Effects of blood-processing protocols on cell-free DNA fragmentomics in plasma: Comparisons of one- and two-step centrifugations.
Author(s): Hu X, Zhang H, Wang Y, Lin Y, Li Q, Li L, Zeng G, Ou R, Cheng X, Zhang Y, Jin X
Publication: Clin Chim Acta, 2024, Vol. 560, Page 119729
PubMed ID: 38754575 PubMed Review Paper? No
Purpose of Paper
This paper compared the size profile, end motifs, and genome distribution of cell-free DNA (cfDNA) and mitochondrial cell-free DNA (mtcfDNA) between matched plasma that was obtained by a single centrifugation (3000 x g) and plasma that was obtained by double centrifugation (1,600 x g followed by 16,000 x g). The size profiles of cfDNA and mtcfDNA were also compared between specimens from patients with SARS-CoV-2, the virus that causes the novel 2019 coronavirus disease (COVID, COVID-19, COVID19), and healthy volunteers. The sensitivity and specificity of the ratio of short to long cfDNA for the detection of SARS-CoV-2 was also compared between plasma samples that were obtained by a single centrifugation or two-step centrifugation.
Conclusion of Paper
No differences in the fragment size profile, Motif Diversity Score (MDS), window protection score (WPS), or average coverage 1000 upstream and downstream of the transcriptional start sites (TSS) and CCCTC-binding factor (CTCF) binding site were observed between plasma obtained by a single centrifugation or two centrifugations. While some small (<0.4%) but significant differences in the frequency of some end-motifs were observed between single- and double-spun plasma, specimens clustered by patient source not centrifugation steps in Principal Component Analysis (PCA). Similarly, the correlations in measured genomic representations (MGRs) were stronger among double-spun than single-spun plasma, but the difference was small compared to differences that occurred between patients. Correlations in end motif frequencies, measured genomic representations (MGRs), and coverage around TSS were stronger between single- and double-spun plasma from the same patient than between patients using either single-spun or double-spun plasma. cfDNA from COVID-19 patients showed a shift toward shorter fragment sizes, regardless of centrifugation method, with a higher percentage of short fragments (<150 bp) and a lower percentage of long fragments (>250 bp) relative to size distributions in healthy patients, regardless of the number of centrifugation steps. The sensitivity and specificity for the detection of COVID-19 of a ratio of short to long fragments of >2.75 was the same for single- and double-spun plasma.
The relative mean abundance of mtcfDNA was significantly higher in single-spun than double-spun plasma of healthy volunteers. Single-spun plasma had a significantly smaller median percentage of short fragments (<100 bp) and a higher percentage long fragments (>100 bp) than double-spun plasma from healthy volunteers or COVID-19 patients.
Studies
-
Study Purpose
This study compared the size profile, end motifs, and genome distribution of cfDNA and mtcfDNA between matched plasma obtained by single centrifugation (3000 x g) and that obtained by double centrifugation (1,600 x g followed by 16,000 x g). The size profiles of cfDNA and mtcfDNA were also compared between specimens from patients with SARS-CoV-2, the virus that causes the novel 2019 coronavirus disease (COVID, COVID-19, COVID19), and healthy volunteers. The sensitivity and specificity of the ratio of short to long cfDNA for the detection of SARS-CoV-2 were also compared between plasma obtained by a single centrifugation and two-step centrifugation. Blood was collected from ten healthy volunteers (age 22-27 years) into EDTA tubes and stored on ice for <4 h before plasma separation. Each specimen was divided into two aliquots and plasma was separated by a single centrifugation at 3000 x g for 15 min at 4°Cor by dual-centrifugation at 1,600 x g for 10 min at 4°C followed by 16,000 x g for 10 min at 4°C To investigate the effects of COVID-19, eight unmatched specimens from patients with COVID-19 were obtained by single (1600 x g for 10 min at 4°C) centrifugation and dual centrifugation (1,600 x g for 10 min at 4°C, followed by 16,000 x g for 10 min at 4°C) were compared to eight randomly selected single- and dual-centrifugation specimens from healthy volunteers; samples were processed as described above. All plasma was stored at -80°C until cfDNA extraction using the automated MagPure Circulating DNA Kit and a MGISP-960 instrument. Sequencing libraries were prepared using the Cell-free DNA Library Prep Kit and pair-end sequenced using the DNBSEQ platform. Sequences were preprocessed using fastp and aligned to the human reference genome (GRCh38/hg38) using BWA before the removal of PCR duplicates and low mapping quality reads. mtcfDNA abundance was calculated based on the number of reads that uniquely aligned to the mitochondrial reference sequence (GRCh38/hg38) per million reads.
