Electronic Letters to:
|
|
eLetters: Electronic letters published:
-
![[Read eLetter]](/icons/misc/sectionDOWN.gif)
The complexities of apoC-III metabolism: more exchangeable results?
- John S. Millar, Raymond C. Boston (4 July 2009)
|
|
|||
|
John S. Millar, Senior Research Investigator University of Pennsylvania, Raymond C. Boston
Send letter to journal:
jsmillar{at}mail.med.upenn.edu John S. Millar, et al.
|
We recently reported that part of the triglyceride-lowering effect of a potent and selective PPAR-alpha agonist was probably due to repression of apoC-III transcription leading to a reduced apoC-III production rate (PR) (1). This was based on studies showing that fibrates reduce plasma triglyceride levels, in part, by decreasing apoC-III expression and production (2-4). Chan et al. recently reported that fenofibrate has no effect on total apoC-III production in humans (5). They instead determined that fenofibrate lowers apoC III levels in plasma by increasing the fractional catabolic rate (FCR). We suggest that the use of an unnecessarily complex model in the analysis of their apoC-III kinetic data may have led to ambiguities in their interpretation of the data. Using their model (figure 1), we simulated plasma apoC-III kinetic data where a 23% change in total apoC-III concentration, similar to the reduction reported by Chan et al. was entirely due to either a reduced PR or an increased FCR (figure 2). If the apoC-III production rate is reduced by 23% by reducing the apoC-III pool size, while all parameters in the model remain fixed to baseline values, the resulting simulated tracer curve is identical to that obtained under baseline conditions. However, if the apoC-III FCR in the baseline model is increased so that the residence time (calculated as 1/FCR) is reduced by 23% then the resulting simulated tracer curve is noticeably higher than the baseline curve. Having established that the baseline and treatment tracer curves are superimposable if the change in apoC-III pool size is entirely due to a decrease in PR, we next sought to determine if a model solution that fit the treatment data existed that included an increase in the apoC-III FCR. This was done by fixing the FCR to the increased value as above and allowing other model parameters to adjust without any constraints. This unconstrained model could reasonably fit to the data under these conditions (figure 3), although the adjustable parameters in the model could not be estimated with any degree of confidence due to the large error (>50%) associated with the estimates. When the parameters in the plasma leucine model were fixed and the downstream apoC-III portion of the model was left unconstrained we found that the model could accommodate a visually acceptable, but not statistically acceptable, fit where either the apoC-III FCR or PR could both completely account for the change in apoC-III pool size. Again, the adjustable parameters in the model could not be estimated with any degree of confidence. We next attempted a similar exercise applying Berman’s minimal change postulate (6). In this case all model parameters for analysis of the apoC III tracer and pool size data from the treatment phase were constrained to the values obtained at baseline and only the apoC-III FCR and PR were adjustable. The result of this constrained analysis (figure 3) is that the FCR during the treatment phase is estimated with a high degree of confidence (<9% error) and is essentially identical to that obtained at baseline. The PR is reduced, accounting for the decrease in the apoC-III pool size. The model used in the analysis by Chan et al. may contain as many as 13 parameters that were allowed to adjust to give the best fit (5). Use of a simpler model to reduce the number of adjustable parameters may marginally decrease the goodness of fit of the model solution to the tracer data but would be expected to improve the confidence of the FCR and PR estimates. In lieu of using a simplified model, constraining the number of adjustable parameters that differ between the baseline and treatment phases of a repeated measures study will also reduce error resulting in more reliable estimates of parameters of interest. John S. Millar, PhD Raymond C. Boston, PhD University of Pennsylvania References: 1. Millar JS, Duffy D, Gadi R, Bloedon LT, Dunbar RL, Wolfe ML, Movva R, Shah A, Fuki IV, McCoy M, Harris CJ, Wang MD, Howey DC, Rader DJ. 2009. Potent and selective PPAR-alpha agonist LY518674 upregulates both ApoA-I production and catabolism in human subjects with the metabolic syndrome. Arterioscler Thromb Vasc Biol. 29:140-6. 2. Staels B, Vu-Dac N, Kosykh VA, Saladin R, Fruchart JC, Dallongeville J, Auwerx J. 1995. Fibrates downregulate apolipoprotein C-III expression independent of induction of peroxisomal acyl coenzyme A oxidase. A potential mechanism for the hypolipidemic action of fibrates. J Clin Invest.;95:705-12. 3. Schoonjans K, Staels B, Auwerx J. 1996. Role of the peroxisome proliferator-activated receptor (PPAR) in mediating the effects of fibrates and fatty acids on gene expression. J Lipid Res. 37:907-25. 4. Malmendier CL, Lontie JF, Delcroix C, Dubois DY, Magot T, De Roy L. 1989. Apolipoproteins C-II and C-III metabolism in hypertriglyceridemic patients. Effect of a drastic triglyceride reduction by combined diet restriction and fenofibrate administration. Atherosclerosis. 77:139-49. 5. Chan DC, Watts GF, Ooi EM, Ji J, Johnson AG, Barrett PH. 2008. Atorvastatin and fenofibrate have comparable effects on VLDL-apolipoprotein C-III kinetics in men with the metabolic syndrome. Arterioscler Thromb Vasc Biol. 28:1831-7. 6. Berman M. 1963. A postulate to aid in model building. J Theor Biol. 4:229-36. |
|||