A methodological study on the process of prenatal optical genome mapping: focusing on cell culture and quality control

Abstract

Background

Optical genome mapping (OGM) has demonstrated significant potential in detecting structural variations (SVs) and has been comprehensively evaluated both retrospectively and prospectively in prenatal diagnosis. However, obtaining an adequate volume of amniotic fluid (AF) samples for OGM remains challenging due to the diverse detection techniques currently employed in prenatal diagnosis, which can limit the applicability of OGM in this setting. This study seeks to explore enhancements in cell culture techniques and quality control processes for prenatal samples when utilizing OGM in prenatal diagnosis.

Results

OGM successfully analyzed 188 AF samples with a minimum input of 0.225 million cells for ultra-high-molecular-weight DNA extraction. The study provides a comprehensive overview of the mass of chorionic villus samples, the volume of AF used for cell culture, the duration of culture, and the cell yields obtained for OGM. It was demonstrated that reducing the number of cells used for DNA isolation may not significantly decrease DNA quality for OGM with optimal cell viability and may even yield better results than those achieved with recommended cell amounts. This suggests that the current QC standards may be overly stringent, and that variant analysis remains feasible for some samples that do not meet these criteria. Based on the variant data analysis of these samples, standards appropriate for prenatal samples were summarized.

Conclusions

The findings of this study indicate that a reduced volume of AF sample or a shortened cell culture duration can be achieved in prenatal OGM, thereby enhancing the feasibility of employing OGM in prenatal diagnosis and potentially benefiting patients. Furthermore, data QC metrics suitable for prenatal samples may be more tolerant than previously recommended, necessitating further investigation with larger cohorts to establish specific QC standards for prenatal samples.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13039-026-00746-7.

Introduction

Congenital birth defects, which affect approximately 3–5% of births and are a leading cause of infant mortality, are significantly attributed to genetic disorders, including chromosomal abnormalities [1–3]. Prenatal diagnosis, primarily through chorionic villus sampling (CVS) and amniocentesis, is crucial for detecting chromosomal abnormalities that cause congenital defects.

Conventionally, karyotyping analysis, chromosomal microarray analysis (CMA), and fluorescence in situ hybridization (FISH) have been extensively employed for detecting chromosomal abnormalities. Despite the possibility of conducting these techniques simultaneously or sequentially for comprehensive assessment, they have certain limitations. Karyotyping analysis is limited by its low resolution and lengthy turnaround time. While CMA detects chromosomal abnormalities with higher resolution than karyotyping, it cannot detect balanced chromosomal rearrangements [4]. FISH is usually performed as a sequential verification method for specific chromosomal abnormalities identified by karyotyping or CMA, with its detection range constrained by the types of probes [5].

As a next-generation cytogenetic technology, optical genome mapping (OGM) can identify chromosomal aberrations as small as 500 bp and detect copy number variations (CNVs) and complex genomic rearrangements within a single assay [6]. OGM has been extensively applied in postnatal diagnosis of constitutional disorders and somatic neoplasms [7–10], and it has also been incorporated into some retrospective or prospective studies of prenatal diagnosis [11–13]. OGM offers the advantage of identifying the orientation and breakpoints of structural variations (SVs) and refining disrupted genes at these regions, thereby enhancing variant interpretation and genetic counseling [13–15]. OGM holds promise and is considered a complementary tool to traditional cytogenetics in prenatal diagnosis.

In the standard OGM protocol, the manufacturer recommends 1.0 million cells (70% viability) for isolating ultra-high-molecular-weight (UHMW) DNA (https://bionano.com/support), and 5–10 mL of amniotic fluid is usually obtained for OGM. Typically, about 20 mL of amniotic fluid (15–20 mL for karyotype analysis and 5–10 mL for chromosomal microarray analysis) is collected for prenatal diagnosis [16], and sometimes additional volume is needed for other necessary tests, such as fluorescence in situ hybridization (FISH) and whole exome sequencing (WES). However, obtaining sufficient amniotic fluid (AF) volume can be challenging due to oligohydramnios in some cases. Moreover, reduced amniotic fluid may increase the risk of orthopedic malformations and respiratory issues in newborns [17–20]. Consequently, the volume of amniotic fluid available for OGM is limited, making the obtained cells particularly valuable. Considering the restricted AF volume and the urgent turnaround time required for prenatal diagnosis, it is essential to optimize the OGM workflow and establish specific settings for prenatal amniocentesis samples, including culture methods, cell quantities, and quality control standards.

