In 1998 Douek and colleagues described a novel approach to measuring thymic output using T-cell receptor excision circles (TRECs) measured in CD4+ and CD8+ T cells. TRECs are byproducts of T-cell receptor gene re-arrangements, whereby sections of excised DNA re-form as stable episomes that do not replicate with mitosis. TRECs present in T cells arising from the thymus are thus 'diluted out' upon T-cell division and can be used as a surrogate marker of thymic output. However, the interpretation of TREC number in lymphocyte subsets is confounded by the influences of T-cell proliferation, intracellular TREC degradation and T-cell death, and it has been criticised as a measurement of thymic function. Here, we describe a method for quantifying TRECs in whole blood. This revised assay provides a representation of 'total peripheral TRECs'. The effects of T-cell proliferation and cell death on the determined absolute TREC level are minimised, providing us with an improved estimation of thymic function over time. In a series of validation experiments our assay has been shown to have excellent intra-assay and inter-assay reproducibility and to generate results consistent with previously published work in healthy controls. We are now using this assay to determine thymic function in the autoimmune disease systemic lupus erythematosus to test our hypothesis that a defect in thymic function contributes to lupus pathogenesis.
To measure the absolute TREC number per unit volume of blood (as a representation of 'total peripheral TRECs'), the amount of DNA extracted from a volume of whole blood must be representative of 'total' DNA in that volume of blood. We have compared four methods of DNA extraction from whole blood for consistency of yield and purity. Figure 1a shows the mean DNA yield from parallel extractions (n = 10) performed after (a) 1 hour, (b) 6 hours and (c) 24 hours from collection for a single sample. The overall mean yield was 32.23 μg/ml ± 2.3 standard deviations (SDs). Two further, separate, samples were similarly prepared in parallel to confirm consistency – (d) 26.8 μg ± 1.97 SD and (e) 26.9 μg ± 2.23 SD (n = 10). To measure TRECs in samples we have used a quantitative real-time PCR (TAQMAN™) assay to amplify a known proportion of the extracted DNA. The total TREC number per sample is determined absolutely from a standard curve (derived from a plasmid containing TREC sequence) and final values are presented as TRECs per millilitre of whole blood. Figure 1a (top axis) shows that the TREC number is reproducibly calculated from parallel extractions (n = 5). The intra-assay coefficient of variation is 1.4% (range 0.8–2.5%) while the inter-assay co-efficient of variation is 2.4% (range 0.6–4.4%). The coefficient of variation is negatively correlated with TREC number (r2 = 0.83 and 0.66, respectively). Using this assay we have also shown that TRECs are reliably detectable in healthy controls (Fig. 1b) and show an age-related decline in keeping with the known reduction in thymic volume over time. TRECs are not detectable in RAJI (human B-cell lymphoma line) or SW48 (human colonic carcinoma cell line) (data not shown). We are now using this assay to determine the TREC number in the autoimmune disease systemic lupus erythematosus to test our hypothesis that a defect in thymic function contributes to lupus pathogenesis.
While measurement of TRECs in T cells and their subsets provides a representation of thymic function in the steady state, in the context of autoimmunity, where abundant auto-antigen provides a constant T-cell-activating stimulus, interpretation of the TREC number within lymphocyte subsets is confounded. By reliably and reproducibly estimating total peripheral TRECs we aim to minimise the impact of T-cell turnover on TREC number in order to gain a better understanding of thymic function in systemic lupus erythematosus.
Figure 1. (a) Lower axis: serial DNA extractions (see text). Bars represent mean of 10 ± standard deviation. Upper axis: T-cell receptor excision circles (TRECs) per millitre of blood from the same samples (mean ± deviation). (b) TRECs per millitre of whole blood declines with age over a range of ~2 logs from birth.
This research is funded by an arc Clinical Research Fellowship awarded to ARL (Grant number 14490)