Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Evaluating the Feasibility of DNA Methylation Analyses Using Long-Term Archived Brain Formalin-Fixed Paraffin-Embedded Samples.

Author(s): Bak ST, Staunstrup NH, Starnawska A, Daugaard TF, Nyengaard JR, Nyegaard M, Børglum A, Mors O, Dorph-Petersen KA, Nielsen AL

Publication: Mol Neurobiol, 2016, Vol. , Page

PubMed ID: 27995571 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study compared DNA integrity of archival FFPE cortical and hippocampal specimens from two patients and an intact DNA control. Post-mortem brains from two patients with schizophrenia were fixed in situ by formalin infusion and brains were removed within 6-65 h and immersed in formalin (non-buffered or buffered) for weeks to months. After which, brains were sliced, sampled for neuropathological examination and embedded in paraffin. FFPE blocks of the hippocampus and cortex of two brains were stored for 33 and 62 years, respectfully. DNA was extracted from two 10 µm thick FFPE sections with the FFPE DNA extraction kit from Diagenode. DNA was quantified using the Quant-iT PicoGreen Double-Stranded DNA (dsDNA) Assay, Quant-iT OliGreen Single-Stranded DNA (ssDNA) assay and by spectrophotometry.  DNA integrity was determined by capillary electrophoresis, the Infinium HD FFPE QC Assay and by real-time PCR. DNA from peripheral blood of 4 unrelated patients (diagnosis not specified) served as a control. 

   Summary of Findings:

   DNA yields were 2-7 fold greater in archival specimens when quantified by Oligogreen and Nanodrop than Picogreen, suggesting that the majority of extracted DNA was single stranded or damaged. The mean fragment length of DNA isolated from FFPE specimens was between 150-180 bp but longer fragments were observed, particularly in the block stored for 33 years. Further, the specimens stored for 33 years yielded more intact DNA than the specimen stored 62 years, but unknown fixation parameters, storage conditions and PMI are confounding variable. The FFPE specimen stored for 62 years had 16-fold fewer functional templates of the longest fragment (160 bp) based on the Sat2 qPCR assay than the specimen stored for 33 years. PCR efficiency improved for FFPE specimens when DNA was diluted by 10-fold or more. Treatment of DNA with a ss-DNA specific nuclease (S1 nuclease) resulted in declines in amplification for both archival FFPE specimens, which is indicative of DNA damage; however, the specimen stored 62 years displayed a larger decline than the one stored 33 years indicating a smaller proportion of double stranded DNA.  Treatment of DNA with dsDNase resulted in a larger decline in amplification in the FFPE specimen stored for 33 years compared to the specimen stored for 66 years, confirming that a larger percentage of the DNA from specimens stored 66 years was single stranded.

   Biospecimens

   • Bodily Fluid - Blood
   • Tissue - Brain

   Preservative Types

   • None (Fresh)
   • Formalin

   Diagnoses:

   • Schizophrenia
   • Not specified
   • Autopsy

   Platform:

   Analyte	Technology Platform
   DNA	Fluorometry
   DNA	Real-time qPCR
   DNA	Automated electrophoresis/Bioanalyzer
   DNA	Spectrophotometry

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Storage	Storage duration	33 years 62 years
   Analyte Extraction and Purification	Nucleic acid digestion	S1 nuclease dsDNase None
   Spectrophotometry Specific	Technology platform	Oligogreen Nanodrop Picogreen Electropherogram PCR
   Real-time qPCR Specific	Targeted nucleic acid	Sat2
   Real-time qPCR Specific	Template/input amount	Undiluted Diluted 10-fold Diluted 30-fold Diluted 100-fold
   Real-time qPCR Specific	Length of gene fragment	67 bp 105 bp 160 bp

   View more

2. Study Purpose

   This study compared DNA methylation assay results obtained with archival FFPE cortical and hippocampal specimens and control specimens, and investigated effects associated with the duration of FFPE block storage and PMI on results of a quality control assay. The majority of the experiments used post-mortem hippocampus and cortex specimens from two patients diagnosed with schizophrenia that had been stored as FFPE blocks for 33 and 62 years, respectfully. Genotyping, sex-determination and the quality control assay were performed on an additional 28 patients that included eight patients diagnosed with schizophrenia, 10 with bipolar disorder and 10 with Alzheimer’s disease. Specimens were fixed in situ by formalin infusion and brains were removed within 6-65 h and immersed in formalin (non-buffered or buffered) for weeks to months. Brains were then sliced, sampled for neuropathological examination and embedded in paraffin. FFPE blocks of the hippocampus and cortex were stored for 33-64.8 years before DNA was extracted from two 10 µm thick FFPE sections using the Diagenode kit. DNA was from frozen cortical and hippocampal tissue collected from a patient diagnosed with Alzheimer’s disease and from four peripheral blood specimens served as controls.

