multiple comparison procedures

#1
How should MCP be applied where the data have "within" and "between" characteristics,
say treatments are A, B, AXB and there are several response variables from the same subjects?
Between the treatments requires protection, but so does the proliferation of response variables