I have been working on genetically engineering an E. coli strain to autotrophically oxidize arsenite into arsenate for bioremediation of arsenic contamination in groundwater. For my research, I have been finding ways to quantitatively measure how much arsenite the bacteria oxidizes over time. After researching ways to quantitatively measure arsenite being oxidized, I have concluded to use the method of Johnson and Pilson (https://www.sciencedirect.com/science/article/pii/S0003267072800059?via%3Dihub).

The arsenate that is formed from arsenite oxidation forms a molybdenum-blue complex with phosphate in the water. As the concentration of arsenate increases, the absorbance value of the water should increase proportionally.

Here was my plan: First, after growing my transgenic bacteria on an amp-plate, I would pipette out various concentrations of cells and arsenite in wells of a 96-well spot plate. I currently plan to use the Fisher Scientific Multiskan FC microplate photometer. I planned on incubating the cells while shaking at 37 C. The main problem I have come across is that my experiment would not be controlled. Over time, the growth of cells will change the absorbance value, not just the molybdenum-blue complex. I need to isolate the variable of just arsenite oxidation.

I have thought of a few possible solutions, but I feel as if I may be missing something crucial. When learning about microplate readers, I know that researchers can use them to measure gene expression. How do these researchers eliminate the variable of cell growth effects on absorbance values?

For one, I have thought of finding a way to filter the cells to one side of the well where they will not affect the absorbance values. Would it be possible for me to cut out a dialysis bag to where it is dipping into the well? In that case, the cells could be suspended in the well, oxidizing arsenite (if my hypothesis is correct).

Anyway, I really want to find a simple method for quantitatively measuring arsenite concentrations over time using a microplate reader without having cell growth affecting absorbance values.


1 Answer 1



Does the photometer allow using different wavelengths of light? Perhaps using a spectrophotometer might give you a control and your desired value, i.e. blue light may be absorbed by your sample, but red might not? The absorbance spectrum of the molybdenum-blue complex is very specific and you could use this to your advantage using spectral photometry.

Photometry + correcting for absorbance due to cell growth

Another approach is to quantify over time how your bacteria affect absorbance as a function of time and then subtract this from your experimental values. For example, in a wellplate, place the following adjacently: use a (i) well without cells, and only arsenite in solution, (ii) cells without arsenite, and (iii) cells with arsenite. This should allow you to quantify quite accurately how much the arsenite, the cells, and how much both affect absorbance.

Of course as you say there may be an interaction between the cells and the arsenite slowing or speeding up cell growth, so that would be a pickle to disentangle. You could try, perhaps in the wellplates (this would be best!), to find out what effect arsenite has on growth and correct your final data this way.

Periodically sampling the solution; removing bacteria from the measurement

Another idea would be to culture your cells in arsenite, and take samples to obtain time-points. For each sample, you filter using small syringe filters (these are cheap and are typically used for sterilizing solutions, i.e. getting rid of bacteria but keeping the solution, given their pore size of ~.1 microns) and then perform photometry for the molybdenum-blue complex. This is typical practice in microbiology.

PCR? An idea.

Last but not least, it's good to remember this is an indirect readout of transgenic gene expression. Similarly, you could use quantitative PCR to quantify the expression of any transgenes - here, you use housekeeping genes as controls for cell number, and the mRNA expression of your transgene as the readout. You would be reading out the expression of mRNA, but I personally don't think it would be more 'indirect' than measuring arsenate concentrations in solution, especially when there are confounding factors and awkwardness in getting a clean measurement. The PCR approach would be quite accurate and likely more reliable.

These are just some initial thoughts - perhaps you could add a comment or we could chat about this if you think it could spark up some concrete ideas!

  • $\begingroup$ Thanks for the advice! I will be collecting data hopefully over a 3 hour course. For the amount of time I have to complete experimentation, I might go with the periodic sampling method that you mentioned. I could try to grow multiple culture tubes on a shaker, and at 10 minute intervals filter out the solutions into the wells (containing LB, amp, sodium arsenite, and potassium phosphate dibasic). Also, I would create several stock solutions with known concentrations of arsenate and arsenite while accounting for absorbance values of LB and amp. I could just calculate all needed concentrations. $\endgroup$
    – HLR
    Mar 4, 2019 at 13:01
  • $\begingroup$ Also, I dont have any arsenate on hand to make a stock solution; I just have arsenite. I know if I have stock solutions, I can calculate the amount of arsenate based on absorbance, but would it be scientifically "acceptable" to have my measurements in absorbance instead of concentration? $\endgroup$
    – HLR
    Mar 4, 2019 at 13:59
  • $\begingroup$ Depends on what you are trying to demonstrate. If it is not feasible to measure arsenate concentration, you can use absorbance as a relative unit! Twice as absorbant doesn't mean you have twice the concentration, but you can compare different bacterial strains; you'll know which one oxidizes arsenite more. Very often in biology (and science generally) we don't have absolute measurements; we can only compare to a control, for instance. PS. It's good to check whether arsenite oxidizes by itself over time (it might not, I have no idea) - you'll have to correct for this if you want solid data. $\endgroup$
    – S Pr
    Mar 4, 2019 at 15:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .