How to design a time course experiment - (Feb/14/2014 )

My question is regarding the design of time course experiment. I wanted to know the expression level of one miRNAs in H1N1 infected A549 Cells at 24, 48, 72 hrs post infection and to compare it with the non infected cells. I will do the experiment in 12 well plate. I am a bit confused whether I design the experiment that I have 3 wells as non infected cells ( no H1N1 added but only cells) and the other other 3 time points in replicates then I will have 9 wells infected with virus and I will collect them at each time point. then in this case I will compare each time points with the 0 point as a control for all times. The other way is to create a 3 wells as non infected fr each time points meanings that i will have 9 wells as non infected ( each 3 wells will be compared with each 3 wells from the 3 time points) and other 9 wells as an infected. then I will compare the 3 replicates of each time points with its 3 well non infected replicate. 

What is the right way ? 

You can do either, but they tell you slightly different things - comparing to T0 means that you can see if the treatment changes something relative to the start, but comparing to samples that are harvested on the same day tells you it isn't just an aging of the cell line effect.

I agree but, In case of miRNAs u need to show somehow the level at the 0 point and If I made 0 point replicate for each time points, then which one to choose to be your uninfected ( baseline)  level of miRNAs ? in my miRNAs I always compare with the non-infected control.

Also in the first case meaning with one replicate of 0 point (control) if you have an aging effect in the time points, you will have it in all time points so the factor is also fixed since we are not measuring an absolute numbers. what is your opinion ?