang Q, Gordon L, Kim J, Elkin C, Pollard
MJ, Richardson P, Rokhsar D, Uberbacher E, Hawkins T, Branscomb E,
and Stubbs L, 2001. Human chromosome 19 and related regions in mouse:
conservative and lineage-specific evolution. Science 293:104­111.
du Sart D, Cancilla MR, Earle E, Mao JI, Saffery R, Tainton KM, Kalitsis P,
Martyn J, Barry AE, and Choo KH, 1997. A functional neo-centromere
formed through activation of a latent human centromere and consisting of
non-alpha-satellite DNA. Nat Genet 16:144­153.
Eder V, Ventura M, Ianigro M, Teti M, Rocchi M, and Archidiacono N,
2003. Chromosome 6 phylogeny in primates and centromere repositioning.
Mol Biol Evol 20:1506­1512.
Eichler EE, 1999. Repetitive conundrums of centromere structure and
function. Hum Mol Genet 8:151­155.
Eichler EE, 2001. Recent duplication, domain accretion and the dynamic
mutation of the human genome. Trends Genet 17:661­669.
Eldridge MD and Close RL, 1993. Radiation of chromosome shuffles. Curr
Opin Genet Dev 3:915­922.
Fan Y, Linardopoulou E, Friedman C, Williams E, and Trask BJ, 2002a.
Genomic structure and evolution of the ancestral chromosome fusion site
in 2q13­2q14.1 and paralogous regions on other human chromosomes.
Genome Res 12:1651­1662.
Fan Y, Newman T, Linardopoulou E, and Trask BJ, 2002b. Gene content
and function of the ancestral chromosome fusion site in human
chromosome 2q13­2q14.1 and paralogous regions. Genome Res
12:1663­1672.
Ferreri GC, Marzelli M, Rens W, and O'Neill RJ, 2004. A centromerespecific
retroviral element associated with breaks of synteny in macropodine
marsupials. Cytogenet Genome Res 107:115­118.
Golfier G, Chibon F, Aurias A, Chen XN, Korenberg J, Rossier J, and
Potier MC, 2003. The 200-kb segmental duplication on human chromosome
21 originates from a pericentromeric dissemination involving human
chromosomes 2, 18 and 13. Gene 312:51­59.
Gratacos M, Nadal M, Martin-Santos R, Pujana MA, Gago J, Peral B,
Armengol L, Ponsa I, Miro R, Bulbena A, and Estivill X, 2001. A
polymorphic genomic duplication on human chromosome 15 is a susceptibility
factor for panic and phobic disorders. Cell 106:367­379.
Haaf T, Warburton PE, and Willard HF, 1992. Integration of human alphasatellite
DNA into simian chromosomes: centromere protein binding and
disruption of normal chromosome segregation. Cell 70:681­696.
Hayman DL, 1977. Chromosome number-constancy and variation. In: The
biology of marsupials. (Gilmore D, ed). London: Macmillan.
Hayman DL, 1990. Marsupial cytogenetics. In: Marsupials from pouches
and eggs: genetics, breeding and evolution of marsupials and monotremes.
(Cooper DW, ed) Melbourne: CSIRO.
Henikoff S, Ahmad K, and Malik HS, 2001. The centromere paradox: stable
inheritance with rapidly evolving DNA. Science 293:1098­1102.
223
Ferreri et al.  Latent Centromeres
Jackson M, 2003. Duplicate, decouple, disperse: the evolutionary transience
of human centromeric regions. Curr Opin Genet Dev 13:629­635.
John B, 1988. Heterochromatin molecular and structural aspects.
Cambridge: Cambridge University Press.
Kirsch JAW and Lapointe F-J, 1995. Resolution of portions of the
kangaroo phylogeny (Marsupialia: Macropodidae) using DNA hybridization.
Biol J Linn Soc 55:309­328.
Knegt AC, Li S, Engelen JJ, Bijlsma EK, and Warburton PE, 2003. Prenatal
diagnosis of a karyotypically normal pregnancy in a mother with
a supernumerary neocentric 13q21­.13q22 chromosome and balancing
reciprocal deletion. Prenat Diagn 23:215­220.
Lo AW, Craig JM, Saffery R, Kalitsis P, Irvine DV, Earle E, Magliano DJ,
and Choo KH, 2001a. A 330 kb CENP-A binding domain and altered
replication timing at a human neocentromere. EMBO J 20:2087­2096.
Lo AW, Magliano DJ, Sibson MC, Kalitsis P, Craig JM, and Choo KH, 2001b.
A novel chromatin immunoprecipitation and array (CIA) analysis identifies
a 460-kb CENP-A-binding neocentromere DNA. Genome Res 11:448­457.
Martin CL, Wong A, Gross A, Chung J, Fantes JA, and Ledbetter DH,
2002. The evolutionary origin of human subtelomeric homologies--or
where the ends begin. Am J Hum Genet 70:972­984.
