Printer Friendly

Noninvasive Prenatal Diagnosis of Single-Gene Disorders by Use of Droplet Digital PCR.

The presence of circulating cell-free DNA (cfDNA)7 of fetal and placental origin in maternal plasma has allowed the development of noninvasive tools to detect fetal genetic abnormalities from a maternal blood draw (1-3). Currently, noninvasive prenatal testing (NIPT) of common aneuploidies (e.g., Down syndrome) is clinically available as a screening test that can be performed as early as week 10 of pregnancy with false-positive rates <0.2% (2, 4, 5) and without the complications related to invasive testing (6, 7). More recently, NIPT has also become commercially available for some genomic micro-deletions (e.g., DiGeorge syndrome, Cri-du-chat syndrome) (8-10).

However, prenatal diagnosis of pregnancies at risk of single-gene disorders still requires the use of invasive techniques, such as amniocentesis or chorionic villus sampling. These methods have a risk of miscarriage, cause higher discomfort, and can be applied during only certain time windows of pregnancy (6, 7). Although commercial development of screening tests for single-gene disorders is difficult because of the low prevalence of each given mutation in the general population (hampering positive predictive values), the development of accurate NIPT could replace invasive testing and become a diagnostic test for parents who are carriers of a mutation and are at high risk of having an affected pregnancy (11). A recent carrier screening study including >350000 individuals has shown that the number of pregnancies at risk of 94 common severe single-gene disorders ranges between 95 and 392 per 100000 depending on ethnic background (12). Remarkably, these values are comparable with the prevalence of common aneuploidies such as trisomy 21 (13).

The development of noninvasive tools for these disorders is important, as it allows patients and doctors to make informed decisions in pregnancies at risk of severe conditions while reducing anxiety related to invasive or postnatal testing. In addition, early treatment is sometimes available for conditions that might otherwise cause irreversible damage to the fetus, such as metabolic disorders or congenital malformations (e.g., dietary treatment or neonatal surgery, respectively) (14--16). Finally, prenatal diagnostics might also prove useful to develop protocols for cord blood collection in view of potential cures of inherited single-gene disorders by using gene-editing techniques on hematopoietic stem cells (17, 18).

Detection of single-gene disorders is straightforward for paternally inherited mutations or common de novo mutations, in which the presence of a mutated allele in maternal plasma can be directly attributed to an affected fetus and not to background cfDNA of maternal origin (19--22). However, most common single-gene disorders are autosomal-recessive because of their deleterious nature; therefore, one must carefully quantify the ratio of mutant to wild-type alleles to genotype the fetus. This problem has been solved in principle by applying the counting principle to high-depth whole-exome sequencing or to full haplotypes (23), but this approach requires the use of sequencing and is more costly than digital PCR. Proof-of-concept studies with digital PCR have been conducted for several autosomal-recessive and X-linked disorders (24-30), but a general method to perform noninvasive diagnosis of these conditions is not yet available. Previous digital PCR studies have been limited in that they have not had large enough single-nucleotide polymorphism (SNP) panels to measure the fetal fraction in the general population, or have not had enough SNP measurements to estimate the error in measurement of fetal fraction.

Here we address these challenges by developing a simple droplet digital PCR (ddPCR) protocol to diagnose autosomal and X-linked single-gene disorders. This protocol is applied directly to the maternal cfDNA sample and does not require a separate maternal genotyping step. An accurate quantification of the fetal fraction is achieved by targeting a panel of 47 high-variability SNPs, and we show that the final measurement error in determining fetal genotype is composed of roughly equal contributions from the error in fetal fraction and the Poisson error because of counting statistics. We show how this method enables diagnosis of recessive single-gene disorders, both when they are because of a mutation shared by both progenitors or result from heterozygous compound mutations (when father and mother carry a different mutation affecting the same gene). Unambiguous results are shown for samples with a fetal fraction as low as 3.6%.

Materials and Methods

SAMPLE COLLECTION AND cfDNA EXTRACTION

Ten blood samples were collected from pregnancies at risk of a single-gene disorder. Samples were collected in cfDNA Streck tubes (3 tubes, approximately 30 mL). Blood was centrifuged at 1600g for 10 min, and the supernatant was centrifuged for an additional 10 min at 16000g to remove cellular debris. Plasma samples were aliquoted in 2-mL tubes and stored at -80 [degrees]C until further processing (cfDNA extraction). Maternal genomic DNA was extracted from the remaining cellular fraction using the Qiagen Blood Mini-kit (200-[micro]L aliquots) and stored for assay validation. Extraction of cfDNA from stored plasma samples was done using the Qiagen Circulating Nucleic Acid kit using the protocol recommended by the manufacturer with the following modifications: We performed an initial centrifugation of plasma for 3 min at 14000 rpm to remove cryoprecipitates; we extended the lysis step to 1 h (as recommended for Streck tubes); and we did not add carrier RNA, as we did not observe an increase in yield. Plasma was processed in batches of 5 mL per Qiagen column and eluted in 50 [micro]L of Tris-EDTA buffer.

QUANTIFICATION OF cfDNA IN PLASMA AND FETAL FRACTION

From the extracted cfDNA (approximately 150 [micro]L in total), we used 8.5 [micro]L (approximately 850 [micro]L of plasma) for a preamplification reaction targeting highly variable SNPs that we used to determine the fetal fraction of each sample. We selected 47 biallelic SNPs that show a high minor allele fraction [(MAF) >0.4] for all 5 super populations of the 1000 Genomes Project (East Asian, European, African, Ad Mixed American, South Asian) and that are not found in regions of structural variation or highly repetitive regions (filtered using UCSC Repeat-Masker and the Database of Genomic Variants). Commercially available SNP genotyping assays (ThermoFisher) were purchased for the selected SNPs (amplicon size <80 bp), as well as separate primers targeting each SNP region (see Table 1 in the Data Supplement that accompanies the online version of this article at http:// www.clinchem.org/content/vol64/issue2). An additional SNP TaqMan assay targeting the ZFXX [8] and ZFY genes in chromosomes X/Y was also included in the assay (31). The size of the SNP panel, threshold MAF, and chromosomal distribution of assays were designed to maximize the probability of making an accurate determination of the fetal fraction across a broad target population (see Fig. 1 in the online Data Supplement).

The preamplification reaction was performed using the TaqMan PreAmp Master Mix (Applied Biosystems, Ref. 4391128) with the pooled 48 primer pairs and the recommended conditions by the manufacturer (reaction volume, 50 [micro]L; final primer concentration, 45 nmol/L each; 11 preamplification cycles). The preamplified DNA was diluted 5 X with Tris-EDTA buffer and stored for ddPCR quantification.

Quantification of the fetal fraction and total amount of cfDNA was performed using ddPCR and standard conditions {reaction volume, 20 [micro]L; final primer [probe] concentration, 900 nmol/L [200 nmol/L]; thermal cycling: [10' 95 [degrees]C, 40X (30" 94 [degrees]C; 1' 60 [degrees]C); 10' 98 [degrees]C], ramp rate, 2 [degrees]C/s}. We used 1 [micro]L of the preamplified DNA for each SNP TaqMan assay reaction and 1 [micro]L of the original cfDNA for each quantification assay (see Tables 1 and 2 in the online Data Supplement). For the quantification assays, we used 2 multiplex assays targeting chromosomes 1, 5, 10, and 14 (see Table 2 in the online Data Supplement). Fetal fraction and quantification ddPCR assays were run in parallel in a single plate.

The amount of cfDNA per milliliter of plasma (in genomic equivalents) is determined as the mean of the 4 quantification assays. For the SNP assays, Poisson-corrected counts are determined as: [N.sub.FAM/VIC] = [N.sub.total] X ln[1--[N.sub.positive]/[N.sub.total]] (Eq. 1), where [N.sub.total] is the total number of droplets and [N.sub.positive] is the number of positive droplets for each channel (FAM or VIC). For each SNP assay, the minor allele fraction is extracted as MAF = min([N.sub.FAM], [N.sub.VIC])/([N.sub.FAM] + [N.sub.VIC]). The fetal fraction ([epsilon]) is determined from the median of all SNPs in which the fetus is heterozygous and the mother is homozygous (0.5% > MAF > 20%) using [epsilon] = 2MAF. Errors are determined as the SD and compared with the Poisson noise expected from the DNA input used in the preamplification reaction ([[delta][epsilon].sub.Poisson] = [square root of (2[epsilon]/[input DNA in preamp])]).

Results on Clinical Samples

For each sample, we initially measured the fetal fraction and total amount of cfDNA as detailed above. From these values, we determined the optimal split of sample between the paternal and maternal mutation, as well as the probability of obtaining an unambiguous result. Assays to detect inheritance of the mutations were designed and validated as described in Section 1 of the online Data Supplement. To test the paternal mutation, we used the amount of cfDNA expected to provide approximately 40 counts for a carrier fetus. This sets the result 6 SD away from the noncarrier case. The remaining sample was used to quantify the imbalance on the maternal mutation. The ddPCR measurements were run using standard conditions and optimal temperatures determined in the validation assays. For each assay, the total number of counts for each allele was determined using Eq. 1. The affected or unaffected status of the fetus was determined using a likelihood ratio classifier with a low threshold of p (X|[H.sub.1])/p(X|[H.sub.0]) = 1/8 and a high threshold of p(X|[H.sub.1])/p(X|[H.sub.0]) = 8, where p(X|[H.sub.1]) is the probability of this result coming from an affected fetus and p(X|[H.sub.0]) is the null hypothesis of a non-affected fetus (26, 27, 32).

Results

CLINICAL PROTOCOL AND VALIDATION OF ASSAYS

In this study, we enrolled pregnant patients who are carriers of mutations causing autosomal-recessive or X-linked disorders. We then followed the experimental protocol depicted in Fig. 1 to test whether the fetus is affected by the disease. For each pregnancy at risk of a known mutation, we designed primers to amplify the region of the mutation and TaqMan probes labeled with different fluorophores against the healthy and mutated allele at risk (i.e., single-nucleotide mutation, insertion, deletion). The assays were validated using genomic DNA (gDNA) of carriers and noncarriers of the mutation in ddPCR experiments (see Section 1 and Fig. 2 in the online Data Supplement). Carrier gDNA was obtained from nucleated cells from maternal blood. Once we enrolled a patient carrying a new target mutation, the limiting factor to perform a diagnostic was the ordering time of the TaqMan assays (approximately 1 week). To be able to validate the assays before maternal blood collection, we developed an alternative approach using mixtures of synthetic DNA fragments and spike in experiments (Fig. 2; see also Section 1 in the online Data Supplement). This allowed us to validate the TaqMan probes before sample collection and reduce the turnaround time of the assay to approximately 1 day.

For each incoming sample, we extracted cfDNA from approximately 30 mL of maternal blood (see Methods section and Fig. 1). We then performed a quantification assay of the total amount of cfDNA and fetal fraction using TaqMan assays targeting 4 genomic markers (cfDNA quantification) and a panel of 47 high-variability SNPs and an X/Y chromosome marker (fetal fraction determination; see Methods section). This information was used to decide whether a determinative result was possible and to determine the optimal split of sample to test the paternally and maternally inherited mutations in compound heterozygous conditions, as well as the CIs of the result.

QUANTIFICATION OF cfDNA AND FETAL FRACTION

We used approximately 7% of each sample to quantify the fetal fraction and total amount of cfDNA (see Methods section). For each sample, we measured the MAF for each SNP in the panel and determined the fetal fraction from the distribution of SNPs that are homozygous for the mother and heterozygous for the fetus, which are found in the range 0.5% < MAF < 15% (Fig. 3A and Methods section). These SNPs are the most informative to determine the fetal fraction, as the measured copies from the fetal allele are not affected by background noise coming from the more abundant maternal alleles. Interestingly, our assay also allowed us to discriminate SNPs that are heterozygous for the mother but homozygous for the fetus, which show a characteristic symmetric peak in the range 35% < MAF < 50% (Fig. 3A). Although we did not use these SNPs to calculate the fetal fraction because of its higher noise, they could also be used to improve the estimate if a reduced SNP panel is used (see Fig. 1 in the online Data Supplement). The total quantification of cfDNA was also obtained for each sample, as well as the sex of the fetus (Fig. 3A, insets). We compared the SD of the fetal fraction measurement with the expected noise from Poisson subsampling (as we are using a limited amount of sample for this measurement), finding good agreement between experimental measurement and theoretical expectation for all samples. This is relevant for samples with a low fetal fraction or total amount of cfDNA, for which additional noise in the fetal fraction measurement would affect the CIs of the diagnostic assay. As expected, the fetal fraction increases with gestational age (Fig. 3A), a result that is also consistently observed for individual SNPs of the panel (see Fig. 4 in the online Data Supplement). Finally, results for 12 different pregnancies show a distribution of maternal and fetal genotypes compatible with the expected values for high-variability SNPs, suggesting that our panel can be used to determine the fetal fraction in populations of different genetic background (Fig. 3B and Methods section).

DIAGNOSIS OF X-LINKED DISORDERS AND AUTOSOMALRECESSIVE DISORDERS

We first addressed the case of X-linked mutations, in which the carrier status of the mother poses a risk for pregnancies carrying a male fetus. We analyzed pregnancies at risk of mutations related to hemophilia A, hemophilia B, and ornithine transcarbamylase (OTC) deficiency. We designed TaqMan assays targeting these mutations (see Table 3 in the online Data Supplement) and validated them as explained in the Methods section. We then ran the validated assay for each sample and counted the Poisson-corrected number of mutated (NM) and healthy (NH) alleles in maternal plasma (Fig. 4, A and B). From the measured fetal fraction of each sample, we determined the ratio of mutated and healthy alleles that we would expect for an affected or an unaffected pregnancy, as well as its associated error (see Section 2.1 in the online Data Supplement) (25). We used this information to compute the expected distributions for an affected or an unaffected pregnancy and to compare them with the experimentally measured ratio (Fig. 4, C and D, blue and green distributions and dotted arrow). The affected or unaffected status of the fetus was determined from the probability of the measurement arising from each distribution using a likelihood ratio classifier (Fig. 4, C and D, and Methods section) (27, 32). Using this approach, we also analyzed 2 pregnancies at risk of OTC deficiency. First, we tested a noncarrier mother at risk because of gamete mosaicism (detected through a previously affected sibling), which we determined to be an unaffected pregnancy (see Fig. 5 in the online Data Supplement). Then we analyzed a pregnancy carrying a female fetus that we determined to be a carrier of the maternal mutation and, therefore, at a partial risk of postneonatal onset (see Fig. 6 in the online Data Supplement).

We next addressed the case of autosomal-recessive mutations in which both mother and father are carriers of the same mutation and, therefore, at a 25% risk of having an affected pregnancy. We analyzed pregnancies at risk of [beta]-thalassemia and mevalonate kinase deficiency (MKD). To perform the assay, we followed the same approach described for X-linked mutations but used the counts and distributions expected for an autosomal-recessive disorder (see Section 2.2 in the online Data Supplement) (24, 26, 30). We ran the assay for each maternal plasma sample and measured the number of mutated and healthy alleles (Fig. 5, A and B). For both samples, the measured ratio was within the CIs for an affected pregnancy (Fig. 5, D and E). The affected status of the MKD case was also confirmed in a sample collected later in pregnancy and with a higher fetal fraction (Fig. 5, C--F). All measurements were also confirmed in postnatal testing and found to agree with the NIPT.

DIAGNOSIS OF HETEROZYGOUS COMPOUND MUTATIONS

Finally, we address the case of single-gene disorders in which each parent carries a different mutation affecting the same gene. We first tested pregnancies at risk of muscle-type acetylcholine receptor (AChR) deficiency (mutations c459dupA and c753_754delAA) and cystic fibrosis (mutations AF508 and W1282X). The latter are the 2 most common mutations for cystic fibrosis in Ashkenazi Jews, with an estimated combined abundance >75% (33). For these conditions, we ran the assay for the paternal mutation using enough sample to observe approximately 40 counts of the mutated allele in an affected pregnancy. To determine this value, we used the combined information of the fetal fraction and total cfDNA abundance in maternal plasma. From Poisson statistics, this sets the expected result for a fetus that is a carrier of the mutation approximately 6 SDs away from a negative result (P < [10.sup.-12]; see Methods section). The remaining sample was used to detect inheritance of the maternal mutation (see Section 2.2 in the online Data Supplement). For each sample, we measured [N.sub.M] and [N.sub.H] for each mutation at risk (Fig. 6, A--D) and determined the genotype of the fetus from the probability of each measurement arising from a carrier or noncarrier using a likelihood ratio classifier (Fig. 6, E-H). Both pregnancies were determined to have an unaffected fetus, although the fetus at risk of AChR deficiency was determined to be a carrier of the maternal mutation, whereas the fetus at risk of cystic fibrosis was determined to be a carrier of the paternal mutation. Using this approach, we also analyzed a pregnancy at risk of GJB2-related DFNB1 nonsyndromic hearing loss (mutations c.71G>A and c.-23 + 1G>A) at week 16 of gestation [fetal fraction, 6.7% (0.5)], which we determined not to be a carrier of the mutated alleles (see Fig. 7 in the online Data Supplement).

Discussion

In this work we developed a direct ddPCR approach to test pregnancies at risk of X-linked and autosomal-recessive single-gene disorders for both single mutations and compound heterozygous mutations. The proposed protocol builds on previous work carried out by other laboratories and our own (25-28) and does not require extensive sample preparation or computational resources, being significantly simpler and more cost-effective than approaches using deep sequencing and maternal haplotyping (29, 34). We show that using this approach, noninvasive prenatal diagnosis can be performed in a clinical laboratory setting in approximately 1 day from sample collection. This is particularly relevant for single-gene disorders, for which samples typically come in sparsely and are rarely at risk for the same mutation. We validated the approach by correctly diagnosing pregnancies at risk of some of the most common point mutations, such as [DELTA]F508 (accounting for >70% of cystic fibrosis cases in Europe) (22, 35), as well as rare metabolic and neuromuscular disorders that had not yet been addressed using noninvasive techniques (e.g., OTC deficiency, MKD, or AChR deficiency). Early diagnosis of these metabolic disorders would improve management of the disease, especially in cases when early onset might lead to the accumulation of metabolites that cause irreversible organ damage and failure (36-38).

The volume of blood collected for this study (30 mL) is sufficient for accurate classification of samples with a fetal fraction down to 4%, which is comparable with current standards for NIPT (39, 40). We have compiled a table of expected test performance as a function of fetal fraction and blood draw (see Table 4 in the online Data Supplement), and these numbers are consistent with what we measure. For instance, in our most challenging sample (patient 4: fetal fraction, 3.6%; Fig. 5B), we used the whole sample and obtained approximately 20000 counts, which agrees with our blood draw (approximately 25--30 mL of blood) and is within a range of expected type I and type II errors of 0.2% to 1%. For certain single-gene disorders, the sensitivity of the assay could be increased by following high-variability SNPs close to the target mutation (using a similar multiplexing approach as the one used here for the fetal fraction determination). This might become a cost-effective approach for common conditions such as cystic fibrosis, although for many mutations, the collection of a moderate volume of blood (2 Streck tubes) should enable a correct classification of samples down to a 5% fetal fraction with false-positive and false-negative rates approximately 0.2% (in line with current standards for trisomy 21).

Finally, we also provide an accurate method to measure the fetal fraction and total amount of cfDNA in plasma samples using a multiplexed SNP panel for ddPCR. This approach is useful to establish CIs in NIPT of autosomal-recessive or X-linked diseases. As shown here, NIPT of these conditions relies on comparing the ratio of mutated and healthy alleles in maternal plasma with the ratios expected for a healthy or affected fetus, as determined from the sample fetal fraction. Overall, the use of an SNP panel instead of a single marker to measure fetal fraction (a) reduces false-positive and -negative rates, (b) reduces sample dropout because of a lack of indicative markers, and (c) simplifies the work flow, as an initial maternal genotyping step is not needed. Moreover, using our SNP panel we show that the errors associated with counting statistics at the mutation site and the errors associated with fetal fraction determination are comparable in size, highlighting the importance of accurately measuring the sample fetal fraction to implement NIPT of single-gene disorders.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following3 requirements: (a) significant contributions to the conception and design, acquisition Of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.

Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:

Employment or Leadership: None declared.

Consultant or Advisory Role: None declared.

Stock Ownership: None declared.

Honoraria: None declared.

Research Funding: J. Camunas-Soler, The Hearst Foundation; L. Hudgins, The Hearst Foundation; S.R. Hintz, The Hearst Foundation. Expert Testimony: None declared.

Patents: J. Camunas-Soler and S.R. Quake, a patent disclosure about this work has been filed with Stanford.

Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or final approval of manuscript.

Acknowledgments: The authors thank Ariana Spiegel, Anna Girsen, and Katie Sherwin for research coordination, as well as Yasemin Dilara Sucu, Thuy Ngo, and Kiran Kocherlota for help in collecting the plasma samples.

References

(1.) Fan HC, Blumenfeld YJ, Chitkara U, Hudgins L, Quake SR. Noninvasive diagnosis of fetal aneuploidy by shotgun sequencing DNA from maternal blood. Proc Natl Acad Sci USA2008;105:16266-71.

(2.) Chiu RW, Chan KC, Gao Y, Lau VY, Zheng W, Leung TY, etal. Noninvasive prenatal diagnosis of fetal chromosomal aneuploidy by massively parallel genomic se quencing of DNA in maternal plasma. Proc Natl Acad Sci USA2008;105:20458-63.

(3.) Fan HC, Quake SR. Detection of aneuploidy with digital polymerase chain reaction. Anal Chem 2007;79: 7576-9.

(4.) Van den Veyver IB. Recent advances in prenatal genetic screening and testing. F1000Res 2016;5:2591.

(5.) Gregg AR, Van den Veyver IB, Gross SJ, Madankumar R, Rink BD, Norton ME. Noninvasive prenatal screening by next-generation sequencing. Annu Rev Genomics Hum Genet 2014;15:327-47.

(6.) Akolekar R, Beta J, Picciarelli G, Ogilvie C, D'Antonio F. Procedure-related risk of miscarriage following amniocentesis and chorionic villus sampling: a systematic re view and meta-analysis. Ultrasound Obstet Gynecol 2015;45:16-26.

(7.) Tabor A, Alfirevic Z. Update on procedure-related risks for prenatal diagnosis techniques. Fetal Diagn Ther 2010;27:1-7.

(8.) Hui L, Bianchi DW. Noninvasive prenatal DNA testing: the vanguard of genomic medicine. Annu Rev Med 2017;68:459-72.

(9.) Dondorp W, de Wert G, Bombard Y, Bianchi DW, Bergmann C, Borry P, et al. Non-invasive prenatal testing for aneuploidy and beyond: challenges of responsible innovation in prenatal screening. Eur J Hum Genet 2015; 23:1438-50.

(10.) Wapner RJ, Babiarz JE, Levy B, Stosic M, Zimmermann B, Sigurjonsson S, etal. Expanding the scope of noninvasive prenatal testing: detection of fetal microdeletion syndromes. Am J Obstet Gynecol 2015;212:332.e1-9.

(11.) Verhoef TI, Hill M, Drury S, Mason S, Jenkins L, Morris S, Chitty LS. Non-invasive prenatal diagnosis (NIPD) for single gene disorders: cost analysis of NIPD and invasive testing pathways. Prenat Diagn 2016;36:636 -42.

(12.) Haque IS, Lazarin GA, Kang HP, Evans EA, Goldberg JD, Wapner RJ. Modeled fetal risk of genetic diseases identified by expanded carrier screening. JAMA 2016;316: 734-42.

(13.) Morris JK, Mutton DE, Alberman E. Revised estimates of the maternal age specific live birth prevalence of Down's syndrome. J Med Screen 2002;9:2-6.

(14.) Stevensson DK, Sunshine P, Benitz WE, Hintz SR, Druzin ML. Fetal and neonatal brain injury: mechanisms, management and the risks of practice. 4th Ed. New York (NY): Cambridge University Press; 2009.

(15.) Krakow D, Lachman RS, Rimoin DL. Guidelines for the prenatal diagnosis of fetal skeletal dysplasias. Genet Med 2009;11:127-33.

(16.) Milunsky A, Milunsky JM. Genetic disorders and the fetus: diagnosis, prevention, and treatment. Hoboken (NJ): John Wiley & Sons, Inc.; 2015.

(17.) Dever DP, Bak RO, Reinisch A, Camarena J, Washington G, Nicolas CE, et al. CRISPR/Cas9 [beta]-globin gene targeting in human haematopoietic stem cells. Nature 2016;539: 384-9.

(18.) Cox DBT, Platt RJ, Zhang F. Therapeutic genome editing: prospects and challenges. Nat Med 2015;21:121-31.

(19.) Chitty LS, Mason S, Barrett AN, McKay F, Lench N, Daley R, Jenkins LA. Non-invasive prenatal diagnosis of achondroplasia and thanatophoric dysplasia: next generation sequencing allows for a safer, more accurate, and comprehensive approach. Prenat Diagn 2015;35:656-62.

(20.) Orhant L, Anselem O, Fradin M, Becker PH, Beugnet C, Deburgrave N, etal. Droplet digital PCR combined with minisequencing, a new approach to analyze fetal DNA from maternal blood: application to the non-invasive prenatal diagnosis of achondroplasia. Prenat Diagn 2016;36:397-406.

(21.) Xiong L, Barrett AN, Hua R, Tan TZ, Ho SS, Chan JK, et al. Non-invasive prenatal diagnostic testing for [beta]-thalassaemia using cell-free fetal DNA and next-generation sequencing. Prenat Diagn 2015;35:258-65.

(22.) Debrand E, Lykoudi A, Bradshaw E, Allen SK. A noninvasive droplet digital PCR (ddPCR) assay to detect paternal CFTR mutations inthe cell-free fetal DNA(cffDNA) of three pregnancies at risk of cystic fibrosis via compound heterozygosity. PLoS One 2015;10:e0142729.

(23.) Fan HC, Gu W, Wang J, Blumenfeld YJ, El-Sayed YY, Quake SR. Non-invasive prenatal measurement of the fetal genome. Nature 2012;487:320-4.

(24.) Lun FM, Tsui NB, Chan KC, Leung TY, Lau TK, Charoenkwan P, etal. Noninvasive prenatal diagnosis of monogenic diseases by digital size selection and relative mutation dosage on DNA in maternal plasma. Proc Natl Acad Sci U SA2008;105:19920-5.

(25.) Tsui NB, Kadir RA, Chan KC, Chi C, Mellars G, Tuddenham EG, etal. Noninvasive prenatal diagnosis of hemophilia by microfluidics digital PCR analysis of maternal plasma DNA. Blood 2011;117:3684-91.

(26.) Gu W, Koh W, Blumenfeld YJ, El-Sayed YY, Hudgins L, Hintz SR, Quake SR. Noninvasive prenatal diagnosis in a fetus at risk for methylmalonic acidemia. Genet Med 2014;16:564-7.

(27.) Barrett AN, McDonnell TCR, Chan KCA, Chitty LS. Digital PCR analysis of maternal plasma for noninvasive detection of sickle cell anemia. Clin Chem 2012;58: 1026-32.

(28.) Chang MY, Kim AR, Kim MY, Kim S, Yoon J, Han JJ, et al. Development of novel noninvasive prenatal testing protocol for whole autosomal recessive disease using pico droplet digital PCR. Sci Rep2016;6:37153.

(29.) Lam KW, Jiang P, Liao GJ, Chan KC, Leung TY, Chiu RW, Lo YM. Noninvasive prenatal diagnosis of monogenic diseases by targeted massively parallel sequencing of maternal plasma: application to [beta]-thalassemia. Clin Chem 2012;58:1467-75.

(30.) New MI, Tong YK, Yuen T, Jiang P, Pina C, Chan KC, et al. Noninvasive prenatal diagnosis of congenital adrenal hyperplasia using cell-free fetal DNA in maternal plasma. J Clin Endocrinol Metab 2014;99:E1022-30.

(31.) Lun FM, Chiu RW, Chan KC, Leung TY, Lau TK, Lo YM. Microfluidics digital PCR reveals a higher than expected fraction of fetal DNA in maternal plasma. Clin Chem 2008;54:1664-72.

(32.) Lo YM, Lun FM, Chan KC, Tsui NB, Chong KC, Lau TK, etal. Digital PCR for the molecular detection of fetal chromosomal aneuploidy. Proc Natl Acad Sci U S A 2007;104:13116-21.

(33.) Bobadilla JL, Macek M, Fine JP, Farrell PM. Cystic fibrosis: a worldwide analysis of CFTR mutations-correlation with incidence data and application to screening. Hum Mutat 2002;19:575-606.

(34.) Chan KC, Jiang P, Sun K, Cheng YK, Tong YK, Cheng SH, et al. Second generation noninvasive fetal genome analysis reveals de novo mutations, single-base parental inheritance, and preferred DNA ends. Proc Natl Acad Sci U SA2016;113:E8159-68.

(35.) Hill M, Twiss P, Verhoef TI, Drury S, McKay F, Mason S, et al. Non-invasive prenatal diagnosis for cystic fibrosis: detection of paternal mutations, exploration of patient preferences and cost analysis. Prenat Diagn 2015;35: 950-8.

(36.) Peciuliene S, Burnyte B, Gudaitiene R, Rusoniene S, Drazdiene N, Liubsys A, Utkus A. Perinatal manifestation of mevalonate kinase deficiency and efficacy of anakinra. Pediatr Rheumatol Online J 2016;14:19.

(37.) Kayser M, de Knijff P. Improving human forensics through advances in genetics, genomics and molecular biology. Nat Rev Genet 2011;12:179 -92.

(38.) Mendez-Figueroa H, Lamance K, Sutton VR, Aagaard-Tillery K, Van den Veyver I. Management of ornithine transcarbamylase deficiency in pregnancy. Am J Perinatol 2010;27:775-84.

(39.) Gil MM, Quezada MS, Revello R, Akolekar R, Nicolaides KH. Analysis of cell-free DNA in maternal blood in-screening for fetal aneuploidies: updated meta-analysis. Ultrasound Obstet Gynecol 2015;45:249-66.

(40.) Gregg AR, Skotko BG, Benkendorf JL, Monaghan KG, Bajaj K, Best RG, et al. Noninvasive prenatal screening for fetal aneuploidy, 2016 update: a position statement of the American College of Medical Genetics and Genomics. Genet Med 2016;18:1056-65.

Joan Camunas-Soler, [1] Hojae Lee, [1] Louanne Hudgins, [2] Susan R. Hintz, [3] Yair J. Blumenfeld, [4] Yasser Y. El-Sayed, [4] and Stephen R. Quake [1,5,6] *

[1] Department of Bioengineering, Stanford University, Stanford, CA; [2] Division of Medical Genetics, Department of Pediatrics, Stanford University, Stanford, CA; [3] Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford University School of Medicine, Stanford, CA; [4] Division of Maternal-Fetal Medicine and Obstetrics, Department of Obstetrics and Gynecology, Stanford University, Stanford, CA; [5] Department of Applied Physics, Stanford University, Stanford, CA; [6] Chan Zuckerberg Biohub, San Francisco, CA.

* Address correspondence to this author at: James H. Clark Center E300, 318 Campus Dr., Stanford CA94305. Fax650-724-5473; e-mail quake@stanford.edu.

Received June 19,2017; accepted October 5,2017.

Previously published online at DOI: 10.1373/clinchem.2017.278101

[C] 2017 American Association for Clinical Chemistry

[7] Nonstandard abbreviations: cfDNA, cell-free DNA; NIPT, noninvasive prenatal testing; SNP, single-nucleotide polymorphism; ddPCR, droplet digital PCR; MAF, minor allele fraction; gDNA, genomic DNA; OTC, ornithine transcarbamylase; MKD, mevalonate kinase deficiency; AChR, acetylcholine receptor.

[8] Human Genes: ZFX, zincfinger protein, X-linked; ZFY, zinc finger protein, Y-linked; GJB2, gap junction protein beta 2; OTC, ornithine carbamoyltransferase.

Caption: Fig. 1. Protocol for noninvasive prenatal diagnostics of single-gene disorders.

Caption: Fig. 2. Validation of diagnostic assays with synthetic spike in controls (g-blocks). Temperature gradient of 1:1 mixtures of synthetic DNA fragments containing the mutant (FAM) and healthy (VIC) allele for mutation c. 835C>T in the OTC gene (Single Nucleotide Polymorphism Database, rs72558455) (A). The optimal temperature for the TaqMan assay in ddPCR experiments is highlighted in red. Spike in controls of the synthetic mutant allele (FAM) in fragmented gDNA of a healthy donor (B). Scatter plots of FAM/VIC fluorescence are shown in Figure 3 of the online Data Supplement. Quantification using ddPCR of varying amounts of spike in synthetic DNA (mutant allele) in a background of fragmented gDNA (approximately 5000 genome equivalents/reaction) from 2 different healthy donors (red, black) (C). Error bars are obtained from Poisson statistics.

Caption: Fig. 3. Quantification of cfDNA and fetal fraction determination. Histogram of the MAF for the 47 SNP assays used to determine the fetal fraction (A). Top (bottom) panel is results from a first (third) term sample of the same pregnancy. The fetal fraction is determined from SNPs that are homozygous for the mother and heterozygous for the fetus (found in the range 0.5% < MAF < 15%) and calculated as 2 x MAF. Agaussian fit to these SNPs is shown in blue. Inset boxes show the (a) quantification of cfDNA in the sample (determined as explained in the Methods section), (b) the fetal fraction and number of informative SNP assays (N), (c) expected error in the fetal fraction (as explained in the Methods section), and (d) sex determination assay. Errors are reported as SD. MAF of the 47 SNP assays for 12 different pregnancies (B). The right panel shows the frequency of each combination of maternal and fetal genotypes. The recovered distributions agree with the expected results for high-variability SNPs (heterozygous mother, approximately 50%; homozygous mother and fetus, about 25%; homozygous mother/heterozygous fetus, approximately 25%).

Caption: Fig. 4. Diagnosis of fetuses at risk of X-linked mutations. Measurement of total counts of mutant (FAM) and healthy (VIC) alleles in maternal plasma using ddPCR for pregnancies at risk of (A) hemophilia A and(B) hemophilia B. Clusters correspond to droplets positive for the mutant allele(blue), the healthy allele (green), both alleles (orange), or none (gray). [N.sub.M] and [N.sub.H] are the Poisson-corrected counts for the mutant and healthy alleles, respectively. Diagnostic test for each sample A and B using a likelihood ratio classifier (see Methods section) (C and D). The dotted arrow corresponds to the measured ratio of mutant allele. The expected distributions for a sample with fetal fraction s and carrying a healthy (affected) fetus are plotted in green (blue). The areas shaded in green and blue correspond to the ratios for which a fetus is determined to be healthy or affected using the ratio classifier (see Methods section). Fetal fraction s is reported as mean [+ or -] SE.

Caption: Fig. 5. Diagnosis of fetuses at risk of autosomal-recessive mutations. Measurement of total counts of mutant (FAM) and healthy (VIC) alleles in maternal plasma using ddPCR for pregnancies at risk of (A) [beta]-thalassemia and (Band C) MKD. Clusters correspond to droplets positive for the mutant allele (blue),the healthy allele (green), both alleles(orange), or none(gray). [N.sub.M] and [N.sub.H] are the Poisson- corrected counts for the mutant and healthy alleles, respectively. Diagnostic test for each sample A-C using a likelihood ratio classifier (see Methods section) (D-F). The dotted arrow corresponds to the measured ratio of mutant allele. The expected distributions for a sample with fetal fraction e and carrying a healthy (affected) fetus are plotted in green (blue). The areas shaded in green and blue correspond to the ratios for which a fetus is determined to be healthy or affected using the ratio classifier (see Methods section). Fetal fraction e is reported as mean [+ or -] SE.

Caption: Fig. 6. Diagnosis of fetuses at risk of combined paternal and maternal mutations for the same gene. Measurement of total counts of mutant (FAM) and healthy (VIC) alleles in maternal plasma using ddPCR for a pregnancy at risk of (A and B) AChR deficiency and (C and D) cystic fibrosis. Panels (A) and (C) correspond to the assay testing inheritance of the maternal mutation; panels (B) and (D) correspond to the assay testing inheritance of the paternal mutation. Clusters correspond to droplets positive for the mutant allele (blue), the healthy allele (green), both alleles (orange), or none (gray). [N.sub.M] and [N.sub.H] are the Poisson-corrected counts for the mutant and healthy alleles, respectively. Diagnostic test for the inheritance of each mutation in A-D using a likelihood ratio classifier (see Methods section) (E-H). The dotted arrow corresponds to the measured ratio of mutant allele. The expected distributions for a sample with fetal fraction s and carrying a healthy (affected) fetus are plotted in green (blue). The areas shaded in green and blue correspond to the ratios for which a fetus is determined to be healthy or affected using a ratio classifier (see Methods section). Fetal fraction s is reported as mean [+ or -] SE.
COPYRIGHT 2018 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Molecular Diagnostics and Genetics
Author:Camunas-Soler, Joan; Lee, Hojae; Hudgins, Louanne; Hintz, Susan R.; Blumenfeld, Yair J.; El-Sayed, Y
Publication:Clinical Chemistry
Date:Feb 1, 2018
Words:6227
Previous Article:Incidental Detection of Maternal Neoplasia in Noninvasive Prenatal Testing.
Next Article:Human Toxicity Caused by Indole and Indazole Carboxylate Synthetic Cannabinoid Receptor Agonists: From Horizon Scanning to Notification.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters