Unbalanced human embryos secrete more hyperglycosylated human chorionic gonadotrophin (hCG-H) than balanced ones

Abstract

Purpose

The aim of this study was to compare the levels of hyperglycosylated human chorionic gonadotropin (hCG-H) secreted from balanced and unbalanced human embryos.

Methods

Single-step culture media samples from 155 good quality embryos, derived from 90 good prognosis patients undergoing intracytoplasmic sperm injection (ICSI), were collected on the fifth day of embryo cultivation. All embryos were tested by next-generation sequencing (NGS) technique. The hCG-H levels in the culture media were evaluated by ELISA kit (Cusabio Biotech, CBS-E15803h) according to the manufacturer’s instructions. Statistical analysis was performed using SPSS v.21 (IBM Corp., Armonk, NY, USA).

Results

The NGS analysis revealed that 36% of the embryos (n = 56) were balanced, and 64% of the embryos were unbalanced (n = 99). The presence of hCG-H was confirmed in all embryo culture media samples but was absent in the negative control. In addition, hCG-H concentration was significantly higher in the culture media from unbalanced embryos compared with the balanced ones (0.72 ± 0.30 mIU/ml vs. 0.62 ± 0.12 mIU/ml, p = 0.02, respectively). Furthermore, the mean levels of hCG-H were significantly increased in the samples from embryos with multiple abnormalities. Finally, the highest levels of hCG-H were expressed from embryos with monosomy of chromosome 11 (1.28 ± 0.04 mIU/ml) and those with trisomies of chromosomes 21 (2.23 mIU/ml) and 4 (1.02 ± 0.35 mIU/ml).

Conclusion

Our results suggest that chromosomal aberrations in human embryos are associated with an increased secretion of hCG-H. However, hCG-H concentration in embryo culture media as a single biomarker is not sufficient for an accurate selection of balanced embryos.

Introduction

Embryonic aneuploidy is a common chromosomal abnormality caused by chromosome segregation errors that has a crucial effect on the success of embryo implantation [1]. Furthermore, it is well known that aneuploid embryos have an increased frequency of growth arrest or miscarriage during the first trimester of pregnancy [2–4]. The aneuploidy is also considered as an important cause of birth defects [5]. Detection and exclusion of aneuploid (chromosomally unbalanced) embryos and the selection of euploid (chromosomally balanced) embryos is therefore essential to ensure better pregnancy and live birth success rate during ART procedures.

Current genetic methods for preimplantation diagnosis are well developed and widely used in ART. Next-generation sequencing (NGS) gives accuracy of more than 99.9% [6]. However, these techniques are invasive, time-consuming, and relatively expensive [7]. Therefore, additional non-invasive methods are needed to select the euploid embryos. One of the potential approaches to pre-assess embryonic euploidy, applying a noninvasive, sensitive, and clinically applicable technique, is the metabolite profiling of embryo culture medium [8]. Previous studies revealed that chromosomal abnormalities in embryos could lead to particular changes in the composition of the media, caused by embryo metabolism and secretion [9].

Human chorionic gonadotropin (hCG) is one of the most commonly studied markers in embryonic development. The hCG molecule has several isoforms that could be distinguished by their varying degrees of glycosylation. The hyperglycosylated isoform of human chorionic gonadotropin (hCG-H) is produced by trophoblast cells from the initial stages of embryo development and also during the first trimester of pregnancy [10–13]. It leads to induction of the extravillous cell proliferation and promotion of the invasion process [14, 15]. Cole et al. suggests that this is achieved by inhibiting TGF-β receptor and, therefore, trophoblast apoptosis [14]. hCG-H is also thought to protect endometrial stromal cells from oxidative stress-related apoptosis [16]. Furthermore, it induces the expression of several molecules expressed by the endometrium favoring the embryo implantation—galectin-3 (LGALS3), homeobox-A10 (HOXA-10), VEGF, and glycodelin [17, 18].

