Determination of Se-Met in human serum

Project Description

Within WP2 of the project TRACEBIOACTIVITY measurement methods for the SI traceable determination of selenomethionine (Se-Met) in the human serum were developed and have to be validated by demonstrating that Se-Met can be measured with sufficient comparability in this matrix. Therefore an international intercomparison is organised for the analysis of Se-Met in a human serum. National metrology laboratories as well as other expert laboratories are invited to participate in this intercomparison.

Final Report 2011-07-04

This international intercomparison was organised to validate the measurement procedures for the SI traceable determination of selenomethionine (Se-Met) in human serum developed within WP2 of the project TRACEBIOACTIVITY.

The intercomparison was carried out using the matrix reference material BCR-637, where the total content of Se is certified but not the content of Se-Met, as the sample material and Se-Met enriched in 76Se as spike material for species-specific isotope dilution mass spectrometry (SS-IDMS).

For the quantification of Se-Met in the serum LNE used single ID-ICP-MS including an inverse step to characterise the spike while LGC and PTB used double ID-ICP-MS. Expanded uncertainties associated with the results for the measurements at all three NMIs were below 8% (k = 2), which is a remarkable achievement considering the difficulties arising from the quantification of Se species in serum at such low levels. The assessment of the expanded uncertainty for the measurement of Se-Met in serum with HPLC-ICP-IDMS was carried out on the basis of the GUM recommendations.

At LNE 5 replicates each of 5 independent samples were analysed for the determination of the uncertainty. LNE found a relative expanded uncertainty of approx. 5 %. The repeatability in the determination of the mass discrimination factor was the most predominant contribution with 20-40 % in case of direct IDMS and 60-70 % for the inverse step.

At LGC, the determination of Se-Met in serum was carried out using hydrolysis with proteolytic enzymes followed by double SS-IDMS with HPLC-ICP-MS. The concentration of the natural standard, the matching of the isotope ratios of sample and calibration blends (exact matching was found to be necessary for the low ppb levels of serum Se-Met) and the blend to blend variation were the major sources of contribution to the overall uncertainty. Having a natural Se-Met standard of a known purity has also been proven essential to achieve traceability in the determination of Se-Met.

At PTB, the same measurement procedure as at LGC was used. As the purity of the Se-Met standard was determined using quantitative NMR resulting in a smaller uncertainty for the purity of the reference Se-Met, the major contributions to the uncertainty arise from the sample preparation, as well as the measured isotope ratios. The results for the uncertainty estimation emphasizes the importance to control the conditions for the enzymatic hydrolysis carefully.

The mass fraction of Se-Met (as Se) obtained by PTB, LGC, LNE and TUBITAK-UME in the BCR-637 human serum were 12.5 +/- 0.57 µg kg-1 Se; 13.6 +/- 1.1 µg kg-1 Se, 13.1 +/- 0.66 µg kg-1 Se and 13.1 +/- 0.84 µg kg-1 Se, respectively. These mass fractions with their expanded uncertainties (k = 2) were found to agree reasonably well.

A joint scientific paper describing the uncertainty estimation, measurement challenges and comparative data by these NMIs is in preparation