A novel pooled-sample multiplex luminex assay for high-throughput measurement of relative telomere length

Abstract

Objectives: Relative telomere length (RTL) is a potential biomarker of aging and risk for chronic disease. Previously, we developed a probe-based RTL assay on Luminex platform, where probes for Telomere (T) and reference gene (R) for a given DNA sample were tested in a single well. Here, we describe a method of pooling multiple samples in one well to increase the throughput and cost-effectiveness.

Methods: We used four different microbeads for the same T-probe and four different microbeads for the same R-probe. Each pair of probe sets were hybridized to DNA in separate plates and then pooled in a single plate for all the subsequent steps. We used DNA samples from 60 independent individuals and repeated in multiple batches to test the precision.

Results: The precision was good to excellent with Intraclass correlation coefficient (ICC) of 0.908 (95% CI 0.856-0.942). More than 67% of the variation in the RTL could be explained by sample-to-sample variation; less than 0.1% variation was due to batch-to-batch variation and 0.3% variation was explained by bead-to-bead variation. We increased the throughput of RTL Luminex assay from 60 to 240 samples per run. The new assay was validated against the original Luminex assay without pooling (r = 0.79, P = 1.44 × 10^-15 ). In an independent set of samples (n = 550), the new assay showed a negative correlation of RTL with age (r = -0.41), a result providing external validation for the method.

Conclusion: We describe a novel high throughput pooled-sample multiplex Luminex assay for RTL with good to excellent precision suitable for large-scale studies.

Figure 1

Schematic representation of the pooled multiplex assay for RTL. Upper pane shows the four 96-well plates, each containing 60 independent test DNA samples (marked by blue color) and the rest of the wells (marked by grey color) were used for “standard” and “control” samples. Samples in each of these four plates were hybridized with designated bead sets. Four different beads for TEL and four different beads for ALK are shown in different colors. The outer surface of all the four beads for TEL contain the same capture probe for TEL (shown in red) and the outer surface of all the four beads for ALK contain the same capture probe for ALK (shown in dark blue). Four different DNA samples in corresponding wells (e.g., well A12) of four plates are shown in different colors as horizontal lines attached to the beads. The lower panel shows the pooling of corresponding wells of all the four plates after hybridization. So in this final reaction plate, each well contains eight different beads. Four have captured TEL region from four different DNA samples and four other beads have captured ALK region from those four DNA samples. This final reaction plate was processed for signal amplification using pre-amplifiers, amplifiers and then StreptAvidin PhycoErythrin (SAPE) as described in method section and then read in Luminex 200 scanner for fluorescence.

Figure 2

Cartoon to illustrate the concept of TQI measurement. Four wells containing Standard and three samples are shown. Each line represents a DNA molecule. Difference in DNA input quantity is shown by the number of molecules in each well. TEL repeats are represented by star (red) and ALK region is represented by round (blue) marker. Median Fluorescent Intensity (MFI) is proportionate to the number of repeats (see standard and sample#1) as well as input quantity (see standard and sample#3). Quantification of TEL and ALK is done in comparison to the standard. TQI is not affected by the input quantity of DNA (see standard and sample#3). Sample#2 shows an example of mixture of DNA molecule of varying number of TEL repeats.

Figure 3

Reproducibility of the new high throughput assay for RTL measurement in a large-scale study. A total of 233 samples were run in duplicates as a quality control measure. The first measurement of TQI is shown on x-axis and the second measurement on y-axis.

Figure 4

Validation of the new high throughput pooled-sample multiplexed assay against the original Luminex assay without sample-pooling. We compared the two measurements of TQI from the same samples (one measurement from the old lower throughput assay shown in x-axis and one from the new higher throughput assay with same bead set and the same DNA as “standard” to generate standard curves for quantification, shown on y-axis).

Figure 5

External validation of the new high throughput pooled-sample multiplexed assay. The blood leucocytes RTL (TQI) shows significant negative correlation with age.