No, you can just directly lyse cells and then do western blot. One thing to watch out is that you should not spin down your sample after sonication. Generally, the pellet that you obtain after spinning after sonication is all the insoluble cell debris, which include chromatin, if you discard this pellet you will discard the bulk of your histone as well. Instead, directly boil your sample after sonication and then run your western blot. Of course, you should check your protein concentration to ensure equal loading, a good loading control for histone is of course total histone H3 in your case.

thanks for your answer. Just out of the curiosity, why a lot of papers would extract histone first and then examine their posttanslational modification?

Later, I decided to use histone isolation kits. It works well

Hey Wen-Jye Lin, I am working on H3K79 me2 on western detection too. I know its been a while for you, but what amount of purified histones did you use to load? I am purifying total histones with acid extraction. Also, did you check for a "linear range" for the signal intensity on H3K9 me2 antibody probing? 

I find that people usually load 20-30microgram of cell lysate but anywhere from 20ng-2microgram of purified histones. So I am confused and your experience would help. Thanks