Limits...
Quantitative variability of 342 plasma proteins in a human twin population.

Liu Y, Buil A, Collins BC, Gillet LC, Blum LC, Cheng LY, Vitek O, Mouritsen J, Lachance G, Spector TD, Dermitzakis ET, Aebersold R - Mol. Syst. Biol. (2015)

Bottom Line: Because the twin study design provides a natural opportunity to estimate the relative contribution of heritability and environment to different traits in human population, we applied here the highly accurate and reproducible SWATH mass spectrometry technique to quantify 1,904 peptides defining 342 unique plasma proteins in 232 plasma samples collected longitudinally from pairs of monozygotic and dizygotic twins at intervals of 2-7 years, and proportioned the observed total quantitative variability to its root causes, genes, and environmental and longitudinal factors.The data further strongly suggest that the plasma concentrations of clinical biomarkers need to be calibrated against genetic and temporal factors.These results therefore have immediate implications for the effective design of blood-based biomarker studies.

View Article: PubMed Central - PubMed

Affiliation: Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, Zurich, Switzerland liu@imsb.biol.ethz.ch aebersold@imsb.biol.ethz.ch.

No MeSH data available.


Related in: MedlinePlus

Dissection of the plasma protein-level variabilityA Histograms of contribution percentage of each biological component (red: heritability; light blue: common environment; dark blue: individual environment; yellow: longitudinal effects; gray: unexplained fraction) determined by a linear mixed model based on the longitudinal twin design. Selected protein names are shown for those clinically assayed proteins with the most heritable levels, for example, APOA, plasma protease C1 inhibitor (IC1), complement C4-A (CO4A), fibrinogen alpha chain (FIBA) and pancreatic triacylglycerol lipase (LIPP), and for the most variable proteins among the cohort, for example, SAA1, APOA, pregnancy zone protein (PZP), trifunctional enzyme subunit alpha (ECHA) and C-reactive protein (CRP).B Examples of protein abundance correlations between MZ and DZ pairs. Fibrinogen beta chain (FIBB), Serum paraoxonase/arylesterase 1 (PON1).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4358658&req=5

fig03: Dissection of the plasma protein-level variabilityA Histograms of contribution percentage of each biological component (red: heritability; light blue: common environment; dark blue: individual environment; yellow: longitudinal effects; gray: unexplained fraction) determined by a linear mixed model based on the longitudinal twin design. Selected protein names are shown for those clinically assayed proteins with the most heritable levels, for example, APOA, plasma protease C1 inhibitor (IC1), complement C4-A (CO4A), fibrinogen alpha chain (FIBA) and pancreatic triacylglycerol lipase (LIPP), and for the most variable proteins among the cohort, for example, SAA1, APOA, pregnancy zone protein (PZP), trifunctional enzyme subunit alpha (ECHA) and C-reactive protein (CRP).B Examples of protein abundance correlations between MZ and DZ pairs. Fibrinogen beta chain (FIBB), Serum paraoxonase/arylesterase 1 (PON1).

Mentions: We took advantage of the longitudinal twin design and utilized a linear mixed model (Nicholson et al, 2011) to systematically partition the variance observed for 342 protein levels. The phenotype variance was decomposed into the biological variance (heritable, shared/individual environmental and longitudinal contributing factors) and the unexplained variance (Fig3A). Even though the twins are adult females who normally do not live in the same household, they generally share more habits and lifestyles than non-twin siblings, which are reflected by the term “shared/ common environment”. The unexplained variance generally accounts for 50% of the detected variance in our data and can be associated with variance not reflected by the experimental design (e.g. short-term protein concentration fluctuations, diet effects, etc.) and technical/experimental variations. The mean proportion of heritability, common environment, individual environment and longitudinal process across all the proteins were estimated to be 13.6, 10.8, 11.6 and 13.6%, respectively, of the total phenotypic variance, that is, 25.4, 20.8, 23.0 and 30.8%, respectively, of the biologically stable variance (the total fraction that is explained by the four biological variance origins under our experimental design). The determined heritability (h2) showed good agreement with the protein abundance correlations in MZ and DZ twins. Examples of apolipoprotein(A), fibrinogen beta chain and serum paraoxonase/arylesterase 1 are illustrated in Fig3B. Notably, heritability and common environment are sometimes combined in twin studies as a “family component”, which explains almost half (i.e. 46.2%) of the biological variance in our results. Table1 lists the five proteins most strongly affected by each biological component (see Supplementary Table S3 for variance decomposition results of all proteins). For example, the plasma level of apolipoprotein(A) was determined to be the most strongly heritable (h2 = 0.6633), a finding that is consistent with previous reports (Boerwinkle et al, 1992; Lopez et al, 2008; Cenarro et al, 2014).