Summary of Findings:
The predominant size peak of cfDNA was 166 bp in both single- and double-spun plasma. Further, the percentages of “short” DNA (60-155 bp) and “long” (170-255 bp) fragments were comparable in single- and double-spun plasma. Analysis of the end motifs found no difference in Motif Diversity Score (MDS). Correlations in end motif frequencies were stronger between single- and double-spun plasma from the same patient (r=0.9952–0.9995) than between patients using either single-spun (r=0.9901–0.9998, P=0.0162) or double-spun plasma (r=0.9891–0.9997, P=0.0095). Of the 256 end-motifs, 166 (64.84%) were not affected by the number of spins, and 90 (35.16%) displayed minimal (<0.4%) but significant differences in abundance. Further analysis found that 8 of the 20 most abundant end motifs were significantly affected by the number of spins, but differences reflected as fold-change were extremely small (0.970-1.029); in Principal Component Analysis based on end motifs, specimens clustered by patient source, not centrifugation steps. The window protection score (WPS) of plasma obtained by single- and double-centrifugation overlapped. The measured genomic representations (MGRs) were stronger between single- and double-spun plasma from the same patient (r= 0.9972–0.9992) than between patients using either single-spun (r= 0.9898–0.9967, P<0.0001) or double-spun plasma (r= 0.9922–0.9969, P<0.0001). Interestingly, the correlation in MGR was stronger among double-spun than single-spun plasma (P=0.035), indicating that centrifugation influenced MGRs, albeit to a smaller extent than occurred between individuals. The average coverage 1000 upstream and downstream of transcriptional start sites (TSS) and CCCTC-binding factor (CTCF) binding sites were similar for cfDNA from single- and double-spun plasma. Coverage around TSS was strongly correlated between single- and double-spun plasma from the same patient (r= 0.86–0.89) than between patients using either single-spun (r= 0.84-0.88), P=0.0006) or double-spun plasma (r= 0.86-0.88, P<0.0001). cfDNA from COVID-19 patients showed a shift toward shorter fragment sizes, regardless of centrifugation method, with a higher percentage of short fragments (<150 bp) and a lower percentage of long fragments (>250 bp) observed in plasma of COVID-19 patients than healthy patients both when plasma was obtained by single- (14.96 % versus 10.74 %, P=0.0002 and 4.43% versus 6.29%, P=0.0207, respectively) and double-centrifugation (14.65% versus 11.93%, P=0.0002 and 3.87% versus 8.02%, P=0.0030, respectively). The sensitivity and specificity of a ratio of >2.75 short to long fragments for the detection of COVID-19 were the same for single- and double-spun plasma (0.875, for all).
The relative mean abundance of mtcfDNA was significantly higher in single-spun than double-spun plasma of healthy volunteers (21.38 versus 14.77, P=0.002), and a similar trend was observed for the plasma of COVID-19 patients (45.45% versus 27.55%, P= 0.1605). The fragment size profile showed a small shift toward larger mtcfDNA in single-spun than double-spun plasma and a significantly smaller median percentage of short fragments (<100 bp) and a higher percentage of long fragments (>100 bp) in single-spun plasma than double-spun plasma from healthy volunteers (35.43% versus 44.11%, P=0.002 and 64.57% versus 55.89%, P=0.002, respectively). Importantly a smaller median percentage of short fragments and a higher percentage of long fragments were also observed in single-spun plasma than double-spun plasma from COVID-19 patients (P<0.01, for both).
Biospecimens
Preservative Types
- Frozen
Diagnoses:
- Pneumonia/Respiratory Infection
- Normal
Platform:
Analyte Technology Platform DNA Next generation sequencing Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Preaquisition Diagnosis/ patient condition COVID-19
Healthy
Biospecimen Aliquots and Components Centrifugation Different number of centrifugation steps compared