I am currently using the classic colorimetric NBT assay to measure intracellular superoxide production in primary cell cultures. Cells are stimulated with PMA, NBT is reduced by superoxide to produce formazan crystals and formazan is dissolved in KOH and DMSO which is read on a plate reader. Based on the protocol that I am trying to adapt, exogenous SOD is added to an additional set of samples stimulated with PMA in the presence of NBT. Essentially you have 2 treatments with 3 technical replicates of each. Each set also has a set of blanks that are identical except for the absence of cells.

1) Cells + NBT + PMA

2) Cells + NBT + PMA + SOD

The authors state that to determine intracellular superoxide, "The absorbance in each well is measured at 630 nm in a microtiter platereader. The OD values measured in both the "blank" wells and those wells containing SOD are subtracted from the OD values measured in the wells with cells, but without SOD. The final nmole concentration of intracellular superoxide is calculated by multiplying the final OD values by 15.87."

From my reading, this would mean that values from treatment #2 are subtracted from treatment #1. I might be missing something, but this doesn't make sense to me given that to my knowledge, SOD is not permeable to the cell membrane and would therefore only scavenge extracellular superoxide. Therefore, if you were to subtract these values, you would be estimating extracellular superoxide.

My question has two parts:

Is my interpretation of this assay/protocol correct? If so, why are the wells containing SOD subtracted from the wells without SOD?

What is the purpose of SOD in this assay?