Biology Reference
In-Depth Information
First, at a local level the physical properties such as
mass density and persistence length of the chromatin
fiber will be influenced by patterns of histone modifica-
tion and by chromatin composition. Networks of long-
range interactions between genes and regulatory elements
will lead to complex looped configurations that directly
affect gene expression. These conformations are between
specific pairs of loci, but it is likely that not all interac-
tions occur at the same time in a single cell. In some
instances, such stochastic interactions between a gene
and a distant strong enhancer can lead to 'jackpot'
expression of the gene in a few cells
[74]
. Thus stochastic
chromatin interactions may be related to the phenomenon
of stochastic gene expression observed at the single cell
level. At a higher level, groups of transcriptionally active
loci located along the same chromosome will associate
with each other, perhaps through association with
subnuclear structures or bodies such as splicing speckles
or transcription factories. Similarly, inactive domains
tend to cluster and associate at the nuclear lamina. These
interactions lead to compartmentalization of chromo-
somes, with inactive regions near the periphery of the
nucleus and active chromatin located more centrally.
Interactions between chromosomes are similarly related
to co-association of active and inactive regions, and by
anchoring of heterochromatic regions at the periphery or
near nucleoli. Thus, the relative positioning of any locus
in the genome will be dictated by its local looping
interactions, by more global positioning of the larger
chromosome domain it resides in near either active or
inactive compartments, and finally by the anchoring of
chromosome to subnuclear structures.
One critical aspect of the model is that the level of
variability in interactions increases and the specificity
decreases with the scale of chromosome organization:
local interactions are highly specific and involve genes
with their respective distal regulatory elements. At the
next level, where groups of active or inactive loci cluster
together, the level of specificity is lower, and which pairs
of active loci are found together is more variable.
Similarly, which inactive regions interact with each other
and with the nuclear periphery can vary significantly
between cells, although in all cells inactive regions are
found near other inactive regions. Finally, the relative
positions of chromosomes is predicted to be dependent
on interactions between active and inactive regions
located on different chromosomes, but
3C-based data indicating that any pair of loci can be in
close spatial proximity in at least some cells. Another
implication of the model is that each relatively large
chromosomal domain is linked through interactions with
many other loci and with subnuclear structure, making it
relatively immobile. Further, the model predicts that
movement of a genomic locus can only occur in coor-
dination with the loci or structures it interacts with, or
would require significant force to disrupt interactions.
Many observations indeed suggest limited movement of
chromosomes, even in response to stimuli that alter the
expression of significant cohorts of genes
[75,76]
.Inter-
estingly, more locally, within
m, chromatin displays
constrained Brownian motion
[77,78]
.Atthislength
scale, which corresponds to up to several hundred kb,
specific long-range interactions between genes and
regulatory elements occur. Possibly the local movement
facilitates the dynamic formation of these specific loop-
ing contacts between nearby genes and elements.
1
m
<
FUTURE CHALLENGES
The detailed view of genome organization provided by new
technologies raises many new questions. How can
a stochastic assembly of the nucleus be consistent with
robust maintenance of gene expression programs and cell
fate? Are the local and specific looping contacts between
genes and nearby elements, e.g., within several hundred kb,
sufficiently reproducible from cell to cell to yield correct
gene expression? Does the local Brownian motion for these
nearby loci ensure that these loci will encounter one
another at an appropriate timescale in every cell? How can
cells rearrange their chromosome structure, e.g., during
differentiation or in response to signals? Does this require
rewiring of all interactions, e.g., as could occur during cell
division? Addressing these questions will require identifi-
cation of the molecular mechanisms that mediate chro-
mosomal anchoring and looping, and experimental
manipulation of networks of chromosomal interactions
combined with analysis of real-time dynamics of chromatin
interactions and gene expression.
REFERENCES
[1] DuPraw EJ. Macromolecular organization of nuclei and chromo-
somes: a folded fibre model based on whole-mount electron
microscopy. Nature 1965;206(982):338
e
43.
[2] Langer-Safer PR, Levine M, Ward DC. Immunological method for
mapping genes on Drosophila polytene chromosomes. Proc Natl
Acad Sci USA 1982;79(14):4381
e
5.
[3] Misteli T. Beyond the sequence: cellular organization of genome
function. Cell 2007;128(4):787
e
800.
[4] Zhao R, Bodnar MS, Spector DL. Nuclear neighborhoods and gene
expression. Curr Opin Genet Dev 2009;19(2):172
e
9.
the precise
pattern of inter-chromosomal
interactions will greatly
differ from cell to cell.
In this model implementation of the same set of
folding principles can lead to dramatic cell-to-cell
differences in the spatial organization of the genome
inside the nucleus. Considerable variability in organiza-
tion is consistent with microscopic observations and
