We assumed that all of this DNA was essential and none of it is junk DNA. However, we know that's not a correct assumption since there are many variants at each centromere. If we were to take the minimum size for each functional centromere, the total amount of essential DNA would be much less (probably <1% of the genome). Many workers are trying to figure out how much DNA is required in order to have a functional centromere. One approach is to identify abnormal chromosomes that segregate normally at mitosis with only a small number of repeats at their centromeres. THEME
Genomes & Junk DNA
Total Junk so far
In the latest issue of PNAS, Murata et al. (2008) looked at small minichromosomes in the plant Arabidposis thaliana. The minichromosomes were produced as a by-product of a transformation experiment that inserted T-DNA into the centromere of chromosome 2.
The five standard chromosomes of Arabidopsis each have centromeres consisting of about 1600 copies of a 180 bp repeat (avergae size 2.7 Mb - 3.0 Mb. The four minichromosomes, α, β, γ, and δ, have centromeres ranging in size from 0.5 Mb to 2.3 Mb. The δ minichromosome appears to segregate normally with only 500 Kb of centromere DNA (about 2800 repeats). This may be close to the minimum size required to assemble a kinetochore.
If this minimum size is true in mammals well—a reasonable assumption—then perhaps only 15-20% of centromere DNA is actually essential and the rest is excess junk DNA produced by unequal cross-overs and DNA replication slippage. Because expansion and contraction of repetitive DNA is unavoidable, there will be considerable variation within a population. Individuals that have close to the minimum amount of DNA at any one centromere will be underrepresented in the population because many of their offspring will have died. Individuals with a large excess of centromere DNA will be overrepresented because their lineages are less likely to encounter lethal deletions. (Provided that there is no fitness penalty for carrying excess DNA.)
Thus, in a certain sense, some of the "excess" centromeric DNA is required as a buffer against the possibility of future deletions. The extra DNA does not contribute to the viability of the individual carrying it but it does contribute to the survival of that individual's offspring. At some point, the potential advantage in terms of offspring survival will become too small to have any influence on the lineage of an individual. This will define the maximum amount of "excess" DNA at the centromere. I wonder if it is possible to model the effect of having extra centromeric DNA?
Murata, M., Yokota, E., Shibata, F. and Kashihara, K. (2008) Functional analysis of the Arabidopsis centromere by T-DNA insertion-induced centromere breakage. Proc. Natl. Acad. Sci. (USA) 105:7511-7516. [PubMed] [doi:10.1073/pnas.0802828105]