Several studies have suggested that measurement of serum hCG-H is a useful marker for prediction of pregnancy failures [19–21]. It was suggested that lower production of hCG-H at the implantation site could possibly lead to inappropriate implantation, placentation, and failure of pregnancy [22]. For this reason, hCG-H was proposed as a biomarker of ectopic pregnancy [23], preeclampsia [24, 25], and gestational trophoblastic diseases [26]. Currently, there are only several studies about the role of hCG as a predictor of embryo quality in IVF and their results are often controversial. Some authors suggested that hCG levels, from day 2 spent embryo culture media, are independent from human embryo developmental status [27]. In contrast, others found a positive correlation between beta-hCG concentration and blastocyst morphological grading and implantation rate [28]. Yang et al. also stated that the concentration of hCG in the culture media is associated with the morphological grade of day 3 embryos [29]. Nevertheless, there are scarce data about the hyperglycosylated isoform of hCG secreted in the embryo culture media. Moreover, it is unknown whether there is a relationship between hCG-H levels secreted by the embryo during the first days of its development and the presence of chromosomal abnormalities.

The aim of the study was to measure and compare the levels of embryonic hCG-H in culture media samples from balanced and unbalanced embryos on day five of their cultivation in vitro.

Materials and methods

All experiments were non-invasive and the women have given a written informed consent for participation in the research. Research protocol and informed consent forms were approved by the institutional review board (IRB) at Nadezhda Women’s Health Hospital and conform to the ethical principles of Declaration of Helsinki for medical research involving human subjects.

Study design

Preimplantation genetic testing for aneuploidy (PGT-A) by next-generation sequencing (NGS) was performed on trophectoderm samples from 155 embryos selected on the base of their quality (Fig. 1). The hCG-H levels in 155 spent culture media samples from 155 good quality day 5 embryos derived from 90 patients were determined. Culture media samples were split into two groups according to the PGT-A results: balanced embryos group and unbalanced embryos group. The hCG-H concentrations were compared between these groups. Fresh culture medium was used as a negative control.

Fig. 1.

Experimental design and overview of the study

Hormone measurements

On day 3 of the menstrual cycle and prior to treatment, blood samples for assay of anti-Müllerian hormone (AMH) and basal follicle-stimulating hormone (FSH) were collected by venipuncture. Serum AMH and FSH concentrations were determined using an electrochemiluminescent (ECLIA) immunoanalyzer (Cobas e411, Roche Diagnostics GmbH, Mannheim, Germany). The results were measured using the immunoanalyzer manufacturer’s instructions. The limit of sensitivity for FSH was 0.1 mIU/mL. The limit of sensitivity for AMH was 0.01 ng/mL. Intra- and inter-assay variations were less than 5% for all parameters.

Oocyte stimulation, retrieval, and preparation

All patients underwent controlled ovarian stimulation by a standard long protocol following pituitary downregulation with GnRH analog (3.75 mg) and subsequent addition of rFSH (300–375 U/Day) until at least three or more follicles had attained a mean diameter of 18 mm. Oocyte retrieval was performed up to 36 h after hCG trigger injection was administered (5000 IU).

After retrieval, cumulus-oocyte complexes were incubated in culture medium (Global for fertilization, LifeGlobal, Connecticut, USA) with 10% protein supplement (LGPS, LifeGlobal, Connecticut, USA) for 3–5 h. The surrounding cumulus cells were removed after exposure in HEPES-buffered medium with hyaluronidase (80 IU/ml, LifeGlobal, Connecticut, USA) and mechanical denudation with a hand Pasteur pipette.

Semen preparation

Patients underwent standard intracytoplasmic sperm injection (ICSI) with fresh sperm. Semen samples were processed by a routine wash and swim-up techniques. Whole semen was washed and pelleted with an equal volume of 37 °C Sperm Washing Medium (Irvine Scientific, CA, USA). By using swim-up method, the semen was performed under 37 °C with Global® for fertilization (LifeGlobal Group, Connecticut, USA) and supplemented with 10% human serum albumin (SAGE, Trumbull, USA). Specimens were incubated at a 45-degree angle at 37 °C in 5% CO₂. The top layer containing motile spermatozoa was washed with 37 °C Sperm Washing Medium.

