Hi there,
I would like to determine the unknown concentration of human insulin in my sample during crystallisation (the concentration range is roughly 0.5 -2 mg/ml, whereas the initial concentration before crystallisation is 2.0 mg/ml).
First, I prepare a standard curve by diluting the stock insulin solution (know concentration) with 0.1 M citrate buffer (pH ~7) to standard samples within the range of 0.1 -0.5 mg/ml (concentration is known). Then, 50 ul of the standard sample is mixed with 500 ul Bradford assay and the solution is filled in PMMA cuvettes and mounted into a uv-vis spectrophotometer. The absorption is read ta 595nm.
To determine the unknown concentration of insulin during crystallisation I dilute the crystallisation sample (after centrifugation and removal of excess solid material) with 0.1 M citrate buffer (pH 7) with a dilution factor of 10 (eg at the onset of crystallisation the concentration in the reactor should be the initial concentration (2 mg/ml), and the concentration of the diluted sample should hence be 0.2 mg/ml). The protocol is the same as described above (50 ul diluted sample with unknown concentration + 500 uL Bradford assay). However, using the standard-curve described above, I end up with a concentration of 0.1-0.15 for the diluted sample, which is significantly lower than I expect since no crystals are observed yet.
I wonder where my mistake lies by using Bradford assay?
Thank you very much for your help in advance
From your description, it looks like there is nothing wrong with your Bradford assay protocol. However, are you sure that you can rule out that microcrystals may form very rapidly and be eliminated by the centrifugation step even though they may not be visible with the naked eye or even a low power magnifyer ? To check for this, you may try to take up whatever pellet is present at the bottom of the centrifuge tube in a small volume of crystallization buffer and look at it under a microscope with several hudredfold magnification.
Pierre Béguin thank you very much for your answer, but do you think the ratio of sample (50uL) to Bradford Assay (500uL) is justified? The original protocol states that at 0.1 ml sample should mixed with 3 ml Bradford Assay (which results in a ratio of 30 between Bradfords assay and sample). However when I am using 50 ul sample + 500 ml Bradford Assay the ratio is "only" 10.
As long as you perform the assay under the same conditions as the standard curve, and your data points fall within the range of the standard curve, I don't see a problem. But if this worries you, you can try 16.6 µl plus 483.3 µl of Bradford reagent. At any rate, beware however that the dose response curve is non-linear with the Bradford assay except for low concentrations of proteins (it flattens agt high concentration. It is therefore not legitimate to draw linear regressions with the standard curve points, nor to extrapolate from data lying outside the range of the standard curve.
Apparently, there is nothing wrong with protocol. Couple of suggestions though. When you are making samples for standard curve, centrifuge those samples too and then set up reactions with Bradford and plot. Another thing is (this is my personal opinion), one should always quantify protein using two different methods. Also, I am not a fan of Bradford reagent for several reasons, important reason being it is not specific for proteins even though the method is very routinely used everywhere. Several factor affect the reaction for example imidazole also reacts with Bradford reagent to some extent to contribute to readings. I would advice quantify protein using another method too and crosscheck where variation is occurring or for that matter whether the Bradford reagent method is really suitable or not.
When we use Bradford method, we add 20ul protein to 980ul reagent and incubate reaction for 10min and record the reading. Also if we have to dilute the sample, instead of diluting the sample we reduce the protein sample and adjust the volume to 1ml with Bradford reagent instead of actually diluting the sample with buffer.
How do you obtain the initial concentration of your insulin? All protein determination methods are relative to a standard substance (frequently BSA or IgG, but you also may use your favourite protein) and follow different principles (e.g. copper chelation, dye binding). Thus, it's difficult to impossible to compare readings obtained with different methods, unless you have established correction factors for your specific protein. The non-linearity of all methods (except weighing with a balance :) ), as Ameya Dipak Bendre already has mentioned, doesn't make things easier.
Does your crystallization solution contain any other compounds, buffers, additives?
Dear Michael G. Weller ,
just citric acid buffer (0.1 M) around a pH of 7, I am not quite sure whether this substance interferes with the Bradford assay tho, but I dissolve insulin in the same buffer for preparing the standard samples
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biochemical Techniques
- Analytical Chemistry
- pH