Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Cell-free DNA methylome analysis for early preeclampsia prediction

Nature Medicine volume 29, pages 2206–2215 (2023)Cite this article

Subjects

Abstract

Preeclampsia (PE) is a leading cause for peripartal morbidity, especially if developing early in gestation. To enable prophylaxis in the prevention of PE, pregnancies at risk of PE must be identified early—in the first trimester. To identify at-risk pregnancies we profiled methylomes of plasma-derived, cell-free DNA from 498 pregnant women, of whom about one-third developed early-onset PE. We detected DNA methylation differences between control and PE pregnancies that enabled risk stratification at PE diagnosis but also presymptomatically, at around 12 weeks of gestation (range 9–14 weeks). The first-trimester risk prediction model was validated in an external cohort collected from two centers (area under the curve (AUC) = 0.75) and integrated with routinely available maternal risk factors (AUC = 0.85). The combined risk score correctly predicted 72% of patients with early-onset PE at 80% specificity. These preliminary results suggest that cell-free DNA methylation profiling is a promising tool for presymptomatic PE risk assessment, and has the potential to improve treatment and follow-up in the obstetric clinic.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: PE-associated cfDNAme changes at time of diagnosis.
Fig. 2: cfDNAme changes presymptomatically predict PE.
Fig. 3: Characteristics of regions with cfDNAme differences predicting PE onset.
Fig. 4: Validation of cfDNAme changes predicting PE onset.
Fig. 5: Complementarity of cfDNAme-based model and an existing PE risk model.

Similar content being viewed by others

Data availability

Methylation sequencing data are available through the European Genome-Phenome Archive under study no. EGAS00001007071. Reuse of these genomic data is possible pending approval by the local UZ Leuven data access committee, which checks informed consent form compliance and ensures that there are no legal impediments. Conflicts are handled by an independent UZ Leuven data access committee advisory board. Publicly available, processed WGBS data aligned to the hg38 genome build used in this study were obtained from the ENCODE data portal with the following accession nos.: ENCFF241AQC, ENCFF122LEF, ENCFF266NGW, ENCFF521DHD, ENCFF923CZC, ENCFF103DNU, ENCFF318AMC, ENCFF477GKI, ENCFF223LJW, ENCFF497IYX, ENCFF513ITC, ENCFF536RSX, ENCFF560SMW, ENCFF684JHX, ENCFF831OYO, ENCFF039JFT, ENCFF477AUC, ENCFF733EFJ, ENCFF842MHJ, ENCFF435SPL, ENCFF489CEV, ENCFF497YOO, ENCFF534RNT, ENCFF811QOG, ENCFF844EFX, ENCFF200MJQ, ENCFF333OHK, ENCFF526PFA, ENCFF550FZT, ENCFF730NQT, ENCFF157POM, ENCFF424XKF and ENCFF455TQO. Published EPIC array data used in this study were acquired from the NCBI Gene Expression Omnibus with the following accession nos.: GSE162984, GSE159526, GSE167998 and GSE181034.

Code availability

Custom code used for data analysis is available through our GitHub page under the following URL: https://github.com/FunctionalEpigeneticsLab/Preeclampsia_cfDNAme.

References

  1. Dubon Garcia, A. et al. Cost-utility of a first-trimester screening strategy versus the standard of care for nulliparous women to prevent pre-term pre-eclampsia in Belgium. Pregnancy Hypertens. 25, 219–224 (2021).

    Article  PubMed  Google Scholar 

  2. Lisonkova, S. & Joseph, K. S. Incidence of preeclampsia: risk factors and outcomes associated with early- versus late-onset disease. Am. J. Obstet. Gynecol. 209, 544.e1–544.e12 (2013).

  3. Wojtowicz, A. et al. Early- and late-onset preeclampsia: a comprehensive cohort study of laboratory and clinical findings according to the new ISHHP criteria. Int. J. Hypertens. 2019, 4108271 (2019).

  4. Roberge, S., Bujold, E. & Nicolaides, K. H. Aspirin for the prevention of preterm and term preeclampsia: systematic review and metaanalysis. Am. J. Obstet. Gynecol. 218, 287–293 (2018).

    Article  CAS  PubMed  Google Scholar 

  5. Rolnik, D. L. et al. Aspirin versus placebo in pregnancies at high risk for preterm preeclampsia. N. Engl. J. Med. 377, 613–622 (2017).

    Article  CAS  PubMed  Google Scholar 

  6. Prasad, S. et al. Performance of Fetal Medicine Foundation algorithm for first trimester preeclampsia screening in an indigenous south Asian population. BMC Pregnancy Childbirth 21, 805 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Wright, D., Wright, A. & Nicolaides, K. H. The competing risk approach for prediction of preeclampsia. Am. J. Obstet. Gynecol. 223, 12–23 (2020).

    Article  CAS  PubMed  Google Scholar 

  8. Zwertbroek, E. F. et al. Performance of the FMF first-trimester preeclampsia-screening algorithm in a high-risk population in the Netherlands. Fetal Diagn. Ther. 48, 103–111 (2021).

    Article  PubMed  Google Scholar 

  9. Bartsch, E., Medcalf, K. E., Park, A. L. & Ray, J. G. Clinical risk factors for pre-eclampsia determined in early pregnancy: systematic review and meta-analysis of large cohort studies. BMJ 353, i1753 (2016).

  10. Matsuo, K. et al. Late postpartum eclampsia: report of two cases managed by uterine curettage and review of the literature. Am. J. Perinatol. 24, 257–266 (2007).

    Article  PubMed  Google Scholar 

  11. Moore-Maxwell, C. A. & Robboy, S. J. Placental site trophoblastic tumor arising from antecedent molar pregnancy. Gynecol. Oncol. 92, 708–712 (2004).

    Article  PubMed  Google Scholar 

  12. Founds, S. A. et al. Gene expression in first trimester preeclampsia placenta. Biol. Res. Nurs. 13, 134–139 (2011).

    Article  PubMed  Google Scholar 

  13. Chaiworapongsa, T., Chaemsaithong, P., Yeo, L. & Romero, R. Pre-eclampsia part 1: current understanding of its pathophysiology. Nat. Rev. Nephrol. 10, 466–480 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Bianchi, D. W. & Chiu, R. W. Sequencing of circulating cell-free DNA during pregnancy. N. Engl. J. Med. 379, 464–473 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Vermeesch, J. R., Voet, T. & Devriendt, K. Prenatal and pre-implantation genetic diagnosis. Nat. Rev. Genet. 17, 643–656 (2016).

    Article  CAS  Google Scholar 

  16. Ashoor, G., Syngelaki, A., Poon, L., Rezende, J. C. & Nicolaides, K. Fetal fraction in maternal plasma cell‐free DNA at 11–13 weeks’ gestation: relation to maternal and fetal characteristics. Ultrasound Obstet. Gynecol. 41, 26–32 (2013).

    Article  CAS  PubMed  Google Scholar 

  17. Lo, Y. D. et al. Presence of fetal DNA in maternal plasma and serum. Lancet 350, 485–487 (1997).

    Article  CAS  PubMed  Google Scholar 

  18. Lehmann-Werman, R. et al. Identification of tissue-specific cell death using methylation patterns of circulating DNA. Proc. Natl Acad. Sci. USA 113, E1826–E1834 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Snyder, M. W., Kircher, M., Hill, A. J., Daza, R. M. & Shendure, J. Cell-free DNA comprises an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164, 57–68 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Wu, Y. et al. Association between levels of total cell-free DNA and development of preeclampsia—a literature review. AJP Rep. 11, e38–e48 (2021).

