The purpose of this paper was to compare the recovery and fragment size of cell-free DNA (cfDNA) extracted using four different kits using pooled cerebrospinal fluid (CSF) collected from Alzheimer’s patients and pooled CSF collected from healthy controls. The authors also compared levels of nuclear and mitochondrial cfDNA among specimens from healthy patients, patients with Alzheimer’s disease and patients with brain cancer.
A standardised methodology for the extraction and quantification of cell-free DNA in cerebrospinal fluid and application to evaluation of Alzheimer's disease and brain cancers.
Takousis P, Devonshire AS, Redshaw N, von Baumgarten L, Whale AS, Jones GM, Fernandez-Gonzalez A, Martin J, Foy CA, Alexopoulos P, Huggett JF, Perneczky R
Purpose
Conclusion
Recovery of the spiked-in ADH fragments was lower when extraction was with the Quick-cfDNA Kit than any of the other three kits and the coefficient of variance (CV) for replicate extractions was lowest when extraction was with the QIAamp CNA Kit or the Quick-cfDNA Kit. While the QIAamp CNA kit showed bias toward recovery of the smallest spike-in fragment, the Norgen Plasma/Serum cfDNA and the NucleoSnap DNA kit showed a bias toward recovery of the longest fragment. The Quick-cfDNA Serum & Plasma Kit had comparable recovery of all three sizes of spike-in fragments. Yields of endogenous mitochondrial (mtDNA-123) and nuclear (AluJ) cfDNA were highest when extraction was with NucleoSnap or Quick-cfDNA Kits and levels were lowest when no extraction was performed. Although no difference in the ratio of mitochondrial to nuclear cfDNA was found among the extraction kits, the ratio was slightly higher in the pooled CSF from healthy controls than Alzheimer’s disease patients.
Mean log transformed mitochondrial and nuclear cfDNA levels were an order of magnitude higher in CSF from brain cancer patients than healthy patients or Alzheimer’s disease patients, and a higher inter-patient variability was also observed in specimens from brain cancer patents. However, extraction efficiency using the QIAamp CNA Kit and the ratio of mean log transformed mitochondrial to nuclear cfDNA were comparable among CSF specimens from healthy patients, patients with Alzheimer’s disease and patients with brain cancer.
Studies (1)
Purpose
The purpose of this study was to compare the recovery and fragment size of cfDNA isolated using four different extraction kits from pooled CSF collected from Alzheimer’s patients and pooled CSF collected from healthy controls. The authors also compared levels of nuclear and mitochondrial cfDNA among specimens from healthy patients, patients with Alzheimer’s disease and patients with brain cancer. CSF from twelve Alzheimer’s disease patients (mean age 75.4 years), ten healthy patients (mean age 57.8 years), and twelve patients with brain tumors (2 primary medulloblastoma, 4 primary astrocytoma; 2 primary glioblastoma, 3 secondary non-small cell lung carcinoma, and a secondary adeno-cancer of unknown origin) were collected using atraumatic cannulas placed in the L3/L4 or L4/L5 intervertebral space into polypropylene tubes. Unless otherwise specified, 115, 461 and 1448 bp fragments of ADH were added to the CSF and cfDNA was extracted using the Qiagen QIAamp Circulating Nucleic Acid Kit and stored at 4°C (immediate use) or -20°C (<1 month). cfDNA was quantified by ddPCR amplification of the ADH spike-in control (115, 461 and 1448 bp), RPPH1 (nuclear), and mtDNA-123 (mitochondrial) and real-time PCR of ALUJ (nuclear). Potential effects of extraction method were investigated by extracting cfDNA in triplicate with the Qiagen QIAamp Circulating Nucleic Acid Kit, Zymo Quick-cfDNA Serum & Plasma Kit, Macherey-Nagel NucleoSnap DNA Plasma Kit, and the Norgen Plasma/Serum Circulating DNA Purification Mini Kit from a pool of CSF collected from Alzheimer’s disease patients and a CSF pool collected from healthy patients and by amplifying directly from CSF (without extraction). Comparisons among patient disease states were conducted using individual CSF specimens.
