WO2016123698A1

WO2016123698A1 - Diagnostic assay for post-transplant assessment of potential rejection of donor organs

Info

Publication number: WO2016123698A1
Application number: PCT/CA2016/050083
Authority: WO
Inventors: Steven GREENWAY; Paul Gordon
Original assignee: Uti Limited Partnership
Priority date: 2015-02-06
Filing date: 2016-02-01
Publication date: 2016-08-11

Abstract

A diagnostic assay for detection of rejection of a donor organ by a recipient comprising the steps of collecting a blood sample from the recipient, separating a cell- free DNA fraction from the blood sample, separating a fraction of fragmented cf DNA from the cell-free DNA fraction, generating a library of the fragmented cfDNA with selected primers. The library is quantified and sequenced. The resulting sequences are mapped to a reference. The mapped sequences are assessed to detect an increase or a decrease in the relative amount of one cfDNA population in a mixture.

Description

DIAGNOSTIC ASSAY FOR POST-TRANSPLANT

ASSESSMENT OF POTENTIAL REJECTION OF DONOR ORGANS

TECHNICAL FIELD

This disclosure relates to diagnostic assays. More specifically, this disclosure pertains to non-invasive diagnostic assays for post-transplant detection of organ rejection by a recipient.

BACKGROUND

Transplant rejection results from recipient immune activation against the donated organ leading to allograft damage. The diagnosis of rejection prompts an immediate change in management and is strongly associated with adverse outcomes. The gold standard for the detection of solid organ transplant rejection is an invasive biopsy and histologic tissue evaluation which has significant limitations both in the performance of the test (e.g. the need for anesthesia in children and the risk of serious complications) and its interpretation (e.g. inter-observer variability, non-specific changes). Furthermore, the timing of biopsies is relatively arbitrary and based on either time post-transplant, which may or may not correlate with increased risk of rejection, or clinical symptoms which may indicate advanced organ dysfunction. Non-invasive tests are safer and allow more frequent monitoring but current tests are generally too non-specific or insensitive for reliable rejection monitoring (McMinn et al. 2014, Biomarkers of acute rejection following cardiac transplantation. Biomark. Med. 8(6):815-832).

The urea cycle disorders are a group of rare but devastating diseases that lead to hyperammonemia which causes neurotoxicity and death. The only long-term treatment that can control elevated ammonia levels is liver transplantation. Liver transplants in young infants are frequently not feasible due to the lack of appropriately-sized grafts and lower success rates in children weighing less than 5 kg due to technical limitations (Nouj aim et al. 2002. Techniques for and outcome of liver transplantation in neonates and infants weighing up to 5 kilograms. J. Pediatr. Surg. 37(2): 159-164). A potential new therapeutic option is hepatocyte transplantation which involves infusing allogeneic liver cells into the portal vein with immune suppression to allow engraftment of functional donor cells and thereby reduce the risk of hyperammonemia and provide a "bridge" to liver transplant (Meyburg and Hoffmann. 2010. Liver, liver cell and stem cell transplantation for the treatment of urea cycle defects. Mol. Genet. Metab. 100 Sl :S77-83). Hepatocyte transplantation in North America is currently conducted as a clinical trial (CCD05; ClinicalTrials.gov identifier NCT01195753) and the conversion of ¹³C-acetate to ¹³C-urea is used to monitor ureagenesis in patients before and after hepatocyte transplantation (Yudkoff et al. 1998. In vivo measurement of ureagenesis with stable isotopes. J. Inherit. Metab. Dis. 21 Sl :21-29). However, this assay does not directly assess whether donor cells are present or reflect the health of the engrafted donor cells and currently in humans there is no direct method for the in vivo detection of donor liver cell engraftment.

