Feasibility of Droplet Digital PCR Analysis of Plasma Cell-Free DNA From Kidney Transplant Patients

Abstract

Increasing research demonstrates the potential of donor-derived cell-free DNA (dd-cfDNA) as a biomarker for monitoring the health of various solid organ transplants. Several methods have been proposed for cfDNA analysis, including real-time PCR, digital PCR, and next generation sequencing-based approaches. We sought to revise the droplet digital PCR (ddPCR)-based approach to quantify relative dd-cfDNA in plasma from kidney transplant (KTx) patients using a novel pilot set of assays targeting single nucleotide polymorphisms that have a very high potential to distinguish cfDNA from two individuals. The assays are capable of accurate quantification of down to 0.1% minor allele content when analyzing 165 ng of human DNA. We found no significant differences in the yield of extracted cfDNA using the three different commercial kits tested. More cfDNA was extracted from the plasma of KTx patients than from healthy volunteers, especially early after transplantation. The median level of donor-derived minor alleles in KTx samples was 0.35%. We found that ddPCR using the evaluated assays within specific range is suitable for analysis of KTx patients' plasma but recommend prior genotyping of donor DNA and performing reliable preamplification of cfDNA.

Keywords: assay evaluation; droplet digital PCR; graft health monitoring; kidney transplantation; minor allele quantification; plasma cell-free DNA.

Figure 1

Comparison of cfDNA isolation from plasma using three different commercially available kits. cfDNA was isolated from 1 ml of plasma using each kit in three replicates. (A) The average yield for each plasma sample. Error bars show the standard deviation of three replicates. (B) Distribution of yield of cfDNA isolation with three kits. Colored shapes show the individual yields of all samples/replicates and boxplots show the average yields across replicates. No significant differences were observed between the yields obtained with the three kits tested. (C) Comparison of fragment sizes. The left panel shows the capillary electrophoresis gel for one of the three replicates of the same sample extracted with three different commercial kits. On the right, the electropherogram of the same gel shows that the fragment size profile is very comparable between the kits, with a major peak around 170 bp representing mainly the mononucleosomal cfDNA fraction and a minor peak around 340 bp (probably dinucleosomal cfDNA). CNA, QIAamp Circulating Nucleic Acid Kit; ME, QIAamp MinElute ccfDNA Kit; MM, MagMAX Cell-Free DNA Isolation Kit.

Figure 2

Yield of plasma cfDNA from kidney transplant (KTx) patients and its donor-derived minor allele content. (A) The KTx cfDNA yield was higher than that of healthy volunteers. Colored shapes show the yields of each sample and boxplots summarize their distribution. P-values of pairwise comparisons using Mann Whitney U-test and Benjamini-Hochberg adjustment are shown. (B) Minor allele content in plasma cfDNA of KTx patients. The detected minor alleles are from the donor, but the zygosity status of the donor DNA is not known, so the exact proportion of donor-derived DNA would be 2-fold in the case of heterozygous donor DNA. For two samples (subject ID 9131761, 6 days after KTx and subject ID 164113, 43 days after KTx), the minor allele proportion was determined using two SNP assays, hence two values for the minor allele proportion of each sample. Error bars represent Poisson confidence intervals. In three cases where the confidence intervals are wider (subject ID 9131761, 6 days after KTx and subject ID 451608, 19 days after KTx), minor allele proportions were determined in non-preamplified cfDNA samples. In the other cases, ddPCR analysis was performed after targeted multiplex PCR preamplification of cfDNA.