What an extinction coefficient assumes

Database pages print an extinction coefficient next to a sequence as if it were a serial number. It is a conversion factor with conditions attached. Absorbance at 280 nm becomes a concentration only when those conditions hold. This explainer is for anyone about to divide an absorbance by a number copied from UniProt or from a cloning report. The decision it supports is whether that division describes the polypeptide in the cuvette. Assay choice in the wider sense is covered in how laboratories estimate protein concentration and in Bradford, BCA and absorbance at 280 nm. A gel still answers a different question, as reading a protein gel sets out.

Composition is an average, not a measurement on your protein

Pace-style calculations assign average molar absorptivities to tryptophan, tyrosine and cystine, then add them up for the sequence. The arithmetic is transparent. The assumptions are easy to skip. The model treats those chromophores as free to absorb in the way the reference proteins did. For the coefficient most people quote, that state is the folded polypeptide in ordinary aqueous buffer, or at least a state in which the chromophores are not quenched, shifted or joined to something the sum ignored. The sequence must be the complete chain you actually have, with disulfides counted as cystine rather than as free thiols if that is what the chosen table expects. There is no extra absorber. The path length is 1 cm unless you correct it. The concentration you compute is the concentration of that polypeptide, not of "total protein" in a lysate.

Those averages were fitted to real proteins, and individual proteins scatter around them. A calculated coefficient is a reasoned estimate. It is not a certificate. If the scientific claim needs a tight molarity, measure the coefficient for that construct under the buffer you use, or show why the estimate is good enough for the decision.

The sequence in the database is often not the chain in the tube

Tags are the everyday break. A hexahistidine tag adds mass and almost no 280 nm absorbance. If you apply the untagged coefficient and the untagged molecular weight, you mis-state both the molarity and the mass concentration of the fused protein. If the tag contains a tyrosine, both the coefficient and the mass change. Fusion partners such as maltose-binding protein or thioredoxin bring their own tryptophans. Use the coefficient calculated for the exact construct, including the linker you left in.

Cleavage makes a mixture. Incomplete removal of a tag means the absorbance is the sum of two species with different masses and sometimes different chromophore counts. One coefficient cannot represent both. A gel shows whether cleavage finished. The absorbance does not.

Truncation and cloning scars do the same quiet damage. A missing N-terminal tryptophan removes a large fraction of the absorbance of a small protein. The database entry for the wild-type gene will not know about your deletion. Recalculate from the translated product you expressed.

Cofactors, nucleic acid and path length

A heme, a flavin, NADH, pyridoxal phosphate or a fluorescent protein chromophore absorbs whether or not the amino-acid sum predicted it. Green fluorescent protein is the teaching case: the chromophore, not tryptophan, dominates the visible colour, and the ultraviolet story is not a plain polypeptide either. Report the absorbance spectrum and handle the cofactor as its own quantity. Do not force a composition coefficient through a coloured elution.

Nucleic acid is the other inflator. DNA and RNA absorb near 260 nm and still contribute at 280 nm. A high A260/A280 ratio means the 280 nm reading is partly nucleic acid. Subtracting with a generic formula is only as good as the formula's assumptions about the contaminant. For a purification that must be nucleic-acid-poor, treat a bad ratio as a failed cleanliness check and clean the sample, then read again.

Path length is part of the unit. Coefficients in reciprocal molar per centimetre assume 1 cm. A microvolume instrument that uses a shorter path must be used as its manufacturer specifies. Writing "A280 = 0.8, so the protein is 0.8 divided by the UniProt number" without a path length is an incomplete sentence. Temperature and buffer refractive effects are smaller, but a blank of a different buffer is not small. Match the blank to the cuvette contents minus protein.

Denaturant, scatter and a gel that disagrees

Inclusion-body work dissolves protein in urea or guanidine. The aromatic residues are then in a different solvent. Using the aqueous folded coefficient without comment overstates how much you know. Some composition tables publish a denatured-chain value. Use that class of number if it matches the solvent, and write the solvent down. After you remove the denaturant, the coefficient may change again if the chain refolds. Activity, not the absorbance, tells you the fold succeeded.

Scatter from aggregates raises apparent absorbance across wavelengths, often more at shorter ones. A rising baseline toward 250 nm is a hint. Clarify the sample before you convert A280 into mass. If a Coomassie gel shows several bands, the coefficient of the target divides an absorbance that belongs to all of them. You have calculated a fictional molarity. Go back to a total-protein assay with a named standard, or purify until one chain dominates.

Research limits

An extinction coefficient is not a purity stamp and not a clinical measurement. Reference-material thinking, of the kind NIST describes for chemical composition, is the right habit: the material, the unit and the method travel together. Biosafety for the protein source is an institutional decision.

Laboratory temperature changes buffer blanks slightly and can shift a sensitive instrument. Record the buffer recipe and the instrument. If the room is far from the temperature at which you prepared the buffer, check pH separately when pH matters to the protein. Do not let a warm cuvette block stand in for a blank.

What an enquiry should carry

If you are asking for a pure protein or for help specifying one, send the exact amino-acid sequence of the construct, including tags and linkers, and say whether a coefficient must be calculated or measured. Note cofactors and whether nucleic-acid limits matter. Reagents and chemicals covers buffer components used around the measurement. Use a quote for the sourcing question. The custom protein expression and purification page is an enquiry reference where construct design can be discussed. It does not mean a laboratory on this site has already expressed the protein.