Cell-free DNA (cfDNA) is released during cellular apoptosis and is found in the blood of all individuals (Tsang et al, 2007, Circulating nucleic acids in plasma/serum. Pathology 39(2): 197-207; Lo et al., 2011, Plasma nucleic acid analysis by massively parallel sequencing: pathological insights and diagnostic implications . J. Pathol. 225(3):318-23). Uniquely, cfDNA in the transplanted patient is derived from both recipient tissues and the donated organ or cells. Higher levels of circulating donor cfDNA in the recipient bloodstream have been found to correlate with rejection events in adult and pediatric heart transplant recipients in retrospective studies (Snyder et al., 2011, Universal noninvasive detection of solid organ transplant rejection. PNAS 108(15):6229-6234) and in prospective studies (De Vlaminck et al, 2014, Circulating cell-free DNA enables noninvasive diagnosis of heart transplant rejection. Sci. Transl. Med. 6(241): 1-8). However, although promising, these studies required a priori knowledge of donor and recipient genotypes and used whole-genome sequencing which remains costly and laborious. This methodology is currently a barrier to widespread use of cfDNA as a biomarker for allograft injury.

The potential benefits of non-invasive tests for detection of potential rejection of transplanted organs by resistance have been explored with a variety of blood biomarkers but such investigations have not been successful (McMinn et al, 2014). Sex-specific markers and assay using a limited number of DNA markers have also been assessed, but these attempts were constrained by their requirements for donor- recipient gender mismatch or donor and recipient genotyping (Sigdel et al, 2013, A rapid noninvasive assay for the detection of renal transplant injury. Transplantation 96(1):97-101 ; Hidestrand et al, 2014, Highly sensitive noninvasive cardiac transplant rejection monitoring using targeted quantification of donor -specific cell-free deoxyribonucleic acid. J. Am. Coll. Cardiol. 2014 Apr 1;63(12): 1224-1226).

SUMMARY

The present disclosure pertains to diagnostic assays and methods for noninvasive post-transplant assessments of the potential rejection of transplanted organs by recipients. The exemplary diagnostic assays and methods disclosed herein do not require a priori genotyping, nor do they require whole-genome sequencing prior to the performance of transplant procedures. The exemplary diagnostic assays require a post-transplant collection of a small amount of plasma from the organ recipient, for example 1 mL to 2 mL samples, and the methods can be rapidly completed

Disclosed herein is an exemplary diagnostic assay and method for non- invasive detection of post-transplant rejection of donor hearts by recipients.

Also disclosed herein is an exemplary diagnostic assay and method for noninvasive determination of post-transplant donor liver cell engraftment for the treatment of metabolic liver disease.

BRIEF DESCRIPTION OF THE FIGURES The present disclosure will be described in conjunction with reference to the following drawings in which:

FIG. 1 is a chart showing the results of whole-genome sequencing of 8 genomic DNA mixtures with the % donor DNA calculated for each sample based on knowledge of the donor and recipient genotypes;

FIG. 2 is a chart showing results for semiconductor sequencing of four genomic DNA mixtures with excellent correlation between expected and observed % donor DNA determined with an exemplary method disclosed herein; FIG. 3(A) is a micrograph of five plasma cfDNA preparations of genomic DNA, and FIG. 3(B) is a micrograph of the same five plasma cfDNA preparations after removal of contaminating genomic DNA with a size-selection step. The left-hand lane in each of the micrographs contains selected standards for reference; FIG. 4 is a chart showing error estimates for proportion of donor cfDNA present in an individual sample based on read depth numbers;

FIG. 5 is a chart showing the proportion of cfDNA present in the blood of a patient with X-linked ornithine transcarbamoylase deficiency before and after infusion of donor hepatocytes at selected post-transplant time intervals; FIG. 6 is a chart showing the proportion of cfDNA present in non-transplant controls, healthy pediatric heart transplant patients, and pediatric heart transplant patients with allograft dysfunction; and

FIG. 7 is a chart showing changes in the percentage of cfDNA levels detected in a single adult heart transplant recipient over a 4-month period. DETAILED DESCRIPTION

The exemplary embodiments of the present disclosure relate to diagnostic assays and methods for assessing the relative amount of a first source and a second source of cell-free DNA (cfDNA) within a mixed sample of cfDNA.

In one embodiment of the present disclosure, the assays and methods may be useful for post-transplant detection and/or assessment of potential rejection of an allografted cell, tissue or organ by a recipient. The diagnostic assays and methods are based on quantification of donor cfDNA isolated from a peripheral-blood sample from the recipient of an allografted cell, tissue or organ.

