[Clyde Schechter]

I predicted that two variables would interact but wasn't sure about the direction. For example, do increasing values of OPC correspond to increasing marginal effects of SR or do increasing values of SR predict increasing marginal effects of OPC?

From a statistical perspective this directional distinction does not exist. If one is true, the other is true, and vice versa, in any interaction model. If you are thinking about which way it runs causally, you can't approach that through regression analysis of observational data--you either need an experimental design, or have to rely on a well-developed theory underlying the mechanisms to draw causal conclusions.