Buccal Mucosa Cells Are Unsuitable As a Control Tissue to Distinguish Somatic Hematopoietic Mutations From Germline Mutations: Nail DNA Is a Suitable Surrogate

Abstract 5156

Blood is an easily obtainable source of genomic DNA (gDNA) for genetic analysis of blood neoplasms. However, germline and non-hematopoietic cell gDNA, essential for distinguishing tumor-specific acquired mutations from constitutional genetic variants, is more difficult to obtain, an issue magnified with the advent of whole genome sequencing. Skin fibroblasts or fibroblasts isolated from marrow biopsies are a source of non-hematopoietic cell gDNA, and in some instances, T-lymphocytes formed before the clonal event have been used as the control tissue. However, such samples require invasive and time-consuming procedures, trained personnel and additional in vitro culturing or other manipulations. Consequently, buccal mucosa cells have been employed as a convenient alternative source of non-hematopoietic cell gDNA. Unfortunately, the appropriateness of buccal mucosa cells for this purpose has come into question in both genomic sequencing studies and bone marrow transplant studies in which the donor genotype was rapidly evident in the recipient's buccal mucosa cell gDNA (Endler et al., Bone Marrow Transplant., 1999).

To determine if buccal mucosa cell gDNA adequately represented the germline genotype in myeloproliferative neoplasms (MPN), we compared JAK2 V617F levels in granulocyte and buccal mucosa cell gDNA using an allele-specific, quantitative real-time polymerase chain reaction assay (Moliterno et al., Blood, 2006). JAK2 V617F, a well documented somatic mutation, was identified in six of six buccal mucosa cell samples and the quantity was 55–100% of that seen in patient granulocyte gDNA. Cytospin preparations failed to reveal non-buccal mucosa cell contamination (>95%) and neither extensive pre-washing of the cells (3 to 4 washes in PBS) nor DNAase treatment before gDNA extraction eliminated or even noticeably reduced the quantity of measurable JAK2 V617F. Furthermore, in keeping with the known presence of circulating JAK2 V617F, DNA isolated from patient cell-free saliva also demonstrated equally high quantities of JAK2 V617F.

In an independent study of a two-year old child with the clinical phenotype of PV and a high JAK2 V617F allele burden, buccal mucosa cells were analyzed to exclude congenital PV and a significant JAK2 V617F allele burden was observed. However, subsequent analysis of gDNA extracted from cultured fibroblasts, isolated peripheral blood lymphocytes and nails were all negative for the JAK2 V617F mutation.

We therefore investigated the potential of using nails as an alternative noninvasive source of gDNA. Using a forensic nail DNA extraction kit (QIAamp DNA Investigator kit, QIAGEN cat. #56504), we found that gDNA from the nails of two JAK2 V617F-positive patients were negative for the JAK2 V617F mutation both by allele-specific quantitative PCR and by a JAK2 V617F-specific pyrosequencing assay. In addition, when we encountered a previously unreported JAK2 mutation, T514M, we were able to demonstrate that it was an acquired mutation by using nail gDNA for direct PCR sequencing, as shown below.

View largeDownload slide

View largeDownload slide

Close modal

While this method is very convenient and suitable for PCR-based studies, it is limited for use in genome-wide studies because the isolated nail gDNA is highly degraded (<400 bp). However, using a NaOH-based technique (Cline et al., J. Forensic Sci., 2003) we have been able to consistently obtain gDNA fragments ranging from 1000 – 3000 bp from either toenails (∼10 ng/mg), or fingernails (∼40 ng/mg). Thus, it should be possible to generate enough gDNA (4–5 μg) with two rounds of nail clippings for genomic sequencing studies. This technique should also lend itself to clonality studies in genetically appropriate female subjects where lymphocytes have been the traditional gDNA source.

We conclude that whether from blood cells, saliva or an unknown source, buccal mucosa cells are heavily contaminated with and inseparable from DNA of hematopoietic cell origin. We also demonstrate that both fingernails and toenails are a suitable source of non-hematopoietic cell gDNA, which can be used to distinguish between acquired and constitutional genetic variants in hematopoietic tissue. In addition, we have identified a protocol that can generate nail gDNA of a potentially suitable quality and quantity for genome-wide sequencing studies.