SUNY at Albany
June 19-23, 2001
We have recently analyzed the dynamics of the nucleosome (1) and tetrasome (2), that is, the thermal fluctuations of these particles between different conformational states. [The~ 0.75 turn tetrasome is a nucleosome sub-particle made of the DNA wrapped around the histone (H3-H4)2 tetramer.] For this, we used an original method involving the reconstitution of the particles on an homologous series of 350-370 bp DNA minicircles, and their relaxation with topoisomerase I (3, 4). This generates equilibrium distributions which contain up to three adjacent topoisomers. In a series of minicircles with sizes differing by 1-2 bp, the particular minicircle always exists with can provide, after relaxation, the appropriate topological constraint, so that the particle on that minicircle, and in any given conformational state, can have a relaxed loop. In contrast, particles on minicircles of even slightly different sizes, or on adjacent topoisomers of the same size, have their loop more or less constrained. The free energy advantage conferred to those "relaxed" particles over the "constrained" particles will favor and enrich the corresponding topoisomer in the equilibrium distribution..
This is illustrated in the above figure, which gives the relative amounts of all topoisomers in the nucleosome and tetrasome equilibria, respectively, as functions of the topoisomer linking number difference, DLk = Lk - Lko (Lko = N/ho, with N being the size, in bp, and ho the DNA helical periodicity under relaxation conditions, in bp/turn). Three bumps are clearly visible in the nucleosome profile, around DLk = -1.7, -1 and -0.5. The first value is that expected for a negative crossing of the entering and exiting DNAs, and from the DNA left-handed wrapping around the histone core; the second value indicates an open configuration with no crossing and broken histone-DNA contacts at the entry-exit, which we had previously shown to exist on DNA minicircles (5); the third value corresponds to a new conformation in which the entering and exiting DNAs make a half-positive crossing (a complete positive crossing would instead result into DLk = 0). This "positive" conformation was independently confirmed by ethidium bromide titration experiments (6). Fluctuations of individual nucleosomes in the chromatin fiber between these conformations explains a number of puzzling observations related to DNA topology in chromatin and its dependence on histone tail acetylation (1). For the tetrasome, two well-separated peaks are observed around DLk = -0.7 and +0.7. These values indicate two conformations : one with the entering and exiting DNAs crossed negatively, the other positively. Again, only the negatively-crossed conformation could have been expected from the DNA known left-handed wrapping around the histone octamer
Despite their resemblance, a basic difference exists between the two transition mechanisms. Whereas the nucleosome transition only involved the loop, with the octamer structure not being significantly altered from one conformation to another, it also involved the protein in the case of the tetrasome. The introduction of a sterical hindrance by covalent linking of a bulky adduct deep into the tetramer (through the oxidation of the centrally-located 110 cysteine in the two H3s) indeed strongly influenced the tetrasome equilibrium conformation. This led to the conclusion that not only the crossing, but also the wrapping polarity was reversed between the two tetrasome conformations, and that this transition was actually driven by a switch of the tetramer itself from a left- to a right-handed proteinaceous superhelix (7, 8). Such a tetrasome chiral transition may play a central role in nucleosome dynamics in vivo, in particular during transcription elongation, but this remains to be firmly established.
References and Footnotes
CNRS et Universités Paris 6 et Paris 7