Methods

Study design

This study was conducted to assess the clinical application value of LRS for CAH gene detection in first-tier and second-tier NBS for CAH, utilizing a population-based retrospective cohort study and a prospective cohort study, respectively. The retrospective cohort included newborns recruited from the Fujian Provincial Newborn Screening Center, China, between January 1 and December 31, 2019, where neonatal 17α-OHP concentrations were measured for first-tier CAH NBS. Infants with positive 17α-OHP screening results underwent second-tier NBS using LRS for CAH gene detection. The prospective cohort included newborns enrolled from Fujian Maternity and Child Health Hospital between May 1 and May 31, 2023. Neonatal 17α-OHP concentration measurement and LRS-based CAH genotyping were performed simultaneously as first-tier NBS for CAH. Meanwhile, the genotypes of CAH patients were analyzed using multiplex ligation-dependent probe amplification (MLPA) and Sanger sequencing. This study was conducted in compliance with the Declaration of Helsinki and approved by the Ethics Review Committee of Fujian Provincial Maternity and Child Health Hospital (approval number: 2023KY017). Written informed consent was obtained from the parents or guardians of all participants after a detailed explanation of the study objectives.

NBS for 17α-OHP level

The workflow of NBS was performed according to the previously described procedure. Briefly, dried blood spot samples were collected within 3–7 days after birth and transported to the NBS center via a cold-chain transportation system. 17α-OHP concentrations were detected using the GSP Neonatal 17α-OHP Progesterone Kit (PerkinElmer, Turku, Finland) with the time-resolved fluorescence method, strictly following the manufacturer’s instructions. A threshold value of 17α-OHP concentration >12 nmol/L was defined as positive for CAH screening in full-term neonates (gestational age at delivery ≥ 37 weeks and birth weight ≥ 2,500 g), while a threshold of >25 nmol/L was set for premature neonates (gestational age at delivery < 37 weeks) or those with low birth weight (<2,500 g). Neonates with positive screening results were recalled, and an additional dried blood spot sample was collected for re-evaluation using the same GSP Neonatal 17α-OHP Progesterone Kit.

Long-read sequencing and data analysis

Heel-prick dried blood spot (DBS) samples were collected from newborns at postnatal days 3–7, followed by cold-chain delivery to the central molecular laboratory within 48 h. A sequential laboratory workflow was implemented, including genomic DNA extraction from DBS punches, long-range PCR amplification of CAH-associated genes, SMRT library preparation, sequencing on the PacBio Sequel II platform, circular consensus sequencing (CCS) read filtering, variant calling, variant classification in accordance with ACMG criteria, and final clinical report generation. The end-to-end laboratory turnaround time (TAT), defined as the interval between DBS receipt by the laboratory and release of the finalized genetic report, was 2 working days. When accounting for the full clinical screening pipeline under routine laboratory throughput, the overall comprehensive TAT reached 7 calendar days.

Genomic DNA was extracted from dried blood spot samples using a mini-nucleic acid extraction kit (Concert Bio, Xiamen, China), strictly following the manufacturer’s protocol. LRS-based CAH genotyping was performed at Berry Genomics Corporation (Beijing, China), as previously reported (Liu et al., 2022). Briefly, six locus-specific primer pairs were synthesized to generate 7.0–8.5 kb amplicons covering CYP21A2, CYP21A1P, CYP11B1, CYP17A1, HSD3B2, and StAR via long-range PCR. A targeted long-range PCR (LR-PCR) assay was further established for full-length amplification of the functional CYP21A2-TNXB haplotype, which enables discrimination against its homologous pseudogene counterpart CYP21A1P-TNXA. Two unique terminal primer pairs were designed against non-homologous flanking sequences of the CYP21A2 promoter and TNXB to prevent off-target cross-amplification of CYP21A1P. The primer sequences used for full-length capture of the 30.8 kb CYP21A2 target amplicon were as follows: Forward Long-F: 5′-CAGTCTCCATGTCSCAAAACACGTTC-3′, Reverse Long-R: 5′-GATGGTGGCATTGAGCAAGGGGCAG-3′. An additional six sets of short LR-PCR primers were designed to cover CYP11B1, CYP17A1, HSD3B2, StAR, and the chimeric junction regions of CYP21A2/CYP21A1P. The design strategy of these primers circumvents cross-reactivity between the 96%–98% homologous exon–intron sequences of CYP21A2 and CYP21A1P. All primers anneal to divergent non-homologous segments between the functional gene and its paralogous pseudogene, ensuring exclusive capture of genuine CYP21A2 alleles for downstream CCS analysis. Subsequently, PCR products were purified, end-repaired, and ligated with unique barcoded adaptors to construct single-molecule real-time (SMRT) bell libraries using the Sequel Binding Kit 2.0 and Internal Control Kit 1.0 (Pacific Biosciences, California, United States). The primed DNA-polymerase complexes were then subjected to SMRT cells and sequenced on the Sequel II platform (Pacific Biosciences, California, United States).

Debarcoded and filtered CCS reads were obtained from the Sequel II platform using CCS software and Lima (included in the Pbbioconda package; Pacific Biosciences, California, United States). These CCS reads were then mapped to the human GRCh38/hg38 reference genome. FreeBayes 1.3.4 was used to identify SNVs and indels. For the analysis of CYP21A2, CYP21A1P, as well as CYP21A1P/CYP21A2 and CYP21A2/CYP21A1P chimeric genes, two custom-constructed reference sequences (CYP21A2-TNXB and CYP21A1P-TNXA) were employed for identification. All variants were classified using two standardized systems: (1) the 2015 ACMG/AMP variant classification guidelines (PMID: 25741868), which stratify variants into pathogenic, likely pathogenic, variant of uncertain significance, likely benign, and benign categories; (2) clinical evidence extracted from the ClinVar public database (https://www.ncbi.nlm.nih.gov/clinvar/) for cross-validation of pathogenic evidence for each annotated variant. Variants detected by the LRS-based genotyping assay that had not been previously reported were validated by Sanger sequencing.

