eukaryotic gene regulation in three dimensions and its ... · eukaryotic gene regulation and its...

COGEDE-548; NO OF PAGES 12

Available online at www.sciencedirect.com

Eukaryotic gene regulation in three dimensions and its impact ongenome evolutionM Madan Babu1, Sarath Chandra Janga1, Ines de Santiago2,3 andAna Pombo2

Recent advances in molecular techniques and high-resolution

imaging are beginning to provide exciting insights into the

higher order chromatin organization within the cell nucleus and

its influence on eukaryotic gene regulation. This improved

understanding of gene regulation also raises fundamental

questions about how spatial features might have constrained

the organization of genes on eukaryotic chromosomes and how

mutations that affect these processes might contribute to

disease conditions. In this review, we discuss recent studies

that highlight the role of spatial components in gene regulation

and their impact on genome evolution. We then address

implications for human diseases and outline new directions for

future research.

Addresses1 MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2

0QH, UK2 MRC Clinical Sciences Centre, Imperial College School of Medicine,

Hammersmith Hospital Campus, Du Cane Road, London W12 0NN, UK3 Instituto Gulbenkian de Ciencia, Oeiras, Portugal

Corresponding author: Madan Babu, M ([email protected])

Current Opinion in Genetics & Development 2008, 18:1–12

This review comes from a themed issue on

Genomes and evolution

Edited by Sarah Teichmann and Nipam Patel

0959-437X/$ – see front matter

# 2008 Elsevier Ltd. All rights reserved.

DOI 10.1016/j.gde.2008.10.002

IntroductionAlthough we observe an amazing diversity in the num-

ber of chromosomes that eukaryotic organisms encode

(e.g. 32 chromosomes in yeast, over 200 in butterflies,

and 46 in humans), they are packaged in similar ways:

each DNA molecule is wrapped around histone proteins

to form nucleosomes, which are then condensed in a

complex hierarchical manner to make up an entire

chromosome (Figure 1). Such an intricate organization

of genetic material within the eukaryotic nucleus pro-

vides ample opportunities to regulate expression of the

encoded genes at many different hierarchical levels. For

instance, eukaryotic transcription is dynamically

regulated at least at three major levels [1,2]. The first

is at the level of DNA sequence (Figure 1) where DNA-

Please cite this article in press as: Madan Babu M, et al. Eukaryotic gene regulation in three d

j.gde.2008.10.002

www.sciencedirect.com

binding proteins (e.g. transcription factors; TFs) associ-

ate with cis-regulatory elements (e.g. TF-binding sites)

to regulate transcription. The second is at the level of

chromatin (Figure 1), which allows segments within a

chromosomal arm to switch between different transcrip-

tional states, that is, those that suppress transcription

(heterochromatin) and those that allow for gene acti-

vation (euchromatin). This involves changes in chro-

matin structure and nucleosome occupancy, both of

which are controlled by the interplay between several

factors such as nucleosome remodeling complexes,

histone modifications, and a variety of repressive and

activating mechanisms [3,4]. The third is at the level of

the entire chromosome (Figure 1) and includes posi-

tioning of chromosomes within the nuclear space (e.g.

closer to the nuclear periphery or next to internal

nuclear compartments) and spatial organization of

specific chromosomal loci within the nucleus, both of

which are known to influence gene expression

[5,6��,7,8,9�].

Several studies have investigated these mechanisms in

detail and have revealed that such processes involve

extensive physical and spatial association between dis-

tantly located genomic elements and widespread cross-

talk between the different levels. Advancements in

molecular techniques and high-resolution imaging

(Table 1) have facilitated investigation of the role of

spatial component in gene regulation and have provided

valuable insights into its importance in gene regulation

[5,6��,7,8,9�]. These studies also raise fundamental ques-

tions: Has the requirement for transcriptional regulation

and its spatial considerations constrained the way in

which genes are organized on chromosomes? If yes, in

what ways does it affect genome evolution? In the first

part of this review, we discuss recent studies that high-

light the importance of spatial components in gene

regulation. We then discuss some studies that address

how the requirements for gene regulation could have

constrained genome evolution. Finally, we discuss

implications for human disease, outline open questions,

and discuss how computational approaches can be help-

ful in investigating the prevalence of spatial regulatory

mechanisms and in understanding their impact on gen-

ome evolution. We wish to stress that the studies dis-

cussed have been carried out in different model systems

and that further research is necessary to assess whether

particular spatial mechanisms are universal or specific to

each system.

imensions and its impact on genome evolution, Curr Opin Genet Dev (2008), doi:10.1016/


http://dx.doi.org/10.1016/j.gde.2008.10.002


mailto:[email protected]


2 Genomes and evolution


Figure 1

Hierarchical organization of eukaryotic genetic material. Each DNA molecule is wrapped into nucleosomes, which form the chromatin and is ultimately

packaged into a chromosome that resides within the nucleus. The first level of regulation includes regulatory elements (e.g. enhancers and insulators),

DNA methylation and DNA structure (e.g. G-quadruplex and Z-DNA). The second level includes post-translational modification of nucleosomes (e.g.

histone methylation and acetylation) and remodeling of nucleosomes (e.g. histone chaperones and ATP-dependent remodeling complexes). The third

level of regulation includes chromosomal organization and the nuclear architecture (e.g. position of various loci and chromosomes within the nucleus).

Long-range interactions involving distalregulatory elementsRegulatory elements in eukaryotes can be spread over

several kilobases away from the associated gene. These

include binding sites for specific TFs, enhancer elements,

locus control regions (LCRs), and insulator elements

(Figure 2a). TF-binding sites are generally close to pro-

moter regions, but enhancer elements, LCRs, and insu-

lator elements can be present far away on the

chromosome and may influence the expression of more

than one gene simultaneously. Enhancers affect expres-

sion of nearby genes, whereas LCRs can affect several

genes that are distantly located within a genomic locus

spanning several kilobases [10]. Insulator elements can

block promiscuous enhancer–promoter interaction or act

as a barrier against the spreading of heterochromatin. The

former class of insulators functions by forming genomic

loops via long-range interactions and the latter class

prevents inappropriate gene expression by recruiting

nucleosome-modifying enzymes [11].

The formation of loops mediated by proteins bound to

specific elements along a chromosome appears to have a

central role in several processes as it can affect the

expression of several genes in a neighborhood [12�].Although only a few loops have been analyzed in detail

and the nature of the molecular forces that maintain them

remains unclear, recent evidence suggests that they are

found in several eukaryotes [10] and that the transcrip-

tional machinery itself could be a molecular tie [13–15]

(Figure 2b). Several studies that have used 3D-FISH,

chromosome conformation capture (3C; see Table 1)

[16��], and its variants 4C, 5C, and 6C [17] and live-cell

imaging [18] support the idea that active transcription

units are in close contact within the nuclear space


j.gde.2008.10.002


[14,19,20]. The results are consistent with a model for

genome organization in which active polymerases cluster

into transcription ‘factories’ bringing together distal genes

(Figure 2b) and where active genes are dynamically

organized into shared nuclear subcompartments [14,18–20]. They are also consistent with these cis-regulatory

elements functioning as insulators, enhancers or LCRs,

depending on their positions relative to other genes.

Interestingly, in a recent study, it has been shown that

specific ‘factories’ produce only a particular kind of tran-

script depending on the promoter type and whether or not

the gene contains an intron [21�]. These results support

the presence of ‘specialized’ transcription factories, which

transcribe specific transcription units depending on pro-

moter type or presence of introns [19,20,21�].

