No instrument reads directly how much protein a solution contains. What is measured is always an indirect quantity: usually the amount of light absorbed at a given wavelength. Turning that reading into a concentration in milligrams requires a conversion rule, and that rule is the standard curve.
A standard curve is a reference line obtained by reading solutions of known concentration. The reading for the unknown sample finds its counterpart on this line. The method looks simple, yet the reliability of the result depends entirely on how the curve was built.
Several approaches are used in protein assays. Some rely on a dye that changes colour when it binds to protein; some depend on copper ions being reduced in the presence of protein; the most direct route is to read the solution's absorbance in the ultraviolet region. Whichever route is chosen, a calibration is needed to convert the reading into a concentration.
How is the standard series prepared?
A reference protein solution of known concentration is prepared first. Serial dilutions of it yield several intermediate concentrations covering the range to be measured. The series is designed to bracket the expected sample value from below and above; if the sample falls outside the curve, the result rests on extrapolation.
The standards and the sample are read in the same solvent, under the same conditions and in the same session. Working with a curve prepared on a different day carries instrument- and environment-related drift into the result. Reading each concentration several times and using the mean is also common practice.
Which protein was used as the standard is put on record as well. The usual choice is a purified reference protein of well-defined concentration. The reference is known not to behave exactly like the proteins in the sample, so the result is read as an equivalent value expressed against the chosen reference.
From curve to equation
The readings are plotted with concentration on the horizontal axis and absorbance on the vertical axis. The best-fitting line through the points is calculated by linear regression. The result is an equation of the following form:
y = slope × x + intercept
Here y is the absorbance read and x the concentration sought. The sample reading is substituted into the equation and solved for x to give the concentration, which is then multiplied by the dilution factor where one applies. An intercept different from zero is normal and usually arises because the blank reading does not come out at exactly zero.
Repeating the measurement is part of reliability too. Taking more than one reading for each standard and sample prevents a single pipetting error from carrying the result away. If the difference between readings is larger than expected, the preparation steps are reviewed before the curve is accepted.
What does the R squared value tell you?
The coefficient of determination shows how well the points fit the calculated line. The closer it is to one, the better the fit. In laboratory practice a high value indicates that the standard series was prepared properly and that the measurement was made within the linear range.
Even so, it is not a sufficient criterion on its own. A high coefficient does not rule out the possibility that the standards were prepared at the wrong concentrations; it only says that the points are consistent with one another. The assessment is therefore also made by looking at the plot of the curve: points curving away at the ends show that the linear range has been exceeded.
| Item | Value in the reference batch report |
|---|---|
| Standard curve equation | y = 0.0011x + 0.1726 |
| Coefficient of determination | R² = 0.9903 |
| Protein concentration found | 28 mg/mL |
| Where the measurement was made | An independent laboratory outside the facility |
The values in the table belong to a single batch. The measurement is repeated for every batch and the result may differ from batch to batch; the report is therefore read as the record of a particular sample, not as a fixed property of the product.
Renewing the curve at regular intervals is part of good practice as well. Working with a curve built long ago means assuming that neither the instrument nor the solutions have changed since that day. The storage conditions and expiry information for the standard solutions are likewise kept on record.
Common pitfalls
- Reading outside the range: If the sample's absorbance exceeds the upper limit of the curve, the result is unreliable; the sample is diluted and the measurement repeated.
- Choice of blank: The blank must be measured with the same solvent as the sample; a different solvent shifts the intercept.
- Interfering substances: Buffer components, detergents and reducing agents can affect the reading, depending on the method.
- Difference in reference protein: The result is expressed against whichever protein the curve was built with; a different reference may give a different value.
- Dilution error: In dilutions made with small volumes, a pipetting error feeds straight into the result.
How the result is reported also matters. The concentration is usually given in milligrams per millilitre; where the total amount of protein is wanted, this value is multiplied by the volume. Alongside the figure, the report includes the method, the reference protein used and the measurement conditions. Without this information, results from two laboratories cannot be compared directly.
Where is the result used?
Protein concentration is more than a figure in its own right: it is the input for subsequent analyses. How much sample to load onto the gel in electrophoresis is decided from this value; the method is described in the article on SDS-PAGE. The physical basis of absorbance measurement was covered in the article on UV-Vis.
The value is also part of the batch record and is reported in the certificate of analysis together with its method; reading the certificate is explained in the article on certificates of analysis. That the measurement is made outside the facility, in an independent laboratory, carries its own weight in assessing the result. Our batch-level results are published on the analysis page.