In this study, we collected 199 prenatal samples, comprising 193 AF samples and 6 chorionic villus samples, for invasive prenatal diagnosis without selecting specific amniocentesis indications. Each sample underwent parallel OGM, CMA, and karyotyping. We investigated the impact of different cell quantities on labeled DNA quality and Molecule Quality Report. Various quality control (QC) levels were examined to determine suitable QC metrics for prenatal samples. This methodological study may offer new insights into cell culturing and QC for prenatal OGM and suggest improvements to the workflow of this emerging technology in prenatal applications.

Methods

Sample collection and processing

The AF specimens were collected from 193 pregnant individuals presenting with gestational ages from 17^+ 1 to 31^+ 1 weeks, alongside chorionic villus samples obtained from 6 pregnant individuals within a gestational window of 12^+ 3 to 14^+ 4 weeks. All samples in this study were collected following prenatal diagnosis at Peking Union Medical College Hospital over the period from September 2022 to September 2024, and were subsequently used for OGM analysis after obtaining informed consent. Notably, sample acquisition was executed irrespective of specific clinical indications prompting amniocentesis. Each participant provided informed consent prior to procurement of either a 20–30 mL aliquot of amniotic fluid or a ~ 20 mg chorionic villus specimen. Concurrently, all participants underwent standard of care (SOC) evaluations encompassing CMA and/or karyotyping. Specifically, a 15 mL amniotic fluid portion was designated for cell culture within karyotyping protocols, while a separate 10 mL volume was reserved for CMA analysis. The residual amniotic fluid from each participant was subsequently allocated to cell culture procedures preceding OGM assessment. In contrast, chorionic villus samples designated for OGM analysis were directly subjected to DNA isolation without preceding cell culture expansion. To maintain unbiased assessment, personnel executing SOC testing and OGM analysis remained mutually blinded to respective results throughout the entire testing duration.

Cell culture for karyotype analysis and OGM

A volume of 5–10 mL of the AF sample underwent centrifugation at 2000 g for 10 min. Subsequently, the cell pellet was resuspended in culture medium and transferred to a T25 culture flask. After an initial five-day incubation period, the supernatant was replaced with fresh medium. All amniotic fluid cells were passaged to ensure the acquisition of a sufficient number of highly active cells. Passaging was performed three days after medium change, and cells were harvested three days after passage growth. The cells were first washed with saline solution and then digested with 0.05% trypsin. After centrifugation, the cell pellet was collected. Before DNA isolation for OGM, cell count and viability were determined. For this, the pellet was resuspended in 1 mL of phosphate-buffered saline (PBS). A 10 µL aliquot was thoroughly mixed with 10 µL of 0.4% trypan blue, and the mixture was transferred to a counting slide. Finally, the slide was inserted into a BIO-RAD TC20™ automated cell counter, which directly provided the total cell count, live cell count, and viability percentage. The residual cells were cryopreserved in dimethyl sulfoxide (DMSO) at −80℃ for long-term storage. For karyotyping purposes, each amniotic fluid sample was cultured in two distinct systems and harvested following a seven-day culture period.

