protein assays

Protein Standard Curves: Setting up

The determination of protein content of an unknown requires an assay in which a measurable quantity is related to protein concentration, preferably in direct proportion. Second, protein determination requires standards for comparison, in order to quantify the results. The most common quantity measured in a protein assay is absorbance of light. By preparing samples with known amounts of protein and comparing their absorbances with that of the unknown, the unknown quantity can be estimated.

A typical standard curve as prepared for the Bradford assay will be used as an example.

Prior to setting up the assay, decide on a protein to use as a standard and determine the range over which the assay is likely to be sensitive. Sensitivity may vary from one batch of reagents to another. For the Bradford assay, immunoglobulin G (IgG) is frequently used as a standard, as is bovine serum albumin (BSA).

Technical note. Different proteins have different amino acid compositions, so that sensitivity of an assay to individual proteins may vary widely. To get the most accurate results, choose a standard that is composed of a mixture of proteins that is as similar as possible to the unknown. For most purposes, a relative amount is good enough. The Bradford assay is much more sensitive to IgG than to BSA, so that with IgG the investigator is likely to overestimate the amount of protein in a sample. With BSA the investigator is likely to underestimate the amount.

Using IgG as a standard a typical sensitivity range is from 20 to 200 micrograms of protein per 5 milliliters of reagent. That is, the result will not be very reproducible below 20 micrograms per 5 ml, and above 200 micrograms per 5 ml the assay becomes saturated.

It is necessary to keep everything as uniform as possible, to avoid bias when comparing unknowns with knowns. The only difference between one standard or unknown and the next should be protein concentration. In order to ensure that uniformity, the same buffer should be used to dilute all standards and samples, and they should all be prepared to the same final volume. A starting volume of 100 microliters is appropriate for the standard Bradford assay.

Preparation of standards

Prepare 2 milligrams/milliliter IgG (2 mg/ml) and have a quantity of buffer on hand, preferably the same buffer that contains the unknown protein.

Technical notes. Since batches of protein vary and one can pipet liquids much more accurately than one can weigh freeze-dried protein, it is best to prepare a large batch of standard protein solution and freeze it in aliquots. That way, throughout your investigations your protein determinations will be consistent. Components of some buffers have been known to interfere with colorimetric and other assays, either by reducing the binding of reagent to protein or by absorbing light in the range used by the assay. The assay descriptions usually indicate what agents may interfere with the assay. However it is a good practice to compare your buffer with protein dissolved in distilled water to look for possible interference, and if at all possible to use the same buffer for all measurements, to rule out any buffer effects.

Colorimetric assays can be prepared in standard disposable culture tubes, although when extremely accurate results are needed you may need to use matched cuvettes for the measurement of absorbance. Prepare one culture tube as a reference - 100 microliters of buffer only. Prepare the standards in cuture tubes as follows - 10 microliters of 2 mg/ml IgG plus 90 microliters of buffer; 20 microliters of 2 mg/ml IgG plus 80 microliters of buffer; likewise, up to 100 microliters of 2 mg/ml IgG with no added buffer. Leave them in a rack at room temperature.

Technical notes. Keep protein assay tubes at room temperature, since you need not be concerned about degradation of the proteins. You need to know the quantity, and the integrity of the proteins is not important. If you chill the tubes they will fog up in the spectrophotometer, producing high variability among your measurements. Do not add color reagent until all unknowns are prepared as well. The exception is if you conduct the same assay repeatedly, then once a standard curve is prepared you need only prepare one standard or two just to make sure that the conditions haven't changed.

To prepare unknowns, start by estimating their concentrations of protein. You need to estimate the range of volumes of each sample that contains amounts of protein that give an absorbance in the useful range of the assay. That is, for the Bradford assay you need the amount of protein in each unknown sample to fall between 20 and 200 micrograms.

For example, 20 to 30 mg/ml is 20 to 30 micrograms per microliter. Therefore, the minimum volume of sample that is likely to give an amount of protein in the sensitivity range is 1 microliter (1 microliter of a 20 mg/ml concentration contains 20 micrograms protein) . The maximum is about 6 microliters (6 microliters of a 30 mg/ml concentration contains 180 micrograms of protein). It is highly recommended that you assay unknowns using two different volumes - one near the minimum, one near the maximum. Then one or the other is likely to fall in the sensitivity range.

