Decrease DNA contamination in the laboratories

It is crucial to avoid contamination in forensic laboratories. Hypochlorite is known to be very efficient in removing any traces of DNA and this chemical is therefore commonly used for cleaning in many DNA laboratories [, , ]. However, hypochlorite can produce poisonous chlorine gases when it is reacting with acidic solutions and key components in several commercial extraction kit [].

In this study, we tested alternative approaches to remove DNA from hard laboratory surfaces. We contaminated clean surfaces with four different concentrations of massively parallel sequencing (MPS) amplicon libraries and left the DNA on the surfaces for 45 min or 24 h, respectively, to simulate a working laboratory environment. The experiment was conducted in a room that had never been used for laboratory work and should be free of PCR amplicons. After cleaning, the surfaces were swabbed and DNA was extracted from the swabs. The DNA was subsequently quantified with real-time PCR using a commercial quantification kit for Illumina MPS libraries.

MPS amplicon libraries were chosen because the adapters are known and therefore primers aligning to these adapters are already designed to amplify all fragments in the pool at once. The adapters used are identical for all fragments for each of the MPS platforms and this can cause contamination between sample preparations. Many laboratories were facing the issue of contamination in the early 1990’s when the PCR technique was at its infancy. It was reported that the extensive amplification of small fragments could lead to false-positive results [, , ]. As a possible solution to the contamination problem was recommended among other things to separate pre- and postPCR physically and to use different pipettes for the two laboratories [] and this has been the procedure in many laboratories since. With MPS, another dimension to the same problem is added given that now sequencing can also possibly lead to contamination in laboratories where the amplicons are already present. With the use of the same adapters every time, each sequencing run can possibly contaminate the next. This means that we need to take extra precautions and carefully clean with chemicals such as bleach that is removing the nucleotides completely.

ForenSeq™ DNA Signature Prep Kit (Illumina) libraries were constructed according to the manufacturer’s protocol. The DNA libraries were quantified using the Qubit 3.0 (Thermo Fisher Scientific) and 10 ng, 1 ng, 500 pg, and 100 pg were pipetted on to a hard surface in clearly separated positions. A clean surface in a room that had never been used for laboratory work was used. Squares of 2 cm² were cut out in paper to mark the positions. Droplets of 10 μL of ForenSeq™ amplicon libraries were added to each of the squares and left for either 45 min or for 24 h. The surfaces in the squares were cleaned using either wipes containing (1) water, (2) water followed by 96% ethanol, (3) 96% ethanol, (4) 10% Klorrent, (5) 3% Klorrent, or (6) the surface was not cleaned. Klorrent is not a pure solution of sodium hypochlorite but contains 30–60% hypochlorite according to the chemical datasheet [], thus, 10% Klorrent and 3% Klorrent contained 3–6% hypochlorite and 0.9–1.8% hypochlorite, respectively. After cleaning, one cotton swab with 20 μL molecular grade water was used for swabbing the surface in each square. Subsequently, the cotton swabs were extracted using the QIAamp^® DNA Blood Mini Kit (Qiagen) and the DNA Purification for Buccal Swabs (Spin Protocol). Finally, the extracts were quantified by real-time PCR using the GeneRead Library Quant kit (Qiagen). The final results were visualised using R version 2.11.0 and the R package ggplot2.

The results are shown in Fig. 1. High concentrations of MPS amplicon libraries were recovered from surfaces that had not been cleaned prior to sample collection. However, DNA was also recovered from cleaned areas. Leaving the libraries 24 h on the surface reduced the amount of amplifiable DNA approximately four times compared to the 45 min incubation, e.g. the concentration of the MPS libraries in the areas cleaned with 96% EtOH were 7.5 pM and 2 pM after 45 min and 24 h incubation. Cleaning the surfaces with ethanol reduced the amount of DNA further and the concentrations were approximately 5 times lower than those obtained from the surfaces that were not cleaned. Water was even more efficient and reduced the amount of amplifiable DNA 100–200 times. Treatment with water or water followed by ethanol gave very similar results which indicated that wiping the surface with ethanol after cleaning with water had very limited effect. Finally, treatment with hypochlorite solutions (both 3% and 10% Klorrent) removed all traces of amplifiable DNA.

All of the cleaning methods used here removed some DNA from the surfaces. Water was clearly a better cleaning solution than ethanol, however, the hypochlorite solutions were superior and removed all traces of amplifiable DNA. In general, the measured concentrations were lower after 24 h incubation compared to the 45 min incubation, which indicates that the DNA was degraded slowly by the heat or the sunlight in the room. However, cleaning of the surface with hypochlorite was needed to remove the contaminating DNA.

This demonstrates how important it is to keep laboratory spaces clean. An accidental spill, of e.g. ForenSeq™ amplicon libraries as the ones used in this study, will be able to contaminate other samples handled in the same space and is likely to be preserved for weeks unless the area is cleaned properly. This also means that contamination may accumulate over time and that the risk of observing contamination will increase with the number of experiments that are conducted in the same space.

Based on these results, it is recommended to use hypochlorite solutions (with 1–2% active hypochlorite) to clean laboratory surfaces in forensic laboratories.

The authors declare no conflict of interest.

The authors would like to thank Anja Ladegaard Jørgensen for laboratory assistance together with Jeppe Dyrberg Andersen and Mikkel Meyer Andersen for helping with R.