  21. Moufarrej, M. N. et al. Early prediction of preeclampsia in pregnancy with cell-free RNA. Nature 602, 689–694 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Gekas, J., Boomer, T. H., Rodrigue, M. A., Jinnett, K. N. & Bhatt, S. Use of cell-free signals as biomarkers for early and easy prediction of preeclampsia. Front. Med. 10, 1191163 (2023).

  23. Rolnik, D. L., da Silva Costa, F., Lee, T. J., Schmid, M. & McLennan, A. C. Association between fetal fraction on cell-free DNA testing and first-trimester markers for pre-eclampsia. Ultrasound Obstet. Gynecol. 52, 722–727 (2018).

    Article  CAS  PubMed  Google Scholar 

  24. Gerson, K. D. et al. Low fetal fraction of cell-free DNA predicts placental dysfunction and hypertensive disease in pregnancy. Pregnancy Hypertens. 16, 148–153 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  25. De Borre, M. & Branco, M. R. in DNA Modifications (eds Ruzov, A. & Gering, M.) 333–348 (Springer, 2021).

  26. Hornbachner, R. et al. MSX2 safeguards syncytiotrophoblast fate of human trophoblast stem cells. Proc. Natl Acad. Sci. USA 118, e2105130118 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Hubert, M. A., Sherritt, S. L., Bachurski, C. J. & Handwerger, S. Involvement of transcription factor NR2F2 in human trophoblast differentiation. PLoS ONE 5, e9417 (2010).

    Article  PubMed  PubMed Central  Google Scholar 

  28. Liang, H. et al. MSX2 induces trophoblast invasion in human placenta. PLoS ONE 11, e0153656 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  29. Haider, S. et al. Transforming growth factor-beta signaling governs the differentiation program of extravillous trophoblasts in the developing human placenta. Proc. Natl Acad. Sci. USA 119, e2120667119 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Chen, C.-P. et al. Decreased junctional adhesion molecule 3 expression induces reactive oxygen species production and apoptosis in trophoblasts†. Biol. Reprod. 107, 1264–1278 (2022).

    PubMed  Google Scholar 

  31. Li, J., Han, J., Zhao, A. & Zhang, G. CircPAPPA regulates the proliferation, migration, invasion, apoptosis, and cell cycle of trophoblast cells through the miR-3127-5p/HOXA7 axis. Reprod. Sci. 29, 1215–1225 (2022).

    Article  PubMed  Google Scholar 

  32. Shin, J. K. et al. Expression of clusterin in normal and preeclamptic placentas. J. Obstet. Gynaecol. Res. 34, 473–479 (2008).

    Article  CAS  PubMed  Google Scholar 

  33. Zhou, Y. et al. Reversal of gene dysregulation in cultured cytotrophoblasts reveals possible causes of preeclampsia. J. Clin. Invest. 123, 2862–2872 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Sitras, V. et al. Differential placental gene expression in severe preeclampsia. Placenta 30, 424–433 (2009).

    Article  CAS  PubMed  Google Scholar 

  35. Wang, H., Jiang, L., Gao, B. & Dong, M. Alteration of serum semaphorin 3B levels in preeclampsia. Clin. Chim. Acta 455, 60–63 (2016).

    Article  CAS  PubMed  Google Scholar 

  36. Van Den Bogaert, K. et al. Outcome of publicly funded nationwide first-tier noninvasive prenatal screening. Genet. Med. 23, 1137–1142 (2021).

    Article  PubMed  Google Scholar 

  37. Gestational hypertension and preeclampsia: ACOG Practice Bulletin, Number 222. Obstet. Gynecol. 135, e237–e260 (2020).

  38. Heylen, L. et al. Ischemia-induced DNA hypermethylation during kidney transplant predicts chronic allograft injury. J. Am. Soc. Nephrol. 29, 1566–1576 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Mendoza, N. et al. Liver epigenome changes in patients with hepatopulmonary syndrome: a pilot study. PLoS ONE 16, e0245046 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Salas, L. A. et al. Enhanced cell deconvolution of peripheral blood using DNA methylation for high-resolution immune profiling. Nat. Commun. 13, 761 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Yuan, V. et al. Cell-specific characterization of the placental methylome. BMC Genomics 22, 6 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Zhang, X. et al. LAMA5 promotes human umbilical vein endothelial cells migration, proliferation, and angiogenesis and is decreased in preeclampsia. J. Matern. Fetal Neonatal Med. 33, 1114–1124 (2020).

