Skip to main content

Pre-analytical processing of plasma and serum samples for combined proteome and metabolome analysis.

Gegner HM, Naake T, Dugourd A, Müller T, Czernilofsky F, Kliewer G, Jäger E, Helm B, Kunze-Rohrbach N, Klingmüller U, Hopf C, Müller-Tidow C, Dietrich S, Saez-Rodriguez J, Huber W, Hell R, Poschet G, Krijgsveld J

Publication: Front Mol Biosci (2022) • Vol 9 • Page 961448PMID: 36605986View on PubMed

Purpose

The purpose of this paper was to compare the metabolome and proteome of plasma obtained from blood after centrifugation delays of up to 24 h at room temperature and 4°C and of serum obtained after centrifugation delays of up to 24 h at room temperature. The metabolome and proteomes of plasma obtained with different centrifugation speeds (2,000 versus 4,000 g) were also compared.

Conclusion

A total of 497 metabolites and 572 proteins were identified. No differences in the metabolome or proteome were observed between plasma obtained by centrifugation at 2,000 g and 4,000 g. Clustering was by patient in principal component analysis (PCA) of both proteomic and metabolomic data, but the separation was more pronounced for the metabolomic than the proteomic analysis. It was possible to classify specimens by patient using partial least square-discriminant analysis (PLS-DA) or sparse PLS-DA (sPLS-DA) with a low classification error. Classification by PLS-DA, based on pre-centrifugation storage duration and temperature, had a higher error rate than classification based on patient.  After FDR correction, sPLS-DA analysis of plasma identified 6 metabolites and 15 proteins affected by pre-centrifugation storage duration, 4 metabolites affected by pre-centrifugation temperature, and 16 metabolites affected by the combination of pre-centrifugation storage duration and temperature.  Importantly, larger differences in the intensities of metabolites in plasma were found when specimens were stored at room temperature than 4°C, and larger differences in protein intensities were observed when specimens were stored at 4°C. In serum, 16 metabolites were found to be affected by pre-centrifugation storage duration after false discovery rate (FDR) correction, demonstrating a progressive increase in the magnitude of change.

As one might expect, signatures that included 20 proteins and metabolites most affected by either storage temperature or duration scored high when evaluated in the same specimen cohort used for the signature’s development.  When applied to another cohort of specimens, the highest signature scores matched the appropriate experimental condition.  Erythrocyte contamination markers were found to be highest in plasma when blood was stored on ice rather than at room temperature prior to plasma isolation. In serum, coagulation contamination was high at baseline; however, while platelet markers increased with delayed centrifugation, coagulation markers did not.

Studies (1)

Purpose

The purpose of this study was to compare the metabolome and proteome of plasma obtained from blood after centrifugation delays up to 24 h at room temperature and 4°C and of serum obtained after centrifugation delays of up to 24 h at room temperature.  Blood was collected from six healthy fasting male volunteers into EDTA S-Monovette tubes and serum S-Monovette tubes. Blood was stored for 0, 2, 4, and 8 h at room temperature (24°C) and 4°C (plasma tubes only) before separation of plasma/serum.  Plasma was separated from tubes stored for 0 h at either temperature by centrifugation at 4000 g for 10 min, and from specimens stored for 0, 2, 4 and 8 h at either temperature by centrifugation at 2000 g for 10 min. Serum was obtained by centrifugation at 400 g for 10 min.  Plasma and serum aliquots were snap-frozen in liquid nitrogen and stored at -80°C. The metabolome was analyzed using the Biocrates MxP Quant 500 Kit with chromatographic separation on a UHPLC-I Class Plus System followed by mass spectrometry on a QTRAP 6500+ System. Proteins were quantified using the bicinchoninic acid assay (BCA) and analyzed using a nanoElute liquid chromatography system coupled with a Time of Flight (TOF) Pro mass spectrometer in data independent acquisition (DIA) mode.

Summary of Findings

A total of 497 metabolites and 572 proteins were identified. Linear models revealed that a large number of features differed between donors. No differences in the metabolome or proteome were identified between plasma obtained by centrifugation at 2000 g or 4000 g. Clustering was by patient in principal component analysis (PCA) of both proteomic and metabolomic data, but the separation was more pronounced for the metabolomic than the proteomic analysis. It was possible to classify specimens by patient using partial least square-discriminant analysis (PLS-DA) or sparse PLS-DA (sPLS-DA) with a low classification error.  When PLS-DA was used to classify patients based on pre-centrifugation storage duration and temperature, there was a higher error rate than when classification was by patient. sPLS-DA analysis identified a list of features that changed with storage duration or temperature.  A total of 41 metabolites and 64 proteins in plasma were affected by pre-centrifugation storage duration, but only 6 and 15, respectively, remained affected after FDR correction.  A total of 17 metabolites and 20 proteins in plasma were affected by pre-centrifugation storage temperature, but only 4 proteins and no metabolites were affected after FDR correction. The combination of pre-centrifugation storage duration and temperature significantly affected 33 metabolites and 60 proteins in plasma, of which 16 metabolites and no proteins were affected after FDR correction.  Importantly, the magnitude of the absolute change of transformed intensities increased with increasing storage duration, with larger differences in the intensities of metabolites found when stored at room temperature than 4°C and larger differences in the protein intensities when stored at 4°C. In serum, 43 metabolites and 99 proteins were found to be affected by pre-centrifugation storage duration, but only 16 metabolites were found to be affected after FDR correction. In serum, a progressive increase in the magnitude of the metabolite change occurred with storage duration.

As one might expect, signatures that included 20 proteins and metabolites most affected by either storage temperature or duration scored high when evaluated in the same specimen cohort used for the signature’s development. Importantly, when signatures were applied to another specimen cohort the highest scores matched the appropriate experimental condition.  Erythrocyte contamination markers were found to be highest when blood was stored on ice prior to plasma isolation compared to storage at room temperature. In serum, coagulation contamination was high at baseline; however, platelet markers increased with delayed centrifugation while coagulation markers did not.

  • Biospecimens

    FluidBlood
    FluidPlasma
    FluidSerum
  • Preservation Types

    Frozen
  • Diagnoses

    Normal
  • Platforms

    ProteinLC-TOF-MS
    Small moleculeLC-MS or LC-MS/MS
    PeptideLC-TOF-MS
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen Aliquots and ComponentsBlood and blood products
    Plasma
    Serum
    Biospecimen Aliquots and ComponentsCentrifugation
    Centrifugation delays investigated
    Multiple speeds compared
    StorageStorage duration
    0 h
    2 h
    24 h
    4 h
    8 h
    StorageStorage temperature
    4°C
    Room temperature (24°C)