Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Flow cytometric sorting coupled with exon capture sequencing identifies somatic mutations in archival lymphoma tissues.

Author(s): Jiang N, Chen C, Gong Q, Shields K, Li Y, Chen Y, Song J, McKeithan TW, Chan WC

Publication: Lab Invest, 2017, Vol. 97, Page 1364-1374

PubMed ID: 28783138 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study investigated the effects of antigen retrieval temperature, duration, and buffer on immunostaining intensity, DNA integrity, and amplificability and investigated if using DNA from sorted cells rather than FFPE sections improved NGS variant discovery. Specimens included a two-year-old tonsil specimen with reactive hyperplasia, a one-and-a-half-year-old lymph node specimen from a patient with follicular lymphoma (FL), and two lymph node specimens from patients with angioimmunoblastomic T-cell lymphoma (AITL). Specimens had been fixed for 4 h to overnight in 10% buffered formalin before paraffin embedding. Fifty micrometer sections were deparaffinized by three ten-minute incubations with xylene and ethanol rehydration. Specimens were then placed in PBS, ground with a Pellet Pestle, and digested with 0.3% collagenase/dispase at 37˚C for 10 min. The cell suspension was placed on ice, filtered through a 70µm filter, and finally centrifuged and resuspended in phosphate buffered saline (PBS) with 1% bovine serum albumin (BSA) two times. The cells were counted using a Countess automated cell counter.  Cells were aliquoted as 1 x 10⁶ cells/tube, pelleted, and resuspended in Tris-EDTA pH8.0. Aliquots were incubated at 85°C for 15 and 25 min, 70°C for 60 min and 80 min, 65°C for 60 min, 55°C for 60 min, 45°C for 60 min, and 37°C for 60 min. Cell suspensions were then labelled with antibodies to cluster of differentiation (CD)3 (ectodomain and cytoplasmic domain), linker for activation of T-cells (LAT), B-cell lymphoma/leukemia 11B (BCL11B), CD4, CD8, programmed death-1 (PD1), CD20, CD79A, paired box 5 (PAX5), and C-X-C Motif Chemokine Receptor 5 (CXCR5), and sorted by flow cytometry. DNA was extracted from sorted cells using the QIAamp DNA FFPE Tissue Kit following a 16 h proteinase K digestion at 65˚C followed by an additional 4 h at 56˚C with fresh proteinase K. DNA integrity was assessed by electrophoresis and real-time PCR amplification of 196 and 295 bp products. NGS was conducted using Illumina HiSeq2500.

   Summary of Findings:

   Fluorescent signal increased with increasing temperature and duration of antigen retrieval with the best results obtained after incubation at 85°C for 25 min or 70 °C for 80 min. PD1 staining was positive in a small number of cells in the case of hyperplasia but was positive in more cells in the other cases. The antibodies to BCL11B and the CD3 ectodomain were not able to effectively distinguish T-cells from B-cells. CD4 staining was too weak to clearly segregate positive and negative cells and the remaining antibodies failed to stain the cells. Antigen retrieval using citrate instead of TE produced comparable staining. While increasing the trypsin/dispase digestion beyond 10 min led to reduced CD79a staining, it had no effect on CD3 staining intensity. 

   As expected, DNA from the FFPE cells was degraded in comparison to DNA from a cell line and only about 12% as much of the 196 bp product and 2-3% as much or the 295 bp product was amplifiable. However, amplification of both products were further decreased and DNA was more degraded in specimens subjected to antigen retrieval in TE at 85˚C for 25 min or 70˚C for 80 min, but was comparable after antigen retrieval in TE at 55-70˚C for 60 min in specimens not subjected to antigen retrieval. Importantly, antigen retrieval in citrate instead of TE at 65˚C for 60 min resulted in slightly less amplifiable DNA.

   The majority of variants found at low frequency in tumor and non-tumor cells of the follicular lymphoma case were found to be alignment artifacts, but 28 true variants were identified, 26 of which have been previously reported in follicular lymphoma or diffuse large B-cell lymphoma.  Similarly, sequencing of CD3+ cells found two mutations in TET2 in the PD1+ cells (tumor cells) of one case and mutations in RHOA, TET2, KMT2C, DYNC1H1, and IDH2 were found to be in a higher percentage of PD1+ cells than negative cells in the other case. Only the DYNC1H1 mutation has not been previously associated with AITL. When the sequencing results were compared between the sorted cells and those with the original block, it was found that the somatic mutation VAFs occurred at half the frequency in the block as in the sorted cells, but the germline and artifactual mutations occurred at similar frequencies and no new mutations were identified in the block.

   Biospecimens

   • Tissue - Tonsil
   • Tissue - Lymph Node

   Preservative Types

   • Formalin

   Diagnoses:

   • Neoplastic - Lymphoma

   Platform:

   Analyte	Technology Platform
   DNA	Real-time qPCR
   DNA	Electrophoresis
   Protein	Flow cytometry
   DNA	Next generation sequencing

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Analyte Extraction and Purification	Antigen retrieval	TE pH 8.0 85°C for 25 min TE pH 8.0 85°C for 15 min TE pH 8.0 70°C for 80 min TE pH 8.0 70°C for 60 min TE pH 8.0 65°C for 60 min TE pH 8.0 55°C for 60 min TE pH 8.0 45°C for 60 min TE pH 8.0 37°C for 60 min Citrate buffer 65˚C for 60 min
   Real-time qPCR Specific	Length of gene fragment	196 bp 295 bp
   Biospecimen Aliquots and Components	Aliquot size/volume	Sorted cells Whole section
   Analyte Extraction and Purification	Protein digestion	0 min Collagenase/dispase at 37˚C for 10 min Collagenase/dispase at 37˚C for 20 min Collagenase/dispase at 37˚C for 30 min

   View more