IVF procedures and sample collection

ICSI was performed 4–6 h after retrieval of oocytes on inverted microscope (Olympus IX71, Japan) with a Hoffman modulation contrast system. Oocytes were placed in separate culture media microdrops (Global for fertilization, LifeGlobal), while sperm cells were put in polyvinylpyrrolidone (PVP) in glass bottom dishes, covered with paraffin oil (LifeGlobal, Connecticut, USA). The injected oocytes were individually cultured under mineral oil, in 20-μL droplets of Global Total® single-step medium (IVFonline, Canada) at 37 °C in an atmosphere of 6% CO₂ and 5% O_2. Embryos were cultured separately, each one in a single microdrop in Petri dishes at 37 °C, 6% CO₂, and 5% O₂ until day 5. The morphology of embryos was evaluated on day 5 by inverted microscope observation (Olympus IX71, Japan). Good quality blastocysts (Day 5 embryos) were defined as having an inner cell mass and trophectoderm with many cells, tightly packed or several cells, loosely grouped.

After 5 days of cultivation, 15–20 μl of the spent culture medium were collected, immediately frozen, and stored at − 80 °C. Only media from good-quality blastocysts was collected and subsequently analyzed.

NGS analysis

Trophectoderm biopsy was carried out on 155 blastocyst stage embryos. DNA was extracted and subjected to whole genome amplification (Sureplex, Illumina, San Diego, USA). To obtain an accurate quantification of the DNA library, quantification of the starting DNA was performed by using a fluorometric-based method specific for duplex DNA – Qubit dsDNA HS Assay Kit (ThermoFisher Scientific, Cat. No. Q32854, Waltham, MA, USA). The library preparation for PGT-A NGS analysis was made according to the VeriSeq PGS protocol (Illumina, San Diego, USA). The SurePlex amplification product was tagmented (tagged and fragmented) by the VeriSeq PGS transposome and amplified using a limited-cycle PCR program, adding index 1 (i7) and index 2 (i5) adapters. AMPure XP beads were used to purify the library DNA and provide a size selection step that removes the short library fragments and primers from the population. The quantity of each library was normalized to ensure more equal library representation in the pooled library. Equal volumes of normalized library were combined and heat-denatured before being transferred to the flow cell for cluster generation and sequencing on MiSeq. Final analysis of the obtained data was performed by BlueFuse Multi version 4.3 (Illumina, San Diego, USA).

During the biopsy of the trophoblast, usually 3–5 cells were aspirated. Mosaicism was clearly visible in VeriSeq PGS through VeriSeq PGSBlueFuse Multi profile. Mosaic calls were made when 30%–80% of the cells were aneuploid. Low-level mosaicism in samples (< 30%) was not called, and mosaicism over 80% was reported as full aneuploidy. These changes were not automatically called unless they reach more than 50% of the level expected for a full gain or loss. In case of mosaicism below 50%, the assessment was made manually assigned for each chromosome with a copy number between 2 and 3 or 2 and 1.

Immunoassay for hCG-H measurement

The concentration of hCG-H was measured in media samples collected at day 5 of embryo development using ELISA kit (Cusabio Biotech, CBS-E15803h, Wuhan, China) according to the manufacturer’s instructions. The absorption was measured on a microplate reader (Beckman Coulter DTX 880 Multimode detector) at 450 nm, and the mean amount of hCG-H was calculated against controls containing medium only.