Technical note. The range of an assay depends on the standard chosen. The best standard is a protein solution of similar composition and known concentration. Otherwise, solutions of bovine serum albumin, immunoglobulins, or other pure proteins are used as standards. An assay can be more or less sensitive to different standards, leading to inconsistencies.

Pipetting small volumes tends to reduce your accuracy. If your samples are so concentrated that you have to pipet under 5 microliters, simply dilute a volume of your sample 10 fold, mix well, and pipet from the diluted solution. For example, to get the equivalent of 1 microliter of unknown protein simply mix 10 microliters of sample with 990 microliters of buffer. Pipet 100 microliters of that dilution into a culture tube. The tube will then contain 1 microliter of undiluted sample.

In your notebook write down exactly how you prepared your standards and unknowns, including volumes used, concentration and type of standard, etc. Prior to adding color reagent you should have:

Once you have all of the tubes prepared it is time to add color reagent and perform the assay.

Performing a protein assay

The standard Bradford assay requires that 5 ml color reagent be added to each tube of 100 microliters reference buffer, protein standard, or unknown. The addition is usually done in a fume hood behind glass, or while wearing a lab coat and safety glasses, since the reagent contains phosphoric acid and is therefore hazardous. Use of a repipet system allows rapid, accurate dispensing of reagent. Reagent should be added to all of the tubes in order at one time. Doing the standard curve first then testing unknowns later could lead to problems if conditions change between assaying standards and assaying unknowns. The color reagent is not completely stable, so comparing samples prepared at different times is not a good idea.

Once color reagent is added the tubes can be mixed by gentle vortexing if necessary. Rapid addition of color reagent usually mixes the samples sufficiently for most applications. The absorbance can be read 3 minutes later. A 5 minute wait is recommended.

The next step is to measure absorbance using a spectrophotometer, and to prepare the standard curve.

Please review the general principles for plotting experimental data and for use of standard curves before trying to plot and interpret a standard curve. Applying the principles to the absorbance data from the Bradford assay, a typical standard curve might look like this:

oops!  Guess you can't 'do' images.  Sorry.

Theoretically, the relationship is linear, however if the data appear to curve you may fit a curve rather than a straight line. At the higher concentrations the absorbance doesn't change very much with protein concentration. In fact, in the example the assay is not useful above 150 micrograms of protein. It isn't reliable below 20 micrograms either (beware of the dangers of extrapolation). So... if one or more of your unknowns gave an absorbance out of the range, repeat the assay for that sample. Hey... it takes 5 minutes - you don't need to repeat the standard curve. You've gone to this much trouble, so don't waste your efforts on a poor guess.

The standard curve gives you the amount of protein in the test tube, not a concentration. You should have recorded all of the pertinent information in your notebook on each unknown, including the volume of undiluted sample in each unknown tube. Calculate the concentration of protein in the undiluted sample by dividing the amount of protein you have determined from the absorbance by the volume of undiluted sample in the unknown. Report concentrations in milligrams per milliter, rounding to the appropriate number of significant figures.

Determining a concentration from duplicate samples

Investigators usually prepare two or more assays for the same unknown, either to obtain better accuracy or to increase the probability that the absorbance measured will be on the linear part of the standard curve. In the latter case, each assay contains a different volume of unknown. That way, one assay might be too concentrated or too dilute, while the other is readable. Sometimes both results fall on the standard curve, however. Here's an example and a suggestion for determining the concentration.

A protein fraction is estimated to have a concentration of from 2 to 10 mg/ml. The investigator prepared one assay tube with 40 microliters unknown, and the other with 10 microliters. Both assays gave absorbances within the useful range of the assay, but after calculating concentrations there was a significant difference between the two. The simplest resolution, which is not the best choice, is to average the two results. If the investigator really doesn't know what he is doing he'll report two different concentrations for the same sample, and the editor will reject his paper. The proper choice is to report the single concentration given by the lower of the two absorbances measured. The rationale is that because the absorbance scale is logarithmic, the results are much more accurate at the low end of the range. Of course, if there is wide disparity between the two results, the investigator might review his/her notes and/or work on pipetting skills.


Copyright and Intended Use
Created by David R. Caprette (caprette@rice.edu), Rice University 18 Aug 1995
Updated 14 May 1997