In finding out importance of predictors using TreeBagger, is correlation among predictors taken care of?

Hello,

I am using TreeBagger OOBPermutedPredictorDeltaError to estimate the importance of the predictors in a classification problem. I am wondering, if there is a strong correlation between two predictors, would they be given equal importance or one predictor would be given a lot more importance than the other?

Also, I guess it always depends, but is there any rule of thumb for the ratio of the number of predictors to the number of instances in Random Forest?

Thank you very much.