In other embodiments of the present disclosure, the assays and methods may be useful for detection and/or assessment of tumor-derived cfDNA isolated from a peripheral-blood sample from a subject. In other embodiments of the present disclosure, the assays and methods may be useful for mitochondria-derived cfDNA isolated from a peripheral-blood sample from a subject.

The exemplary diagnostic assays and methods disclosed herein address at least some of the problems associated with previously published work that focused on isolating genomic DNA (gDNA) or cfDNA from recipients' blood samples and subsequently using whole-genome sequencing to detect increasing amounts of one sample in a mixture, for example as taught by Snyder et al. (2011). However, those approaches are expensive, rely on genome-wide knowledge of both the donor and recipient genotypes and require intensive bioinformatics using both open-source software and custom programs written in-house.

Definitions:

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

The term "about" as used herein refers to an approximately +/-10% variation from a given value. It is to be understood that such a variation is always included in any given value provided herein, whether or not it is specifically referred to.

The term "allele" refers to a different form of a given gene that is caused by a mutation and that is located in a similar position as a non-mutant form of the same gene within a host's chromosome.

The term "allograft" refers to a cell, tissue or organ that is transplanted into a recipient where the donor and recipient are of the same species but have a different genetic makeup. The term "amplicon" refers to a sequence of deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) that are the source or product of an amplification or replication reaction or process that is either synthetic or naturally occurring. The term "autosome" refers to a pair of non-sex chromosomes that have the same form.

The term "BAM file" is a binary computer file that contains sequence alignment data. The terms "cfDNA" and "cell free DNA" are used interchangeably herein to refer to DNA that is found outside of a host's cell.

The term "gDNA" as used herein refers to genomic DNA that is chromosomal DNA within a host's cell.

The term "heterozygous" refers to a scenario when two different alleles of a gene are present on both chromosomes of chromosome pair.

The term "homozygous" refers to a scenario when identical alleles of a gene are present on both chromosomes of a chromosome pair.

The terms "locus" and the plural form "loci" refer to location of a gene or sequence of DNA within a host's chromosome. The term "primer" as used herein refers to a relatively short sequence of nucleic acids that are used as a starting template for DNA synthesis.

The terms "SNP" and "single nucleotide polymorphisms" are used interchangeably herein to refer to the genetic differences between hosts as represented by a comparative differences of one nucleotide within a given length of DNA. The exemplary diagnostic assays and methods disclosed herein are based on the detection of donor cfDNA in blood samples collected from transplant recipients by the use of an ion semiconductor sequencing system. One non-limiting example of an ion semiconductor sequencing system is the ION PGM^® semiconductor sequencing system (ION PGM is a registered trademark of Life Technologies Corp., Carlsbad, CA, USA). The ion semiconductor sequencing system can sequence a limited panel of 121 highly polymorphic SNPs that were originally developed for forensic identification of individual DNA samples. Relatively deep sequencing (at least 1,500-fold coverage) of selected high-quality autosomal SNPs (on average 42 SNPs) from this panel enabled the accurate detection of the relatively rare donor cfDNA molecules, thereby enabling rapid and accurate quantification of the proportion of donor alleles in a sample by the use of bioinformatics. Use of donor cfDNA as direct biomarkers for allograft injury removes the need for (i) a priori genotyping, (ii) whole-genome sequencing and (iii) intensive bioinformatics. The use of ion semiconductor sequencing facilitates library construction enabling the preparation of between 8 to 16 samples per day with current instrumentation and technologies. Use of ion semiconductor sequencing in the diagnostic assays and methods disclosed herein also equals or surpasses turnaround time for routine clinical biopsy interpretation because ion semiconductor sequencing can be completed within 4 hours as compared to the number of days currently required for whole-genome sequencing. Use of a small panel of SNPs contributes to the lower cost of the exemplary diagnostic assays disclosed herein since fewer reads are required to obtain sufficient coverage to accurately distinguish between a donor's cfDNA and the recipient's cfDNA. According to an embodiment of the present invention, an exemplary method comprises steps wherein the first step is selection of autosomal loci that are homozygous in the recipient's sample contributor. A locus is considered homozygous if high-throughput sequencing shows a consensus amongst more than 75% of the data with minimum base quality of 30 aligned reads at the given position in an input BAM file. Loci with more than 5% ambiguous DNA bases are also excluded for quality control purposes.

The second step is use of the non-consensus data at recipient homozygous loci collectively to estimate a donor's contribution through a statistical procedure known as Mixture Modeling with linear constraints. For a human sample with two contributors, biology dictates that the donor model should have three non-consensus signals i.e., (i) same base as recipient 100% of the time, (ii) same base as recipient 50% of the time, and (iii) same base as recipient 0% of the time. The "normalmixMMlc" method from the open source R package "Mixtools" (http://www.inside- r.org/packages/cran/mixtools) can be used to generate the models. Mixture modeling requires simulation, and therefore can yield different results for the same input data on different software runs. The mean and standard deviation of 10,000 simulations may be taken as the final result and the overall calculation will only takes a few minutes. The noise estimate is calculated using all samples run on the same chip.

The third step is addition of biologically-relevant constraints to the simulation parameters in order to improve the estimates and yield consistent results between runs. Since the panel consists of independent, highly genetically variable loci, a 1 :2: 1 intensity ratio is used based on random probability for the three signals i.e. (i) noise, (ii) heterozygous single-nucleotide polymorphisms (SNPs), and (iii) homozygous SNPs. A constrained ratio for signal standard deviations (e.g. not applicable 2:4) is also employed. The fourth step is deriving an estimate of the donor's contribution by taking the average of (i) the mean of the third signal from simulation results, and (ii) two times the mean of the second signal. The first signal (roughly equivalent to instrument noise) is subtracted from the donor estimate. The confidence interval of the third signal is used as a worst-case estimate of the accuracy of the model. It is preferable to use the Gaussian ("normal") distribution yielded more accurate results than the gamma distribution or various expectation maximization methods available in the Mixtools package.

According to another embodiment, an exemplary diagnostic assay comprises the steps of collecting a peripheral blood sample from a recipient of a transplanted donor organ, or alternatively, a recipient of an infusion of cells from a donor. Plasma is separated and recovered from the blood sample, for example, by centrifugation. If so desired, the separated plasma may be further clarified and or purified by one or more additional separation steps exemplified by centrifugation. The separated plasma may be stored, if so desired, at about -80° C before proceeding with the subsequent steps. Next, a cfDNA fraction is isolated from the plasma, after which, a fraction of fragmented cfDNA in the range of about 100 bp to about 300 bp is separated from the cfDNA fraction and the contaminating gDNA is removed. Then, a library of amplicons is generated with selected primers and quantified. The quantified library of amplicons is used to create template beads for sequencing. The resulting sequences are mapped to the human reference genome. The mapped sequences are assessed to detect an increase or a decrease in an amount of cfDNA in the blood sample. The following examples are included to illustrate some exemplary embodiments of the present disclosure.

EXAMPLE 1

Whole-genome sequencing of genomic DNA mixtures Two commercially-available DNA samples that have been extensively characterized as part of the HapMap Project (http : //hapmap . ncbi . nlm. nih. gov/) were used to replicate the use of whole-genome sequencing to distinguish two separate DNA populations in a mixture following the methods taught by Snyder et al. (2011). Genomic DNA (gDNA) were purchased from the Coriell Institute for Medical Research (Camden, NJ, USA) and sample NA07348 was designated as the "recipient" and NA10830 was designated the "donor". Eight samples were created with increasing concentrations of donor NA10830 DNA (0, 1, 1.5, 2, 2.5, 3, 4 and 8%). These sample concentrations were chosen based on previous work showing that donor cell-free DNA is present at baseline between 0-1% and increases with rejection to 3-6% (Snyder et al, 2011). The 8 samples then underwent whole-genome sequencing on a Life Technologies SOLID^® 5500x1 sequencer (SOLID is a registered trademark of Life Technologies Corp.) in 75+35 bp paired-end mode. Sequencing output was mapped to the GRCh37 human genome reference using the workflow "genomic.resequencing.pe" program in the LIFESCOPE^® 2.5 Genomic Analysis Software (LIFESCOPE is a registered trademark of Life Technologies Corp.). Resultant BAM files were then genotyped using the GATK Unified Genotyper software (version 2.5.2). Expected sample genotypes and SNP locations were downloaded from the HapMap Project site. The inventors focused on SNPs that were homozygous in the recipient to avoid any ambiguity or assumptions in identifying donor DNA. SNPs that were heterozygous or homozygous for the alternate allele were assigned to the donor.

Semiconductor sequencing of genomic DNA mixtures

Four mixtures of gDNA with increasing amounts of NA10830 (1, 2, 4 and 8%) as the 'donor' were mixed with NA07348 representing the 'recipient' using the ION PGM^® semiconductor sequencing platform. Genomic DNA mixtures were used for the ION AMPLISEQ^® library construction (ION AMPLISEQ is a registered trademark of Life Technologies Corp.) following the guidelines as specified for the HID-ION AMPLISEQ^® identity panel (Life Technologies Corp.). The prepared and barcoded library was quantified using the QUBIT^® assay (QUBIT is a registered trademark of Qubit Digital Ltd., London, UK) and the TAPESTATION^® assay (TAPESTATION is a registered trademark of Agilent Technologies Inc., Santa Clara, CA, USA), diluted to 20 pM, pooled and then used by the ION ONETOUCH^® 2 system (ION ONETOUCH is a registered trademark of Life Technologies Corp.) for the creation of template beads which were then loaded onto an Ion 316 semiconductor chip for sequencing on the ION PGM^® in 200 bp mode. The resultant sequences were mapped to the GRCh37 reference sequence using the OEM base caller with default parameters, and OEM tmap software using the stagel and map4 modes.

There was good correlation (R² = 0.98) between the expected and observed donor fractions with reasonable discrimination between the samples using the whole- genome sequencing (FIG. 1). Consequently, an alternative methodology disclosed herein was developed that was initially tested using similar mixtures of HapMap gDNA. In the initial proof-of-concept experiment, the method disclosed herein was able to accurately detect increasing amounts of an individual DNA sample in a gDNA mixture (FIG. 2) but at a fraction of the time and cost of the whole-genome sequencing method. There was excellent (R² = 0.999) correlation between the expected and measured donor DNA concentrations.

Then, the isolation steps for recovering cfDNA from plasma were optimized. Several different centrifugation speeds for the separation of plasma from whole blood were tested and, in general, found equivalence in cfDNA yield regardless of centrifugation speed used. Using a benchtop microfuge simplified isolation of plasma from the smaller blood volumes obtained from pediatric patients. Greatest and most reliable yields of cfDNA from plasma were obtained using the QIAAMP^® Circulating Nucleic Acid kit (QIAAMP is a registered trademark of Qiagen GMBH Corp., Holden, Fed. Rep. Germany). However, the presence of gDNA complicated estimates of cfDNA yield for all kits tested and in order to remove contaminating gDNA from before proceeding onto library construction, a size-selection step was added to remove large DNA molecules (FIG. 3). As shown in FIG. 4, estimates of donor DNA plateaued at approximately 1,500 reads per base. Therefore, since approximately 350 Mb of sequencing reads are obtained using the medium-sized 316 PGM chip and on average 42 SNPs are used per sample, 6-10 samples per chip can be accommodated.

EXAMPLE 2 An exemplary diagnostic assay and method according to the embodiments disclosed herein were used to monitor and assess the engraftment of infused donor liver cells in a male infant patient with X-linked ornithine transcarbamoylase deficiency (OTC) who received donor hepatocytes at a dose of 0.3 x 10⁹ viable human liver cells per kg body weight via a portal venous catheter using the procedures disclosed in the CCD05 trial (ClinicalTrials.gov identifier NCT01195753).