    Article  CAS  PubMed  Google Scholar 

  43. Ma, L., Zhang, Z., Dong, K. & Ma, Y. TWIST1 alleviates hypoxia-induced damage of trophoblast cells by inhibiting mitochondrial apoptosis pathway. Exp. Cell. Res. 385, 111687 (2019).

    Article  CAS  PubMed  Google Scholar 

  44. Yang, S., Tong, W. & Li, Y. The mechanism of downregulation of Twist1 inhibiting trophoblast invasion and aggravating the development of preeclampsia. Front. Surg. 9, 862716 (2022).

    Article  PubMed  PubMed Central  Google Scholar 

  45. Lu, X. et al. Twist1 is involved in trophoblast syncytialization by regulating GCM1. Placenta 39, 45–54 (2016).

    Article  CAS  PubMed  Google Scholar 

  46. Gille, J., Moore, D. & Sedgwick, C. Placental infarction: a sign of pre-eclampsia in a patas monkey (Erythrocebus patas). Lab. Anim. Sci. 27, 119–121 (1977).

    CAS  PubMed  Google Scholar 

  47. Palmer, A., London, W., Sly, D. & Rice, J. Spontaneous preeclamptic toxemia of pregnancy in the patas monkey (Erythrocebus patas). Lab. Anim. Sci. 29, 102–106 (1979).

    CAS  PubMed  Google Scholar 

  48. Redman, C. W., Sargent, I. L. & Staff, A. C. IFPA Senior Award Lecture: making sense of pre-eclampsia – two placental causes of preeclampsia? Placenta 35, S20–S25 (2014).

    Article  PubMed  Google Scholar 

  49. Aneman, I. et al. Mechanisms of key innate immune cells in early- and late-onset preeclampsia. Front. Immunol. 11, 1864 (2020).

  50. Burton, G. J., Redman, C. W., Roberts, J. M. & Moffett, A. Pre-eclampsia: pathophysiology and clinical implications. BMJ 366, I2381 (2019).

  51. Raymond, D. & Peterson, E. A critical review of early-onset and late-onset preeclampsia. Obstet. Gynecol. Surv. 66, 497–506 (2011).

    Article  PubMed  Google Scholar 

  52. von Dadelszen, P., Magee, L. A. & Roberts, J. M. Subclassification of preeclampsia. Hypertens. Pregnancy 22, 143–148 (2003).

  53. Wadhwani, P., Saha, P. K., Kalra, J. K., Gainder, S. & Sundaram, V. A study to compare maternal and perinatal outcome in early vs. late onset preeclampsia. Obstet. Gynecol. Sci. 63, 270–277 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

  54. O’Gorman, N. et al. Multicenter screening for pre‐eclampsia by maternal factors and biomarkers at 11–13 weeks’ gestation: comparison with NICE guidelines and ACOG recommendations. Ultrasound Obstet. Gynecol. 49, 756–760 (2017).

    Article  PubMed  Google Scholar 

  55. Wright, D., Syngelaki, A., Akolekar, R., Poon, L. C. & Nicolaides, K. H. Competing risks model in screening for preeclampsia by maternal characteristics and medical history. Am. J. Obstet. Gynecol. 213, e61–e62 (2015).

    Article  Google Scholar 

  56. Kim, S. K. et al. Determination of fetal DNA fraction from the plasma of pregnant women using sequence read counts. Prenat. Diagn. 35, 810–815 (2015).

