There are some bioinformatical considerations showing that there is a significant correlation of some between histone code and alternative splicing; but what are the hypothesis about the underlying mechanisms? Is there any experimental insight in the problem?
According to literature, histone marks seem to be affecting splicing both directly and indirectly.
The indirect regulation appears to be due to biophysical hindrance, where a certain chromatin structure would cause Pol-II slowdown, allowing the splicing machinery to do its job.
The direct regulation seems to be essentially dependent on H3K36me3 and H3K4me1, marks that are recognized by MRG15, which in turn recruits the splicing regulator polypyrimidine tract-binding protein to splice sites.