Optical genome mapping

UHMW DNA isolation, DNA labeling and chip loading

Cells and chorionic villus samples, in varying quantities, were subjected to DNA isolation. UHMW DNA was extracted utilizing the Bionano Prep SP Blood and Cell Culture DNA Isolation Kit and the Bionano Prep SP Tissue and Tumor DNA Isolation Kit, adhering to the Bionano Prep SP Amnio and CVS Culture DNA Isolation Protocol and and Bionano Prep Frozen Cells DNA Isolation Protocol as per the manufacturer’s guidelines (Bionano Genomics, San Diego, CA, USA) UHMW DNA was extracted utilizing the Bionano Prep™ Blood and Cell Culture DNA Isolation Kit, adhering to the Bionano Prep SP Amnio and CVS Culture DNA Isolation Protocol as per the manufacturer’s guidelines (Bionano Genomics, San Diego, CA, USA). AF samples were stratified into four distinct groups based on cell count and viability metrics: Group 1 comprised samples with cell counts exceeding one million and viability above 70% (in line with manufacturer recommendations); Group 2 included samples with cell counts between 0.5 million and one million and viability over 70%; Group 3 consisted of samples with cell counts below 0.5 million but viability still above 70% (designated as risk group Ⅰ); and Group 4 included samples with viability under 70% (categorized as risk group Ⅱ) (Table 1). DNA quantification was executed using the Qubit dsDNA BR Assay Kit (Thermo Fisher Scientific) in conjunction with a Qubit 4.0 Fluorometer (Thermo Fisher Scientific). The UHMW DNA concentration was controlled within the range of 39–150 ng/µl, with a CV constrained to ≤ 0.3. Subsequently, fluorescent labeling of the UHMW DNA was accomplished using the DLE-1 enzyme via the DLS DNA Labeling Kit (Bionano Genomics, San Diego, USA). The labeled DNA concentration was maintained between 4 and 16 ng/µl, with a CV controlled at ≤ 0.3. Finally, the labeled DNA was loaded onto a Saphyr chip for linearization and imaging utilizing the Saphyr instrument.

Table 1.

Wet lab QC performance of AF samples grouped in different cell amounts and viability

Group (cell amounts; cell viability)	Total case	gDNA concentration		Labeled DNA concentration
		≥ 39ng/uL	< 39ng/uL	4-16ng/uL	out of 4-16ng/uL
Group 1: ≥1*106; ≥70% a	7	7	0	7	0
Group 2: [5*105, 1*106); ≥70% b	128	114	14	123	5
Group 3: <5*105; ≥70% c	14	13	1	14	0
Group 4: viability < 70% d	44	42	2	39	5
Total	193	176	17	183	10

a: manufacture’s recommended cell amount and viability; b: current optimal cell amount and viability; c: risk group Ⅰ; d: risk group Ⅱ

OGM metrics data

Analytical quality control metrics encompassed the N50 molecule-length parameter, label density (LD), map rate, and effective coverage. The manufacturer’s recommended threshold for the average filtered N50 was set at > 150 kbp, with a specific target of 230 kbp. Label density should be maintained between 14 and 17 per 100 kbp. Bionano’s recommended map rate threshold for human samples was ≥ 70%. For effective coverage, the threshold value was established at ≥ 80× to facilitate de novo assembly generation. Samples that failed to meet these recommendations were categorized into two QC groups, namely level A and level B, based on map rate. Samples exhibiting a map rate exceeding 60% were classified as level A, whereas those with a map rate below 60% were classified as level B. The targeted throughput for each sample was set at 800 Gb.

De Novo assembly and variant calling

The de novo assembly was performed using Bionano Solve software v.3.7 (https://bionanogenomics.com), and OGM-specific pipelines were managed via Bionano Access software v.1.7 (https://bionanogenomics.com). De novo assembly of the genome was generated from single molecules, with direct alignment of consensus maps to the reference human genome (GRCh38). Germline SVs, including insertions, duplications, deletions, inversions and translocations, were identified by analyzing differences in the alignment of labels between the sample consensus map and the reference map. Furthermore, CNVs and aneuploidies were detected using a coverage-based algorithm.

Data analysis

For data analysis, variants were filtered based on the following criteria: (a) manufacturer-recommended confidence thresholds were applied: insertion/deletion = 0, inversion = 0.7, duplication = −1, translocation = 0.05, aneuploidy = 0.95, and CNV = 0.99; (b) retaining variants meeting specified size criteria; (c) applying the GRCh38 SV mask to exclude SVs in hard-to-map regions; (d) excluding variants with ≥ 1% frequency in the OGM control sample SV database; (e) size thresholds: SV size ≥ 1500 kb, CNV size ≥ 200 kb; (f) SV Overlap Gene Precision: 3 kb, CNV Segment Overlap Gene Precision: 500 kb and (g) excluding variants not overlapping with disease-associated genes.