The genomic loops involving regulatory elements are

dynamic, depend on the transcriptional status of a gene,

vary between cell types in the same organism and may

involve several proteins. The beta-globin locus in mouse

is the most studied one and involves the Hbb-b1 gene

(which encodes beta-globin), its LCR and the Eraf gene

(encoding an alpha-globin-stabilizing protein) on the

same chromosome. This LCR is thought to nucleate a

chromatin hub which correlates with expression of globin-

related genes. It has been confirmed, by 3C, 4C, and

RNA-TRAP that the contacts between Hbb-b1, the

LCR, and Eraf are seen only in erythroid nuclei (in which

all three are transcribed) but not in brain cell nuclei (in

which Hbb-b1 is inactive) [14,22�,23]. Moreover, the

contacts that Hbb-b1 makes with other genomic regions

depend on its transcriptional activity; in erythroid nuclei,

80% of contacts are with other active genes, but, in brain

cells, this falls to only 13% [22�]. Interestingly, the LCR

region itself is also transcribed and this might even be





Eukaryotic gene regulation and its impact on genome evolution Madan Babu et al. 3


Please cite this article in press as: Madan Babu M, et al. Eukaryotic gene regulation in three dimensions and its impact on genome evolution, Curr Opin Genet Dev (2008), doi:10.1016/

j.gde.2008.10.002

Table 1

Experimental and computational approaches to study eukaryotic transcriptional regulation.

Experimental

approaches

Description

ChIP-chip Chromatin-bound proteins are covalently linked to DNA by using an in vivo crosslinking agent such as formaldehyde

(histones can be detected in unfixed chromatin preparations in native ChIP). Chromatin is then sheared and

immunoprecipitated (ChIP) using an antibody for a native protein, a tagged version, or a specific post-translational

modification. Reversal of the crosslink releases the bound DNA, allowing the enrichment of specific DNA fragments,

whose identity is determined by hybridization to a microarray (chip).

ChIP-seq In ChIP-seq experiments, the immunoprecipitated DNA is directly sequenced using high-throughput sequencing

technologies (e.g. Solexa or 454). The sequences are then computationally mapped back to the reference genome.

Fragments that were bound by the protein will be more abundant and sequenced several times, providing a direct

measure of enrichment.

DamID The DNA-binding protein of interest is fused to an E. coli protein, Dam. Dam methylates the N6 position of the adenine

in the sequence GATC, which is expected to occur once in every �256 bases. Upon binding DNA, the Dam protein

preferentially methylates adenine in the vicinity of binding. The DNA is digested by DpnI and DpnII restriction enzymes,

which cleave within the nonmethylated GATC sequence, and remove fragments that are not methylated. The remaining

methylated fragments are amplified by selective PCR and quantified using a microarray.

RNA-TRAP Newly made transcripts are detected in crosslinked cells by RNA-FISH using biotinylated probes and

probe–RNA–chromatin complexes are amplified with tyramide or directly immunoprecipitated, before PCR analyses.

Chromosome

conformation

capture (3C)

3C is used to determine which DNA sequences lie close together in 3D space in fixed cells. This typically involves

fixation to crosslink DNA sequences that lie next to each other (usually through DNA–protein–DNA links), before

cutting with a restriction enzyme, dilution, and ligation at low concentration. This favors the ligation of pairs of DNA

sequences that are crosslinked after which the reversing of crosslinks allows the ligated DNA to be detected by PCR.

4C 4C technology [chromosome conformation capture on chip (3C-on-chip) or circular chromosome conformation capture

(circular-3C)] allows for an unbiased genome-wide search for DNA loci that contact a given locus.

5C Chromosome conformation capture carbon copy (5C) is a massively parallel technique, which involves mapping physical

interactions between genomic elements and sequencing or microarray analysis of the ligated end products of the 3C

technique. 3C typically converts physical chromatin interactions into specific ligation products, which are quantified

using high-throughput microarrays or quantitative DNA sequencing using 454-technology as detection methods.

6C Combines ChIP for a specific chromatin bound protein with 3C-based methods to correlate specific long-range

chromatin interactions with the presence of a specific bound protein.

FISH Fluorescent in situ hybridization (FISH) detects specific DNA sequences and localizes them on cytogenetic preparations of

chromosomes or interphase cell nuclei. Cells are hypotonically swollen and dropped on glass slides before hybridization,

such that fine structural details might be lost. It uses tagged probes amplified from specific DNA fragments up to single

chromosomes, to detect the target sequences. The genomic regions bound by the probe are visualized by

fluorescence microscopy.

3D-FISH A modified FISH procedure that improves the preservation of 3D nuclear structure (3D-FISH), important for spatial

mapping of the position of specific genomic sequences within the interphase nuclei. This technique can be slow as it

requires imaging of multiple image stacks on a small number of nuclei and 3D reconstruction. It can also be

combined with protein and RNA localization.

Cryo-FISH A modified FISH procedure that uses ultra-thin cryosections from sucrose-embedded fixed samples. Sections are 100 to

200-nm thick. Preservation of ultrastructure is optimized, signal-to-noise ratios are improved and imaging artifacts are

minimized. It is ideal for imaging short-range interactions between specific loci or their associations with specific landmarks

with higher resolution and faster data collection. Specific cells in their tissue context can be easily investigated.

Single molecule

imaging

(fluorescence

microscopy)

In single molecule imaging (SMI) of live cells, the molecules of interest are conjugated with fluorophores and introduced

into cells. The behavior of multiple fluorescent molecules in cells is then visualized using high-sensitivity video microscopy.

The observables in SMI are the position or movement of the fluorescent spots, the fluorescence intensity of individual

spots, the fluorescence spectrum or color of individual spots, and the number and distribution of the spots.

Lac-binding-site

array

In this approach, in vivo visualization of chromatin dynamics is based on lac repressor recognition of direct repeats of

the lac operator. The method allows tagging of specific chromosomal sites and thus in situ localization in vivo. Detection

by light microscopy, using GFP–lac–repressor fusion proteins or immunofluorescence, can be complemented by higher

resolution electron microscopy using immunogold staining. This method facilitates the investigation of interphase

chromosome dynamics, as well as chromosome segregation during cell division in organisms that lack cytologically

condensed chromosomes.

Computational

approaches

Description

Boolean modeling In qualitative modeling, kinetic processes are simulated by tracking over discrete time, the state of the system, defined in

terms of a coarse range for each variable. The weak specification of such models conserves computer resources needed

to explore the space of possible behaviors. Moreover, it provides high-level predictions applying to a whole family of

systems. Although simulation of qualitative models can be fast, even a rough exploration of parameter space can become

intractable as the size of the system increases, highlighting the need for increasing computer resources and methods

to accelerate the parameters’ search space. For genes that are naturally found in only two states (e.g. on or off), the

trade-off in accuracy may not be high. On the contrary, simple models can, in some cases, predict behaviors that

are far from reality.

www.sciencedirect.com Current Opinion in Genetics & Development 2008, 18:1–12





Table 1 (Continued )

Computational

approaches

Description

Deterministic

modeling

Deterministic modeling falls into the class of quantitative models. The most popular formalism is the deterministic

ordinary differential equations (ODEs) which, when extended to model space, is referred to as partial differential

equations (PDEs). Each equation in a set typically represents the rate of change of a species’ continuous

concentration as a sum or product of, more or less, empirical terms. This accounts for the effect of biological

events on the concentration of the species. By definition, the initial state of the system in a deterministic model

uniquely sets all future states. As analytical solutions seldom exist, numerical solutions need to be computed

(once for each set of parameter values and initial conditions explored).

Stochastic modeling Molecular interactions involving a small number of objects in a large volume are intrinsically random and cellular

behavior itself sometimes seems to reflect this randomness. Indeed, occurrences of ‘noise’ have been found to be

exploited by cells — for instance, to survive a variety of environmental changes or to increase sensitivity in signal

transduction processes. To model such stochastic systems, two main methods are used. The first comprises using

stochastic differential equations (SDEs; derived from ODEs by adding noise terms to the equations), the solutions for

which can be numerically obtained either by computing many trajectories (Monte Carlo methods) or approximating

their probability distribution and then calculating statistical measures (such as mean and variance). The second is

an exact method which can cope with different reaction time-scales or spaces. Within this approach, molecules are

modeled individually and reaction events are calculated by their probability.

