Scholarlink

Background <p>Congenital heart disease (CHD) is the most common birth defect worldwide, affecting approximately 1% of live births. Improving the prenatal diagnosis of CHD remains an urgent clinical priority. In this study, we aimed to screen and validate amniotic fluid methylation biomarkers from clinical cases for the prenatal diagnosis of CHD.</p> Methods <p>A total of 135 amniotic fluid samples (80 cases and 55 controls) collected during the second trimester were included and divided into three independent cohorts. Cohort I was used to screen for differentially methylated regions (DMRs) by whole-genome bisulfite sequencing (WGBS) during the discovery phase. The above candidate DMRs were detected using target bisulfite sequencing (TBS) in an independent cohort II. The final biomarker set was selected according to adjusted <i>P</i>-values, sequencing depth, group differences, and relevance to cardiac development and key metabolic pathways. The model was built using individual biomarker cut-off values. Its predictive performance and generalizability were evaluated in an independent cohort III.</p> Results <p>WGBS and unsupervised clustering analysis revealed differential methylation patterns in amniotic fluid DNA between the CHD and normal groups. The 52 DMRs (adj <i>P</i> &lt; 0.01) were identified by screening with the above factors. Then, the analysis of cohort II via TBS identified 25 differentially methylated sites. From these, a predictive model was constructed using the four hypermethylated sites with the greatest group differences, all located within a single intron of the <i>PCNT</i> gene. Validation in an independent cohort III identified 11 differentially methylated sites, with an 8-site overlap with cohort II. Notably, 7 of the overlapping sites were associated with <i>PCNT</i>, highlighting strong reproducibility. This model showed consistent and decent performance in the cohort III (specificity: 95.65%, sensitivity: 70.45%), the cohort II (specificity: 73.08%, sensitivity: 83.33%), as well as in simple congenital heart disease (SCHD) (specificity: 83.67%, sensitivity: 76.09%) and complex congenital heart disease (CCHD) sample sets (specificity: 83.67%, sensitivity: 75.00%).</p> Conclusions <p>DNA methylation profiles in amniotic fluid differ significantly between fetuses with CHD and normal fetuses. The methylation of <i>PCNT</i> was suggested to be associated with the pathogenesis of CHD. Furthermore, the four-marker methylation panel identified in this study demonstrated decent efficacy in distinguishing fetuses with CHD, highlighting its promise as a new adjunctive diagnostic approach for prenatal CHD in high-risk pregnant women.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Identification of DNA methylation biomarkers in amniotic fluid for prenatal detection of congenital heart disease (CHD)

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月26日

查看全文
  • Weilun Zuo,
  • Jianwei Rao,
  • Yaqin Ma,
  • Shiyu Sun,
  • Xiaoqin He,
  • Han Yang,
  • Jiali Cao,
  • Qichang Wu,
  • Huiming Ye

摘要

Background

Congenital heart disease (CHD) is the most common birth defect worldwide, affecting approximately 1% of live births. Improving the prenatal diagnosis of CHD remains an urgent clinical priority. In this study, we aimed to screen and validate amniotic fluid methylation biomarkers from clinical cases for the prenatal diagnosis of CHD.

Methods

A total of 135 amniotic fluid samples (80 cases and 55 controls) collected during the second trimester were included and divided into three independent cohorts. Cohort I was used to screen for differentially methylated regions (DMRs) by whole-genome bisulfite sequencing (WGBS) during the discovery phase. The above candidate DMRs were detected using target bisulfite sequencing (TBS) in an independent cohort II. The final biomarker set was selected according to adjusted P-values, sequencing depth, group differences, and relevance to cardiac development and key metabolic pathways. The model was built using individual biomarker cut-off values. Its predictive performance and generalizability were evaluated in an independent cohort III.

Results

WGBS and unsupervised clustering analysis revealed differential methylation patterns in amniotic fluid DNA between the CHD and normal groups. The 52 DMRs (adj P < 0.01) were identified by screening with the above factors. Then, the analysis of cohort II via TBS identified 25 differentially methylated sites. From these, a predictive model was constructed using the four hypermethylated sites with the greatest group differences, all located within a single intron of the PCNT gene. Validation in an independent cohort III identified 11 differentially methylated sites, with an 8-site overlap with cohort II. Notably, 7 of the overlapping sites were associated with PCNT, highlighting strong reproducibility. This model showed consistent and decent performance in the cohort III (specificity: 95.65%, sensitivity: 70.45%), the cohort II (specificity: 73.08%, sensitivity: 83.33%), as well as in simple congenital heart disease (SCHD) (specificity: 83.67%, sensitivity: 76.09%) and complex congenital heart disease (CCHD) sample sets (specificity: 83.67%, sensitivity: 75.00%).

Conclusions

DNA methylation profiles in amniotic fluid differ significantly between fetuses with CHD and normal fetuses. The methylation of PCNT was suggested to be associated with the pathogenesis of CHD. Furthermore, the four-marker methylation panel identified in this study demonstrated decent efficacy in distinguishing fetuses with CHD, highlighting its promise as a new adjunctive diagnostic approach for prenatal CHD in high-risk pregnant women.