Bacterial “gain of function” experiments. What could possibly go wrong?