Mefford HC and Trask BJ, 2002. The complex structure and dynamic
evolution of human subtelomeres. Nat Rev Genet 3:91­102.
Muller S, Stanyon R, Finelli P, Archidiacono N, and Wienberg J, 2000.
Molecular cytogenetic dissection of human chromosomes 3 and 21
evolution. Proc Natl Acad Sci USA 97:206­211.
O'Neill RJ, Eldridge MDB, and Metcalfe CJ, 2004. Centromere dynamics
and chromosome evolution in marsupials. J Hered in press.
O'Neill RJ, O'Neill MJ, and Graves JA, 1998. Undermethylation associated
with retroelement activation and chromosome remodelling in an interspecific
mammalian hybrid. Nature 393:68­72.
O'Neill RJW, Eldridge MDB, and Graves JAM, 2001. Chromosome
heterozygosity and de novo chromosome rearrangements in interspecific
mammalian hybrids. Mamm Genome 12:256­259.
Pevzner P and Tesler G, 2003a. Genome rearrangements in mammalian
evolution: lessons from human and mouse genomes. Genome Res
13:37­45.
Pevzner P and Tesler G, 2003b. Human and mouse genomic sequences
reveal extensive breakpoint reuse in mammalian evolution. Proc Natl Acad
Sci USA 100:7672­7677.
PostlethwaitJH, WoodsIG, Ngo-Hazelett P, Yan YL, KellyPD, Chu F,Huang
H, Hill-Force A, and Talbot WS, 2000. Zebrafish comparative genomics and
the origins of vertebrate chromosomes. Genome Res 10:1890­1902.
Potier M, Dutriaux A, Orti R, Groet J, Gibelin N, Karadima G, Lutfalla G,
Lynn A, Van Broeckhoven C, Chakravarti A, Petersen M, Nizetic D,
Delabar J, and Rossier J, 1998. Two sequence-ready contigs spanning the
two copies of a 200-kb duplication on human 21q: partial sequence and
polymorphisms. Genomics 51:417­426.
Rens W, O'Brien PC, Fairclough H, Harman L, Graves JA, and FergusonSmith
MA, 2003. Reversal and convergence in marsupial chromosome
evolution. Cytogenet Genome Res 102:282­290.
Rhoades MM and Dempsey E, 1966. The effect of abnormal chromosome
10 on preferential segregation and crossing over in maize. Genetics 53:
989­1026.
Rofe RH, 1979. G-banding and chromosomal evolution in Australian
marsupials. Ph. D. thesis. Adelaide: University of Adelaide.
Singer MF, 1982. Highly repeated sequences in mammalian genomes. Int
Rev Cytol 76:67­112.
Stankiewicz P and Lupski JR, 2002. Molecular-evolutionary mechanisms for
genomic disorders. Curr Opin Genet Dev 12:312­319.
Sullivan BA and Willard HF, 1998. Stable dicentric X chromosomes with
two functional centromeres. Nat Genet 20:227­228.
Thomas JW, Schueler MG, Summers TJ, Blakesley RW, McDowell JC,
Thomas PJ, Idol JR, Maduro VV, Lee-Lin SQ, Touchman JW, Bouffard
GG, Beckstrom-Sternberg SM, and Green ED, 2003. Pericentromeric
duplications in the laboratory mouse. Genome Res 13:55­63.
Tusie-Luna MT and White PC, 1995. Gene conversions and unequal
crossovers between CYP21 (steroid 21-hydroxylase gene) and CYP21P
involve different mechanisms. Proc Natl Acad Sci USA 92:10796­10800.
Ventura M, Archidiacono N, and Rocchi M, 2001. Centromere emergence
in evolution. Genome Res 11:595­599.
Ventura M, Mudge JM, Palumbo V, Burn S, Blennow E, Pierluigi M,
Giorda R, Zuffardi O, Archidiacono N, Jackson MS, and Rocchi M, 2003.
Neocentromeres in 15q24­26 map to duplicons which flanked an ancestral
centromere in 15q25. Genome Res 13:2059­2068.
Ventura M, Weigl S, Carbone L, Cardone MF, Misceo D, Teti M,
ĎAddabbo P, Wandall A, Bjorck E, de Jong PJ, She X, Eichler EE,
Archidiacono N, and Rocchi M, 2004. Recurrent sites for new centromere
seeding. Genome Res 14:1696­1703.
Warburton PE, 2004. Chromosomal dynamics of human neocentromere
formation. Chromosome Res 12:617­626.
Willard HF, 1990. Centromeres of mammalian chromosomes. Trends
Genet 6:410­416.
Williams BC, Murphy TD, Goldberg ML, and Karpen GH, 1998.
Neocentromere activity of structurally acentric mini-chromosomes in
Drosophila. Nat Genet 18:30­37.
Corresponding Editor: Shozo Yokoyama
224
Journal of Heredity 2005:96(3)