Monte Carlo

simulation

Monte Carlo methods are a class of computational algorithms that rely on repeated random sampling to compute their

results. Monte Carlo methods are often used when simulating physical and mathematical systems. Because of their

reliance on repeated computation and random or pseudo-random numbers, Monte Carlo methods are most suited for

computer simulations and tend to be used when it is infeasible or impossible to compute an exact result with a

deterministic algorithm. There is no single Monte Carlo method; instead, the term describes a large and widely

used class of approaches.

Multiscale modeling Multiscale modeling refers to the modeling of a system at several levels of detail to increase the accuracy and

representation of the system as close to reality. For instance, modeling of a chromatin unit, a nucleosome, using a

simplified model for rapid discrete molecular dynamics simulations and an all-atom model for detailed structural

investigation, would correspond to this class of modeling.

Statistical

correlations

Statistical models search for patterns in experimental data. Correlation, regression, and cluster analysis are all powerful

statistical tools that can identify relationships among measured variables that probably are not attributable to chance.

Statistics is also a powerful tool for uncovering the prevalence of a phenomena and evidence for potentially new

mechanisms.

Spatial modeling Spatial modeling takes into account that biological processes take place in heterogeneous and highly structured

environments regulating cellular processes in both space and time. While recent technological advances are addressing

the dearth of spatial data, theoretical advances are improving computational methods, making it now possible to

simulate spatio-temporal models of biological processes in coarse-grained or realistic geometries.

Kinetic modeling Kinetic modeling supports quantitative hypothesis testing by first translating a diagram into a mechanistic kinetic model.

Diagrams typically consist of molecules, complexes, cellular locations, and processes. As molecules and complexes can

exist in several locations, it is often necessary to define several states for a single molecule — each state is a set of

chemical species in a physical place.

Comparative

genomics

Comparative genomics permits addressing questions at the sequence level both within and across organisms and their

variations across diverse phylogenetic groups. Evolutionary aspects of several cellular elements from genes, regulatory

elements, organellar macromolecular complexes to their chromosomal organization can be addressed using

computational genome-scale approaches.

Network-based

approaches

In a network approach, objects are represented as nodes and interactions between objects are represented as links.

This permits representation of genome-scale information in a convenient way to identify interesting topological features.

One of the features is the presence of hub nodes which are objects connected to an extremely high number of other

objects in a system. With the amount of data from several high-throughput technologies, representing interactions

between biological molecules as networks has provided us with a general framework to address fundamental

biological questions at a systems level. Examples of molecular interactions represented as networks include

protein–DNA (transcription network), protein–RNA (post-transcriptional network), protein–protein (post-translational

network, signaling, and protein complexes), and protein–metabolite (metabolic network) interactions.

required for its function [24]. A range of TFs has been

implicated in mediating genomic loops in the case of the

Hbb-b1 locus and these include Eklf (erythroid Kruppel-

like factor; aka Klf1), Gata1 (GATA-binding protein 1),

and the zinc-finger protein Fog1 (aka Zfpm1) [25,26].

Inter-chromosomal interactionsWhile long-range regulatory interactions involving loci

from the same chromosome have been known for

sometime, it is only recently that inter-chromosomal regu-

latory interactions (trans-interactions) were discovered


j.gde.2008.10.002


(Figure 2c). Inter-chromosomal interactions may involve

enhancer elements and genes from different chromosomes

and can be cell-type-specific. The first example of a trans-interaction between chromosomes in mammals (identified

by 3C and FISH) is the association between the T helper 2

cytokine locus on mouse chromosome 11 and the promoter

of IFN-gamma gene on chromosome 10 in the nuclei of

naıve CD4+ T-cells [27]. This interaction was proposed to

hold the two loci in a poised state and might facilitate a

quicker response upon T-cell activation to differentiate

into Th1 and Th2 lineages by expression of either gene-







Figure 2

Features of genome architecture in 3D. (a) The DNA is shown as a black line, a gene is represented as an arrow and the different classes of regulatory

elements are shown in various shapes and colors. Insulator elements (blue rectangles) block spread of heterochromatin (red circles) and prevent

inappropriate interaction between enhancers (green oval) and unrelated genes. Enhancers can facilitate regulation of nearby genes that may still be a

few kilobases away. Locus control region (gray oval) can bring genomic loci that are several kilobases away close to each other to coordinate gene

expression. The bottom panel shows various aspects of the spatial component in eukaryotic gene regulation. The nucleus is shown in the center. (b)

Different active regions of the same or different chromosomes can associate with the same transcription factory (yellow domain). (c) Enhancers from

one chromosome may regulate the expression of genes present on another chromosome via inter-chromosomal interactions. (d) Chromosomes

occupy defined volumes within the nucleus, called chromosomal territories, which are depicted in different colors with significant intermingling mostly

at the edges. (e) Genetic material residing near the nuclear periphery has been correlated with gene silencing and gene activity. One theme that stands

out is that regions of the chromosome that interact with the lamina and the nuclear inner membrane are largely inactive in both mammals and yeast,

whereas loci that interact with the components of the nuclear pore appear to be transcriptionally active.

locus. Another interesting example is the regulation of

olfactory receptor genes. Dual RNA and DNA FISH

revealed that the expression of a specific olfactory gene

is accompanied by inter-chromosomal or intra-chromoso-

mal interactions between the active gene and a genomic

region on chromosome 14 containing an enhancer

sequence, referred to as the H element [28].

Recently, two different studies showed that inter-chro-

mosomal interactions may involve several factors and can

be induced upon exposure to specific stimuli or upon viral


j.gde.2008.10.002


infection. The dynamics of gene association with tran-

scription factories was investigated during immediate

early gene induction in mouse B lymphocytes and was

shown to result in a rapid relocation of the Myc proto-

oncogene on chromosome 15 to the same factory that

transcribed the Igh gene located on chromosome 12 [29].

The study on the investigation of the interferon (IFN-

beta) gene-locus upon viral infection reported that the

stochastic and monoallelic expression of the IFN-beta

gene depends on inter-chromosomal associations with

distinct genetic loci that could mediate binding of the







transcription factor NF-kappaB to the IFN-beta enhan-

cer, thus triggering transcription from this allele [30��].

Another prominent example comes from the mammalian

X-chromosome inactivation process, which involves a

specific trans-association of the X-inactivation center

(Xic) between the X chromosomes during early devel-

opment. Female cells carry two X chromosomes, one of

which is mostly silenced so that expression levels of X-

linked genes are comparable to those in male cells.

Recent studies detected a transient colocalization of

the Xic of the homologous chromosomes that precedes

the initiation of inactivation of one of the two chromo-

somes [31��] and that the chromatin insulator protein

(CTCF) is involved in mediating this interaction [32�].CTCF also colocalizes with cohesin at specific sites in

human and mouse chromosomes [33] and raises the

possibility that protein–chromatin interaction involving

genomic loci from different chromosomes could possibly

stabilize, at least transiently, a network of inter-chromo-

somal interactions within the cell nucleus.

Chromosomal territories, movement, andnuclear organizationDespite the growing evidence on inter-chromosomal

interactions, it is known that chromosomes occupy terri-

tories with preferred and nonrandom positions in the

nucleus of mammalian cells, the so-called chromosome

territories (CTs) [34,35] (Figure 2d). FISH experiments

have revealed the relocation of chromatin domains con-

taining activated genes to substantial distances outside

their CT, suggesting that positional organization of chro-

matin domains within the nucleus could impinge on the

regulation of gene expression. This finding, together with

the observation of extensive intermingling of DNA from

different chromosomes [36�] raises the issues of how and

why genes move relative to their CTs and whether the

looping out regulates, or is regulated by transcriptional

activity. Evidences in favor of a role for looping out in the

regulation of gene expression has come from studies that

show the colocalization of genes in the nucleus for coex-

pression or coregulation (reviewed in [6��]). Active genes

on decondensed chromatin loops extend outside CTs and

can colocalize both in cis and in trans at sites within the

nucleus to share the same transcription factories [14,29] or

to sites adjacent to splicing-factor-enriched speckles [37�]or Cajal bodies [38]. Extensive relocalization of large

genomic regions in response to gene activation can

depend on actin [37�,38] and myosin [37�], suggesting

that intranuclear movements of genomic regions are, at

least in some cases, more directed than previously

thought [39].

