Organization of chromatin in the interphase mammalian cell.

Abstract

The use of imaging techniques has become an essential tool in cell biology. In particular, advances in fluorescence microscopy and conventional transmission electron microscopy have had a major impact on our understanding of chromatin structure and function. In this review we attempt to chart the conceptual evolution of models describing the organization and function of chromatin in higher eukaryotic cells, in parallel with the advances in light and electron microscopy over the past 50 years. In the last decade alone, the application of energy filtered transmission electron microscopy (EFTEM), also referred to as electron spectroscopic imaging (ESI), has provided many new insights into the organization of chromatin in the interphase nucleus. Based on ESI imaging of chromatin in situ, we propose a 'lattice' model for the organization of chromatin in interphase cells. In this model, the chromatin fibers of 10 and 30nm diameter observed by ESI, produce a meshwork that accommodates an extensive and distributed interchromosomal (IC) space devoid of chromatin. The functional implications of this model for nuclear activity are discussed.

Full text

Page 1

Review
Organization of chromatin in the interphase mammalian cell
Hesam Dehghania, Graham Dellairea, David P. Bazett-Jonesa,b,*
aProgramme in Cell Biology, The Research Institute, The Hospital for Sick Children, 555 University Avenue, Toronto, Ont., Canada M5G 1X8
bDepartment of Biochemistry, University of Toronto, Medical Sciences Building, 1 King’s College Circle, Toronto, Ont., Canada M5S 1A8
Received 20 August 2004; revised 11 October 2004; accepted 12 October 2004
Abstract
The use of imaging techniques has become an essential tool in cell biology. In particular, advances in fluorescence microscopy and
conventional transmission electron microscopy have had a major impact on our understanding of chromatin structure and function. In this
review we attempt to chart the conceptual evolution of models describing the organization and function of chromatin in higher eukaryotic
cells, in parallel with the advances in light and electron microscopy over the past 50 years. In the last decade alone, the application of energy
filtered transmission electron microscopy (EFTEM), also referred to as electron spectroscopic imaging (ESI), has provided many new
insights into the organization of chromatin in the interphase nucleus. Based on ESI imaging of chromatin in situ, we propose a ‘lattice’ model
for the organization of chromatin in interphase cells. In this model, the chromatin fibers of 10 and 30 nm diameter observed by ESI, produce a
meshwork that accommodates an extensive and distributed interchromosomal (IC) space devoid of chromatin. The functional implications of
this model for nuclear activity are discussed.
q 2004 Elsevier Ltd. All rights reserved.
Keywords: Nucleus; Chromatin; Chromosome; Interphase; Models; Electron spectroscopic imaging; Imaging techniques
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
2. Pioneering ultrastructural studies of chromatin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96
3. Electron spectroscopic imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
4. Functional organization of DNA in the nucleus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1. Chromosome territories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2. Interchromosomal (IC) space and perichromatin structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4.3. Mind the [resolution] gap: the hazards of inferring chromatin ultrastructure from fluorescence microscopy . . 101
4.4. The ‘lattice’ model of interphase chromatin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
4.5. Regulation of gene transcription in the chromatin lattice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
99
99
5. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
0968-4328/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.micron.2004.10.003
Micron 36 (2005) 95–108
www.elsevier.com/locate/micron
* Corresponding author. Address: Department of Biochemistry, University of Toronto, Medical Sciences Building, 1 King’s College Circle, Toronto, Ont.,
Canada M5S 1A8. Fax: C1 416 813 2235.
E-mail address: dbjones@sickkids.ca (D.P. Bazett-Jones).

Page 2

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
1. Introduction
Close to a half century has passed since the first
experiments were performed that addressed the question
of how chromatin is organized in the nucleus. Much of the
work was considered to be of peripheral interest because of
the descriptive nature of microscopy at the time. In the last
few years, however, the importance of chromatin organiz-
ation and its remodeling during the regulation of nuclear
events such as the transcription, replication and repair of
DNA, has finally been recognized.
As in all other aspects of cell biology, the study of
chromatin has largely benefited from the use of imaging
techniques. Consequently, our understanding of chromatin
and nuclear organization has also benefited from advances
in microscopy. In this review we will trace the development
and conceptual evolution of models describing chromatin
organization that is applicable to a large variety of
eukaryotes. These models include those describing the
filamentous structure of chromatin, its territorial subnuclear
organization, and the perichromatin structures found in the
interchromosomal (IC) space with which chromatin inter-
acts. We will emphasize the impact that electron
microscopy (EM) has had in this field as well as the
advantages offered by electron spectroscopic imaging (ESI)
over traditional transmission electron microscopy (TEM).
We also propose a ‘lattice’ model to describe the
organization of chromatin in situ within interphase cells
based on observations using ESI. We conclude with a
discussion of the functional implications of the lattice model
and the potential of ESI for future work in chromatin and
nuclear biology.
2. Pioneering ultrastructural studies of chromatin
Before the 1950s, methods for the study of chromatin
were limited to light microscopy, enzymatic digestion, and
solvent extraction (Olins and Olins, 2003). Since that time,
electron microscopy has revolutionized our understanding
of much of cell biology, including how chromatin is
organized in the cell nucleus. The filamentous nature of
chromatin was first observed in the 1950s and 1960s. This
paradigm was followed in the 1970s with the discovery of
the nucleosome, the repeating chromatin subunit, based on
purified components analyzed by both biochemical and
electron microscopy techniques. Finally, through the 1980s
to the present, fluorescence microscopy and ultrastructural
studies of higher order organization, particularly the
observation of chromosome territories, have contributed to
a greater appreciation of the importance of the structural
compartmentalization of the nucleus.
In the 1950s and 1960s, several pioneering electron
microscopy studies were carried out, including work on
lampbrush chromosomes of amphibian oocytes (Mirsky and
Ris, 1951), grasshopper spermatocytes (De Robertis, 1956),
frog erythrocytes (Davies and Spencer, 1962), and newt
erythrocytes (Gall, 1963). Although these studies were
revolutionary, the work suffered from limits in spatial
resolution, the inability to obtain thin sections, and
inadequate fixation techniques. Nonetheless, De Robertis
(1956), for example, showed that chromosomes and
chromatin in grasshopper spermatocytes are composed of
a filamentous macromolecular component. Because ultra-
microtomy had not yet advanced sufficiently, a ‘squash’ or
‘spread’ technique was used in most electron microscopy
experiments. For example, blood cells were spread into a
very thin layer on the surface of water and the remains of the
hypotonically ruptured cells were picked upon carbon-
coated grids and air dried (Gall, 1963). With these methods,
chromatin of interphase cells appeared as cylindrical fibers
with an average diameter of 40–60 nm (Fig. 1A). Whereas
digestion with ribonuclease did not have an effect,
deoxyribonuclease digestion completely destroyed these
fibers (Wolfe, 1965). Another study by Wolfe showed that
10 nm fibers were present in un-fixed nuclei. These fibers
could be stabilized by fixation so that their diameter was not
affected by spreading on water (Wolfe and Martin, 1968). In
all, the results of close to two decades of research on the
ultrastructure of chromatin could perhaps be summarized in
this sentence from Ris and Kubai (1970); “the 100 A˚fiber
demonstrated by both [these] approaches. must represent
the basic unit of inactive chromatin”.
The next decade of electron microscopy studies were
centered on the structure of chromatin fibers and the
repeating subunits along their length. Electron microscopy
of thin sections offixed nuclei revealed threads and granules
(Davies and Small, 1968). In rat liver, nuclear regions with
condensed and dispersed chromatin threads were observed,
where the average width ranged from 10 to 19 nm (Olins
Woodcock et al., 1976b). (D) EDTA-regressive staining of rat liver cells (reproduced with permission of Academic Press from Monneron and Bernhard, 1969).
Ribosomes (r), nucleolus (nu), interchromatin granules (ig), perichromatin granules (/), perichromatin fibrils (–j /). (E) and (F) Chromosome painting in
human lymphocyte nuclei shown for chromosome 1 (reproduced with permission from Lichter et al., 1988). (G) and (H) Chinese hamster ovary cells in G1, 1 h
(G) and 2 h (H) after mitosis. Large arrows point to condensed chromatin and small arrows point to chromonema fibers (reproduced from Belmont and Bruce,
1994 with permission from The Rockefeller University Press).
"
H. Dehghani et al. / Micron 36 (2005) 95–10896

Page 3

Fig. 1. A gallery of chromatin images from key experiments between 1963 and 1994. (A) Fibers from red-cell nucleus of the newt spread on water surface
(reproduced with permission from Gall, 1963). (B) Linear arrays of spherical chromatin particles (n bodies) about 70 A˚in diameter prepared from isolated rat
thymus and chicken erythrocyte nuclei (reproduced with permission from Olins and Olins, 1974). (C) Chromatin fragments prepared from brief digestion with
DNase II or micrococcal nuclease and separated on sucrose gradients and negatively stained with 1% uranyl formate. From left to right (3,4, and 5), single,
double, and triple arrows point to monomer, dimer, and trimer peaks, respectively, from gradients containing 0.5 M NaCl (reproduced with permission from
H. Dehghani et al. / Micron 36 (2005) 95–10897

