Valentina Fermanelli
Mathematical Sciences, Chalmers University of Technology and the University of Gothenburg
Objectives: Apolipoproteins are a family of proteins attached to the surface of lipoproteins, the particles that facilitate the transportation of the lipids in the blood. The apolipoproteins are distributed among lipoprotein subtypes and are involved in the regulation of the lipoprotein kinetics. Apo-E is needed for the lipoprotein to be removed by the liver. ApoC-III inhibits the lipoprotein lipase. ApoA4 has an unknown function, but it has been supposed that it can be involved in regulating the food intake.
The plasma kinetics of lipoproteins are well studied. However, secretion and clearance of the apolipoproteins is less studied.
Biological aim: We aimed at investigating the relations between the production rate, elimination rate and concentrations of these three apolipoproteins. We aimed also at finding a correlation structure with additional biological parameters.
Modeling aim:We developed non-linear mixed effects models for each apolipoprotein as well as a combined model including all tree proteins. We wanted to check whether there would be substantial differences in the estimation of the parameters and whether the biological conclusion would be affected by the modeling approach.
Methods: Labelled leucine was subministered as bolus injection in 20 non-diabetic individuals. Peptide based proteomics was used to extract time series data of tracer enrichment in each apolipoprotein and GC/MS was used to measuring the tracer enrichment of free leucine in the blood. A linear ODE kinetic model was used to describe the process of formation of the apolipoproteins. Parameters was calculated separately for each apolipoprotein and in a combined model. The multicompartmental model for each apolipoprotein consists of 10 compartments: four compartments constitute the leucine kinetics, 5 compartments form a delay representing synthesis and secretion of the apolipoprotein, where the leucine is used as a building block of the apolipoprotein, and a last compartment, representing the apolipoprotein kinetics in the blood. The calculations of the kinetics parameters using a population approach were performed in Monolix. The correlation structure among the parameters has been analyzed using R.
Results: Biological results: The pool sizes and apoE, apoC-III and apoA4 correlated stronger with their secretion than with their clearance. Both apoE and apoC-III, which are known to be important for triglyceride rich lipoproteins metabolism, were closely linked to measures of plasma TG metabolism. ApoA4, on the other hand, were linked to HDL metabolism.
Comparison of the models: No biologically relevant difference could be detected for the parameters estimated with the two different models (combined estimation of the three apolipoproteins versus the separate estimation for each apolipoproteins). The correlation structure is also preserved.
Conclusions: Two models have been used to analyze longitudinal data of ApoE, ApoC3 and ApoA4. Both have yielded a similar correlation structure among the biologically relevant parameters. The correlation structure is biologically reasonable.
References:
[1] Feingold KR, Grunfeld C. Introduction to Lipids and Lipoproteins. [Updated 2018 Feb 2]. In: Feingold KR, Anawalt B, Boyce A, et al., editors. Endotext [Internet]. [2] Compartmental model of leucine kinetics in humans, C. Cobelli, M. P. Saccomani, P. Tessari, G. Biolo, L. Luzi, and D. E. Matthews, American Journal of Physiology-Endocrinology and Metabolism 1991 261:4, E539-E550.
[3] Dick C. Chan, Minh N. Nguyen, Gerald F. Watts, P. Hugh R. Barrett; Plasma Apolipoprotein C-III Transport in Centrally Obese Men: Associations with Very Low-Density Lipoprotein Apolipoprotein B and High-Density Lipoprotein Apolipoprotein A-I Metabolism, The Journal of Clinical Endocrinology & Metabolism, Volume 93, Issue 2, 1 February 2008, Pages 557–564, https://doi.org/10.1210/jc.2006-2676
[4] Effect of fenofibrate and atorvastatin on VLDL apoE metabolism in men with the metabolic syndrom, Esther M. M. Ooi, Theodore W. K. Ng, Gerald F. Watts, Dick C. Chan, P. Hugh R. Barrett, J Lipid Res. 2012 Nov; 53(11): 2443–2449. doi: 10.1194/jlr.P029223
[5] Monolix version 2018R1. Antony, France: Lixoft SAS, 2018. http://lixoft.com/products/monolix/
[6] RStudio Team (2015). RStudio: Integrated Development for R. RStudio, Inc., Boston, MA URL http://www.rstudio.com/.
Reference: PAGE 28 (2019) Abstr 9180 [www.page-meeting.org/?abstract=9180]
Poster: Methodology - Other topics