For heterozygous carriers summarized in Tables 1, 2, the phenotypic annotations (SW, SV, NC) assigned to each CYP21A2 variant indicate the disease subtype observed in biallelic affected individuals carrying the same pathogenic allele. Such labels solely reflect the intrinsic residual 21-hydroxylase activity and pathogenic potency of each individual variant, and do not correspond to any clinical phenotype in heterozygous asymptomatic neonates. Subtype categorization of all detected CYP21A2 variants was implemented in accordance with the Endocrine Society clinical practice guideline (Speiser et al., 2018) and the EMQN consensus recommendations for molecular testing of 21-hydroxylase deficiency (Baumgartner-Parzer et al., 2020). Variant pathogenicity and subtype assignments were further cross-validated against the ClinVar public variant database and published genotype–phenotype correlation data from Chinese CAH cohorts (Liang et al., 2025). Consistent with the autosomal recessive inheritance pattern of CAH, heterozygous individuals carrying only one pathogenic allele retain normal adrenal steroidogenesis and remain clinically unaffected.

Gene | Variant type | Nucleotide variation | Phenotype | Carrier allele | Allele number | Frequency % (n = 20)
CYP21A2 | SNV/Indel | c.293-13C>G | SW/SV | Y | 4 | 20 (4/20)
​ | ​ | c.293-39_293-38delinsGG | SW/SV | Y | 3 | 15 (3/20)
​ | ​ | c.371C>T | NC | Y | 2 | 10 (2/20)
​ | ​ | c.1280G>A | SW | Y | 1 | 5 (1/20)
​ | ​ | c.45C>T | - | N | 1 | 5 (1/20)
​ | Deletion chimera | CYP21A1P/A2-CH-3 | SW | Y | 1 | 5 (1/20)
​ | ​ | CYP21A1P/A2-CH-1 | SW | Y | 1 | 5 (1/20)
​ | Duplication | CYP21A2/A1P | - | N | 2 | 10 (2/20)
​ | ​ | CYP21A2 | - | N | 1 | 5 (1/20)
​ | Complex mutations | c.[844G>T,923dup, 955C>T] | SW | Y | 1 | 5 (1/20)
​ | ​ | c.[293-13C>G, 332_339del,518T>A, 710TA,713T>A] | SW | Y | 1 | 5 (1/20)
​ | ​ | c.[-126C>T,-113G>A] | NC | Y | 1 | 5 (1/20)
​ | ​ | c.[710T>A,713T>A,719T>A] | SW | Y | 1 | 5 (1/20)
CYP11B1 | SNV/Indel | c.456C>G | - | - | 1 | -
​ | ​ | c.1122-20A>G | - | - | 1 | -
STAR | SNV/Indel | c.179-14G>A | - | - | 3 | -
HSD3B2 | SNV/Indel | c.308-8G>A | - | - | 2 | -
​ | ​ | c.13T>A | - | - | 2 | -

Gene | Nucleotide variation | Phenotype | Carrier allele | Allele frequency % (n = 88) | Carrier rate (n = 88)
CYP21A2 | c.371C>T | NC | Y | 0.524 (22/4200) | 1:95 (22/2,100)
c.293-13C>G | SW/SV | Y | 0.190 (8/4200) | 1:262.5 (8/2,100)
c.[-126C>T,-113G>A] | NC | Y | 0.143 (6/4200) | 1:350 (6/2,100)
c.844G>T | NC | Y | 0.119 (5/4200) | 1:420 (5/2,100)
c.955C>T+TNXB/A-CH-2 | - | N | 0.095 (4/4200) | 1:525 (4/2,100)
TNXA/B-CH-1 | SW | Y | 0.095 (4/4200) | 1:525 (4/2,100)
c.-126C>T | NC | Y | 0.071 (3/4200) | 1:750 (3/2,100)
c.1069C>T | SW | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
c.188A>T | NC | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
c.913G>A | NC | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
c. [955C>T,1069C>T] | SW | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
CYP21A1P/A2-CH-1 | SW | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
c.844G>T+CYP21A2/A1P-CH-7 | NC | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
c.-103A>G | NC | Y | 0.048 (2/4200) | 1:1,050 (2/2,100)
c.518T>A | SV | Y | 0.048 (2/4200) | 1:1050 (2/2,100)
CYP21A2/A1P-CH-8 | - | N | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.-113G>A | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.955C>T | SW | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.1450dup | SW | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
TNXA/B-CH-1 | SW | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.[-126C>T,-113G>A,-110T>C,-103A>G] | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.[-126C>T,-113G>A,-110T>C] | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.518T>A+CYP21A2/A1P-CH-7 | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.1306C>T | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.[-126C>T,-113G>A,-110T>C,-103A>G,92C>T,293-13C>G,332_339del,518T>A, E6_cluster,923dup] | SW | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.[-113G>A,-110T>C,-103A>G] | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c. [188A>T,208G>T] | SV | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.[-126C>T,-113G>A,518T]>A | SV | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.371C>T,c.518T>A | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.374G>A | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.1070G>A | SV | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
c.[371C>T,1099C>T] | NC | Y | 0.024 (1/4200) | 1:2,100 (1/2,100)
CYP17A1 | c.1466delT | - | - | 0.024 (1/4200) | 1:2,100 (1/2,100)
​ | c.1226C>G | - | - | 0.024 (1/4200) | 1:2,100 (1/2,100)
STAR | c.235G>T | - | - | 0.024 (1/4200) | 1:2,100 (1/2,100)