Statistical analysis

All data were analyzed by SPSS v.21 (IBM Corp., Armonk, NY, USA). Descriptive parameters and patients’ characteristics were reported as mean ± SD. Two-tailed t test was applied for the variables where data were normally distributed. Mann–Whitney U test was used to compare balanced vs. unbalanced embryos’ hCG-H secretion. p < 0.05 was considered statistically significant. Sensitivity, specificity, and the estimate of the area under the curve (AUC) were calculated based on the receiver operating characteristic (ROC) curve analysis. A cut-off was referred to as optimal when the point classifies most of the cases correctly (balanced or unbalanced embryos). The optimal cut-off was determined based on the minimal absolute value of the difference between the sensitivity and specificity values. The obtained cut-point was verified by second approach, known as the point closest-to-(0,1) corner in the ROC plane which defines the optimal cut-off as the point minimizing the Euclidean distance between the ROC curve and the (0,1) point.

Results

NGS results and basal characteristics

In this study, 155 spent culture media from 90 patients with NGS results were used for hCG-H detection. The basal characteristics of the patients and retrieved embryos are represented in Table 1. There were no significant differences between the group of balanced and unbalanced blastocysts in most of the analyzed parameters (BMI, AMH, FSH, retrieved oocytes, and embryos male to female ratio). Only patient’s age was significantly higher in the group with unbalanced embryos (Table 1). As might be expected, the aneuploidy rate was higher in the embryos from the older women.

Table 1.

Characteristics of the patients and selected embryos in both groups (balanced/unbalanced embryos)

Patient characteristics	Total	Patients with balanced embryos Mean ± SD	Patients with unbalanced embryos Mean ± SD	p value
Number of patients*/embryos	90/155	43/56	68/99
Age (years)	38.83 ± 4.29	37.38 ± 4.57	39.62 ± 3.97	0.05
BMI (kg/m2)	23.11 ± 3.5	23.56 ± 3.7	22.93 ± 3.4	0.42
Basal AMH (ng/ml)	2.78 ± 1.52	2.92 ± 1.31	2.57 ± 1.62	0.55
Basal FSH (mIU/ml)	7.44 ± 3.91	7.23 ± 4.06	7.56 ± 3.78	0.61
Retrieved oocytes	11.85 ± 6.04	10.02 ± 5.38	12.39 ± 6.72	0.33
Embryos characteristics		Balanced embryos	Unbalanced embryos
Embryo quality	Good	Good	Good	–
Embryos male : female ratio	1.06	0.93	1.14	0.18

*Patients with more than one embryo who had balanced and unbalanced embryos were included in both groups

NGS analysis revealed that 36% of the embryos (n = 56) were balanced, and 64% of the embryos were unbalanced (n = 99) (Table 2). The unbalanced embryos comprised embryos with cytogenetic monosomies (n = 22), trisomies (n = 27), presence of both types of cytogenetic aberrations (monosomies and trisomies) (n = 26), and mosaic embryos (n = 24) (Table 2). Detected aneuploidies were various including different chromosomes (1, 2, 3, 4, 6, 7, 8, 9, 11, 13, 15, 16, 18, 19, 20, 21, and 22). Most commonly affected chromosomes were 21 (n = 12; 16.00%), 22 (n = 11; 11.66%), and 16 (n = 11; 14.66%). Mosaic events occurred predominantly with the participation of chromosomes 3 (n = 4; 16.66%), 16 (n = 3; 12.00%), and 5 (n = 3; 12.00%).

Table 2.

Overview of the human embryo ploidy status and the hCG-H culture media concentration from embryos included in this study

