Colorimetric Determination of Tungsten and Molybdenum in Biological Samples

Materials and reagents

Biological materials

1. Native tungsten-containing proteins in the cytosolic extract of the anaerobic bacterium Eubacterium limosum were separated anaerobically using anion exchange chromatography and further purified using size exclusion chromatography (SEC) by a Bio-Rad Enrich SEC 650 (10 × 300) column using 25 mM Tris-HCl, pH 7.6, and 100 mM NaCl buffer. These fractions contain the tungsten-binding protein, Tub, which is a hexamer of 7 kDa subunits and binds a total of eight tungsten oxyanions [8]. Sixteen column fractions from the SEC column were used in the tungsten assay.

Reagents

1. Cetylpyridinium chloride (CPC) (Sigma-Aldrich, catalog number: C0732)

2. Deionized water (diH₂O)

3. Ethanol, absolute (Decon Laboratories, catalog number: 2716)

4. 70% nitric acid (Fisher Chemical, catalog number: EW-88050-70)

5. Quercetin hydrate (Sigma-Aldrich, catalog number: 337951)

6. Sodium fluoride (Sigma-Aldrich, catalog number: S1504)

7. Sodium molybdate dihydrate (Sigma-Aldrich, catalog number: M1651)

8. Sodium tungstate dihydrate (Sigma-Aldrich, catalog number: S0765)

Solutions

1. 1 mM quercetin in ethanol

2. 10 mM CPC

3. 1 M nitric acid

4. 0.1% sodium fluoride

5. 100 μM sodium tungstate

6. 100 μM sodium molybdate

7. 2.5× nitric acid mix (see Recipes)

8. 1× nitric acid mix (see Recipes)

9. Ethanol blank mix (see Recipes)

10. Quercetin assay mix (see Recipes)

Recipes

1. 2.5× nitric acid mix

2. 1× nitric acid mix

3. Ethanol blank mix

4. Quercetin assay mix

*Note: Quercetin will begin to precipitate out of the solution more than 30 min after the formation of the mixture.

Laboratory supplies

1. 1.5 mL microcentrifuge tubes (Heathrow Scientific, catalog number: AXYRESV25S)

2. Plate-sealing film (EXCEL Scientific, Inc., catalog number: EZP-100)

3. 96-well assay microplates (Corning, catalog number: 9017)

4. Polypropylene 96-well plate (Elkay, catalog number: MICR-PVX)

5. PYREX^® 50 mL round media storage bottles (Corning, catalog number: 1395-50)

6. PYREX^® 100 mL round media storage bottles (Corning, catalog number: 1395-100)

7. Universal pipette tips, 200 μL (VWR, catalog number: 76322-516)

8. Universal pipette tips, 300 μL (VWR, catalog number: 76322-520)

9. Universal pipette tips, 1,250 μL (VWR, catalog number: 76322-138)

10. 50 mL polypropylene conical tubes (Geiner Bio-One, catalog number: 227 261)

Equipment

1. Centrifuge (Beckman Coulter Microfuge 20, catalog number: B31600)

2. Microplate reader (Molecular Devices, model: SpectraMax^® 190)

3. Vortex mixer (Vortex-Genie 2, model: G-560)

4. P20 Pipetman (Gilson, catalog number: F123600)

5. P200 Pipetman (Gilson, catalog number: F123601)

6. P1000 Pipetman (Gilson, catalog number: F123602)

7. 5–100 μL 8-channel electronic pipette (Eppendorf, Xplorer^®, catalog number: 4861000120)

8. 15–300 μL 8-channel electronic pipette (Eppendorf, Xplorer^®, catalog number: 4861000147)

9. 20–1,200 μL 8-channel electronic pipette (Eppendorf, Xplorer^®, catalog number: 4861000163)

Software and datasets

1. SoftMax^® Pro 4.8, by Molecular Devices

2. Microsoft Excel

Procedure

A. Preparation of protein samples

1. Assemble up to 16 protein samples to test per plate.

2. Transfer 200 μL of each protein sample to labeled centrifuge tubes.

3. Add 50 μL of 2.5× nitric acid mix to each centrifuge tube.

4. Vortex tubes, seal, and incubate at 65 °C overnight.

5. Centrifuge tubes at 20,000× g for 20 min.

6. Transfer 200 μL of the supernatant from each tube to wells on a 96-well holding plate.

B. Preparation of standard

1. Prepare tungsten or molybdenum standard solutions by diluting 100 μM stock solution to produce concentrations of 0, 1, 2.5, 5, 7.5, 10, 15, and 25 μM.

2. Add 50 μL of 2.5× nitric acid mix to each standard solution.

3. Vortex solutions and transfer 200 μL of each solution to wells on a 96-well holding plate.

C. Plate assay

1. Transfer 40 μL of each sample and each standard to four wells on an optical 96-well plate.

2. Transfer 160 μL of ethanol blank mix to 1 out of 4 wells for each sample/standard.

3. Transfer 160 μL of quercetin mix to the three remaining wells for each sample/standard.

4. Place a plate seal on the plate and shake in a plate shaker to mix.

5. Remove the seal (see Troubleshooting 1) and measure absorbance at 419 nm in a plate reader.

Data analysis

The absorbance from the ethanol blank well of each sample is subtracted from the absorbances of the other three wells. The absorbance of the 0 μM tungstate in the standard is used as a quercetin blank and is subtracted from all absorbance values. The standard absorbance values are plotted against tungstate concentration to produce a standard curve (Figure 2). From the standard curve, an equation is generated by which the tungstate concentration in each sample well can be calculated and averaged.

