Quantifying cell free DNA in urine: comparison between commercial kits, impact of gender and inter-individual variation

Cell-free DNA is released from cells at cell death and also through active cell secretion [1–3]. cfDNA is readily available in various body fluids, such as blood plasma and urine [4–6]. One advantage of urine over blood is non-invasive sampling at the study probands’ or patients’ own choice of time, without the need for medical staff. Urine sampling could improve proband recruitment success and compliance, and avoids potential ethical problems in studies that involve children or patients in great pain. In contrast to blood cfDNA, there are insufficient data on the quantity and variability of cfDNA in urine. This fundamental information is required for designing urine cfDNA studies.

Here, we benchmarked two new commercial urine cfDNA isolation kits from Norgen and PerkinElmer with respect to cfDNA yields, replicability of yields from the same probands’ urine at sequential intervals, and potential differences between males and females. We chose these specific urine kits because the Norgen blood plasma cfDNA isolation kits showed the highest yields in a benchmark [7] and the PerkinElmer plasma cfDNA kit yields in our lab were even higher (data not shown).

Informed consent was obtained from all participants, consisting of healthy volunteers and cancer patients. The University Hospital of Schleswig-Holstein's ethics committee approved the patient information sheet and the consent form used for the study (B327/10, D470/14). All patients included in the study gave written informed consent to donate their samples to the biobank for research use. The research was conducted according to the principles of the Declaration of Helsinki. Initially, we collected samples from cancer patients and healthy volunteers in Urine Collection and Preservation Tubes (15 cc) (Norgen Biotek, Cat. 18120) and isolated cfDNA using the PerkinElmer kit and the Norgen kit. The cfDNA yields ranged from 0 to 86.5 ng cfDNA from 5 ml of urine (data not shown). It was unclear whether this variability was of biological nature.

Therefore we systematically compared the kits by examining the reproducibility of urine cfDNA yield from eight healthy volunteers (four females and four males) on 5 different days. They were informed that the second morning urine was required, as the first morning urine was reported to contain more degraded cfDNA [8]. 100 ml sterile disposible urine cups were distributed. Urine donation took place between 9am and 10am on each day of donation. No preservation agents were used. Urine processing started between 10am and 11am on the same day. The urine samples were centrifuged twice to remove cellular matter, first at 200 × g (10 min) followed by 1800 × g (10 min). Equal volumes from the four female or four male samples were pooled to obtain an average result for female or male donors without the cost of isolating each individual's cfDNA. 10 ml of each pool was used for Norgen-based cfDNA isolation and 4 ml for PerkinElmer-based isolation. Additionally, on the fifth day of donation, cfDNA was isolated from each individual's urine sample (F1–F4 and M1–M4). All isolations were performed according to manufacturers’ instructions. Length distribution and concentration of isolated DNA were analyzed using the Agilent TapeStation, according to the manufacturer's instructions. A DNA size range of 100 to 1000 bp was manually selected in the TapeStation Analysis Software A.01.04 to obtain the cfDNA concentration.

Figure 1 compares electropherograms between isolates from the same pooled female and male urine samples, respectively. Figure 2A and Table 1 summarize the cfDNA yield ranges and differences between kits in the female and male pools on five donation days. The mean yields of cfDNA (ng per ml of urine ± standard deviation) were 4.72 ± 3.53 (Norgen) and 14.83 ± 11.16 (PerkinElmer) for female pools and 0.83 ± 0.39 (Norgen) and 2.04 ± 0.33 (PerkinElmer) for male pools. The yield difference between kits was significant for male pools (p = 0.008, Wilcoxon test). The yield difference between genders was significant for the Norgen kit (p = 0.016, Wilcoxon test) and when both kits were considered (p = 0.003, Wilcoxon test). Figure 2B and Table 2 indicate yield ranges and kit-based differences for the eight probands’ individual urine samples that were isolated on the fifth day. The mean yields of cfDNA (ng per ml of urine ± standard deviation) were 0.98 ± 0.40 (Norgen) and 2.27 ± 1.17 (PerkinElmer) for female individuals and 1.46 ± 0.67 (Norgen) and 1.99 ± 1.02 (PerkinElmer) for male individuals. Finally, Table 3 summarizes the isolation cost and processing times.

In conclusion, the bead-based method was twice as fast as the column-based method and tended to yield more cfDNA per ml of urine. Larger sample numbers may lead to greater clarity. The urine cfDNA length profiles (Figure 1) suggest that the PerkinElmer kit is more efficient at capturing short DNA. Short DNA is of scientific interest, as it is present in blood plasma as cfDNA. In blood plasma, cfDNA lengths peak at about 165 nucleotides with a minor peak at about 1000 nucleotides [9]. In urine, we observed prominent DNA fractions longer than 165 nucleotides. These fractions of longer DNA fragments possibly arise from the epithelia of the urinary tract, after shedding and subsequent lysis of epithelial cells. Our gender-related differences in cfDNA yields correspond to a previous study on genomic DNA in urine [10,11]. Higher DNA yields in female urine have been previously reported [11–13] indicating that urine from females contained more epithelial cells than that from males.

To sum up, there is vast variation in DNA yields between different individuals and even for the same individuals on different days. When designing a study, we recommend ample amounts of urine to be collected. For NGS with 50 ng of cfDNA, we recommend collecting 60 ml urine for the PerkinElmer kit (15 extractions × 4 ml) or 70 ml for Norgen (7 extractions × 10 ml using the Midi kit or 3 × 30 ml using the Maxi kit).