	Number of cases	hCG-H levels, mIU/ml Mean ± SD	Min–max
Balanced embryos	56	0.62 ± 0.12	0.36–0.95
Unbalanced embryos	99	0.72 ± 0.30	0.36–2.23
Monosomies	22	0.70 ± 0.22	0.37–1.31
Trisomies	27	0.69 ± 0.33	0.36–2.11
Both monosomies and trisomies	26	0.73 ± 0.30	0.47–1.97
Mosaic	24	0.76 ± 0.34	0.54–2.23

hCG-H detection

The presence of hCG-H was confirmed in all culture media samples while it was absent in negative controls containing medium only. A statistically significant difference was found when comparing the levels of hCG-H in culture media samples from balanced (mean hCG-H 0.62 ± 0.12 mIU/ml) versus unbalanced embryos (mean hCG-H 0.72 ± 0.30 mIU/ml) (p = 0.02) (Fig. 2a). As can be seen in Fig. 2b, hCG-H spent media concentrations were significantly lower in the group of balanced embryos compared with mosaic embryos (p = 0.03) and embryos with combined monosomies and trisomies (p = 0.05) (Table 2). In addition, hCG-H levels among unbalanced embryos with monosomies, trisomies, both monosomies and trisomies, and with mosaicism were not significantly different (p > 0.25).

Fig. 2.

Box plots representing (a) hCG-H spent media concentration in the groups of balanced (n = 56) and unbalanced (n = 99) embryos and (b) hCG-H spent media concentration from embryos with cytogenetic monosomies (n = 22), trisomies (n = 27), combined monosomies and trisomies (n = 26), and mosaic embryos (n = 24). The rectangles in the box plots show the second and third quartiles, with the line inside representing their median. The lower and upper quartiles are drawn as lines outside the box

The highest values of hCG-H were observed in cases with monosomy of chromosome 11 (1.28 ± 0.04 mIU/ml, n = 2) and chromosome 15 (0.92 mIU/ml, n = 1) and also in the culture media from embryos with trisomies of chromosomes 21 (2.23 mIU/ml, n = 1), 9 + 19 (2.11 mIU/ml, n = 1), and 4 (1.02 ± 0.35 mIU/ml, n = 4).

When receiver operating curve (ROC) analysis was performed to find the optimum cut-off value to predict embryo euploidy/aneuploidy, the area under the curve (AUC) was 0.63, CI 95%, 0.54–0.72) (Fig. 3). The optimum cut-off value for hCG-H was 0.64 mIU/ml (sensitivity 61.1%, specificity 64.8%). In addition, all spent culture media with hCG-H concentration above 0.95 mIU/ml were from unbalanced embryos only (n = 8; 8% of the unbalanced embryos).

Fig. 3.

ROC curve for prediction of embryo euploidy/aneuploidy using hCG-H concentration in embryo spent culture media as a prognostic factor. Area under the ROC curve (95% CI), 0.63 (0.54–0.72). ROC, receiver operating characteristic

Discussion

In an earlier study, we found that embryo’s secretion of hCG-H is not significantly influenced by maternal age and morphological quality of human blastocysts [30]. On the other hand, chromosomal sex of human embryos could have an effect on the secretion of hCG-H. Comparing hCG-H secretion from 29 embryos, we revealed that the female embryos could produce more variable quantities of hCG-H compared with the male ones [31]. The groups of balanced and unbalanced embryos, in the present study, did not have significant difference in the male to female ratio. In this way, the potential confounding effect of gender was excluded (Table 1).

Based on previous studies confirming the predictive power of hCG-H for successful implantation and live birth, we were able to test whether embryo secretion of hCG-H during the first 5 days of development is associated with the presence of genetic abnormalities. Our results demonstrated that unbalanced chromosomal translocations in embryos are associated with relatively higher secretion of hCH-H into spent culture medium. The measured hCG-H level was capable of differentiating embryos with unbalanced chromosomal translocations from those with balanced karyotype. However, the obtained relatively low sensitivity (56%) has shown that hCG-H secretion is not enough as an efficient biomarker to be applied as a single tool for selection of embryos without genetic abnormalities.