    Article  CAS  PubMed  Google Scholar 

  57. Brown, M. A. et al. Hypertensive disorders of pregnancy: ISSHP classification, diagnosis, and management recommendations for international practice. Hypertension 72, 24–43 (2018).

    Article  CAS  PubMed  Google Scholar 

  58. Tranquilli, A. L., Brown, M. A., Zeeman, G. G., Dekker, G. & Sibai, B. M. The definition of severe and early-onset preeclampsia. Statements from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Pregnancy Hypertens. 3, 44–47 (2013).

    Article  PubMed  Google Scholar 

  59. Roberts, J. M. & Escudero, C. The placenta in preeclampsia. Pregnancy Hypertens. 2, 72–83 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Galle, E. et al. DNA methylation-driven EMT is a common mechanism of resistance to various therapeutic agents in cancer. Clin. Epigenetics 12, 27 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  61. Court, F. et al. Genome-wide parent-of-origin DNA methylation analysis reveals the intricacies of human imprinting and suggests a germline methylation-independent mechanism of establishment. Genome Res. 24, 554–569 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Hannum, G. et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol. Cell 49, 359–367 (2013).

    Article  CAS  PubMed  Google Scholar 

  63. Kunde-Ramamoorthy, G. et al. Comparison and quantitative verification of mapping algorithms for whole-genome bisulfite sequencing. Nucleic Acids Res. 42, e43 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Krueger, F. & Andrews, S. R. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics 27, 1571–1572 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2013).

  66. Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).

  67. Gu, Z., Eils, R. & Schlesner, M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32, 2847–2849 (2016).

    Article  CAS  PubMed  Google Scholar 

  68. Supek, F., Bošnjak, M., Škunca, N. & Šmuc, T. REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS ONE 6, e21800 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank T. Van Brussel and D. Lambrechts (VIB, KU Leuven) for help with high-throughput sequencing; the Obstetrics and Gynaecology unit (UZ Leuven) for help with sampling blood and placental tissue; members of the Leuven Center for Human Genetics for cfDNA extraction, storage and supply; the Leuven Genomics Core for sequencing; and all women who agreed to participate. Computing was performed at the Vlaams Supercomputer Center. This study was supported by Research Foundation – Flanders (FWO), grant nos. 1524119 N (to B.T. and J.R.V.), S003422N (to J.R.V. and B.T.) and G0B4822N and G0C7519N (to B.T.); by KU Leuven Bijzonder Onderzoeksfonds, grant no. 3M180296 (to B.T.); and by a KU Leuven C1 grant (to J.R.V. and B.T.). M.D.B. and M.A. are supported by a Research Foundation – Flanders (FWO) fellowship and B.T. by a BOFZAP mandate.

Author information

Author notes

  1. These authors contributed equally: Huiwen Che, Qian Yu.

Authors and Affiliations

  1. Laboratory for Functional Epigenetics, Department of Human Genetics, KU Leuven, Leuven, Belgium

    Marie De Borre, Huiwen Che, Qian Yu, Kobe De Ridder, Mika Van Den Ackerveken, Mio Aerden, Eva Galle & Bernard Thienpont

  2. Center for Human Genetics, University Hospitals Leuven, University of Leuven, Leuven, Belgium

    Marie De Borre, Leen Vancoillie, Mio Aerden, Jeroen Breckpot, Koen Devriendt & Joris Robert Vermeesch

  3. Department of Obstetrics and Gynaecology, University Hospitals Leuven, Leuven, Belgium

    Lore Lannoo & Kristel Van Calsteren

  4. Faculty of Medicine and Life Science, Hasselt University, Hasselt, Belgium

    Pauline Dreesen

  5. Department of Obstetrics and Gynaecology, St Jan’s Hospital, Bruges, Belgium

    Joachim Van Keirsbilck

  6. Department of Gynaecology, Ziekenhuis Oost-Limburg, Genk, Belgium

    Wilfried Gyselaers

Authors
  1. Marie De Borre
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Huiwen Che
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Qian Yu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Lore Lannoo
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Kobe De Ridder
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Leen Vancoillie
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Pauline Dreesen
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Mika Van Den Ackerveken
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Mio Aerden
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Eva Galle
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Jeroen Breckpot
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Joachim Van Keirsbilck
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Wilfried Gyselaers
    View author publications

    You can also search for this author in PubMed Google Scholar

  14. Koen Devriendt
    View author publications

    You can also search for this author in PubMed Google Scholar

  15. Joris Robert Vermeesch
    View author publications

    You can also search for this author in PubMed Google Scholar

  16. Kristel Van Calsteren
    View author publications

    You can also search for this author in PubMed Google Scholar

  17. Bernard Thienpont
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

B.T., J.B., K.D., K.V.C., J.R.V. and M.D.B. conceptualized and designed the study. B.T. and M.D.B. supervised the project. M.D.B., K.V.C., L.L. and M.A. provided clinical samples and patient data. P.D. and W.G. assembled validation cohort 1 and provided patient data. J.V.K. assembled validation cohort 2 and provided patient data. L.V. oversaw cfDNA extraction. E.G., B.T., M.V.D.A. and M.D.B. developed experimental protocols for DNA extraction up to and including capture enrichment. M.D.B. executed experimental protocols including capture enrichment. H.C., Q.Y. and K.D.R. developed computational protocols. H.C. and Q.Y. performed bioinformatics analysis of bisulfite-converted, capture-enriched sequencing data. M.D.B., B.T., H.C., Q.Y. and K.D.R. contributed to the interpretation of results. B.T., H.C. and M.D.B. wrote the manuscript. All coauthors reviewed the manuscript.

Corresponding author

Correspondence to Bernard Thienpont.

Ethics declarations

Competing interests

B.T., M.D.B., Q.Y., H.C. and K.V.C. are listed as inventors on a patent application covering part of the work described in this paper. The other authors declare no competing interests.

Peer review

Peer review information

Nature Medicine thanks Peng Jin, Liona Poon and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Anna Maria Ranzoni, in collaboration with the Nature Medicine team.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Characteristics of cfDNA at PE diagnosis.

a-c. Boxplots showing available cfDNA concentrations estimated following library preparation for NIPS on cfDNA from PE (n = 40) and control pregnancies (n = 36) (a), fetal fraction estimated from NIPS (b), and mean insert size in cfDNAme libraries (c). Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. d-e. Comparison of DNA methylation levels as measured using target enrichment bisulfite sequencing, for unmethylated DNA from HCT116 cells that are knock-out for DNMT1 and DNMT3A (‘unmethylated DNA’), and the same DNA, in vitro methylated using M.SssI (‘methylated DNA’). Shown are a dot (d) and a box plot without outliers (e) comparing methylation levels in both samples at n = 34,735 loci. Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range. f-g. Comparison of cfDNA methylation levels as measured by target enrichment bisulfite sequencing (x axis), with target enrichment enzymatic methylome sequencing (f) or with whole genome bisulfite sequencing (g) (y axis) at n = 34,735 loci. h. PE-associated changes in unnormalized cfDNAme levels analyzed using target capture bisulfite sequencing. Regions with more (red) and less (blue) 5mC at P < 0.05 and 5 % FDR adjustment are highlighted. i. Mean cytosine methylation level as estimated from cfDNAme analysis on cfDNA from PE (n = 40) and control pregnancies (n = 36). Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. j. Methylation in blood (n = 32) and placental tissue (n = 37) at regions showing hypomethylation (blue), no change (grey) or hypermethylation (red) in cfDNA samples from PE pregnancies, as per panel h. Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. k. Comparison of 5mC levels estimated from oxidative bisulfite sequencing of cfDNA, and 5hmC levels estimated from the difference between regular and oxidative bisulfite sequencing of cfDNA (n = 13; see Methods). l. Overlap between regions detected to be hypomethylated, hypermethylated or both in blood DNA and cfDNA (grey) or in placental DNA and cfDNA (blue). m. Gene ontology enrichment of genes associated with cfDNAme differences in PE. P values by a two-sided Wilcoxon rank-sum test (a, i), a two-sided Welch’s t-test (b, c), a two-sided Pearson’s product-moment correlation (f, g), a two-sided moderated t-test (h), a one-tailed hypergeometric test (l) or a two-sided Fisher Exact test (m).

Extended Data Fig. 2 Example cfDNAme profiles at PE diagnosis.

a-h. Shown are means ± S.E.M. of methylation levels (y axis) in the cfDNA from PE cases at diagnosis (n = 64, blue), and from gestation-age-matched controls (n = 38, grey), as measured using target enrichment bisulfite sequencing at loci of interest relative to their genomic positions in base pairs (bp) according to human genome build hg19 (x axis). Positions of CpG dinucleotides are indicated by a ‘+’, with nearby genes and relevant chromosomes being named in the x axis label. i. Sequencing read pileup plot at chromosome 11, position 133,994,748 to 133,995,104 near JAM3, for representative cfDNA samples from a control (top) and a PE case (bottom). Methylated and unmethylated cytosines as shown as red and blue squares, respectively, and reads mapping to the top and bottom strand as light and darker grey.

Extended Data Fig. 3 DNA methylation changes in placenta and blood.

a-b. Changes in DNAme levels from placenta (a) and blood (b) as measured using target capture bisulfite sequencing. Regions with more (red) and less (blue) methylation at P < 0.05 and at a 5 % false discovery rate (FDR; panel b) are highlighted. c-d. DNAme heatmap constructed using unsupervised hierarchical clustering of the 200 most significantly differentially methylated regions as defined in panels a and b, respectively. e-h. Performance of the classifier developed for cfDNAme analyses (Fig. 1f), stratified for patients that did (n = 54) or did not (n = 10) receive steroids for fetal lung maturation prior to cfDNA sampling (e-f), or applied to DNAme levels estimated in placenta (g) or blood (h). Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. AUC: area under the curve, cfDNA: cell-free DNA, PE: preeclampsia, SCS: secondary caesarean section. Two-sided P values by a moderated t-test (a-b) or Wilcoxon rank sum test (f).

Extended Data Fig. 4 Characteristics of cfDNA and methylation changes therein, for the discovery cohort.

a-c. Boxplots of control pregnancies (grey, n = 124) and pregnancies with future PE (blue, n = 75), showing: fetal fraction as estimated from NIPS (a), cfDNA concentrations estimated following library preparation for NIPS (b), and mean insert size in cfDNAme libraries (c). Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. d. Venn diagrams depicting significantly hyper- (red) and hypomethylated (blue) regions in PE cases in cfDNA at time of diagnosis (top) and at the end of trimester 1 (bottom, control). 91 hypo- and 115 hypermethylated regions overlap in both data sets. e. Boxplots of cfDNAme levels near genes of interest in controls (grey, n = 124) and PE cases (blue, n = 75). Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. Two-sided P values by Wilcoxon rank-sum test (a) or Welch’s t-test (b, c).

Extended Data Fig. 5 example cfDNAme profiles in the first trimester.

a-h. Shown are means ± S.E.M. of methylation levels (y axis) in the cfDNA from PE cases before the onset of symptoms, in the first trimester (n = 75, blue), and from matched controls (n = 124, grey), as measured using target enrichment bisulfite sequencing at loci of interest, relative to their genomic positions in base pairs (bp) according to human genome build hg19 (x axis). Positions of CpG dinucleotides are indicated by a ‘+’, with nearby genes and relevant chromosomes being named in the x axis label. i. Sequencing read pileup plot of a region in JAM3 at chromosome 11, position 133,994,748 to 133,995,104, for representative cfDNA samples from a control pregnancy (top) and a PE case (bottom). Methylated and unmethylated cytosines as shown as red and blue squares, respectively, and reads mapping to the top and bottom strand as light and darker grey.

Extended Data Fig. 6 cfDNAme profiling to predict PE pregnancies.

a. Performance of the classifier developed for cfDNAme analyses at the time of PE diagnosis (Fig. 1f) applied to DNAme analyses of cfDNA extracted at the end of the first trimester. b. DNA methylation differences between PE (n = 2) and control samples (n = 2), as measured with target enrichment bisulfite sequencing or whole genome bisulfite sequencing. Shown are values for hypermethylated (n = 203) and hypomethylated loci (n = 135). c. Scores of the cfDNAme model developed from blood plasma sampled at the end of the first trimester (Fig. 2d), stratified according to the gestational age of sampling. Two-sided P values by the Wilcoxon rank-sum test. d-i. Performance of the classifier developed for cfDNAme analyses at end of trimester 1 (Fig. 2d), when exploratorily stratifying pregnancies according to the gestational age of sampling (d), the clinical history of PE (e), the body mass index (BMI, f), the severity of PE (g) and parity (h). Boxes in b and c show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. PE: preeclampsia; cfDNA: cell-free DNA; AUC: area under the curve.

Extended Data Fig. 7 cfDNAme predicts normal and preeclamptic pregnancies in validation.

a-c. Boxplots of the control pregnancies (grey; n = 136) and pregnancies with future PE (blue; n = 61) included in the validation cohorts, showing: cfDNA concentrations estimated following library preparation for NIPS (a), mean insert size in cfDNAme libraries (b), and fetal fraction as estimated from NIPS (c). Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. d. ROC curves showing performance of the cfDNAme-based model developed from blood plasma sampled at the end of the first trimester (Fig. 2d), applied to both center 1 and 2. e. Scores of the cfDNAme model developed from blood plasma sampled at the end of the first trimester (Fig. 2d), applied to the validation cohort and stratified according to the gestational age of sampling. Boxes show median, first, and third quartile; whiskers extend until the largest/smallest value or at most until 1.5 interquartile range, after which observations are represented as individual data points. f-j. Performance of the classifier developed for cfDNAme analyses at end of the first trimester (Fig. 2d) and applied to the validation cohort, when stratifying pregnancies according to the gestational age of sampling (f), the clinical history of PE (g), the body mass index (BMI, h), the severity of PE (i) and parity (j). Two-sided P values by Wilcoxon rank-sum test (a, e) or Welch’s t-test (b, c).

Extended Data Fig. 8 cfDNAme score per gestational age and performance of the classifier by clinical history of PE in discovery and validation cohorts combined.

a. cfDNAme score in PE cases (blue, n = 136) and controls (grey, n = 260), averaged per gestational age. Lines with shaded areas show outcomes of linear regression, two-sided P value by Pearson’s product-moment correlation test. b. The performance of the classifier developed for cfDNAme analyses at end of the first trimester (Fig. 2d) and applied to the entire cohort (discovery and validation combined), when stratifying pregnancies according to the clinical history of PE Power calculations for this comparison can be found in Supplementary Table 10. c. Estimated sensitivity and specificity of PE risk prediction based on the presymptomatic cfDNAme model (left), maternal and pregnancy-related characteristics (middle), and the combined model (right). Shown are estimates with error bars indicating the 95 % confidence interval. AUC: area under the curve; cfDNA: cell-free DNA; PE: preeclampsia.

Extended Data Table 1 Pregnancies with a cfDNA sample at PE diagnosis
Full size table
Extended Data Table 2 Full validation cohort - Pregnancies with cfDNA sampling at end of first trimester
Full size table

Supplementary information

Reporting Summary

Supplementary Tables

Supplementary Tables 1–14.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

De Borre, M., Che, H., Yu, Q. et al. Cell-free DNA methylome analysis for early preeclampsia prediction. Nat Med 29, 2206–2215 (2023). https://doi.org/10.1038/s41591-023-02510-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41591-023-02510-5

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing