Ultra‐performance liquid chromatography‐tandem mass spectrometry method for analysis of tau in human cerebrospinal fluid without the need of immunocapture

Magdalena Korecka,Andreas Leinenbach,Johan Gobom,Vincent Delatour,Francois Becher,Kaj Blennow,Henrik Zetterberg,Josef Pannee,Katarina Hofving,Leslie M. Shaw
DOI: https://doi.org/10.1002/alz.040373
2020-12-01
Abstract:Abstract Background Reliable tau protein measurement in CSF is required for detection of Alzheimer’s disease pathology when combined with Aβ42/40 measurement. Tau is present at a very low concentration in CSF, has 6 different isoforms and many forms with post‐translational modification including methylation, glycosylation and phosphorylation. Immunoassays are sensitive enough to measure low tau concentration however poor linearity and lack of specificity in regards to tau’s isoforms have been reported. Thus, mass spectrometry is a candidate alternative technology. Additionally, the possibility of avoiding immunocapture would be important for a protein with multiple isoforms. Method Here we describe our efforts to develop an LC‐MS/MS (triple quadrupole) assay for quantification of tau in CSF. Sample preparation includes selective protein precipitation, neutralization and on‐filter digestion with trypsin without the need of immunocapture. Following trypsinization the peptide mixture is washed/centrifuged through the filters, injected onto the analytical column and eluted with water/ acetonitrile/0.1% formic acid gradient. For quantification we use the tryptic peptide ‐ GAAPPGQK [(156‐163, m/z‐ 363.26/526.28 (quantifier ion)], which is present in all tau isoforms and does not contain a phosphorylation site, and whose quantity is directly proportional to tau concentration. The same peptide labelled with 15N is used as the internal standard (IS) (m/z‐ 368.19/267.14 and 368.19/533.28, quantifier ions). Result The method is still under development however several parameters have been established: surrogate matrix (0.5% serum in water), amount of IS added at the beginning of sample preparation (4ng/mL), lower (350pg/mL) and upper (30000pg/mL) limit of measuring interval, analytical recovery of human peptide from the filters (67‐83%), duplicates precision (5.0 3.9%). No background peaks found at the retention time of the analyte/IS from a matrix‐matched double blank sample confirms good assay specificity. Collection of a full set of imprecision/accuracy data using surrogate matrix and human CSF are underway. Conclusion The assay will be assessed for tau quantification in human CSF by evaluating each parameter of the full validation process. This newly developed and validated assay will be important for direct measurement of tau in CSF samples, and may be applied as a candidate reference measurement procedure for CSF tau in standardization projects on this biomarker.
What problem does this paper attempt to address?