FAST-SeqS: a simple and efficient method for the detection of aneuploidy by massively parallel sequencing

Abstract

Massively parallel sequencing of cell-free, maternal plasma DNA was recently demonstrated to be a safe and effective screening method for fetal chromosomal aneuploidies. Here, we report an improved sequencing method achieving significantly increased throughput and decreased cost by replacing laborious sequencing library preparation steps with PCR employing a single primer pair designed to amplify a discrete subset of repeated regions. Using this approach, samples containing as little as 4% trisomy 21 DNA could be readily distinguished from euploid samples.

Figure 1. Comparison of observed and predicted distributions of FAST-SeqS amplification products.

(A) A density plot of the expected distribution of fragment lengths, with peaks at 124 and 142 bp. (B) A density plot of the actual tag counts obtained in eight normal plasma DNAs. The 124 bp fragments are preferentially amplified compared to the 142 bp fragments, likely due to an amplification bias towards smaller fragments. Inset: polyacrylamide gel of a representative FAST-SeqS sequencing library. Note: the amplification products contain an additional ∼120 bp of flanking sequence to facilitate sequencing (Table S2). (C) The average representation of the most frequently observed L1 retrotransposon subfamilies in eight normal plasma samples. Roughly 97% of uniquely aligning tags arise from positions representing only seven L1 retrotransposon subfamilies. (D) A detailed examination of the average number of observed positions per chromosome from eight normal plasma DNAs compared with the number predicted by RepeatMasker for each of the seven L1 retrotransposon subfamilies noted in (C). Error bars in each panel depict the range.

Figure 2. Demonstration of FAST-SeqS reproducibility among different samples, sequencing instruments, and sequencing depth.

FAST-SeqS was performed on eight normal plasma DNA samples, their corresponding matched peripheral blood white blood cell (WBC) DNA, and on the splenic or WBC DNA of an additional eight unrelated individuals. The eight samples within each experiment constituted the reference group (see ‘Materials and Methods’ section) from which the plotted z-scores were calculated. No autosome in any sample had a z-score outside the range of −3.0 and 3.0 (dotted lines). Despite 3-fold less sequencing of the splenic or WBC samples, the z-scores (range: −2.2 to 2.1) were similar to those obtained from the plasma (range: −2.1 to 1.9) and matched WBC DNA samples (range: −2.2 to 1.9).

Figure 3. Accurate discrimination of euploid DNA samples from those containing trisomic DNA.

(A) Comparison of z-scores from patients with trisomy 21 (n = 4), trisomy 18 (n = 2), and trisomy 13 (n = 1) with eight normal spleen or peripheral blood white blood cell (WBC) DNAs. The z-scores displayed represent the relevant chromosome for the comparison. The maximum z-score observed for any of the compared normal chromosomes was 1.9 (chr13). (B) Control WBC DNA was analyzed alone (n = 2) or when mixed with DNA from a patient with trisomy 21 at 5% (n = 2), 10% (n = 1), or 25% (n = 1) levels. A tight correlation existed between the expected and observed fractions of extra chromosome 21 (r = 0.997 by Pearson correlation test, n = 6). (C) Control WBC DNA was analyzed alone (z-score range: −0.8 to 1.3) or when mixed with DNA from a patient with trisomy 21 at 4% (z-score range: 4.5 to 7.2) or 8% (z-score range: 8.9 to 10.) levels. Each experiment in (C) was performed in quadruplicate.