I got the answer D
He got the answer C
We both used a variation of the matlab tool to get our P values. My hunch is that there was a mixup in the way we look at "test for no interaction" and that one of us mixed up the factor order.
When I saw what the OP had written, I thought: did he expect us to do the computing for him? Or is there an other trick that I simply miss? Actually I can not do interaction plot in my head. But even then it would be diffcult to compute the p-values.