Single-cell analysis comparing early-stage oocytes from fresh and slow-frozen/thawed human ovarian cortex reveals minimal impact of cryopreservation on the oocyte transcriptome.

Abstract

Study question: Does the slow-freezing and thawing process have a negative impact on the transcriptome of oocytes isolated from early-stage human follicles compared to fresh controls?

Summary answer: The transcriptional profiles of fresh and frozen/thawed oocytes did not cluster separately, indicating undetectable differences between the two groups when compared to within-donor heterogeneity.

What is known already: Previous studies using histological analysis of follicle morphology, density, and stage distribution in slow-frozen/thawed human ovarian cortex compared to fresh controls showed no differences between the two groups. Clinical cases reported in the past 10 years have demonstrated that transplanted slow-frozen/thawed and fresh ovarian cortex restored normal serum FSH levels and regular menstrual cycles by 5 months. However, the slow-frozen and thawed tissue resulted in lower rates of pregnancies and live births, albeit not statistically significant.

Study design, size, duration: We utilized single-cell RNA-sequencing (scRNAseq) of 144 human oocytes isolated from cadaver ovaries obtained from three donors.

Participants/materials, setting, methods: Human ovarian cortex from three healthy premenopausal donors 16, 18, and 27 years old was cut into squares measuring 10 × 10 × 1 mm3 and either slow-frozen and thawed or processed fresh. First, using a novel method for isolating live oocytes from primordial and primary follicles, the ovarian cortex squares were fragmented with a McIlwain tissue chopper and enzymatically digested. Next, oocytes were mechanically denuded under a dissection microscope and placed individually into wells containing lysis buffer for scRNAseq. Lysed single oocytes were subjected to library prep using the seqWell PlexWell rapid single-cell RNA protocol. Pooled libraries were subjected to 150-bp paired-end sequencing on the NovaSeq6000 Illumina platform. In total, we sequenced 144 oocytes-24 oocytes isolated fresh and 24 oocytes isolated after slow-freezing and thawing from each of the three donors. Additionally, we performed histological analysis of fresh and frozen/thawed ovarian cortex tissue from all three donors using hematoxylin and eosin staining and analyzed morphology, follicle density, and follicle stage distribution differences between fresh and cryopreserved ovarian cortex.

Main results and the role of chance: The histological analysis revealed no differences in follicle stage distribution or follicle morphology between conditions, with the percentage of normal follicles in fresh and frozen/thawed tissue, respectively, as 86.7% and 91.0% for Donor 1, 91.7% and 92.5% for Donor 2, and 96.1% and 91.1% for Donor 3. The follicle density per mm3 in fresh and frozen/thawed tissue, respectively, was 279.4 and 235.8 for Donor 1, 662.2 and 553.5 for Donor 2, and 55.8 and 71.4 for Donor 3. The difference in follicle density was not statistically significant between fresh and frozen/thawed conditions for Donors 2 and 3, and significant (P = 0.017) for Donor 1. The stromal cell densities in fresh and frozen/thawed tissue, respectively, were 0.014 in both conditions for Donor 1, 0.014 and 0.016 for Donor 2, and 0.013 and 0.014 for Donor 3. There was no statistically significant difference in stromal cell density between conditions in Donor 1 and Donor 3, though it was statistically significant (P ≤ 0.001) for Donor 2. The transcriptional profiles of fresh and frozen/thawed oocytes did not cluster separately, suggesting insignificant differences between the two groups. However, at the group mean level, there was a small shift between the fresh and frozen/thawed oocytes and the shifts were parallel across the three donors. In this comparison, fresh oocytes were enriched for gene ontology terms related to chromosome segregation and mitosis, whereas frozen/thawed oocytes were enriched for terms related to wound response, cAMP signaling, and extracellular matrix organization.

Large scale data: Datasets available on Zenodo.org. DOI: https://zenodo.org/records/13224872.

Limitations, reasons for caution: In this study, we only sequenced the oocytes isolated from early-stage follicles due to technical challenges collecting and sequencing the somatic cells surrounding the oocytes. Investigating the transcriptomic changes after freezing and thawing in the somatic cells would need to be studied in the future. Additionally, we built RNAseq libraries immediately after thawing focusing on the immediate changes. Investigation of the effects that manifest at later timepoints, either in culture or upon implantation in an animal model, may reveal additional effects of the freeze/thaw process on the transcriptome.

Wider implications of the findings: The only clinically approved method of fertility preservation for prepubertal cancer patients and adult patients who cannot delay cancer treatment is ovarian tissue cryopreservation. Investigation of cryopreservation-induced changes in follicles at all stages is critical to further our understanding of the safety and efficacy of using these tissues for fertility preservation in the clinic. Our study is the first to analyze transcriptomic changes between individual fresh and slow-frozen/thawed human oocytes collected from early-stage follicles. To accomplish this, we developed a novel method for dissociating both fresh and frozen/thawed human ovarian cortex to obtain live denuded oocytes from early-stage follicles. Our findings provide insights into the use of cryopreserved tissue and follicles for fertility preservation efforts.

Study funding/competing interest(s): This work was funded by National Institutes of Health (NIH) R01HD099402, Career Training in Reproductive Biology (CTRB) Training Grant National Institutes of Health (NIH) T32 to Jordan Machlin, National Institutes of Health (NIH) F31-HD106626 and National Institutes of Health (NIH) T31H-D079342 to Andrea Jones, National Institutes of Health (NIH) T32-GM70449 to D. Ford Hannum, and The Chan Zuckerberg Initiative Grant CZF2019-002428. We have no conflicts of interest to declare.