Standard of care (SOC) testing

Karyotype analysis

The cell harvesting, slide preparation, Giemsa staining, and subsequent analysis were performed following the institute’s standard operating procedures. Chromosomal aberrations were annotated using the International System for Human Cytogenetic Nomenclature (ISCN) guidelines.

Chromosomal microarray analysis

Genomic DNA (gDNA) extraction was performed using a QIAamp DNA Mini Kit (Qiagen, Germany). Chromosomal microarray analysis utilizing a single-nucleotide polymorphism (SNP) array (Affymetrix CytoScan 750 K) was conducted to detect CNVs and the data was analyzed by using Chromosome Analysis Suite (ChAS) software (https://www.thermofisher.cn). Genome coordinates were based on the hg19/GRCh37 human reference genome.

Statistical analysis

All the data were presented as mean ± SE and were analyzed using GraphPad Prism 6.0 software (GraphPad Software, CA, USA). Comparisons between two groups were performed using Student’s t-test and continuous measurements were analyzed using repeated ANOVA. A p value of < 0.05 indicated significant difference.

Results

General outcomes of prenatal samples for OGM analysis

The primary objective was to evaluate the quantity and quality of UHMW DNA obtained from 193 cultured amniocyte samples and 6 chorionic villus samples. Each chorionic villus sample, weighing approximately 10 mg, was subjected to DNA isolation. For AF samples, each had a volume of approximately 10 mL, with an average culture duration of 13 days. An average of 1.675 million cells (ranging from 0.225 to 4.54 million) were harvested per AF sample. The harvested cells, with viability ranging from 22% to 100%, were used for UHMW DNA isolation, achieving an average DNA concentration of 77.08 ng/µL (ranging from 12.25 to 207 ng/µL). Some samples outside the recommended concentration ranges were included in subsequent procedures with caution. Except for two samples with DNA quantities of less than 500 ng, approximately 560 ng of each DNA sample (both AF and chorionic villus samples) was subjected to labeling, resulting in an average labeled DNA concentration of 9.552 ng/µL. (Fig. 1) Six samples (five AF samples and one chorionic villus sample) failed to yield reliable variant data, while 193 prenatal samples were successfully analyzed.

Fig. 1.

Workflow of OGM with prenatal samples. AF: amniotic fluid; UHMW DNA: ultra-high-molecular-weight DNA; SV: structural variant; CNV: copy number variant

Impact of various cell amounts and viability on OGM technical performance

The impact of different cell amounts and viability on the technical performance of OGM was evaluated through wet lab QC and data QC, with the results presented in Tables 1 and 2. A sample was designated as a QC failure if any one of the data QC parameters fell outside the specified range. Initially, samples with cell amounts more than one million and viability over 70% (Group 1) were selected for DNA isolation as recommended. However, the data QC pass rate was only 57.14%, which was lower than expected. Subsequently, the remaining AF samples were categorized into groups based on varying cell amounts and viability to determine the optimal conditions for AF samples in OGM. The group with cell amounts between 0.5 million and one million and viability above 70% (Group 2) achieved a data QC pass rate of 81.25%, which was higher than that of the group meeting the recommended cell amounts and viability. Risk group Ⅰ (Group 3: cell amount < 5*10⁵; viability ≥ 70%) achieved a data QC pass rate of 71.43%, while risk group Ⅱ (Group 3: cell viability < 70%) showed a lower pass rate of 55%. (Table 1) These results indicate that, under optimal cell viability conditions, reducing the number of cells used for DNA isolation may not significantly affect DNA quality for OGM and might even yield better outcomes than using the recommended cell amounts. Additionally, cell yield and viability across different gestational weeks were also studied, showing that gestational age had no significant influence on cell yield and cell viability. (Table S1, S2 and Figure S1)

Table 2.