The spatial organization of CTs in mammalian cells can

be described by their radial positioning relative to the

center of the nucleus, as was recently done for the three-

dimensional (3D) map of all chromosomes in human male


j.gde.2008.10.002


fibroblast nuclei [40�]. The radial positioning of chromo-

somes correlates with their gene density in spherical

nuclei such as lymphocytes [41] and is evolutionarily

conserved [42,43] but nevertheless tissue-specific to a

certain degree [44]. Gene-rich regions tend to occupy

more interior positions, while gene-poor and late-replica-

tion regions tend to be associated with the nuclear per-

iphery [41,42]. In addition, similar nonrandom chromatin

arrangements with respect to the local gene density or GC

content have been observed for different cell types (e.g.

fibroblasts, bone-marrow cells, and cell lines) from several

eukaryotic lineages such as amphibians, reptiles, birds,

and mammals [42,45]. Surprisingly, a detailed analysis of

the position of chromosomes in mouse lymphocytes has

shown that chromosomes are more likely to form ‘heter-

ologous’ neighborhoods, where homologous chromo-

somes are preferentially separated from each other

[46], which might facilitate more extensive trans-inter-

actions between heterologous chromosomes.

Although the general patterns appear to be evolutionarily

conserved, they are nevertheless dynamic and are altered

during cellular differentiation. Changes in the transcrip-

tional program of a cell correlate with specific changes in

the organization of individual CTs, at the level of inter-

mingling, CT volume and radial position during lympho-

cyte activation [47], possibly reflecting an adaptation to

the new transcriptional program. Similarly, other recent

studies have shown that the architecture of CTs changes

during differentiation (e.g. human adipocyte differen-

tiation [48] and mouse T-cell differentiation [49]).

Association of genomic loci with the nuclearperipheryIn metazoan nuclei, the nuclear envelope is underlaid by

a continuous meshwork of lamins and lamin-associated

proteins, which preferentially associate with inactive

chromatin regions and facilitate chromatin organization

[50]. Pickersgill et al. [51] characterized the regions that

interacted with the nuclear lamina in Drosophila melano-gaster and showed enrichment for gene-poor regions and

repressed genes. More recently, Guelen et al. [52��]mapped the interaction sites of the entire genome with

the nuclear lamina components in human fibroblasts and

described over 1300 lamina-associated-domains (LAD)

which were again enriched for genes with low expression

levels. Although an association of silenced genes with

the nuclear periphery is demonstrated, what was unclear

from these studies was whether the requirement for

gene repression causes association or if repression is

an effect of association with the nuclear lamina. Exper-

imentally induced repositioning of human chromosomal

regions to the nuclear periphery in Finlan et al. [53�]suggests a causative role of the nuclear periphery in

suppressing the expression of some (but not all) genes

as repositioning to the periphery is still compatible with

active transcription. Another study investigating the







Figure 3

The interplay between spatial component in gene regulation and genome evolution, normal physiology, and disease states. *Although several factors

(e.g. functional relatedness of genes, replication origin sites, etc.) are known to influence genome organization, several studies clearly provide support

that transcriptional regulation is an important factor. +Position effect is defined as a change in the level of gene expression brought about by an

alteration in the position of the gene relative to its normal chromosomal environment, but not associated with an intragenic mutation or deletion.

consequences of repositioning the immunoglobulin loci

in mouse fibroblasts to the nuclear periphery supports

the notion that such molecular interactions may be a

mechanism to limit the accessibility to proteins that

facilitate recombination or transcription [54�].

While the nuclear periphery has been generally associ-

ated with repressed genes, several studies have shown a

correlation with active genes being associated with com-

ponents of the nuclear pore complexes (NPCs), which

serve as gates for the transport of molecules between the

nucleus and cytoplasm. ChIP experiments in yeast for

NPC components revealed an enrichment for active

genes [55]. Several inducible genes such as INO1,

HXK1, GAL1, GAL2, and HSP104 become stably posi-

tioned at the nuclear periphery when activated and

remain there after the transcription is shut off [55–58].

In the case of Gal1 and Ino1, the relocalization to pores

was found to be dependent upon the SAGA acetyl-trans-

ferase complex [58,59]. In humans and Drosophila, the

MSL complex can recruit transcriptionally active loci to

the nuclear pore [60], although another study revealed

that the association of silent genes is just as likely as for

active genes [61]. Most of these results have to be

reconciled with the observation that many (if not most)


j.gde.2008.10.002


transcribed genes, in both yeast [62] and mammals [63],

do not associate stably with pores. Despite this, one

common theme that is emerging is a tendency for the

inner nuclear membrane to be associated with less active

genes, whereas the NPCs tend to associate with tran-

scriptionally active loci, at least in yeast, possibly in order

to facilitate efficient transport of mRNAs (Figure 2e).

Implications for genome evolutionGiven that eukaryotic gene regulation involves several

events, including the spatial association between differ-

ent genomic loci and micro-compartmentalization within

the nuclear space to be coordinated in space and time, this

raises a fundamental question: have the spatial processes

involved in gene regulation constrained genome evol-

ution? If so, has this affected (i) the way in which genes

are organized on the chromosomes (ii) the evolution of

gene expression patterns between species and (iii) the

evolution of gene regulatory networks underlying specific

phenotypes? There are now several examples which

provide evidence that all of this has happened.

Computational analyses of genome sequence data and

gene expression datasets have revealed a general prin-

ciple that is applicable to several organisms: gene order is







nonrandom and genes with similar or coordinated expres-

sion profiles tend to be clustered. In particular, genes

which are coexpressed in a cell cycle dependent manner,

during specific developmental stages, in a tissue-specific

manner, or across different conditions tend to cluster

together on chromosomes [64,65,66��,67�]. For instance,

a recent analysis of the differentiation of hematopoietic

progenitor cells has shown a significant tendency for

coexpressed genes to be proximal, suggesting that proxi-

mity in the form of chromosomal gene distribution may

be important for organizing the genome to coordinate

gene regulation during cellular differentiation [66��].Investigation of the spacing of genes on yeast chromo-

somes revealed that regulatory elements outside coding

regions could affect intergenic distances in yeast [68].

Interestingly, genes in such coexpressed clusters tend to

be more conserved as clusters in other genomes than

random sets of proximal genes, suggesting that natural

selection might have kept them together for reasons such

as requirement to maintain the right expression levels of

clustered genes that encode for interacting proteins

[64,67�,69,70]. However, a neutral model of chromosomal

evolution also operates wherein neighboring genes with

smaller intergenic distances tend to stay closer together

while gene pairs with much larger intergenic distances

tend to be shuffled simply because they are more likely to

be involved in a recombination event [71]. In addition,

higher order regulatory mechanisms such as opening up of

chromatin regions may be involved in the coexpression of

proximal genes [72–74]. Gierman et al. [75�] showed that

identical green fluorescent protein (GFP) reporter con-

structs integrated at different chromosomal positions in

the human genome obtain expression levels that corre-

spond to the activity of the integration domain, high-

lighting the importance of genomic context in gene

expression. Remarkably, a recent study [76��], which

demonstrates the evolutionary significance of such a

phenomenon, discovered that duplication and change

in genomic neighborhood of SUN in tomato plants

increased its expression relative to that of the ancestral

copy. This event has resulted in an enormous variation in

fruit shape and has been selected for in today’s domesti-

cated tomato plants. Two recent reviews discuss

mutations that alter the genomic context or that result

in a change in ploidy, thereby resulting in evolution of

regulatory networks underlying key morphological traits

[77,78].

