I'm measuring fluorescence intensity of the product in an enzyme reaction in 37 celsius degree. I set a kinetic measurement in 1 hour. The fluorescence intensity of the BLANK ( contains only buffer and substrate ) decreases by time and its graph is not a straight base line. I'm thinking if other samples I'm measuring has the same change as the Blank solution, so should I normalise the data by dividing the blank's fluorescence intensity ?
Difficult to answer categorically. As a general rule, substracting the blank value is the correct thing to do. On the other hand, since a positive result will alter the chemistry, it may also alter the rate of drift.
I would suggest doing a series of tests with a dilution series of an enzyme solution of known activity and checking whether background subtraction improves the data or not.
Either way, make sure you explicitly state whether background subtraction was applied and how when you publish your results.
Generally, I would subtract the baseline trace from the sample with no enzyme from the trace for the full reaction. The decrease in fluorescence is probably due to photobleaching of the substrate. The amount of photobleaching of the remaining substrate should be essentially the same for the full reaction, as long as the proportion of substrate that is converted to product is small, such as 10% or less, which should be the case if you are measuring the initial rate of the reaction.
You may also observe photobleaching of the product. If this is the case, then you may be able to reduce it by reducing the amount of light to which the sample is exposed. In a plate reader or flash-lamp-type spectrofluorometer, this would be done by reducing the number of flashes and/or taking more widely spaced time points. In a spectrofluorometer with continuous illumination, this would be done by reducing the excitation slit width.
It depends on your requiremets. Generally, you do not need to normalize it.
Hi Phuong,
I agree with Andrew and Adam's answers and will add a couple other hypotheses worth considering. In general, you are looking for the reason your blank signal has unexpectedly decreased.
1. photobleaching - This will be an exponential decay with time, though it can look linear if the rate is slow relative to your time window. As Adam mentioned, you want to decrease the amount of light reaching your sample. Narrowing the excitation slit will do this. If you want to maintain the same signal level, match this with a wider emission slit. If you don't have control over slits, you can put a neutral-density filter between the excitation source and the sample.
2. instrument drift - I used to work for an instrument manufacturer and drift is just a difficult thing to minimize. Warm up your instrument for longer than the manufacturer recommends -- I suggest at least 1 hour. You may need to do a measurement to completely initialize the instrument.
3. material translocation - you have enzymes in solution, and proteins to some degree like to be at interfaces. Your instrument is examining the middle of the cuvette volume, but with time your proteins will migrate from the bulk solution to the air-water interface or cuvette-water interface to some extent. I have seen papers showing this on the 1 hour timescale. If you can add a stir-bar to your cuvette, this could help maintain solution homogeneity. If there are long times between measurements, maybe take the cuvette out and mix it by inversion between measurements.
If you suspect hypotheses 1 or 2, and cannot totally eliminate the baseline artifact, then subtract the blank. If you think you have material translocation and cannot mix the solution... good luck :)
Hi Phuong Nghiem Thanh It is first important to find out that the decreased fluorescence is not due to any solubility issue or aggregation problem of the components in the mixture. You can verify this by measuring both absorbance and fluorescence for each sample. Also observe your samples visibly. This can be one posibility.
- Badges
- Science method