Quantitative variability of 342 plasma proteins in a human twin population.

Liu Y, Buil A, Collins BC, Gillet LC, Blum LC, Cheng LY, Vitek O, Mouritsen J, Lachance G, Spector TD, Dermitzakis ET, Aebersold R - Mol. Syst. Biol. (2015)

Dissection of the plasma protein-level variabilityA Histograms of contribution percentage of each biological component (red: heritability; light blue: common environment; dark blue: individual environment; yellow: longitudinal effects; gray: unexplained fraction) determined by a linear mixed model based on the longitudinal twin design. Selected protein names are shown for those clinically assayed proteins with the most heritable levels, for example, APOA, plasma protease C1 inhibitor (IC1), complement C4-A (CO4A), fibrinogen alpha chain (FIBA) and pancreatic triacylglycerol lipase (LIPP), and for the most variable proteins among the cohort, for example, SAA1, APOA, pregnancy zone protein (PZP), trifunctional enzyme subunit alpha (ECHA) and C-reactive protein (CRP).B Examples of protein abundance correlations between MZ and DZ pairs. Fibrinogen beta chain (FIBB), Serum paraoxonase/arylesterase 1 (PON1).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4358658&req=5

fig03: Dissection of the plasma protein-level variabilityA Histograms of contribution percentage of each biological component (red: heritability; light blue: common environment; dark blue: individual environment; yellow: longitudinal effects; gray: unexplained fraction) determined by a linear mixed model based on the longitudinal twin design. Selected protein names are shown for those clinically assayed proteins with the most heritable levels, for example, APOA, plasma protease C1 inhibitor (IC1), complement C4-A (CO4A), fibrinogen alpha chain (FIBA) and pancreatic triacylglycerol lipase (LIPP), and for the most variable proteins among the cohort, for example, SAA1, APOA, pregnancy zone protein (PZP), trifunctional enzyme subunit alpha (ECHA) and C-reactive protein (CRP).B Examples of protein abundance correlations between MZ and DZ pairs. Fibrinogen beta chain (FIBB), Serum paraoxonase/arylesterase 1 (PON1).
Mentions: We took advantage of the longitudinal twin design and utilized a linear mixed model (Nicholson et al, 2011) to systematically partition the variance observed for 342 protein levels. The phenotype variance was decomposed into the biological variance (heritable, shared/individual environmental and longitudinal contributing factors) and the unexplained variance (Fig3A). Even though the twins are adult females who normally do not live in the same household, they generally share more habits and lifestyles than non-twin siblings, which are reflected by the term “shared/ common environment”. The unexplained variance generally accounts for 50% of the detected variance in our data and can be associated with variance not reflected by the experimental design (e.g. short-term protein concentration fluctuations, diet effects, etc.) and technical/experimental variations. The mean proportion of heritability, common environment, individual environment and longitudinal process across all the proteins were estimated to be 13.6, 10.8, 11.6 and 13.6%, respectively, of the total phenotypic variance, that is, 25.4, 20.8, 23.0 and 30.8%, respectively, of the biologically stable variance (the total fraction that is explained by the four biological variance origins under our experimental design). The determined heritability (h2) showed good agreement with the protein abundance correlations in MZ and DZ twins. Examples of apolipoprotein(A), fibrinogen beta chain and serum paraoxonase/arylesterase 1 are illustrated in Fig3B. Notably, heritability and common environment are sometimes combined in twin studies as a “family component”, which explains almost half (i.e. 46.2%) of the biological variance in our results. Table1 lists the five proteins most strongly affected by each biological component (see Supplementary Table S3 for variance decomposition results of all proteins). For example, the plasma level of apolipoprotein(A) was determined to be the most strongly heritable (h2 = 0.6633), a finding that is consistent with previous reports (Boerwinkle et al, 1992; Lopez et al, 2008; Cenarro et al, 2014).

Bottom Line: Because the twin study design provides a natural opportunity to estimate the relative contribution of heritability and environment to different traits in human population, we applied here the highly accurate and reproducible SWATH mass spectrometry technique to quantify 1,904 peptides defining 342 unique plasma proteins in 232 plasma samples collected longitudinally from pairs of monozygotic and dizygotic twins at intervals of 2-7 years, and proportioned the observed total quantitative variability to its root causes, genes, and environmental and longitudinal factors.The data further strongly suggest that the plasma concentrations of clinical biomarkers need to be calibrated against genetic and temporal factors.These results therefore have immediate implications for the effective design of blood-based biomarker studies.

View Article: PubMed Central - PubMed

Affiliation: Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, Zurich, Switzerland liu@imsb.biol.ethz.ch aebersold@imsb.biol.ethz.ch.

No MeSH data available.


Related in: MedlinePlus