Looking into details, we found out that embryos with more than one aberration (monosomy and trisomy), even with mosaicism, were able to secrete significantly more hCG-H (Table 2). Probably, the increased number of affected chromosomes with aberrations leads to significant changes in hCG-H secretion. Another question that has to be considered was monosomy/trisomy of which chromosomes were associated with an increased hCG-H secretion. We observed the highest hCG-H levels in embryos with monosomy of chromosome 11 and chromosome 15. In a previous study, a total of 775 genes were found to be differentially expressed in monosomy 11 blastocysts and 280 genes in monоsomy 15 blastocysts compared with euploid controls [32]. Moreover, chromosomal location of these differentially expressed genes was not restricted to the chromosome involved in the error. The authors also showed a significant enrichment for necrosis, cell death, apoptosis, and decreased cell proliferation in both groups of monosomy blastocysts compared with controls. A higher association to cell death was observed for monosomy 11 compared with monosomy 15 blastocysts. Our data suggests that monosomy of chromosome 11 and chromosome 15 has a significant effect also on hCG-H expression. In addition, we found relatively higher hCG-H levels in the culture media from embryos with monosomy 11 (1.28 ± 0.05 mIU/ml) in comparison with those with monosomy 15 (0.92 mIU/ml). However, these results should be confirmed with further studies and a larger number of samples and embryos, respectively.

The possibility that hCG-H imbalance originates from chromosome abnormalities and modulates differentiation of the cytotrophoblast cells has been suggested by other authors [22]. Indeed relatively high serum hCG levels at mid-trimester pregnancies is usually associated with Down’s syndrome and preeclampsia [24, 33, 34]. Other research also revealed that trisomy 21 is associated with elevated levels of maternal serum total hCG and β-hCG [35–37]. Our data have shown that the changes in hCG-H expression could be found also in the earlier stages of embryo development in cases of trisomy 21. However, the elevated levels were observed only in one case while other embryos with trisomy 21 had relatively lower hCG-G concentration in the spent culture media (0.63 ± 0.12; n = 3).

The observed changes in hCG-H concentration between balanced and unbalanced embryos in our study were not more than six times and usually were several times lower (Table 2). Although it is not fully clear how unbalanced chromosomal aberrations could lead to changes in hCG-H secretion, our results suggest that certain chromosomal aberrations have an effect on hCG-H synthesis. It was proposed that hCG initiates implantation by promoting the invasive process and certain quantity is necessary for successful implantation. Hyperglycosylated hCG is made by placental cytotrophoblast cells at the time of implantation in order to initiate invasive activities [22]. It has been shown that blocking hCG-H production, or inactivating hCG-H with antibodies, leads to diminished cytotrophoblast cell invasion [38, 39]. As a result, inappropriate, either decreased or increased production of hCG-H at the implantation site could possibly lead to problems with implantation and placentation, and subsequently to failure of pregnancy [22]. Based on our findings, probably the increased secretion of hCG-H above certain limits in unbalanced embryos is associated with the mentioned negative effects.

In conclusion, our results suggest that certain chromosomal abnormalities in human embryos are associated with an increased secretion of hCG-H. Those characteristics are novel and are valuable for future application of hCG-H as a non-invasive marker for the selection of normal and balanced embryos for embryo transfer in IVF. However, the future studies should take into account the secretion of additional factors, for better estimation of the risk of unbalanced chromosome translocations.

Author contributions

D.P., D.N., and G.S. conceived the experiment; D.P., R.G., K.N., M.V, I.R., M.P., M.S., R.S., and S.H. conducted the experiment; D.P., D.N., R.G., R.S., F.S., and S.H. analyzed the results. D.P., D.N., and R.G. wrote the main manuscript text and prepared tables and figures. R.S., S.H. D.P, R.G., and G.S. edited the manuscript and made its final revision. All authors critically reviewed and approved the final version of the manuscript.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflicts of interest.

Ethical approval

The protocol of the study and the informed consent forms have been approved by the IRB committee of the Nadezhda University Hospital and conforms to the ethical principles of the Declaration of Helsinki for medical research involving human subjects.

Informed consent

All experiments were non-invasive and the women have given a written informed consent for participation in the research.