Data QC performance of AF samples grouped in different cell amounts and viability

Group (cell amounts; cell viability)	Total case	Recommended data QC metrics			Final variant data output (rate)
		Pass (rate)	Unpass (rate)
			Level A	Level B
Group 1: ≥1*106; ≥70% a	7	4(57.14%)	1(14.29%)	2(28.57%)	4(57.14%)
Group 2: [5*105, 1*106); ≥70% b	128	104(81.25%)	23(17.97%)	1(0.78%)	127(99.22%)
Group 3: <5*105; ≥70% c	14	10(71.43%)	4(28.57%)	0(0)	14(100%)
Group 4: viability < 70% d	44	24(54.55%)	19(43.18%)	1(2.27%)	43(97.73%)
Total	193	142	47	4	188

a: manufacture’s recommended cell amount and viability; b: current optimal cell amount and viability; c: risk group Ⅰ; d: risk group Ⅱ

Some samples slightly unqualified May generate analyzable data

To determine the appropriate OGM QC standards for prenatal samples, variant calling was performed on all prenatal cases. According to the recommended QC metrics parameters in OGM, 73.6% (142/193) of AF samples passed data QC metrics and generated credible variant data for subsequent analysis. However, 46 (23.8%) AF samples that were slightly below the QC thresholds also produced credible data for subsequent variant analysis. (Table 2, Figure S3) This suggests that the recommended QC standards might be somewhat overly stringent for prenatal samples, and that some samples not fully meeting the criteria can still yield analyzable data. Subsequently, QC results not meeting the manufacturer’s recommendations were categorized into two levels: an average map rate higher than 60% was designated as level A, and lower than 60% as level B. Except for one AF sample, all level A samples generated credible variant data, while none of the level B samples produced analyzable data. The lowest labeled DNA concentration that still yielded a run meeting all data QC is 3.09 ng/µL (belongs to Group 2). (Table 2) For CVS samples, all level A QC samples were successfully analyzed, whereas those from level B QC failed to generate credible variant data (results not shown). Considering one AF sample from level A QC failed analysis due to low label density, which showed many apparently false CNVs, other QC parameters were also evaluated. Based on the characteristics of samples that ultimately yielded analyzable data in this study and the reported OGM QC for hematologic malignancies [21], some data QC parameters were summarized for prenatal samples. N50 ≥ 150kbp longer than 200 kb, Label Density/100kbp of 13–21 and Average map rate ≥ 60% were recommended. (Table S5)

Comparison of OGM results with those of SOC methods in variant detection

Among the total of 193 successfully analyzed prenatal cases (including AF and CVS samples), OGM diagnosed variants in 137 cases that were previously identified by karyotype or CMA. In 74 of these cases, OGM analysis provided more detailed information regarding specific regions, orientations, modes of rearrangement, and fused genes, offering deeper insights into structural variations. There were 50 cases where neither OGM nor SOC methods detected any variants. The consistency rate between OGM and SOC methods was 96.9% (187/193). However, OGM did not identify chromosomal anomalies related to low-degree mosaicism, regions of homozygosity (ROH), variants adjacent to centromeres, or marker chromosomes composed of heterochromatin in six samples, as detected by SOC methods. Probably, Low-degree mosaicism (10% and 22%) was omitted due to a longer cell culture period than that in karyotype analysis, and the detection of ROH was limited by the size of ROH and the amounts of SVs in the samples.

Discussion

Prenatal diagnosis is a highly effective method for preventing birth defects and is typically offered to pregnant women who have an elevated risk of fetal chromosomal abnormalities. This risk may be indicated by factors such as advanced maternal age, high-risk results from serological screening or non-invasive prenatal testing, fetal ultrasound abnormalities, and others [22–25]. Karyotyping analysis, CMA, and FISH are commonly used to detect chromosomal abnormalities in prenatal diagnosis. However, the low resolution of karyotyping limits its ability to identify some CNVs and cryptic translocations. Meanwhile, CMA is unable to detect balanced chromosomal rearrangements. OGM offers significant advantages in detecting location, orientation, and chromosomal origin of variations with higher resolution than karyotyping. Moreover, OGM is capable to identify more complex structural variations, suggesting that it has great potential for application in prenatal diagnosis. Based on a comprehensive review of the existing literature about prenatal OGM, most previous studies have focused on the consistency of variant analysis between OGM and SOC testing [13, 14]. In contrast, few studies have specifically investigated the AF cell culture conditions or cell quality suitable for OGM. Furthermore, most established QC parameters were originally developed from postnatal samples. This study explored the practical requirements for applying OGM to prenatal samples.

