You need to measure the phosphorylated form of AKT, the total level of AKT and your loading control (ie, tubulin, actin, GAPDH, etc). You should measure the denistometry for each band separately.

1. phospho-AKT/loading control

2. total AKT/loading control

Then you take the ratio of the two ratios

3. (phospho-AKT/loading control = 1)/ (total AKT/loading control = 2).

This "ratio of ratios" will give you the increase of activation of AKT. Report data for #3 as phospho-AKT/total Akt. This will indicate if you have real increased activation (increase in phosphorylated AKT when total AKT levels stay the same) OR increase in total expression (increase in both phosphorylated and total AKT levels).

Both.  First normalize phospho-AKT and total AKT values to your loading control (Actin, tubulin), then look at the phospho/total ratio using the normalized values.

Actin, or other structure or housekeeping protein, should be used as a loading control for both phospho-AKT and total AKT expression.  Quantifying the protein and loading equal quantities of protein for each sample is also recommended, in addition, for normalizing data.   

If you are looking at the differences in phospho-AKT expression resulting from some experimental condition (e.g. PTEN knockdown), it would be good to also show the expression of total AKT in order to distinguish between a change in AKT expression (change in transcription/translation) vs a change in the phosphorylation status of AKT.    

For example, if phospho-AKT increases and total AKT remains constant, or decreases slightly, then you could argue that your experimental stimulus increases phosphorylation of AKT.  However, if phospho-AKT and total AKT both increase (or decrease), and the ratio of phospho/total remains constant, than it could be argued that your experimental condition increases (or decreases) expression of AKT, without changing its phosphorylation.

Hope this helps.

You need to measure the phosphorylated form of AKT, the total level of AKT and your loading control (ie, tubulin, actin, GAPDH, etc). You should measure the denistometry for each band separately.

1. phospho-AKT/loading control

2. total AKT/loading control

Then you take the ratio of the two ratios

3. (phospho-AKT/loading control = 1)/ (total AKT/loading control = 2).

This "ratio of ratios" will give you the increase of activation of AKT. Report data for #3 as phospho-AKT/total Akt. This will indicate if you have real increased activation (increase in phosphorylated AKT when total AKT levels stay the same) OR increase in total expression (increase in both phosphorylated and total AKT levels).

I agree. Normalise to total Akt and then also a loading control such as actin. 

Wouldn't (pAkt/ GAPDH) divided by (Akt/GAPDH) be the same as pAkt/ Akt? (assuming the levels were checked on the same gel by stripping and re-probing)

Poornemaa is correct. Simple math:

PAKT=1

AKT=2

GAPDH=3

(1/3)/(2/3)=.5=(1/2) or pAKT/AKT

Question you should be asking is does akt expression change? Does AKT/GAPDH change per condition? If the answer is no then you just do pAKT/AKT to ascertain phosphorylation state.

If AKT expression changes relative to GAPDH then you have a different observation completely. Though you can probably still quantify pAKT/AKT, you just need to make sure you understand that AKT total expression is also changing among your conditions, and this is at least partially responsible for any change in apparent phospho state.

1. p-protein/loading control

2. t-protein/loading control

3. (p-protein/loading control)/(t-protein/loading control)

If the same loading control is used for 1 and 2, is it equivalent to p-protein/t-protein?

Should only 3 be shown in the figure?

In theory, you would not need to use the loading control in the calculation of the gel is evenly loaded AND you can detect both forms on the same membrane. However, I don't believe it's truly possible to detect both forms on the same membrane, as their molecular weights are exactly the same, and detect in the same area of the membrane. Thus, you have to run two separate gels and therefore have two separate loading controls for each membrane, as loading may not be 100% accurate between both gels.

@DevonMason You would be correct IF you assume that everything is loaded at the same amount - which it might not be. Personally, I think using 32P or 33P is a much better approach, albeit a lot more cumbersome.

Do I have to take care of exposure time or antibody dilution regarding pAkt and total Akt (on different Blots), when I detect the loading control seperately (different exposure time, diferent picture than pAkt and Akt)? Or do I need to detect the loading control in the same picture with the same exposure time together with pAkt or Akt?