Peripheral whole blood samples (about 2 mL) were drawn into cfDNA blood collection tubes (Streck, Omaha, NE, USA) prior to the infusion of the donor liver cells and then at 1 week, 1 month, 3 months and 6 months post-infusion. The blood samples were centrifuged at 13,000 rpm at 4° C for 15 min, after which, the supernatant plasma was centrifuged again at 13,000 rpm at 4° C for 15 min. The supernatant was collected and stored frozen at -80° C until used.

Cell-free DNA was isolated from 1-2 ml of thawed plasma supematent using the QIAAMP^® Circulating Nucleic Acid kit (Qiagen). Several commercially-available kits were also tested including the QIAAMP^® MinElute Virus Spin Kit (Qiagen Canada, Montreal QC, CA), NUCLEOSPIN^® Blood (Macherey-Nagel GmbH, Montreal QC, CA) and the Plasma/Serum Cell-Free Circulating DNA Purification Mini Kit (Norgen Biotek, Welland, ON, CA) for plasma cfDNA yield. Isolated cfDNA was quantified using a QUBIT^® fluorometer (Life Technologies Corp) and a TAPESTATION^® 2200 D1000 tape (Agilent Technologies Inc., Santa Clara, CA, USA). Contaminating gDNA was removed by centrifugation of the cfDNA with magnetic AMPURE^® XP beads (AMPURE is a registered trademark of Beckman Coulter Inc., Brea, CA, USA) to produce a sample containing fragmented cfDNA of 100-300 bp. Size-selected cfDNA (5-10 ng) was then used to generate an ION AMPLISEQ^® library following the guidelines specified for the HID- ION AMPLISEQ identity panel. The prepared and barcoded library was then quantified using the QUBIT^® and TAPESTATION^® assays, diluted to 20 pM, pooled, and then used by the ION ONETOUCH^® 2 system for the creation of template beads which were then loaded onto an Ion 316 semiconductor chip for sequencing on the ION PGM^® in 200 bp mode. The resultant sequences were mapped to the GRCh37 reference sequence using the OEM base caller with default parameters, and OEM tmap software using the stage 1 and map4 modes.

Donor cfDNA was quantified at serial time points post-infusion to provide a measure of engraftment success (FIG. 5). There was a large increase in the amount of donor DNA detected at 1 week post-transplant infusion and then levels steadily declined to reach a steady-state of approximately 0.8% (at least two-fold higher than background noise represented by pre-infusion levels) by 3 months post-transplant. The presence of detectable donor cfDNA does correlate with lower post-infusion ammonia levels and patient freedom from requiring dialysis. The data in FIG. 5 show that background levels of donor DNA prior to donor hepatocyte infusion were low (as expected) followed by a rapid increase in % donor DNA detected at 1-week post-infusion. This may represent the early death of donor hepatocytes that did not successfully engraft into the recipient liver. This loss of cells continued at 1 month post-transplant but appear to be largely complete by 3 months post-infusion when % donor DNA reached a nadir. Stable engraftment appeared to be complete by 3 months based on similar levels of % donor DNA seen at 6 months post- infusion. This stable low level of % donor DNA after 3 months presumably represents basal apoptosis of engrafted and functional donor cells. This assay offers a mechanism for assessing donor hepatocyte engraftment and can be correlated with patient clinical parameters.

EXAMPLE 3

An exemplary diagnostic assay and method according to the embodiments disclosed herein were used to monitor and assess the presence of a donor's cfDNA in blood samples collected from pediatric heart transplant recipients (first study) and an adult heart transplant recipient (second study). This study was approved by the Conjoint Health Research Ethics Board at The University of Calgary (ID E-25116).

In the first study, peripheral whole blood samples (about 2 mL) were drawn from two healthy pediatric subjects (identified as "CI" and "C2" in FIG. 6) and from three pediatric subjects that received transplanted donor hearts (identified as ΉΤ ', "HT2", "HT3" in FIG. 6), at the times indicated for a clinical endomyocardial biopsy, into cfDNA blood collection tubes (Streck).

