It’s all about perspective. Like a funhouse mirror that makes you look fat or skinny depending on how it’s curved, it seems there’s some controversy on the subject of nucleosome positioning based on how results get interpreted. So, are nucleosomes stable or do they move around? In this contributed piece, Istvan Ladunga from the University of Nebraska, takes a shot at resolving the whole issue through statistical analysis.
Scientific minds can draw opposing conclusions from very similar experiments. One recent Nature paper claimed the “…DNA-encoded nucleosome organization of a eukaryotic genome…” while another high-profile publication stated the exact opposite: “…histone-DNA interactions are not the major determinant of nucleosome positions. ”
In order to find a resolution to this heated controversy our group dug beneath the beautiful cherry-picked published images, and did a comprehensive analysis of the statistics from twenty six published experiments. Sure, the images that we ended up with were much less attractive than some idealized nucleosome diagram, but we should never compromise science for beautiful images, should we?
We discovered that only a minority of nucleosomes, or “barriers,” serve as Archimedean fixed points in the chromatin universe. These barriers guard regulatory regions and isolate the chromatin of individual genes. Instead, we saw that most other nucleosomes are not anchored “immovably” to the DNA. After all, if nucleosome positions were primarily “encoded in the DNA” as some suggest, then what roles would the massive array of histone chaperones and remodeling enzymes play? One of the many such roles is to evict nucleosomes, giving way to the transcriptional apparatus. Our new meta-analysis of 26 experiments showed that the sequence of DNA causes weak and fuzzy positioning in genes. This weakness helps to evict nucleosomes, and these effects combined with remodeling enzymes and transcription, causes nucleosomes to frequently be rebuilt at varying loci.
The messy-looking images and the statistics show that nucleosome remodeling is a highly probabilistic event. So how could old-fashioned deterministic logic (and prose) describe them? What kind of “maps” can illustrate “mountains” (nucleosomes) that are sliding or being recreated far away from their previous foothills? Maybe we ought to follow the example of physicists, and think of certain nucleosomes as more like probabilistic clouds similar to electrons.
Big thanks to Dr.Ladunga for this contribution. You can read up on the Ladunga lab at their website. Check out the full nucleosome forecast at PLos One, September 2010