Skip to main content

NGS-based accurate and efficient detection of circulating cell-free mitochondrial DNA in cancer patients.

Liu Y, Zhou K, Guo S, Wang Y, Ji X, Yuan Q, Su L, Guo X, Gu X, Xing J

Publication: Mol Ther Nucleic Acids (2021) • Vol 23 • Page 657-666PMID: 33575112View on PubMed

Purpose

This paper investigated the effects of delayed centrifugation and storage of blood on circulating cell-free mitochondrial DNA (ccf-mtDNA) copy number and fragment size and compared ccf-mtDNA copy numbers, fragment size, and variant detection in tumor and plasma specimens among patients with hepatitis, hepatocellular carcinoma (HCC), or colorectal cancer (CRC) and healthy controls. Additionally, the authors investigated the effects of probe amount, probe length, hybridization temperature, and number of amplification cycles used in the capture-based mtDNA sequencing protocol on mapping rate, sequencing depth, and variant detection.

Conclusion

Storage of blood for up to 24 h at room temperature or 4°C had no effect on ccf-mtDNA copy number and fragment size but significant declines in ccf-mtDNA copy number and fragment size occurred when plasma was stored at room temperature or 4°C.

Plasma had much lower mean ccf-mtDNA copy number than fresh tumor specimens. Plasma ccf-mtDNA copy number was higher in healthy patients than those with hepatitis, HCC, or CRC. The peak fragment size of plasma ccf-mtDNA was smaller than that of nuclear cfDNA (90 bp versus 167 bp). The median size of the plasma ccf-mtDNA was larger in patients with hepatitis than from healthy or HCC patients and in patients with CRC than HCC. The ccf-mtDNA break points were evenly distributed throughout the mt genome, regardless of patient diagnosis. Optimization experiments found the highest mapping rate and sequencing depth when capture-based mtDNA sequencing was conducted using 0.8 ng of 350 bp probes per 200 ng of library, hybridization was performed at 65°C, and using 20 amplification cycles. Compared with the initial protocol, the optimized protocol resulted in a higher mapping rate, mRNA depth, and number of heteroplasmic variants but did not affect detection of homoplasmic ccf-mtDNA variants. Importantly, 92% of the variants identified using the optimized procedure were confirmed in replication. Using the optimized protocol to analyze plasma from HCC patient, ccf-mtDNA were only detectable in tumor (not adjacent tissue or PBMC) but none of these tumor-specific mutations were identified in plasma from CRC patients. The heteroplasmy levels were very strongly correlated between plasma ccf-mtDNA and tumor mtDNA of the HCC patients.

Studies (2)

Purpose

This study compared ccf-mtDNA copy numbers, fragment size, and variant detection in tumor and plasma specimens among patients with hepatitis, HCC, or CRC and healthy controls. Additionally, the authors investigated the effects of probe amount, probe length, hybridization temperature, and number of amplification cycles used in the capture-based mtDNA sequencing protocol on mapping rate, sequencing depth, and variant detection. K2EDTA blood was collected obtained from 10 healthy patients, 10 patients with hepatitis, 10 patients with HCC, and 20 patients with colorectal carcinoma. Matched fresh tumor specimens were obtained from the 10 CRC and 5 HCC patients. Plasma was obtained by centrifugation at 4°C at 1600 rpm for 10 min followed by recentrifugation of the supernatant at 4°C at 10,000 rpm for 15 min. ccf-DNA was extracted from plasma using the QIAamp Circulating Nucleic Acid Kit. DNA was isolated from the PBMC (isolation method not specified) and tumor and normal adjacent tissue of 5 HCC and 10 CRC patients using the EZNA DNA Kit. DNA quantity was evaluated using a Qubit 3.0 fluorometer and Agilent 2100 Bioanalyzer and mean size evaluated by bioanalyzer. WGS libraries were constructed using the NEB Ultra v2 Kit and pair-end sequenced using a HiSeq X Ten platform. Capture-based mtDNA sequencing was performed by mixing WGS library with biotinylated mtDNA capture probes (0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.4, 0.8, 1.6, 3.2, 6.4, or 10 ng) of different lengths (150, 250, or 350 bp) and hybridization at 55°C or 65°C and amplified for 16, 20, or 24 cycles before sequencing on an Illumina HiSeq X Ten platform using PE 150.

Summary of Findings

Plasma had much lower mean ccf-mtDNA copy number than fresh tumor specimens (5.05 versus 935.40, P<0.001). Plasma ccf-mtDNA copy number was higher in healthy patients than those with hepatitis, HCC, or CRC (P<0.05, all). The peak fragment size of plasma ccf-mtDNA was 90 bp, which was smaller than the 167 bp found for nuclear cfDNA. The median size of the plasma ccf-mtDNA was larger in patients with hepatitis than from healthy or HCC patients (P<0.05, both) and in patients with CRC than HCC (P<0.05). The ccf-mtDNA break points were evenly distributed throughout the mt genome, regardless of patient diagnosis. Optimization experiments found the enrichment efficient as assessed by mapping rate and sequencing depth were highest when sequencing was conducted with 0.8 ng of probes per 200 ng of library and were significantly higher using probes of 350 bp versus 150 (P<0.001 and P<0.01, respectively) or 250 bp (P<0.001 and P<0.01, respectively), hybridization at 65°C versus 55°C (P<0.001, both), and using 20 amplification cycles rather than 16 (P<0.001, both) or 24 (P<0.001, both). Plasma ccf-mtDNA copy number was strongly correlated with enrichment efficiency (r=0.952, P<0.001). Compared with the initial protocol, the optimized protocol resulted in a higher mapping rate (P<0.001), mRNA depth (P<0.0001), and number of heteroplasmic variants (P<0.0001) but did not affect detection of homoplasmic ccf-mtDNA variants. Importantly, 92% of the variants identified using the optimized procedure were confirmed in replication. Further, most variants detected only by using the optimized procedure had a heteroplasmic level of ≤10%. Using the optimized protocol to analyze plasma from HCC patient, ccf-mtDNA were only detectable in tumor (not adjacent tissue or PBMC), but none of these tumor-specific mutations were identified in plasma from CRC patients. The heteroplasmy levels were very strongly correlated between plasma ccf-mtDNA and tumor mtDNA of the HCC patients (ρ = 0.761, P<0.001).

  • Biospecimens

    FluidBlood
    FluidPlasma
    TissueLiver
    TissueColorectal
  • Preservation Types

    None (Fresh)
  • Diagnoses

    NeoplasticCarcinoma
    Hepatitis
    Normal
    NeoplasticNormal Adjacent
  • Platforms

    DNAAutomated electrophoresis/Bioanalyzer
    DNANext generation sequencing
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    PreaquisitionDiagnosis/ patient condition
    Colorectal cancer
    HCC
    Healthy
    Hepatitis
    Biospecimen AcquisitionBiospecimen location
    Plasma
    Tumor
    Next generation sequencing SpecificNucleic acid amplification
    Different number of amplification cycles applied
    Next generation sequencing SpecificTemplate modification
    Different amounts of capture probe used
    Different size capture probes used
    Hybridized at different temperatures