   Summary of Findings:

   Methylation levels determined by fluorometry were 2-3 fold higher in archival FFPE specimens than frozen specimens. Archival FFPE specimens also displayed higher levels of hydroxymethylation than frozen specimens. Methylation of TSH2B and LINE-1 was observed in the FFPE specimen stored 33 years but not the FFPE specimen stored 66 years. Hydroxymethylated DNA immunoprecipitation revealed that hydroxymethylated DNA was enriched relative to unmethylated DNA in the FFPE specimens. When DNA isolated from archival FFPE specimens was restored and analyzed by Infinium methylation bead arrays, 100-1000 fold more probes failed than when fresh DNA from blood was used, and only one of the four samples from FFPE specimens had the expected bimodal distribution. Similarly, bisulfate sequencing revealed methylation in 12 of 19 hippocampal FFPE specimens stored for 62 years and 15 of 16 hipppocampal FFPE specimens stored 33 years, indicating that methylation remains detectable in some archival FFPE specimens. The authors report that pyrosequencing of fragments longer than 90 bp was not possible for archival FFPE specimens due to inefficient amplification.  Although HISTH3 methylation was comparable in archival FFPE specimens and freshly frozen specimens (93- 94% versus 86%), the authors report experimental outliers in all pyrosequencing runs. Pyrosequencing was able to reproducibly identify the genotype in 74% of the 30 archival FFPE specimens evaluated, and discrepancies in the remaining 26% were attributed to poor DNA quality. For the majority of FFPE specimens, patient sex was properly assigned based on amelogenin pyrosequencing, but there was insufficient information to determine gender for two specimens and the findings were inconsistent among replicates for another specimen.  Similarly, concordance between observed (based on CpG island methylation on the X chromosome)and expected sex was present for 13 of the 30 FFPE specimens, while information was inadequate  for the remaining 17 specimens. Of the 17 FFPE specimens for which observed sex could not be determined, one assay predicting the correct sex for 12 specimens while no informative data was generated for the remaining 5 specimens.  A two-step quality control (QC) assay was evaluated that was based on the ratio of Sat2 67 bp amplicons in FFPE specimens relative to those observed in frozen specimens and the ratio of the Sat2 67 bp to 105 bp amplicons in the FFPE specimen. The thresholds were >1% as much amplification of the FFPE specimens as observed for the frozen specimen and at least 10% as many copies of the 105 bp amplicon as the 67 bp amplicon.  Of the 30 FFPE specimens screened using the QC assay, 17 passed, 12 performed well, and 5 failed. One specimen that failed the QC assay performed well in the methylation assays. Significant correlations were observed between the duration of FFPE  block storage and QC assay results  (r=not specified, P<0.0002), and the ratio of the Sat2 67 bp amplicon in FFPE relative to frozen specimens of more than 1% (r=0.371, P=0.004), but not the ratio of the two Sat2 amplicons. The authors report that results of the QC assay were not correlated to PMI or patient age.

   Biospecimens

   • Bodily Fluid - Blood
   • Tissue - Brain

   Preservative Types

   • Formalin
   • None (Fresh)
   • Frozen

   Diagnoses:

   • Schizophrenia
   • Alzheimer's Disease
   • Bipolar Disorder
   • Not specified
   • Autopsy

   Platform:

   Analyte	Technology Platform
   DNA	Bisulfite conversion assay
   DNA	DNA sequencing
   DNA	DNA microarray
   DNA	Immunoassay

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Preaquisition	Postmortem interval	6-65 h
   Preaquisition	Patient age	49.7-89 years
   Biospecimen Preservation	Type of fixation/preservation	Formalin (buffered) Formalin (unbuffered) Frozen
   Storage	Storage duration	33-64.8 years
   Analyte Extraction and Purification	Nucleic acid digestion	S1 nuclease dsDNase None
   Bisulfite conversion assay Specific	Technology platform	Illumina beadarray Pyrosequencing Clonal bisulfate sequencing Immunoprecipitation

   View more