Not only is this title a painful reminder of a mistake I made in my exam this year, but simply wrong since centromeres will still be included in discussion! So, just to clarify, centrosomes are the organelles from which microtubules (MTs)* arise from, whilst centromeres are regions on chromosomes that attach to MTs via the kinetochore (more on kinetochores below!).
*If you are not sure what MT’s are, think of them as a series of ropes that can grow and shrink and are used for intracellular movement
Why are the names so complicatingly* similar?
To help myself remember names, I find it helps to know where they derive from. Centro- of the Latin centrum meaning central and -mere from the Ancient Greek méros meaning ‘part’, centromeres thus comprise the central part of a chromosome. Although this is not always the case..(see below). The -some of centrosome comes from the Greek sōma meaning ‘body’ and thus translates as central bodies. This sort of makes sense if you remember that during interphase centrosomes are a central hub of a cell for organising the MTs. To make matters worse, centrosomes are actually composed of two centrioles (-ole meaning small parts, used similarly in arteriole) that are at right angles to each other along with some pericentriolar material (which contains the proteins required for MT formation).
*I am aware this is not a word, but it should be
Back to centromeres
Centromeres are incredibly important for cell division. Cell division must be performed correctly to avoid aneuploidy (inheriting an abnormal number of chromosomes in a cell) which can cause many genetic diseases including cancer. The key question then is what defines a centromere… is it genetic or epigenetic?
Arguably the best defined centromeres come from the Brewer’s yeast (S.cerevisiae) – they were also not surprisingly the first eukaryotic centromere isolated. All of these centromeres have three conserved DNA sequences termed centromere DNA element I, II, and III (CDEI, CDEII, CDEIII). These elements are enough to provide mitotic stability to plasmids. They are better known as ‘point centromeres’ since they are so small – in total ~120bp.(3) This is in stark contrast to the ‘regional centromeres’ seen in human cells that span Mb of DNA. Broadly these larger centromeres can be split into two regions – centromeric chromatin(CH) lies at the very centre where MT-attachment points form. This is flanked by pericentric heterochromatin(PCH) which mediates sister-chromatid cohesion due the increased density of cohesin (a ring complex that clamps the sister chromatids together). Both these regions are composed of arrays of tandem repeats and are less well defined genetically.
See reference (1) below for a great illustration of centromere structure and organisation in different organisms.
One protein found conserved from yeast to mammals is CENPA, a histone H3 variant. Histones are the proteins that bind to DNA providing the first level of DNA compaction. Histones typically group as octamers composed of two copies of H2A, H2B, H3 and H4. Different histone variants are just one way chromatin can be epigenetically defined. CENPA is recruited to centromeres and self-propagates to define the area. Studies in the other famous yeast, S.pombe, show that an RNAi may also be involved in CENPA recruitment.
Interestingly, despite its importance, mammalian cells can surpass a 90% reduction in CENPA levels! Complete loss, however, is lethal.
It is most likely that centromeres are defined through a combination of both genetic and epigenetic features.
If you think of chromosomes you probably think of the X-shaped kind, the metacentric chromosomes (well at least I do), but not all chromosomes take on that iconic shape. In fact, the shape of chromosomes are defined by the position of their centromeres. Mice have all acrocentric chromosomes – centromeres near one end of the chromosome – making them look like a funny pair of scissors. Sub-metacentric fall in between the two, whilst telocentric, as the name suggests, have centromeres at the very end forming a ‘V’ shape when the chromatids are paired.
If they made chromosome-shaped crisps, I’d buy them.
Kinetochore is truly an excellent name – it is just great fun to say and one I have yet got confused with any other. There is an excellent series of seminars discussing recent experiments to understand the function and composition of kinetochores on iBiology given by Dr Sue Biggins (https://www.youtube.com/watch?v=cFwBuYgsqMw). Basically though, kinetochores are a complex of proteins that associate with centromeres and bind MTs. Different types of kinetochore-MT attachments can occur (Figure 1) – proper segregation of the genome in mitosis can only occur through bioriented, amphitelic attachments. A lot of current work is focusing on the way that kinetochores form amphitelic attachments instead of syntelic and mechanisms that avoid the latter from occurring.
A model for an error-correction model was recently presented; it suggests that kinetochore-MT detachment requires many phosphorylation events of kinetochore proteins, providing a delay time to allow kinetochores to form stable, high-tension amphitelic attachments where detachment is prevented (2). The phosphorylation is performed by Aurora B kinase which de-stabilises attachments allowing for further attempts for biorientation.
By having a delay before detachment, it allows for a transition between low and high-tension kinetochore-MT attachments and thus stabilisation. Mathematical analyses of budding yeast kinetics (S.cerevisiae again) including MT dynamics and kinase/phosphatase antagonism have now shown that the requirement of multiple phosphorylation events for detachment provides such delay. Budding yeast however have one MT attached per kinetochore whereas we have up to 20, so further work will need to be performed to see if similar mechanisms work in humans.
Centromeres and centrosomes are thus equally important and dependent on each other to perform their cellular functions.
P.s Just to make thing slightly more complicated, budding yeast technically don’t have centromeres, they have spindle pole bodies!
(2) Tubman, Emily S. et al. Stochastic modelling yields a mechanistic framework for spindle attachment error correction in budding yeast mitosis Cell Systems , Volume 4 , Issue 6 , 645 – 650 (2017)
(3) JS Verdaasdonk. Centromeres unique chromatin structures that drive chromosome segregation Nature reviews molecular cell biology 12, 320-332 (2011)