Although these studies have unambiguously revealed the

existence of chromosomal domains that contain several

genes with similar expression pattern (coexpressed

genes), clustering of coexpressed genes need not always

imply regulation by the same TF because coexpressed

genes may be clustered because of several reasons such as

mechanisms involving chromatin remodeling, transcrip-

tional read-through, regulation of genes by the same TF

or regulation by different TFs in the same transcription-


j.gde.2008.10.002


ally active euchromatin domain [72,79�,80,81]. A recent

study by Janga et al. [82��] investigated if the targets of

TFs display positional clustering on yeast chromosomes.

Although the targets of TFs cluster on chromosomes,

such clusters contained only �3 genes within windows of

20 genes, suggesting that all of the above-mentioned

mechanisms may contribute to domains of coexpression.

The existence of a higher order organization of genes

across and within chromosomes that is constrained by

transcriptional regulation was also demonstrated, as target

genes (TGs) of TFs were found to be encoded in a highly

ordered manner both across and within the 16 S. cerevisiaechromosomes. In particular, the TGs of a majority of TFs

show a strong preference to be encoded on specific

chromosomes, TGs of a significant number of TFs display

a strong preference (or avoidance) to be encoded in

regions containing particular chromosomal landmarks

such as telomeres and centromeres, and TGs of most

TFs are positionally clustered within a chromosome.

Therefore, specific organization of genes that allows for

more efficient control of transcription within the nuclear

space has been selected during evolution. Taken

together, all these studies suggest that transcriptional

regulation and spatial constraints could be an important

driving force for genome organization (Figure 3).

Implications in disease and agingWhile the above-mentioned phenomena of genome evol-

ution operate on the evolutionary time-scale and

mutations which result in genome rearrangements get

fixed only when they occur in germ cells, occurrence of

such mutations in somatic cells also take place and result

in undesirable outcomes, such as cancer (Figure 3). For

instance, the observations that some chromosomes are

next to each other in specific chromosomal territories

[36�,83,84�], and that regions from different chromosomes

involving proto-oncogenes and regulatory elements for

highly expressed genes come together in space [29,85]

have all been correlated with specific translocation events

predominant in well-known cancers. Similarly, the fact

that several eukaryotic genomes contain repeat elements

coupled with the fact that chromosomes are not randomly

organized within the nucleus facilitates nonallelic hom-

ologous recombination when double strand breaks are

introduced in repetitive regions [86]. In fact, this might

even be a way to generate diversity in a population of

individuals as structural rearrangements of genetic

material can contribute to gene expression divergence,

genomic variation, and speciation [87].

Apart from cancer, mutations in nuclear lamina proteins

result in altered genome organization and have also been

implicated in specific human disorders collectively

termed as laminopathies. The diseases range from

Emery–Dreifuss muscular dystrophy to the accelerated

aging seen in Hutchinson-Gilford Progeria syndrome

[88]. Interestingly, lamins have been implicated in the







aging process; cells harboring mutant lamin proteins have

altered nuclear morphology and display increased geno-

mic instability because of perturbation of the DNA

damage response and repair pathway [89,90]. For an

excellent account of how the disruption of long-range

control of gene expression could result in disease, please

see a recent review by Kleinjan and van Heyningen

[91��].

Conclusions and outlookAlthough it is clear that the spatial components have been

exploited in eukaryotic gene regulation, its prevalence in

a genome, its occurrence in different organisms and how it

affects genome organization and evolution are only now

being addressed in detail. Future experiments from sev-

eral model systems coupled with analysis of genome-wide

datasets [92,93] are likely to provide insights into the

prevalence and universality of these mechanisms and

their impact on genome evolution.

Experimental advances in microscopy (e.g. multicolor

immuno-FISH [94] and thin section cryo-FISH [95])

and innovative techniques (e.g. magnetic nanoparticles

[96]) could allow to accelerate such studies in different

model organisms, cell types, different stages of develop-

ment such as cellular differentiation and in disease cell

types. With advances in high-throughput techniques,

such as ChIP-seq and variants of 3C, it is now possible

to investigate the interaction between different loci

within the nucleus at high resolution (Table 1). Such

studies will not only provide a deeper understanding of

the principles that govern long-range interactions but will

also facilitate generation of a physical map of the network

of interaction between different genomic loci within the

eukaryotic nucleus.

This is an exciting time for computational biologists who

aim to understand the impact of spatial component in

gene regulation on genome organization, evolution of

gene expression, and regulatory networks. New directions

in this area include objective treatment of the data by

developing statistical models at various levels of resol-

ution within the hierarchical framework of gene regula-

tion, 3D network representations of interacting loci in

different cell types, computational procedures for inte-

grating diverse datasets and comparative analysis across

different organisms [15,97–102] (Table 1). Such advances

can have direct applications in genetic engineering exper-

iments and gene therapy. For instance, describing the

map of spatial interactions between genomic loci in

higher eukaryotes will have implications in gene therapy

and in rationally identifying suitable sites to incorporate

reporter genes while producing transgenic organisms.

While the advances in experimental and computational

approaches provide new roads to address these problems,

we believe that a combined approach will accelerate our


j.gde.2008.10.002


efforts and take us closer towards understanding the

contribution of the spatial component for gene regulation

in different cell types, disease states, development, and

ultimately genome evolution.

AcknowledgementsMMB and SCJ acknowledge MRC-LMB. SCJ acknowledges support fromCCT and MMB thanks Darwin College and Schlumberger Ltd for generoussupport. AP and IS thank the MRC-CSC for support. IS acknowledgesfunding from Fundacao para a Ciencia eTecnologia (SFRH/BD/33205/2007). We thank K Weber, A Wuster, R Janky, H Braberg, D Jani, DRhodes, A Emili, M Murthy, V Pisupati, A Cristino, L Costa, S Michnick, ATravers, and A Klug for providing comments on this manuscript.

References and recommended readingPapers of particular interest, published within the period of review,have been highlighted as:

� of special interest�� of outstanding interest

1. Lee TI, Young RA: Transcription of eukaryotic protein-codinggenes. Annu Rev Genet 2000, 34:77-137.

2. van Driel R, Fransz PF, Verschure PJ: The eukaryotic genome: asystem regulated at different hierarchical levels. J Cell Sci2003, 116:4067-4075.

3. Millar CB, Grunstein M: Genome-wide patterns of histonemodifications in yeast. Nat Rev Mol Cell Biol 2006, 7:657-666.

4. Razin SV, Iarovaia OV, Sjakste N, Sjakste T, Bagdoniene L,Rynditch AV, Eivazova ER, Lipinski M, Vassetzky YS: Chromatindomains and regulation of transcription. J Mol Biol 2007,369:597-607.

5. Schneider R, Grosschedl R: Dynamics and interplay of nucleararchitecture, genome organization, and gene expression.Genes Dev 2007, 21:3027-3043.

6.��

Fraser P, Bickmore W: Nuclear organization of the genome andthe potential for gene regulation. Nature 2007, 447:413-417.

This is a timely review on the impact of nuclear organization on generegulation.

7. Pombo A, Branco MR: Functional organisation of the genomeduring interphase. Curr Opin Genet Dev 2007, 17:451-455.

8. de Laat W, Grosveld F: Inter-chromosomal gene regulation inthe mammalian cell nucleus. Curr Opin Genet Dev 2007,17:456-464.

9.�

Misteli T: Beyond the sequence: cellular organization ofgenome function. Cell 2007, 128:787-800.

This review discusses the functional relevance of spatial and temporalgenome organization at three hierarchical levels: the organization ofnuclear processes, the higher order organization of the chromatin fiber,and the spatial arrangement of genomes within the cell nucleus.

10. Dean A: On a chromosome far, far away: LCRs and geneexpression. Trends Genet 2006, 22:38-45.

11. Dorman ER, Bushey AM, Corces VG: The role of insulatorelements in large-scale chromatin structure in interphase.Semin Cell Dev Biol 2007, 18:682-690.

12.�

O’Sullivan JM, Tan-Wong SM, Morillon A, Lee B, Coles J, Mellor J,Proudfoot NJ: Gene loops juxtapose promoters andterminators in yeast. Nat Genet 2004, 36:1014-1018.

