Populations of microorganisms exhibit wide variability in protein expression that can lead to differences in shape, structure, and behavior. In this talk I will describe the effect of expression variability on the behavior of bacteria in a temperature gradient as an example, and provide a quantitative relationship between the two. I will then turn my attention to understanding the expression variability in microorganisms in general. In this regard, I will describe a recent study, in which we have showed that the protein content in populations of microorganisms including bacteria and yeast, measured under a broad range of biological realizations, displays a universal distribution when scaled properly. This universality indicates that the observed variability cannot reflect any specific molecular or cellular mechanisms, and suggests that some buffering process masks these details and induces the universality. I will also show that the buffering process that induces the universality is the rich dynamics of protein expression in single-cells. Finally, I will briefly discuss the possible consequences of our findings to population dynamics and fitness.