The DNA binding and bending activities of truncated tail-less HMGB1 protein are differentially affected by Lys-2 and Lys-81 residues and their acetylation.

The role of lysines 2 and 81 as target sites for acetylation in full-length HMGB1 and truncated tail-less protein, respectively, has been studied by mutation analysis for the abilities of these proteins to bind and bend DNA. The DNA bending ability of truncated tail-less HMGB1 containing Lys-2 mutated to alanine does not differ from that of the wild-type protein, while the same mutation of Lys-81 reduced the bending capacity of the mutant protein. These data demonstrate that Lys-81 is critical for the DNA bending ability of truncated HMGB1. Such a conclusion is further confirmed by the experiments carried out with CBP-acetylated proteins: acetylation of Lys-2 in mutant protein K81/A81 alleviated DNA bending and induced DNA end-joining. On the contrary, the acetylation of Lys-81 in the mutant K2/A2 enhanced the bending potential of HMGB1∆C. Regarding the ability of HMGB1 to specifically bind bent DNA, the individual mutations of either K2 or K81 as well as the double mutation of both residues to alanine were found to completely abolish binding of truncated tail-less HMGB1 to cisplatin-modified DNA. We conclude that unlike the case with the bending ability of truncated HMGB1, where Lys-81 has a primary function, Lys-2 and Lys-81 are both critical for the protein's binding to cisplatin-modified DNA. The mutation K2/A2 in full-length HMGB1 and acidic tail removal induce the same conformational changes. Any further substitutions at the acetylable lysines in the truncated form of HMGB1 do not have an additional effect.

Interplay between in vitro acetylation and phosphorylation of tailless HMGB1 protein.

The postsynthetic acetylation of HMGB1 protein and its truncated form affects significantly its properties as "architectural" factor - recognition of bent DNA and bending of short DNA fragments. We created mutants at the target sites (lysines 2 and 81) in the tailless HMGB1 modified by the histone acetyltransferase CBP. The results show that there is no preferential site for the enzymatic activity of CBP and both lysine moieties are modified independently. Our findings for the first time demonstrate the link between the acetylation and phosphorylation of HMGB1DeltaC in vitro. The PKC phosphorylation prior to acetylation inhibits the CBP activity 40-60% for the truncated form and its mutants. The effect of the CBP acetylation on the phosphorylation level turns out to be much more prominent. In the case of HMGB1DeltaC modified at Lys 2 and Lys 81 prior to PKC treatment background phosphorylation is detected. If only one of the lysines is modified the inhibitory effect decreases.