This study reports that two long genes of Saccharomyces cerevisiae,FMP27 and SEN1, exist in a looped conformation, effectively bringingtheir promoter and terminator regions close together in space.

13. Grimaud C, Bantignies F, Pal-Bhadra M, Ghana P, Bhadra U,Cavalli G: RNAi components are required for nuclear clusteringof Polycomb group response elements. Cell 2006, 124:957-971.

14. Osborne CS, Chakalova L, Brown KE, Carter D, Horton A,Debrand E, Goyenechea B, Mitchell JA, Lopes S, Reik W et al.:Active genes dynamically colocalize to shared sites ofongoing transcription. Nat Genet 2004, 36:1065-1071.







15. Marenduzzo D, Faro-Trindade I, Cook PR: What are themolecular ties that maintain genomic loops? Trends Genet2007, 23:126-133.

16.��

Dekker J, Rippe K, Dekker M, Kleckner N: Capturingchromosome conformation. Science 2002, 295:1306-1311.

A landmark work that describes a molecular approach to detect thefrequency of interaction between any two genomic loci.

17. Simonis M, Kooren J, de Laat W: An evaluation of 3C-basedmethods to capture DNA interactions. Nat Methods 2007,4:895-901.

18. Muller WG, Rieder D, Karpova TS, John S, Trajanoski Z,McNally JG: Organization of chromatin and histonemodifications at a transcription site. J Cell Biol 2007,177:957-967.

19. Pombo A, Jones E, Iborra FJ, Kimura H, Sugaya K, Cook PR,Jackson DA: Specialized transcription factories withinmammalian nuclei. Crit Rev Eukaryot Gene Expr 2000, 10:21-29.

20. Pombo A, Jackson DA, Hollinshead M, Wang Z, Roeder RG,Cook PR: Regional specialization in human nuclei:visualization of discrete sites of transcription by RNApolymerase III. EMBO J 1999, 18:2241-2253.

21.�

Xu M, Cook PR: Similar active genes cluster in specializedtranscription factories. J Cell Biol 2008, 181:615-623.

This study demonstrates that transcription factories specialize in produ-cing a particular type of transcript depending on promoter type andwhether or not the gene contains an intron.

22.�

Simonis M, Klous P, Splinter E, Moshkin Y, Willemsen R, de Wit E,van Steensel B, de Laat W: Nuclear organization of active andinactive chromatin domains uncovered by chromosomeconformation capture-on-chip (4C). Nat Genet 2006, 38:1348-1354.

This study establishes 4C technology as a tool to study nuclear archi-tecture. It demonstrates that chromosomes fold into areas of activechromatin and areas of inactive chromatin. Both active and inactivegenes were shown to be engaged in long-range intra-chromosomaland inter-chromosomal interactions.

23. Carter D, Chakalova L, Osborne CS, Dai YF, Fraser P: Long-rangechromatin regulatory interactions in vivo. Nat Genet 2002,32:623-626.

24. Ho Y, Elefant F, Liebhaber SA, Cooke NE: Locus control regiontranscription plays an active role in long-range geneactivation. Mol Cell 2006, 23:365-375.

25. Drissen R, Palstra RJ, Gillemans N, Splinter E, Grosveld F,Philipsen S, de Laat W: The active spatial organization of thebeta-globin locus requires the transcription factor EKLF.Genes Dev 2004, 18:2485-2490.

26. Vakoc CR, Letting DL, Gheldof N, Sawado T, Bender MA,Groudine M, Weiss MJ, Dekker J, Blobel GA: Proximity amongdistant regulatory elements at the beta-globin locus requiresGATA-1 and FOG-1. Mol Cell 2005, 17:453-462.

27. Spilianakis CG, Lalioti MD, Town T, Lee GR, Flavell RA:Interchromosomal associations between alternativelyexpressed loci. Nature 2005, 435:637-645.

28. Lomvardas S, Barnea G, Pisapia DJ, Mendelsohn M, Kirkland J,Axel R: Interchromosomal interactions and olfactory receptorchoice. Cell 2006, 126:403-413.

29. Osborne CS, Chakalova L, Mitchell JA, Horton A, Wood AL,Bolland DJ, Corcoran AE, Fraser P: Myc dynamically andpreferentially relocates to a transcription factory occupied byIgh. PLoS Biol 2007, 5:e192.

30.��

Apostolou E, Thanos D: Virus infection induces NF-kappaB-dependent interchromosomal associations mediatingmonoallelic IFN-beta gene expression. Cell 2008, 134:85-96.

The authors report that the stochastic and monoallelic expression of theIFN-beta gene depends on inter-chromosomal associations with threeidentified genetic loci that could mediate binding of the transcriptionfactor NF-kappaB to the IFN-beta enhancer, thus triggering enhanceo-some assembly and activation of transcription from this allele.

31.��

Augui S, Filion GJ, Huart S, Nora E, Guggiari M, Maresca M,Stewart AF, Heard E: Sensing X chromosome pairs before X


j.gde.2008.10.002


inactivation via a novel X-pairing region of the Xic. Science2007, 318:1632-1636.

The authors demonstrate homologous associations driven by a novel X-pairing region (Xpr) of the Xic and propose that they enable cells to sensewhen more than one X chromosome is present and to coordinatereciprocal Xist/Tsix expression.

32.�

Xu N, Donohoe ME, Silva SS, Lee JT: Evidence that homologousX-chromosome pairing requires transcription and Ctcfprotein. Nat Genet 2007, 39:1390-1396.

This study shows that pairing of X-chromosomes requires Ctcf bindingand cotranscriptional activity of Tsix and Xite.

33. Parelho V, Hadjur S, Spivakov M, Leleu M, Sauer S, Gregson HC,Jarmuz A, Canzonetta C, Webster Z, Nesterova T et al.: Cohesinsfunctionally associate with CTCF on mammalian chromosomearms. Cell 2008, 132:422-433.

34. Cremer T, Cremer M, Dietzel S, Muller S, Solovei I, Fakan S:Chromosome territories — a functional nuclear landscape.Curr Opin Cell Biol 2006, 18:307-316.

35. Meaburn KJ, Misteli T: Cell biology: chromosome territories.Nature 2007, 445:379-781.

36.�

Branco MR, Pombo A: Intermingling of chromosome territoriesin interphase suggests role in translocations andtranscription-dependent associations. PLoS Biol 2006, 4:e138.

The authors provide evidence for extensive chromosome interminglingand suggest that local chromatin conformation and gene transcriptioninfluence the extent with which chromosomes interact and affect theiroverall properties, with direct consequences for cell type specific genomestability.

37.�

Chuang CH, Carpenter AE, Fuchsova B, Johnson T, deLanerolle P, Belmont AS: Long-range directional movement ofan interphase chromosome site. Curr Biol 2006, 16:825-831.

This study identifies chromatin movements of an inducible locus from thenuclear periphery to the nuclear interior, and shows that they depend onactin and nuclear myosin I.

38. Dundr M, Ospina JK, Sung MH, John S, Upender M, Ried T,Hager GL, Matera AG: Actin-dependent intranuclearrepositioning of an active gene locus in vivo. J Cell Biol 2007,179:1095-1103.

39. Kumaran RI, Thakar R, Spector DL: Chromatin dynamics andgene positioning. Cell 2008, 132:929-934.

40.�

Bolzer A, Kreth G, Solovei I, Koehler D, Saracoglu K, Fauth C,Muller S, Eils R, Cremer C, Speicher MR et al.: Three-dimensionalmaps of all chromosomes in human male fibroblast nuclei andprometaphase rosettes. PLoS Biol 2005, 3:e157.

In this study, the authors developed tools that allow the quantitative 3Dpositional mapping of all chromosomes simultaneously. They presentunequivocal evidence for a probabilistic 3D order of prometaphasechromosomes, as well as of CTs in nuclei of quiescent (G0) and cycling(early S-phase) human diploid fibroblasts (46, XY).