In the second study, peripheral whole blood samples (about 2 mL) were drawn from the adult heart transplant recipient over a two-month period commencing three weeks after receiving the donor heart. The blood samples were collected at routine clinically-indicated surveillance endomyocardial biopsies (refer to FIG. 7)

For both studies, the whole blood samples were centrifuged at 13,000 rpm at 4° C for 15 min after which, the plasma was removed and centrifuged again at 13,000 rpm at 4° C for 15 min. The supernatant was collected and stored frozen at -80° C until used. Cell-free DNA was isolated from 1-2 ml of thawed plasma supematent using the QIAAMP^® Circulating Nucleic Acid kit (Qiagen). Several commercially-available kits were also tested including the QIAAMP^® MinElute Virus Spin Kit (Qiagen Canada, Montreal QC, CA), NUCLEOSPIN^® Blood (Macherey-Nagel GmbH, Montreal QC, CA) and the Plasma/Serum Cell-Free Circulating DNA Purification Mini Kit (Norgen Biotek, Welland, ON, CA) for plasma cfDNA yield. Isolated cfDNA was quantified using a QUBIT^® fluorometer (Life Technologies Corp) and a TAPESTATION^® 2200 D1000 tape (Agilent Technologies Inc., Santa Clara, CA, USA).

Contaminating gDNA was removed by centrifugation of the cfDNA with magnetic AMPURE^® XP beads (AMPURE is a registered trademark of Beckman Coulter Inc., Brea, CA, USA) to produce a sample containing fragmented cfDNA of 100-300 bp. Size-selected cfDNA (5-10 ng) was then used to generate an ION AMPLISEQ^® library following the guidelines specified for the HID- ION AMPLISEQ^® identity panel. The prepared and barcoded library was then quantified using the QUBIT^® and TAPESTATION^® assays, diluted to 20 pM, pooled, and then used by the ION ONETOUCH^® 2 system for the creation of template beads which were then loaded onto an Ion 316 semiconductor chip for sequencing on the ION PGM in 200 bp mode. The resultant sequences were mapped to the GRCh37 reference sequence using the OEM base caller with default parameters, and OEM tmap software using the stage 1 and map4 modes. In the first study, relatively low levels of plasma donor cfDNA were seen for all heart transplant patients (i.e., "HTl", "HT2", "HT3" in FIG. 6). This was expected since none of the three recipients showed evidence of acute cellular rejection (ACR) or antibody-mediated rejection (AMR). All three heart recipients were classified as "ACR OR" and "pAMR 0" using the ISHLT grading criteria (Stewart et al, 2005, Revision of the 1990 working formulation for the standardization of nomenclature in the diagnosis of heart rejection. J. Heart Lung Transpl. 24(11): 1710-1720; Kobashigawa et al. 2011. Report from a consensus conference on antibody-mediated rejection in heart transplantation. J. Heart Lung Transpl. 30(3):252-269). However, the levels of donor cfDNA were significantly higher (p=0.02 by Student t-test) than the background noise seen in the non-transplant controls (FIG. 6).

In the second study, the levels of cfDNA observed in the adult donor heart recipient remained low, stable, and did not differ significantly between biopsies graded OR or 1R (none or mild acute cellular rejection with a single focus of lymphocytes) and negative for antibody -mediated rejection (FIG. 7). The data generated in the two studies demonstrate that low levels of % donor

DNA can be detected even in the absence of allograft dysfunction or rejection and that these levels are distinct from non-transplant controls (representing background noise of the assay). One concern that has been raised is that levels of cfDNA may fluctuate over time. In the serial samples collected from the single adult donor heart recipieint in the second studies indicates that this should not be a major problem. C Biopsy results showing either none (ACR OR) or mild (ACR 1R) rejection do not appear to have significant differences in levels of donor cfDNA (FIG. 7). This finding is consistent with the clinical practice of treating ACR OR and 1R as equivalent and not treating or altering management. In summary, the methods disclosed herein can be applied to any situation wherein a mixture of cfDNA molecules can be derived from samples collected from subjects. The relative contribution of each of the sources of the mixture of cdDNA may also be determined. It is also within the scope of the present disclosure for use of the methods disclosed herein for detection of mutations present in cancerous tumours. Tumours undergo cellular apoptosis and release nucleic acids into the bloodstream. Cell-free DNA derived from samples isolated from patients with cancer, particularly solid tumours, will enable identification of DNA mutations present in the tumour but absent from the DNA of the patient. Identification of these pathologic mutations will aid in cancer diagnosis and monitoring for post-treatment recurrence of cancerous growths. Tumour sequencing will identify the pathologic driver mutations that are unique to the tumour. Gene-specific or mutation-specific panels can then be created for a particular patient or for a particular type of malignancy. Cell-free DNA from the patient can then be extracted and the relative proportion of tumour cell-free DNA determined using our method. Levels of circulating tumour cell-free DNA may correlate with tumour burden and assess response to treatment.