Figure 2. Example standard curves for molybdenum (blue) and tungsten (red). Data points were corrected for background absorbance and zeroed against their respective 0 μM standards. Results show an average of three replicates. Error bars show calculated standard deviation. Limit of detection calculated from equation, (3.3/σ_regression)/slope, is 0.9 μM for W and 0.8 μM for Mo.

Validation of protocol

Native tungsten-containing proteins from Eubacterium limosum were separated using SEC, and the resulting fractions were tested using both the quercetin-linked assay protocol and ICP-MS. The latter technique was used to determine the accuracy of the colorimetric assay. For both procedures, three replicates were analyzed, and the results were averaged. The W concentrations for the quercetin assay matched well with the ICP-MS W values and resulted in a similar shape of the W peak from the SEC samples (Figure 3). Specifically, for fractions 16 and 17, the quercetin assay results correspond to those from ICP-MS analysis, with the former yielding concentrations that were 89% and 120%, respectively, of those determined by ICP-MS (Table 1). A tungsten concentration below ~0.25 μM in any fraction as determined by ICP-MS was not reliably detected with the quercetin assay. Nevertheless, the results from the quercetin assay correctly identified W-containing proteins in column chromatography fraction. The correlation between the two protocols was quantified with the production of a linear regression model using the results from each protocol. The model shows an R² confidence value of 0.9373 (Figure S3).

Figure 3. Quercetin assay and ICP-MS results. Tungsten concentrations measured using the quercetin assay (red) and ICP-MS (black) [8] on SEC fractions containing the tungsten-storage protein, Tub. The ICP-MS protocol is described in Schut et al. [9]. For both the quercetin assay and ICP-MS, three replicates were averaged. Error bars show calculated standard deviation.

Table 1. Results for tungsten-containing fractions. Tungsten concentrations (μM) from fractions 15–20 calculated using the colorimetric quercetin assay and ICP-MS (control) along with percent recovery. For full sample set results, see Table S1.

General notes and troubleshooting

General notes

1. Due to the similarity in the chemical properties of molybdate and tungstate, this assay cannot distinguish between them. In samples with both molybdenum and tungsten, this assay will determine the sum of the concentrations of Mo and W in the sample.

2. Optimal W or Mo concentration range for the assay is from 1.0 μM to 20 μM. The detection limit calculated from the calibration curves (Figure 2) is ~0.8 μM for Mo and ~0.9 μM for W.

3. The presence of yellow-colored compounds that have absorbance near 420 nm in the samples may affect the calculated tungsten content. This issue can be mitigated by accounting for background interference using the ethanol blank mix.

Troubleshooting

Problem 1: Standard results produce poor linear regression line.

Possible cause: Bubbles in the wells have warped absorbance data.

Solution: After removing the plate seal, check to make sure no bubbles are present in the wells by gently tapping the plate against a flat surface.

Supplementary information

The following supporting information can be downloaded here:

1. Figure S1. The effect of protein, iron, and flavin on the molybdenum standard curve

2. Figure S2. The effect of protein, iron, and flavin on the tungsten standard curve

3. Figure S3. Correlation between quercetin assay and ICP-MS results

4. Table S1. Full 16-sample set results for tungsten determination

Acknowledgments

This work was funded by a grant from the National Institutes of Health (GM 136885).

This protocol is adapted from the method originally described by El-Sayed et al. [1].

Competing interests

The authors declare no competing interests.

Ethical considerations

No human or animal subjects were used in this protocol.

References

1. El-Sayed, A. A., Saad, E. A., Mohamed Ibrahime, B. M. and Tarek Mohamed Zaki, M. (2000). Flavonol Derivatives for Determination of Cr(III) and W(VI). Microchim Acta. 135: 19–27.
2. Bednar, A., Mirecki, J., Inouye, L., Winfield, L., Larson, S. and Ringelberg, D. (2007). The determination of tungsten, molybdenum, and phosphorus oxyanions by high performance liquid chromatography inductively coupled plasma mass spectrometery. Talanta. 72(5): 1828–1832.
3. Hagedoorn, P. L., van't Slot, P., van Leeuwen, H. P. and Hagen, W. R. (2001). Electroanalytical Determination of Tungsten and Molybdenum in Proteins. Anal Biochem. 297(1): 71–78.
4. Wang, J. (1992). Catalytic—adsorptive stripping voltammetric measurements of ultratrace levels of tungsten. Talanta. 39(7): 801–804.
5. Vandecasteele, J. P. and Burris, R. H. (1970). Purification and Properties of the Constituents of the Nitrogenase Complex from Clostridium pasteurianum. J Bacteriol. 101(3): 794–801.
6. Clark, L. J. and Axley, J. H. (1955). Molybdenum Determination of Soils and Rocks with Dithiol. Anal Chem. 27(12): 2000–2003.
7. Mohamed Zaki, M. T. and El-Sayed, A. A. (1992). Use of quercetin in an improved method for molybdenum determination in waste water, silicate rocks and diverse alloys. Mikrochim Acta. 106: 153–161.
8. Shao, N., Dayong, Z., Schut, G. J., Poole, F. L., Coffey, S. B., Donaghy, A. P., Putumbaka, S., Thorgersen, M. P., Chen, L., Rose, J., et al. (2024). Storage of the Vital Metal Tungsten in a Dominant SCFA-Producing Human Gut Microbe Eubacterium limosum and Implications for Other Gut Microbes.
9. Schut, G. J., Thorgersen, M. P., Poole, F. L., Haja, D. K., Putumbaka, S. and Adams, M. W. W. (2021). Tungsten enzymes play a role in detoxifying food and antimicrobial aldehydes in the human gut microbiome. Proc Natl Acad Sci USA. 118(43): e2109008118.