41. Kupper K, Kolbl A, Biener D, Dittrich S, von Hase J, Thormeyer T,Fiegler H, Carter NP, Speicher MR, Cremer T et al.: Radialchromatin positioning is shaped by local gene density, not bygene expression. Chromosoma 2007, 116:285-306.

42. Neusser M, Schubel V, Koch A, Cremer T, Muller S: Evolutionarilyconserved, cell type and species-specific higher orderchromatin arrangements in interphase nuclei of primates.Chromosoma 2007, 116:307-320.

43. Mora L, Sanchez I, Garcia M, Ponsa M: Chromosome territorypositioning of conserved homologous chromosomes indifferent primate species. Chromosoma 2006, 115:367-375.

44. Parada LA, McQueen PG, Misteli T: Tissue-specific spatialorganization of genomes. Genome Biol 2004, 5:R44.

45. Federico C, Scavo C, Cantarella CD, Motta S, Saccone S,Bernardi G: Gene-rich and gene-poor chromosomal regionshave different locations in the interphase nuclei of cold-blooded vertebrates. Chromosoma 2006, 115:123-128.

46. Khalil A, Grant JL, Caddle LB, Atzema E, Mills KD, Arneodo A:Chromosome territories have a highly nonsphericalmorphology and nonrandom positioning. Chromosome Res2007, 15:899-916.







47. Branco MR, Branco T, Ramirez F, Pombo A: Changes inchromosome organization during PHA-activation of restinghuman lymphocytes measured by cryo-FISH. ChromosomeRes 2008, 16:413-426.

48. Kuroda M, Tanabe H, Yoshida K, Oikawa K, Saito A, Kiyuna T,Mizusawa H, Mukai K: Alteration of chromosome positioningduring adipocyte differentiation. J Cell Sci 2004,117:5897-5903.

49. Kim SH, McQueen PG, Lichtman MK, Shevach EM, Parada LA,Misteli T: Spatial genome organization during T-celldifferentiation. Cytogenet Genome Res 2004, 105:292-301.

50. Akhtar A, Gasser SM: The nuclear envelope and transcriptionalcontrol. Nat Rev Genet 2007, 8:507-517.

51. Pickersgill H, Kalverda B, de Wit E, Talhout W, Fornerod M, vanSteensel B: Characterization of the Drosophila melanogastergenome at the nuclear lamina. Nat Genet 2006, 38:1005-1014.

52.��

Guelen L, Pagie L, Brasset E, Meuleman W, Faza MB, Talhout W,Eussen BH, de Klein A, Wessels L, de Laat W et al.: Domainorganization of human chromosomes revealed by mapping ofnuclear lamina interactions. Nature 2008, 453:948-951.

The authors map the interaction sites across the human genome to thenuclear lamina and show that LADs represent repressive chromatindomains that organize chromosomes within the nucleus.

53.�

Finlan LE, Sproul D, Thomson I, Boyle S, Kerr E, Perry P, Ylstra B,Chubb JR, Bickmore WA: Recruitment to the nuclear peripherycan alter expression of genes in human cells. PLoS Genet 2008,4:e1000039.

This study uses in vivo targeting of two subchromosomal regions to thenuclear periphery and shows that the radial position within the nucleuscan influence the expression of some, but not all, genes. It argues thatrelocalization of genes relative to the peripheral zone of the nucleus couldbe used by metazoans to modulate the expression of selected genesduring development and differentiation.

54.�

Reddy KL, Zullo JM, Bertolino E, Singh H: Transcriptionalrepression mediated by repositioning of genes to the nuclearlamina. Nature 2008, 452:243-247.

Using 3D DNA-immuno-FISH, the authors demonstrate repositioning ofchromosomal regions to the nuclear lamina that is dependent on break-down and reformation of the nuclear envelope during mitosis. Theypropose that such molecular interactions may be used to compartmen-talize and to limit the accessibility of immunoglobulin loci to transcriptionand recombination factors.

55. Casolari JM, Brown CR, Komili S, West J, Hieronymus H, Silver PA:Genome-wide localization of the nuclear transport machinerycouples transcriptional status and nuclear organization. Cell2004, 117:427-439.

56. Brickner DG, Cajigas I, Fondufe-Mittendorf Y, Ahmed S, Lee PC,Widom J, Brickner JH: H2A, Z-mediated localization of genes atthe nuclear periphery confers epigenetic memory of previoustranscriptional state. PLoS Biol 2007, 5:e81.

57. Taddei A, Van Houwe G, Hediger F, Kalck V, Cubizolles F,Schober H, Gasser SM: Nuclear pore association confersoptimal expression levels for an inducible yeast gene. Nature2006, 441:774-778.

58. Cabal GG, Genovesio A, Rodriguez-Navarro S, Zimmer C,Gadal O, Lesne A, Buc H, Feuerbach-Fournier F, Olivo-Marin JC,Hurt EC et al.: SAGA interacting factors confine sub-diffusionof transcribed genes to the nuclear envelope. Nature 2006,441:770-773.

59. Brickner JH, Walter P: Gene recruitment of the activated INO1locus to the nuclear membrane. PLoS Biol 2004, 2:e342.

60. Mendjan S, Taipale M, Kind J, Holz H, Gebhardt P, Schelder M,Vermeulen M, Buscaino A, Duncan K, Mueller J et al.: Nuclear porecomponents are involved in the transcriptional regulation ofdosage compensation in Drosophila. Mol Cell 2006, 21:811-823.

61. Brown CR, Kennedy CJ, Delmar VA, Forbes DJ, Silver PA: Globalhistone acetylation induces functional genomicreorganization at mammalian nuclear pore complexes. GenesDev 2008, 22:627-639.

62. Gartenberg MR, Neumann FR, Laroche T, Blaszczyk M,Gasser SM: Sir-mediated repression can occur independently


j.gde.2008.10.002


of chromosomal and subnuclear contexts. Cell 2004,119:955-967.

63. Janicki SM, Tsukamoto T, Salghetti SE, Tansey WP,Sachidanandam R, Prasanth KV, Ried T, Shav-Tal Y, Bertrand E,Singer RH et al.: From silencing to gene expression: real-timeanalysis in single cells. Cell 2004, 116:683-698.

64. Purmann A, Toedling J, Schueler M, Carninci P, Lehrach H,Hayashizaki Y, Huber W, Sperling S: Genomic organization oftranscriptomes in mammals: coregulation andcofunctionality. Genomics 2007, 89:580-587.

65. Burhans DT, Ramachandran L, Wang J, Liang P, Patterton HG,Breitenbach M, Burhans WC: Non-random clustering of stress-related genes during evolution of the S. cerevisiae genome.BMC Evol Biol 2006, 6:58.

66.��

Kosak ST, Scalzo D, Alworth SV, Li F, Palmer S, Enver T, Lee JS,Groudine M: Coordinate gene regulation during hematopoiesisis related to genomic organization. PLoS Biol 2007, 5:e309.

The author’s analysis reveals that there is a significant tendency forcoregulated genes to be proximal and suggest that proximity in the formof chromosomal gene distribution and homolog association may be thebasis for organizing the genome for coordinate gene regulation duringcellular differentiation.

67.�

Hurst LD, Pal C, Lercher MJ: The evolutionary dynamics ofeukaryotic gene order. Nat Rev Genet 2004, 5:299-310.

An excellent review discussing the various forces that shape gene order ineukaryotic chromosomes.

68. Hermsen R, ten Wolde PR, Teichmann S: Chance and necessityin chromosomal gene distributions. Trends Genet 2008,24:216-219.

69. Singer GA, Lloyd AT, Huminiecki LB, Wolfe KH: Clusters ofco-expressed genes in mammalian genomes are conservedby natural selection. Mol Biol Evol 2005, 22:767-775.

70. Teichmann SA, Veitia RA: Genes encoding subunits of stablecomplexes are clustered on the yeast chromosomes: aninterpretation from a dosage balance perspective. Genetics2004, 167:2121-2125.

71. Poyatos JF, Hurst LD: The determinants of gene orderconservation in yeasts. Genome Biol 2007, 8:R233.

72. Zhan S, Horrocks J, Lukens LN: Islands of co-expressedneighbouring genes in Arabidopsis thaliana suggesthigher-order chromosome domains. Plant J 2006,45:347-357.

73. Ferrari F, Bortoluzzi S, Coppe A, Basso D, Bicciato S, Zini R,Gemelli C, Danieli GA, Ferrari S: Genomic expression duringhuman myelopoiesis. BMC Genomics 2007, 8:264.

74. Cohen BA, Mitra RD, Hughes JD, Church GM: A computationalanalysis of whole-genome expression data revealschromosomal domains of gene expression. Nat Genet 2000,26:183-186.

75.�

Gierman HJ, Indemans MH, Koster J, Goetze S, Seppen J,Geerts D, van Driel R, Versteeg R: Domain-wide regulation ofgene expression in the human genome. Genome Res 2007,17:1286-1295.

It is shown here that identical GFP reporter constructs integrated at 90different chromosomal positions obtain expression levels that corre-spond to the activity of the domains of integration. 3D-FISH showedthat active domains of integration have a more open chromatin structurethan integration domains with weak activity. On the basis of theseobservations, the authors propose the existence of a novel domain-wideregulatory mechanism that, together with transcription factors, exerts adual control over gene transcription.

76.��

Xiao H, Jiang N, Schaffner E, Stockinger EJ, van der Knaap E: Aretrotransposon-mediated gene duplication underliesmorphological variation of tomato fruit. Science 2008,319:1527-1530.

An excellent study which shows that a retrotransposon-mediated changein genomic context of a key regulator, SUN, resulted in an alteredexpression level relative to that of the ancestral state, thereby resultingin variation in tomato fruit shape.

77. Doebley JF, Gaut BS, Smith BD: The molecular genetics of cropdomestication. Cell 2006, 127:1309-1321.







78. Otto SP: The evolutionary consequences of polyploidy. Cell2007, 131:452-462.

79.�

Ebisuya M, Yamamoto T, Nakajima M, Nishida E: Ripples fromneighbouring transcription. Nat Cell Biol 2008, 10:1106-1113.

The authors show that intensive transcription at one locus frequently spillsover into its physical neighboring loci. Their results suggest that tran-scriptional activation has a ripple effect, which may be advantageous forcoordinated expression.

80. Batada NN, Urrutia AO, Hurst LD: Chromatin remodelling is amajor source of coexpression of linked genes in yeast. TrendsGenet 2007, 23:480-484.

81. Semon M, Duret L: Evolutionary origin and maintenance ofcoexpressed gene clusters in mammals. Mol Biol Evol 2006,23:1715-1723.

82.��

Janga S, Collado-Vides J, Madan Babu M: Transcriptionalregulation constrains the organization of genes on eukaryoticchromosomes. Proc Natl Acad Sci U S A 2008, 105:15761-15766.

The authors present evidence for the existence of a higher order orga-nization of genes across and within chromosomes that is constrained bytranscriptional regulation. They suggest that specific organization ofgenes that allowed for efficient control of transcription within the nuclearspace has been selected during evolution.

83. Parada LA, McQueen PG, Munson PJ, Misteli T: Conservation ofrelative chromosome positioning in normal and cancer cells.Curr Biol 2002, 12:1692-1697.

84.�

Brianna Caddle L, Grant JL, Szatkiewicz J, van Hase J, Shirley BJ,Bewersdorf J, Cremer C, Arneodo A, Khalil A, Mills KD:Chromosome neighborhood composition determinestranslocation outcomes after exposure to high-dose radiationin primary cells. Chromosome Res 2007, 15:1061-1073.

It is shown that chromosomal territory (CT) neighborhoods compriseheterologous chromosomes, within which inter-CT distances directlyrelate to translocation partner choice. Their findings demonstrate thatinterphase chromosome arrangement is a principal factor in genomicinstability outcomes in primary lymphocytes, providing a structural con-text for understanding the biological effects of radiation exposure, and themolecular etiology of tumor-specific translocation patterns.

85. Nikiforova MN, Stringer JR, Blough R, Medvedovic M, Fagin JA,Nikiforov YE: Proximity of chromosomal loci that participate inradiation-induced rearrangements in human cells. Science2000, 290:138-141.

86. Argueso JL, Westmoreland J, Mieczkowski PA, Gawel M,Petes TD, Resnick MA: Double-strand breaks associated withrepetitive DNA can reshape the genome. Proc Natl Acad Sci U SA 2008, 105:11845-11850.

87. Coghlan A, Eichler EE, Oliver SG, Paterson AH, Stein L:Chromosome evolution in eukaryotes: a multi-kingdomperspective. Trends Genet 2005, 21:673-682.

88. Gruenbaum Y, Margalit A, Goldman RD, Shumaker DK, Wilson KL:The nuclear lamina comes of age. Nat Rev Mol Cell Biol 2005,6:21-31.


j.gde.2008.10.002


89. Liu B, Wang J, Chan KM, Tjia WM, Deng W, Guan X, Huang JD,Li KM, Chau PY, Chen DJ et al.: Genomic instability inlaminopathy-based premature aging. Nat Med 2005,11:780-785.

90. Scaffidi P, Gordon L, Misteli T: The cell nucleus and aging:tantalizing clues and hopeful promises. PLoS Biol 2005, 3:e395.

91.��

Kleinjan DA, van Heyningen V: Long-range control of geneexpression: emerging mechanisms and disruption in disease.Am J Hum Genet 2005, 76:8-32.

An excellent and timely review providing an account of several humandiseases which arise owing to mutations that disrupt long-range controlof gene expression.

92. Babu MM, Balaji S, Iyer LM, Aravind L: Estimating the prevalenceand regulatory potential of the telomere looping effect in yeasttranscription regulation. Cell Cycle 2006, 5:2354-2363.

93. Unneberg P, Claverie JM: Tentative mapping of transcription-induced interchromosomal interaction using chimeric ESTand mRNA data. PLoS ONE 2007, 2:e254.

94. Zinner R, Teller K, Versteeg R, Cremer T, Cremer M: Biochemistrymeets nuclear architecture: multicolor immuno-FISH for co-localization analysis of chromosome segments anddifferentially expressed gene loci with various histonemethylations. Adv Enzyme Regul 2007, 47:223-241.

95. Pombo A: Advances in imaging the interphase nucleus usingthin cryosections. Histochem Cell Biol 2007, 128:97-104.

96. Kanger JS, Subramaniam V, van Driel R: Intracellularmanipulation of chromatin using magnetic nanoparticles.Chromosome Res 2008, 16:511-522.

97. Segal E, Fondufe-Mittendorf Y, Chen L, Thastrom A, Field Y,Moore IK, Wang JP, Widom J: A genomic code for nucleosomepositioning. Nature 2006, 442:772-778.

98. Dodd IB, Micheelsen MA, Sneppen K, Thon G: Theoreticalanalysis of epigenetic cell memory by nucleosomemodification. Cell 2007, 129:813-822.

99. Bon M, Marenduzzo D, Cook PR: Modeling a self-avoidingchromatin loop: relation to the packing problem, action-at-a-distance, and nuclear context. Structure 2006, 14:197-204.

100. Balaji S, Iyer LM, Babu MM, Aravind L: Comparison oftranscription regulatory interactions inferred from high-throughput methods: what do they reveal? Trends Genet 2008,24:319-323.

101. Mans BJ, Anantharaman V, Aravind L, Koonin EV: Comparativegenomics, evolution and origins of the nuclear envelope andnuclear pore complex. Cell Cycle 2004, 3:1612-1637.

102. Boeger H, Griesenbeck J, Kornberg RD: Nucleosome retentionand the stochastic nature of promoter chromatin remodelingfor transcription. Cell 2008, 133:716-726.





eukaryotic gene regulation in three dimensions and its ... · eukaryotic gene regulation and its...

Documents