Page 4

and Olins, 1972). Woodcock (1973) described fibers that
ranged from 10 to 30 nm in diameter, and variation in fiber
diameter was the result of irregular coiling or folding of
single chromatin fibers rather than through being multi-
stranded. In 1974, Olins and Olins, studying purified
chromatin, reported linear arrays of spherical chromatin
particles (n bodies) about 7 nm in diameter separated by
connecting 1.5 nm strands (Fig. 1B). These particles were
later named ‘nucleosomes’ (Oudet et al., 1975). Woodcock
and colleagues (1976a), also examining purified material,
presented evidence suggesting that these particles are
repeats of deoxyribonucleoprotein and are not preparation
artifacts. This work was followed by the elegant demon-
stration that the linker DNA is DNase sensitive releasing the
nucleosome subunits when digested (Fig. 1C). The first
hints of hierarchy in the organization of chromatin would
later come from work describing the divalent cation- and
salt-dependent compaction of these bead-like nucleoprotein
structures (Griffith and Christiansen, 1978), which we now
refer to as 10 nm nucleosome fibers.
The discovery of chromosome territories (reviewed in
Section 4.1) would precede yet another decade of dramatic
change in ultrastructural models of chromatin within cells.
Fluorescence microscopy of chromosomes, probed by
sequence-specific probes showed that DNA of a single
chromosome is restricted to a discrete location within the
interphase nucleus (Fig. 1E and F). At the resolution limit of
the fluorescence microscope, an impression was created that
chromosomes exist in a very condensed form and ina tightly
enclosedvolume.However,
microscopy revealed different levels of chromatin packing
and substantial chromatin-free space throughout the
nucleus. Belmont and Bruce in 1994 using non-ionic
detergents in polyamine or divalent cation buffers (contain-
ing 80 mM KCl and 20 mM NaCl) and uranyl/lead staining,
followed de-condensation of chromosomes during G1 of the
cell cycle. In early G1 and late G1/early S phases, chromatin
fibres, which they termed chromonema fibres, were
observed; one class of which was 100–130 nm in diameter,
and another of 60–80 nm in diameter (Fig. 1G and H,
respectively). In this chromonema model, an approximately
100 nmdiameter chromosomefiberfoldsintoa200–300 nm
diameter prophase chromatid, which in turn coils into the
metaphase chromosome structure (Belmont et al., 1999).
Conventional TEM of these fibres, which was used to
characterize the compaction of DNA from interphase to
metaphase chromosomes, is facilitated by molecular stains
that provide sufficient contrast between relatively thick
fibres and the background of the nucleoplasm. Under the
conditions used for chemical fixation, much of the soluble
nucleoplasmic protein is extracted, thus providing sufficient
contrast of the remaining chromatin fibers. In interphase,
however, the majority of chromatin exists in a decondensed
(euchromatic) state. As a result, euchromatin has little
contrast in electron micrographs of samples prepared using
the typical fixation and staining conditions with uranyl
transmissionelectron
acetate alone, or in combination with lead citrate. The use of
energy filtered transmission electron microscopy (EFTEM),
also referred to as electron spectroscopic imaging (ESI), can
overcome this problem when imaging interphase chromatin
by deriving contrast in electron micrographs from the
elemental content (nitrogen and phosphorous) of the
chromatin itself (Bazett-Jones and Hendzel, 1999; Dellaire
et al., 2004; Nisman et al., 2004).
3. Electron spectroscopic imaging
The principle behind electron spectroscopic imaging
(ESI) is electron energy loss spectroscopy, where incident
electrons lose specific amounts of energy based on
excitations and ionizations of the specimen’s atoms
(Fig. 2). To form energy-filtered images, an instrument
that acts both as an electron spectrometer and as a lens is
required. ESI images are formed with electrons that have
lost specific amounts of energy as a result of inner electron
shell ionization events. Since only electrons that have
interacted with the specimen are used to create an image, the
result is a dark field image. The contrast, therefore, is very
high, even without the use of heavy atom contrast agents.
Particularly when post-phosphorus edge images are
recorded, the ‘endogenous stain’ effect of the phosphorus
signal from the DNA backbone is responsible for the ability
to readily visualize 10 nm chromatin (Dellaire et al., 2004).
In spite of the ability of ESI to define the high-resolution
biochemical composition of the nucleus, it has not been
adopted as a widespread technique due to the increased cost
of the imaging filter and the lack of expertise in obtaining
and analyzing the energy loss data. Imaging of chromatin in
situ with ESI has led us to propose an alternative to the
chromonema model. This model and the impact of ESI on
our understanding of nuclear organization in general, will be
addressed later in this review.
ESI is particularly well suited for structural studies of the
cell nucleus. Phosphorus and nitrogen levels obtained from
element-specific maps can be used to delineate protein—
from nucleic acid-based structures (Fig. 2). Besides the
information from these two elements, the continuing
development of specific metal tags to detect endogenous
or exogenously expressed proteins and antibodies will
provide schemes for detecting multiple proteins simul-
taneously (i.e. multiplex detection) (Malecki et al., 2002;
Nisman et al., 2004). Recently, we have applied quantum
dots as probes for ESI. Quantum dots are luminescent nano-
crystals that can be composed of a variety of chemical
elements. Their fluorescent properties can be exploited in
fluorescence microscopy, and they are ideally suited for
imaging with ESI, since the unique elemental composition
of quantum dots provides the opportunity for multiple
detection of proteins by element-specific imaging. Even
with fluorescence labels alone, that is without EM-specific
tags, the identity of certain subnuclear structures can often
H. Dehghani et al. / Micron 36 (2005) 95–108 98

