whereas Ecdysozoa such as flies and worms would have
evolved a different type of connection inwhich CEN3 is dispen-
sable. Of note, whereas Cen3p is more concentrated in the
centrosome of vertebrate cells [58], Cen2p is most abundant in
the nucleus and the cytoplasm [6] and participates also in the
nucleotide excision repair reaction [59]. Intriguingly, nucleotide
excision repair after UV-irradiation is the only known function
that is impaired upon knocking out all centrin genes in DT40
chicken lymphoma B cells [60]. However, this function does
not require the calcium-binding capacity of Cen2p, whereas its
recruitment to the centrosome and its binding to the centroso-
mal centrin-binding POC5 does [6,61]. The centrosomal
calcium-dependent functions may be complemented in this
cell line by other members of the Calmodulin superfamily.
10. On the centrosome as signalling centre
An appealing and relatively recent role ascribed to the centro-
some is one in the integration and coordination of signalling
pathways (see article by Erich Nigg and collaborators) [62],
which had been suggested by the analogous and more estab-
lished role of the yeast SPB. Many aspects, including long
suspected links between centrosomes and DNA damage
response pathways, have begun to be unravelled and will
likely be deciphered further in the years to come. Could there
even be something in common between the centrosome as a sig-
nalling centre and the cilium being used in major signalling
transduction cascades (see article by Maxence Nachury) [63]?
The molecular and functional similarities between cilia and
immune synapses in which centrosome repositioning at the
plasma membrane is critical (see article by Jane Stinchcombe
and Gillian Griffith) [64] further lends support to the notion
that the centrosome functions as a signalling hub in many
biological contexts.
The centrosome may also act as a signalling centre in a
setting where it is usually perceived as a mere source of micro-
tubules. The ability of centrosomes to propel the associatedmale
pronucleus towards the centre of large marine or amphibian
eggs in a matter of minutes fascinated the early students of fer-
tilization and development. How can centrosomes and the
microtubules they nucleate sense egg size and shape to reach
the cell centre in due time to be coordinated with mitotic
entry? Microtubules possess an autonomous ability to reach
the geometric centre of a given volume [65], but
in vivo
biochemi-
cal waves emanating from centrosomes are important also to set
the timing of cell division (see article by TimMitchison and col-
laborators) [66]. The idea that centrosomes can set gradients of
enzymatic activities is not new, but the field is at a stage
where these ideas can bemodelled and testedwith the appropri-
ate experimental approaches. The fact that the Golgi apparatus
can also act as an MTOC in vertebrate cells (see article by Rosa
Rios) [67] adds another layer of complexity to this topic in
offering yet another potential source of signalling.
11. On the mother –daughter asymmetry
It is now well established that the conservative duplication of
centrioles and of the yeast SPBs, resulting in an old and a new
unit, contributes to asymmetric cell division and stemness (see
article by Jose Reina and Tano Gonzalez) [68]. But why should
there be two centrioles per centrosome instead of one, as is the
case in the yeast SPB? In animal cells, the capacity of the
daughter centriole to nucleate microtubules and to guide pro-
centriole assembly occurs well before microtubules are
anchored on sub-distal appendages of the mother centriole to
form an aster or permit docking at the plasma membrane via
distal appendages to take place to grow a primary cilium. One
possible benefit of such a time delay could be to introduce con-
siderable flexibility into the design of the centrosome organelle.
In this way, both free and anchored microtubules can be pro-
duced independently, considering that the inter-centriolar
distance can reach 20
m
m in some cells [69]. Regulating this dis-
tance could be part of differentiation programmes that set where
microtubules are operating in a given cell and thus contribute to
facilitate tissue organogenesis or response to extracellular cues.
Having such a time delay between the biogenesis of the two
units imposes a slow differentiation process, with the distinct
control of centriole length as well as the timely control of disen-
gagement of the two centrioles at mitotic exit, after the two
diplosomes have separated at the G2/M transition (see the
article by Elmar Schiebel and collaborators) [70].
12. On centrosome and disease
The links between centrosomes and disease are as old as the field
itself, with Boveri’s first observations with polyspermic eggs that
lead to multipolar divisions and aneuploidy, and it has taken
over a century to clarify some of the tenets of this connection.
Whereas it now appears clear that centrosome dysfunctions
can favour tumour onset (see article by Susana Godinho and
David Pellman) [71], it will be important to figure out in each
type of tumour whether this is by promoting aneuploidy, as
Boveri postulated, by promoting tissue destabilization and inva-
sion, through cell polarity defects or perhaps a combination of
these effects. Other diseases associatedwith centrosome dysfunc-
tions have a more recent history but nonetheless an important
impact on human health. Among these diseases, it will be impor-
tant to address, for example, why the brain can be exclusively
affected by some mutations in centrosomal components that
lead to microcephaly, whereas other tissues are seemingly
spared (see article by Fanni Gergely and collaborators) [72].
13. On removing centrioles
Although usually very stable, centrioles probably have a finite
lifetime in most cells and disappear in a stereotyped manner in
specific cell types. This is the case during oogenesis in most
metazoan organisms, and such disappearance is critical to
ensure that the newly fertilized embryo is endowed with a
single pair of centrioles, which is delivered by the sperm [73].
Centriole loss can also take place in somatic cells, as for
example during mammalian skeletal myogenesis, when myo-
blasts fuse into myotubes [74]. Whether there is a common
mechanism in both cases is not known, nor is it known whether
such disappearance recapitulates in reverse the sequence of
events occurring during centriole assembly. Could there be a
common theme between these two cell types that explains
why they both lose their centrosome?
14. Concluding remarks
One may wonder why the centrosome has ever evolved in
metazoans if other multicellular organisms such as higher
rstb.royalsocietypublishing.org
Phil. Trans. R. Soc. B
369
: 20130452
6
on October 7, 2014
rstb.royalsocietypublishing.org
Downloaded from
Book Recommendation
