Control of (pre)-analytical aspects in immunoassay measurements of metabolic hormones in rodents

Abstract

The measurement of circulating hormones by immunoassay remains a cornerstone in preclinical endocrine research. For scientists conducting and interpreting immunoassay measurements of rodent samples, the paramount aim usually is to obtain reliable and meaningful measurement data in order to draw conclusions on biological processes. However, the biological variability between samples is not the only variable affecting the readout of an immunoassay measurement and a considerable amount of unwanted or unintended variability can be quickly introduced during the pre-analytical and analytical phase. This review aims to increase the awareness for the factors 'pre-analytical' and 'analytical' variability particularly in the context of immunoassay measurement of circulating metabolic hormones in rodent samples. In addition, guidance is provided how to gain control over these variables and how to avoid common pitfalls associated with sample collection, processing, storage and measurement. Furthermore, recommendations are given on how to perform a basic validation of novel single and multiplex immunoassays for the measurement of metabolic hormones in rodents. Finally, practical examples from immunoassay measurements of plasma insulin in mice address the factors 'sampling site and inhalation anesthesia' as frequent sources of introducing an unwanted variability during the pre-analytical phase. The knowledge about the influence of both types of variability on the immunoassay measurement of circulating hormones as well as strategies to control these variables are crucial, on the one hand, for planning and realization of metabolic rodent studies and, on the other hand, for the generation and interpretation of meaningful immunoassay data from rodent samples.

Keywords: ELISA; assay performance; assay validation; endocrine factors; immunoassay; insulin; metabolism; mice; multiplex; plasma hormones; variability.

Figure 1

Example illustrating the effect of alternating the blood sampling site within short intervals (tail vein puncture vs retrobulbar sinus) on measured plasma insulin concentrations in five anesthetized C57BL/6J mice (A). Blood sampling sites were alternated from mouse to mouse to avoid the introduction of a systematic error. (B) Shows the expected absence of effects on measured plasma insulin concentrations in another cohort of anesthetized C57BL/6J mice when blood was sampled within short intervals from the same site (in this case from the retrobulbar sinus).

Figure 2

Comparison of measured plasma insulin concentrations in samples collected from the tail vein with or without prior isoflurane anesthesia in C57BL/6J mice. The anesthetic procedure before blood sampling was alternated from mouse to mouse to avoid the introduction of a systematic error. In this example, measured plasma insulin concentrations were significantly (P < 0.05) lower when blood was collected under isoflurane anesthesia but showed a tight and significant non-parametric correlation.

Figure 3

Plasma insulin was measured in the same 38 mouse samples by two different immunoassays (multiplex and singleplex). Measured concentrations showed a significant variability depending on the immunoassay. (A) Depicts the results using a Passing-Bablok regression and (B) shows the corresponding Bland-Altman plot for this assay comparison.