Detection of Murine DNA in FFPE Columns
Using Qiagen's QIAamp DNA FFPE Tissue Kit (Qiagen, Crawley, UK) as directed in the manufacturer's instructions, we recently observed that DNA extracted from formalin-fixed paraffin-embedded (FFPE) prostate cancer tissue from the UK, Thailand and Korea (supplied from non commercial sources as detailed in 
), occasionally contained mouse DNA. This murine contamination was detected by PCR using primers IAP forward and IAP reverse (), specifically designed to amplify IAP sequences. We used TaqGold polymerase (Applied Biosystems, Warrington, UK) for the PCR under described conditions 
Sequences for IAP were obtained from elution buffer processed through several empty (control) columns used directly as supplied by Qiagen. Multiple PCR water controls (not processed through the columns) were consistently negative, indicating that the PCR buffers and polymerase enzyme used in the PCR were clean and that mouse contamination had not been introduced during the reaction (data not shown).
It was not possible to determine if the kit buffers were contaminated with mouse DNA, as they inhibited the IAP PCR when they were added to the positive control, namely DNA from McCoy murine fibroblast cell line (ECACC #90010305), (data not shown). To investigate the prevalence of murine DNA signals in FFPE columns we looked at a total of 68 “mock-extracted” column eluates following manufacturer's instructions, but without adding any tissue sample. We detected signals for IAP in 11 of these 68 columns assayed (16.1%) (, upper panel and data not shown). The PCR product shown in , lane 10 was cloned into plasmid pPCR4TOPO (Invitrogen, Paisley, UK), sequenced and identified as IAP.
Amplification of contaminating DNA from empty columns of the QiaAmp FFPE Tissue Kit.
Detection of XMRV and pMLV Sequences in FFPE Columns
As recently described 
DNA extracted from IAP positive prostate cancer tissue occasionally also gave rise to pMLV sequences and XMRV-specific PCR products, the latter defined by the 24 bp deletion in the leader/gag
region of XMRV 
. When nested PCR was carried out on the eluate of the 11 IAP-positive mock-extracted columns, a fragment with a sequence identical to XMRV was amplified from one of them. The nested PCR conditions to detect XMRV sequences have been documented 
. First-round reaction conditions were the same as described for the amplification of IAP sequences, but with primers XMRV Forward outer and XMRV Reverse outer (). In order to clone and sequence the PCR product, the second round again made use of the XMRV Forward outer and XMRV Reverse outer primers, as these bind outside the XMRV-specific 24 bp deletion. The XMRV infectious clone, VP62 plasmid DNA, constituted the positive control. Using less stringent annealing conditions (50°C instead of 55°C), a PCR with primers XTP1 and MLV reverse outer, which targets the XMRV gag/pro/pol
open reading frame (ORF), amplified further products (, middle panel). Under the same conditions, a multiplex PCR with the four primers XMRV-R, XMRV Forward outer, XMRV Reverse outer and 1154R () that bind to the long terminal repeat, the leader/gag
sequence and the gag
ORF, also produced various amplicons (, lower panel). Using published primers to the XMRV env
, three out of ten columns tested produced an amplification product (not shown). At least four “no template” water controls were included in each experiment.
Several of these amplicons were cloned and sequenced. A GenBank database search indicated contaminating sequences of human and murine origin. These included two leader/gag regions displaying the 24 nt deletion described for XMRV (one sequence being identical, one having two mismatches and a 1 bp insertion, respectively), five leader/gag regions of pMLVs, all displaying a 9 bp deletion, one gag/pol ORF of pMLV or XMRV (8 mismatches and 10 mismatches out of 599 bp, respectively), and one sequence of the env gene of a murine enogenous polytropic retrovirus. In addition one DNA sequence had homology to mouse chromosome 11, two other DNA sequences showed homology to human chromosome 6 and 10. Two further sequences were of unspecified origin. Although these were not mapped to any particular gene, they are nevertheless an indication of contaminating DNA from a variety of sources. Alignments of leader/gag and env sequences obtained from the columns and VP62, the reference XMRV infectious clone (GenBank accession no. EF185282.1), are shown in .
Alignment of the infectious molecular clone of XMRV, VP 62, and sequences obtained from mock eluted columns.
Detection of Murine DNA in Other Columns
Columns of the QIAamp DNA Mini Kit were also mock-extracted as described in the manufacturer's instructions, but no IAP sequences were identified in the eluate. In a further experiment, carried out by a different operator in a separate laboratory, a column from the QiaAmp DNA Investigator Kit was dismantled. After soaking the DNA binding membrane in elution buffer overnight, samples of this elution buffer were amplified in a real time PCR (RT-PCR) specific for IAP. The RT-PCR was carried out using the Quantitect probe kit (Qiagen, UK) as per manufacturer's instructions, using the primers and probes described in . Reactions were in volumes of 10µl to which 2.5µl eluate was added and were amplified in a Biorad CFX96 thermal cycler. Qiagen 2x Quantitect probe mastermix was used (Qiagen, UK) with 2.5 pmol of each primer IAP for and IAP rev. In addition, 2.5 pmol IAP PROBE was added to each reaction. Cycling conditions were one cycle of 95°C 15 seconds followed by 60 cycles of 94°C 15 seconds, 60°C 15 seconds. As a positive control, DNA from McCoy cells was used. At least 6 “no template” controls were set up in each RT-PCR. The control sample of unexposed elution buffer remained negative, but an IAP-specific signal from elution buffer exposed to the column pieces was observed. Upon cloning, the sequence of the amplicon was confirmed to be IAP (data not shown). It is worth noting that QIAamp Ultraclean Production (UCP) Pathogen columns which are certified to be free of contaminating microbial DNA yielded no amplification product for IAP or MLV-related sequences in 50 “mock-extracted” columns.