In accordance with the manufacturer’s recommended OGM protocol, a minimum of 1.0 million cells with viability of 70% is required for UHMW DNA isolation. This requirement can pose a practical challenge, as most AF obtained from amniocentesis is typically allocated to SOC testings. Prior studies have indicated that as few as 0.4 million cells can suffice for successful downstream OGM analysis [11], and DNA quantity from samples with a cell count as low as 0.24 million was sufficient with normal QC metrics [13]. In this study, four groups of AF samples were analyzed during DNA extraction to evaluate the impact of varying cell amounts and viability on OGM QC. It was found that samples with cell amounts below the recommended level exhibited superior QC performance, possibly because samples with higher cell counts contained more AF impurities, such as fetal fat and protein. Based on this study, the optimal cell amount range is between 0.5 and 1.0 million, with cell viability exceeding 70%. These parameters can be applied to clinical samples. Samples with cell viability below 70% showed a low data QC pass rate, offering a reference for QC risk estimation of AF samples in OGM. Overall, with sufficient cell viability, the cell count requirements for AF samples or the duration of cell culture can be reduced, which implies a shorter reporting turnaround time for prenatal OGM.

The data QC of 193 AF samples was evaluated to assess the QC risk of prenatal samples with varying quality in OGM analysis. In this study, 46 samples that slightly fell below the data QC thresholds still generated analyzable data, indicating that the data QC metrics for prenatal samples could potentially be more lenient than previously suggested, as mentioned in a previous study that samples slightly below the QC thresholds were successful for method evaluation [13]. None of the four samples with a map rate below 60% yielded analyzable data, while analyzable results were successfully generated in 188 out of 189 samples (99.5%) with a map rate ≥ 60%. Therefore, data with Map rate < 60% was not recommended for analysis.

Ultimately, it was demonstrated that AF samples with cell numbers below one million achieved a variant data output rate exceeding 97%. However, aside from the group with optimal cell amounts and viability, the sample sizes in the other groups were relatively small. In this study, the laboratory work was performed by highly experienced operators, so the success rates—particularly for low-input or low-viability samples—might be higher than what less experienced laboratories would achieve. Therefore, performance for samples outside the optimal cell-count/viability range may be lower in routine settings. Furthermore, this study explored the QC standards applicable to prenatal samples and summarized certain QC parameters based on the variant analysis performances of these prenatal samples. Nonetheless, a larger cohort is necessary to validate the QC standards specifically for prenatal samples in OGM.

In conclusion, this study successfully conducted OGM analysis on 188 samples with a minimum input of 0.225 million AF cells for UHMW DNA extraction. Optimal ranges for cell quantity and viability were preliminarily established. The optimal cell amount range was determined to be between 0.5 and 1.0 million cells with viability exceeding 70%. It should be noted that the cell culture duration in this study was set as 13 days. In routine practice, the target cell yield could be achieved with a shorter culture duration. Therefore, further research is still needed to determine the optimal culture duration for prenatal OGM applications. Moreover, the QC standards for prenatal samples in OGM could potentially be more lenient, and some data QC parameters were recommended based on this study. However, further studies based on larger cohorts are warranted to refine these standards and enhance the workflow of this emerging technology in prenatal diagnosis.

Author contributions

N.H. coordinated the project and designed the experiment, and revised the manuscript. X.Y. participated in the data analysis and drafted the manuscript. K.Y., M.L., J.Z., H.Z. and Y.W. executed different parts of the research. Q.Q., X.Z. and Y.J. directed the critical discussion of the manuscript. All authors approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Funding 

The research was supported by the National High Level Hospital Clinical Research (2022-PUMCH-B-076) and and the National Key Clinical Specialty Construction Project (U114000).

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Peking Union Medical College Hospital (HS-2979). All study participants consented to undergo genetic evaluations and signed written informed consent.

Consent for publication

The results/data/figures in this manuscript have not been published elsewhere, nor are they under consideration by another publisher.

Authorship

The corresponding author has read the journal policies and submit this manuscript in accordance with those policies.

Third party material

All of the materials are owned by the authors.

Competing interests

The authors declare no competing interests.