and should be used in the modified assays discussed in Note 4. However, the quality of the dye is not critical for routine protein determination using the method described in this chapter. The data presented in Fig. 1 were obtained using Coomassie Brilliant Blue G ( C.I. 42655; product code B-0770, Sigma-Aldrich ).
6. Whenever possible the protein used to construct the calibration curve should be the same as that being determined. Often this is impractical and the dye response of a sample is quantified relative to that of a “generic” protein. Bovine serum albumin (BSA) is commonly used as the protein standard because it is inexpensive and readily available in a pure form. The major argument for using this protein is that it allows the results to be compared directly with those of the many previous studies that have used bovine serum albumin as a standard. However, it suffers from the disadvantage of exhibiting an unusually large dye response in the Bradford assay, and thus, may underestimate the protein content of a sample. Increasingly, bovine γ- globulin is being promoted as a more suitable general standard, as the dye binding capacity of this protein is closer to the mean of those proteins that have been compared (Table 2) . Because of the variation in response between different proteins, it is essential to specify the protein standard used when reporting measurements of protein amounts using the Bradford assay.
7. Generally, it is preferable to use a single new disposable polystyrene semimicrocuvette that is discarded after a series of absorbance measurements. Rinse the cuvette with reagent before use, zero the spectrophotometer on the reagent blank and then do not remove the cuvette from the machine. Replace the sample in the cuvette gently using a disposable polyethylene pipet.
8. The standard curve is nonlinear because of problems introduced by depletion of the amount of free dye. These problems can be avoided, and the linearity of the assay improved, by plotting the ratio of absorbances at 595 and 450 nm (15). If this approach is adopted, the absolute optical density of the free dye and dye–protein complex must be determined by measuring the absorbance of the mixture at each wavelength relative to that of a cuvette containing only water (and no dye reagent). As well as improving the linearity of the calibration curve, taking the ratio of the absorbances at the two wavelengths increases the accuracy and improves the sensitivity of the assay by up to 10-fold (15).
9. For routine measurement of the protein content of many samples the microassay may be adapted for use with a microplate reader (7,16). The total volume of the modified assay is limited to 210 μL by reducing the volume of each component. Ensure effective mixing of the assay components by pipetting up to 10 μL of the protein sample into each well before adding 200 μL of the dye reagent. If a wavelength of 595 nm cannot be selected on the microplate reader, absorbance may be measured at any wavelength between 570 nm and 610 nm. However, absorbance measurements at wavelengths other than 595 nm will decrease the sensitivity of response and may increase the minimum detection limit of the protocol.
10. For studies on the use of the Bradford assay in analyzing glycoproteins, see Note 9 in Chapter 3.