Summary of Findings
Recovery of the three differentially sized spiked-in ADH fragments was lower when extraction was with the Quick-cfDNA Kit compared to the other three kits evaluated (P<0.001), but no other differences in recovery were found among the kits. The QIAamp CNA Kit had higher recovery of the 115 bp fragment than the 461 and 1448 bp fragments (78% versus 68% and 67%, respectively, P<0.001). Recovery of all three differentially sized fragments was comparable using the Quick-cfDNA Serum & Plasma Kit (68%). In contrast, the Norgen Plasma/Serum cfDNA Kit and the NucleoSnap Kt displayed biases toward the longest fragments (Norgen Kit: 25% for 1448 bp, 16% for 461 and 8% for 115 bp, p < 0.01; and NucleoSnap Kit: 79% for 1448- and 461 bp fragments and 69% for the 115 bp fragments, p < 0.001, respectively) . The CV for replicate extractions was low when extraction was with the QIAamp CNA Kit (4-9%) and Quick-cfDNA Kit (4-8%) but was higher when extraction was with the NucleoSnap Kit or the Norgen Plasma/Serum cfDNA Kit (5-19% and 27-40%, respectively).
Yield of endogenous mitochondrial cfDNA was lower when the Norgen Plasma/Serum cfDNA Kit was used for extraction compared to either the NucleoSnap or Quick-cfDNA Kit (P<0.001 and P<0.05), and when the QIAamp CNA Kit was used compared to the NucleoSnap Kit (P=0.03). Similarly, AluJ (nuclear) levels were lowest when extraction was with the Norgen Plasma/Serum cfDNA Kit (P<0.001) and lower when extracted with the QIAamp Kit than the NucleoSnap or Quick-cfDNA Kits (P<0.001 and P=0.01, respectively). In contrast to AluJ, RPPH1 nuclear cfDNA was only detected in 50% of extracts. Although no difference in the ratio of mitochondrial to nuclear cfDNA was found among the extraction kits evaluated, the ratio was slightly higher in the CSF pool from healthy controls than the CSF pool from Alzheimer’s disease patients (P<0.05). Although nuclear cfDNA could be amplified directly from CSF, levels were much lower than when extraction was performed, and PCR inhibition occurred even after sample dilution.
Extraction efficiency was comparable using the QIAamp CNA Kit for specimens from healthy patients, Alzheimer’s disease patients and brain cancer patients. Mean log transformed mitochondrial and nuclear cfDNA levels were an order of magnitude higher in CSF from brain cancer patients than healthy patients or Alzheimer’s disease patients (P<0.001, both). The inter-patient variability in cfDNA levels was also higher in specimens from brain cancer patients than healthy controls or patients with Alzheimer’s disease. Levels of mitochondrial (mtDNA-123) and nuclear (ALUJ) cfDNA were very strongly correlated (R2=0.7, P<0.0001) and the ratio of mean log transformed mitochondrial to nuclear cfDNA was comparable among CSF from healthy patients, patients with Alzheimer’s disease and patients with brain cancer.
Biospecimens
FluidCerebrospinal FluidPreservation Types
FrozenDiagnoses
NeoplasticCarcinomaNeoplasticOtherAlzheimer's DiseaseNormalPlatforms
DNADigital PCRDNAReal-time qPCRPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Preaquisition Diagnosis/ patient condition Alzheimer’s diseaseBrain CancerHealthyAnalyte Extraction and Purification Analyte isolation method Macherey-Nagel NucleoSnap DNA Plasma kitNo extractionNorgen Plasma/Serum Circulating DNA Purification Mini KitQiagen QIAamp Circulating Nucleic Acid KitZymo Quick-cfDNA Serum & Plasma KitDigital PCR Specific Length of gene fragment 115 bp1448 bp461 bpDigital PCR Specific Targeted nucleic acid ADHmtDNA-123Real-time qPCR Specific Targeted nucleic acid ALUJ