It is also within the scope of the present disclosure for use of the methods disclosed herein for detection of mitochondrial DNA mutations. A panel specific for mitochondrial genes can be created and then the relative proportion of diseased and healthy mitochondria can be determined using our method. This may have utility in monitoring disease exacerbations or progression in mitochondrial diseases.

Claims

1. A diagnostic assay for assessing a contribution of a first source and a contribution of a second source of cell-free DNA (cfDNA) to a mixed cfDNA fraction, the assay comprising steps of:

collecting a blood sample from a subject;

separating the mixed cfDNA fraction from the blood sample;

separating a fraction of fragmented cfDNA from the mixed cfDNA fraction; generating a library of the fraction of fragmented cfDNA with one or more selected primers;

quantifying said library;

sequencing said library to prepare a set of one or more sequences;

mapping the set of one or more sequences to a reference assembly;

assessing the mapped one or more sequences to detect an increase or a decrease in an amount of cfDNA in the blood sample; and

assigning the mapped one or more sequences to either the first source or the second source.

2. The diagnostic assay according to claim 1, further comprising steps of:

selecting at least two homozygous autosomal loci in the subject, wherein the subject is the first source of cfDNA; and

using non-consensus data at the at least two homozygous autosomal loci to estimate the second source's contribution to the fraction of fragmented cfDNA.

3. The diagnostic assay according to claim 1, further comprising steps of:

adding biologically-relevant constraints for assessing a contribution of each of noise, heterozygous single-nucleotide polymorphisms, and homozygous single- nucleotide polymorphisms in the mapped one or more sequences; and deriving an estimate of the second source's contribution to the fraction of fragmented cfDNA.

4. The diagnostic assay of claim 1, wherein the sequencing of the library is performed using an ion semiconductor sequencing system.

5. The diagnostic assay of claim 1, wherein the step of separating the fraction of fragmented cfDNA from the mixed cfDNA fraction further comprises a step of removing genomic DNA (gDNA) by a step of centrifuging with magnetic beads.

6. The diagnostic assay of claim 5, wherein the fraction of fragmented cfDNA is about 100 to about 300 base pairs in length.

7. The diagnostic assay of claim 1, wherein the step of separating the mixed cfDNA fraction from the blood sample comprises a first centrifugation step and a second centrifugation step of a supernatant produced by the first centrifugation step, wherein a second supernatant produced by the second centrifugation step comprises the mixed cfDNA fraction.

8. The diagnostic assay of claim 7, wherein the first and second centrifugation steps are performed at about 4° C, for about 15 minutes at about 13,000 rpm.

9. Use of the diagnostic assay of claim 1, wherein the first source is a recipient of an allograft and the second source is a donor of the allograft.

10. The use of claim 9, wherein the allograft is selected from a group consisting of a plurality of donor cells, donor tissue and a donor organ.

11. The use of claim 10, wherein the plurality of donor cells is a plurality of donor hepatocytes.

12. The use of claim 9, wherein the recipient is a pediatric subject and the allograft is a donor heart.

13. The use of claim 9, wherein the recipient is a non -pediatric subject and the allograft is a donor heart.

14. Use of the diagnostic assay of claim 1, wherein the first source is a subject and the second source is a tumour.

15. The use of claim 12, wherein the tumour is a solid tumour.

16. Use of the diagnostic assay of claim 1, wherein the first source is one or more healthy mitochondria and the second source is a one or more unhealthy mitochondria.