Date Published: March 28, 2017
Publisher: Public Library of Science
Author(s): Gabriel Dumitrescu, Andras Komaromi, Olav Rooyackers, Maria Klaude, Christina Hebert, Jan Wernerman, Åke Norberg, Reza Khodarahmi.
The possibility of using two different isotopomers, for the incorporation of isotopically labeled amino acids, was explored to enable longitudinal studies of de novo synthesis of two export liver proteins, albumin and fibrinogen. The agreement of the synthesis rates between the two different labels was evaluated along with the reproducibility of repeated experiments using different time intervals. Healthy volunteers were studied in a standardized fed state. Protocol A (n = 10) involved two measurements 48 hours apart. Protocol B (n = 6) involved three measurements at baseline and five hours and then seven days after the initial measurement. De novo synthesis of albumin and fibrinogen by the incorporation of D5-phenylalanine or D8-phenylalanine were measured using the flooding dose technique. Albumin and fibrinogen were isolated from plasma using standard techniques. Fractional and absolute synthesis rates were calculated. Repeated measurements employing the two isotoptomers showed good agreement for albumin fractional synthesis rate after 48 hours (p = 0.92) and after 7 days (p = 0.99), with a coefficient of variation of 5.9% when using the same isotopic label. For fibrinogen, the coefficient of variation for the fractional synthesis rate employing the same isotopic label was 16.6%. Repeated measurements after 48 hours and seven days showed less agreement although there was no statistical difference (P = 0.32 and P = 0.30 respectively). Repeated measurement after five hours showed a statistical significant difference for the fractional synthesis rate of fibrinogen (p = 0.008) but not for albumin (p = 0.12). Repeated measurements of albumin de novo synthesis more than 48 hours apart show acceptable agreement using either one or two different isotopic labels. For fibrinogen the larger intra-individual scatter necessitates larger study groups to detect changes in longitudinal studies. Repeated measurements within 48 hours need to be validated further.
The changes in plasma concentrations of several liver export proteins in relation to surgical procedures, inflammation and infectious states are well known. However, the mechanisms behind these changes are less well understood. Altered synthesis rates or degradation rates are obvious possibilities as well as changes in plasma volume and a redistribution of the proteins into different compartments.
The characteristics of the healthy volunteers are given in Table 1. In addition characterization and some results related to gender are given in the electronic supplement in S1 Table.
The goal of the present study was to use two different isotopic labels of phenylalanine to determine the synthesis rates of the liver export proteins albumin and fibrinogen. The results show that measurements 48 hours or more apart give comparable results on a group level, while measurements five hours apart show interference that make this shorter period unsuitable for repeated measurements. The similar patterns of results for the two different plasma proteins suggests there is a sufficient basis for this conclusion. Whether a time period somewhere in between 5 and 48 hours would be sufficient to allow for reliable repeated measurements is outside the scope of this study. In longitudinal studies with repeated measurements, an interval of more than 48 hours is recommended, in particular when more than two measurements are planned. For future studies, there may be a choice between less frequent sampling, or to investigate parallel groups at different times. The latter alternative will of course necessitate a much larger number of involved subjects. The coefficient of variation for repeated measurements one week apart is most likely attributable to a combination of physiologic variability together with analytical and protocol related factors, such as sample preparation, response to feeding etc. This underlines the need to standardize these factors as much as possible when repeated measurements for paired comparisons are used. The inter-individual variability has been highlighted by several investigators related to feeding status, body composition, age, gender, etc [18–20]. The influence of gender in a post hoc analysis of data in the present study, presented in S1 Table, showed a higher value of P-fibrinogen in females, accompanied by a higher absolute synthesis rate. Among possible confounders, the females were older [21–23] and had a lower hematocrit, while body mass index and plasma volume per kg body weight were similar in both gender groups.