Page 5

be determined by correlative fluorescence microscopy and
ESI. Correlative microscopy provides the advantage of
being able to obtain, with a high success rate, the
ultrastructural detail in complexes that are temporally or
spatially rare (Ren et al., 2003).
4. Functional organization of DNA in the nucleus
4.1. Chromosome territories
Biochemical models of nuclear function are primarily
based on diffusion mechanisms. Consequently, the nucleus
has been treated as little more than a membrane-bound bag
containing a randomly organized genome in a nucleoplasm
consisting of freely mobile regulatory factors. Such trans-
acting factors seek out cis-regulatory elements in the DNA,
whether for initiating transcription, replication, repair or
recombination, through diffusion mechanisms. The simpli-
city of such models is attractive, but is now recognized to be
naive. One of the first clues that the nucleus has a high
degree of internal organization was the observation of
chromosome territories. Early observations by Rabl in the
1880s (Rabl, 1885) hinted at a defined spatial organization
of the genome, and the subject was re-visited in the 1960s
and 1970s (Comings, 1968; Wischnitzer, 1973; Stack,
1977). Interphase chromosomes appeared to be attached to
the nuclear membrane at sites of initiation of DNA synthesis
and occupied distinct subnuclear territories. Cremer and
colleagues (1982) micro-irradiated certain regions of the
nucleus of fibroblastoid Chinese hamster cells by laser UV
resulting in the incorporate radioactive 3H at the sites of de
novo DNA synthesis. These labeled regions in G1 phase
could easily be followed by autoradiography after 20–60 h
in later phases of the cell cycle. Although this technique was
useful to study the dynamics of chromatin subdomains in
the interphase nucleus (Cremer et al., 1993), it could not be
used to define whole chromosomes in situ or to confirm the
existence of chromatin territories. An elegant proof-concept
for the territorial model would later come from key
experiments by Lichter and colleagues (1988) (Fig. 1E
and F). In this study, the application of fluorescence in situ
hybridization (FISH) using chromosome-specific DNA
probes coupled with fluorescence microscopy (also called
‘chromosome painting’) revealed that individual chromo-
somes in human fibroblast cells occupied focal territories in
the interphase nucleus (Fig. 3), confirming Rabl’s model.
Later using the same technique, chromosome territories
(also called ‘chromatin
for all human chromosomes (Boyle et al., 2001).
domains’) weredetected
Fig.2. Electron-specimeninteractionsin the electronmicroscopeand theconceptofESI.To showthe abilityof ESIin detectionofproteinsand nucleicacidsin
biological specimens, the spectroscopic filtering of electrons from ionization of nitrogen and phosphorus are shown. Briefly, electrons collide with atoms in the
specimen. This causes production of X-rays which could be used to provide analytical information. Some of the incident electrons scatter and are blocked by
the objective aperture, providing bright-filed contrast. Some transmitted electrons that have lost energy enter the spectrometer, where they are separated
according to their energy and used to create energy-filtered image.
H. Dehghani et al. / Micron 36 (2005) 95–10899