Midkine characterization in human ovaries: potential new variants in follicles

OBJECTIVE

To characterize the growth factor midkine (MDK) in the human ovary to determine whether MDK is produced locally within the ovary, examine whether different ovarian cell types are more likely to produce MDK, and determine whether there are any stage-specific variations during follicle growth. Previous studies have revealed that MDK potentially affects human follicle growth and oocyte maturation. Proteomic analyses in follicular fluid (FF) have identified MDK to functionally cluster together and follow a similar expression profile to that of well-known proteins involved in ovarian follicle development. Midkine has not yet been characterized in the human ovary.

DESIGN

Descriptive study.

SETTING

University Hospital.

PATIENTS

The study included samples from 121 patients: 71 patients (aged 17-37 years) who underwent ovarian tissue cryopreservation provided granulosa cells (GC), cumulus cells, ovarian cortex, medulla tissue, and FF from small antral follicles (SAF); and 50 patients (aged 20-35 years) receiving in vitro fertilization treatment provided FF from preovulatory follicles before and after induction of final follicle maturation.

INTERVENTIONS

None.

MAIN OUTCOME MEASURES

MDK relative gene expression was quantified using a real-time quantitative polymerase chain reaction in cumulus cells, GC, and medulla tissue. Additionally, immunostaining and western blotting assays were used to detect MDK protein in the ovarian cortex, which contains preantral follicles, SAF, and medulla tissue. Furthermore, enzyme-linked immunosorbent assay analyses were performed to measure the concentration of MDK in FF aspirated from SAF and preovulatory follicles both before and 36 hours after inducing the final maturation of follicles.

RESULTS

Immunostaining and reverse transcription-quantitative polymerase chain reaction revealed a more prominent expression of MDK in GC compared with other ovarian cell types. Intrafollicular MDK concentration was significantly higher in SAF compared with preovulatory follicles. In addition, different molecular weight species of MDK were detected using western blotting in various ovarian sample types: GC and FF samples presented primarily one band of approximately 15 kDa and an additional band of approximately 13 kDa, although other bands with higher molecular weight (between 30 and 38 kDa) were detected in medulla tissue.

CONCLUSIONS

This is the first time that MDK has been immunolocalized in human ovarian cells at the protein level and that potentially different MDK variants have been detected in human FF, GC, and ovarian medulla tissue. Future studies are needed to sequence and identify the different potential MDK variants found to determine their functional importance for ovary and oocyte competence.

P-517 Validating reference gene expression stability in human ovarian follicles, oocytes, cumulus cells, ovarian medulla, and ovarian cortex tissue

Study question

How does the expression stability of commonly used reference genes (RG) vary in different human ovarian cells and tissues?

Summary answer

The RG stability differed among human ovarian cell types and tissues, but an RG with high stability was identified for each cell and tissue type.

What is known already

The expression of RGs used to normalize RT-qPCR may vary between different tissues, cell types, and experimental conditions. Hence, selecting the most appropriate RGs is critical in any experimental design to interpret data generated by RT-qPCR with the best accuracy. Human ovarian cells are phenotypically very different and often only available in limited amounts. In several animal species RG expression stability has been validated in oocytes and other ovarian cells, however, the suitability of a single universal RG in the different human ovarian cells and tissues has not been determined.

Study design, size, duration

This is an experimental study performed at a university hospital from January 2021 to September 2021.

Participants/materials, setting, methods

The freely available NormFinder software was used to analyze the expression stability of five commonly used RGs (GAPDH, B2M, RPLP0, ACTB, and PPIA) in human oocytes (n = 160), preantral follicles (n = 160), cumulus cells (n = 13), ovarian medulla (n = 8), and ovarian cortex tissue (n = 60). Samples were collected from 29 patients (aged 28 years on average; range 14–36) undergoing unilateral oophorectomy and ovarian tissue cryopreservation for fertility preservation.

Main results and the role of chance

The Normfinder software identified ACTB as the best RG for oocytes and cumulus cells; and B2M for medulla tissue and isolated follicles. Overall, comparisons of the cycle threshold (Ct) values demonstrated a wide variation among the RGs within the same group of samples (P < 0.05), especially for oocytes and preantral follicles with normalized RNA. The genes ACTB and RPLP0 showed the highest levels of expression and PPIA the lowest levels of expression in all types of samples, except for cortex tissue, where PPIA had the highest level of expression and B2M the lowest. These results infer that different results could be obtained when using different RGs for data normalization. The combination of two RGs only marginally increased stability, indicating that using a single validated RG would be sufficient when the available testing material is limited. For cultured ovarian cortex culture, GAPDH or ACTB were found to be the most stable genes depending on culture conditions.

Limitations, reasons for caution

Since human oocytes are scarce and contain a small amount of RNA, only five RGs were evaluated. Moreover, only germinal vesicle (GV) oocytes and preantral follicles (60-80 µm) were tested. Future studies may validate other genes and investigate the effect of oocyte maturation and follicle growth on RG stability.

Wider implications of the findings

Our findings highlight the importance of validating RGs for each cell type or tissue and culture condition. Hence, our results can be of use as guidance for future studies involving gene expression analyses in human ovarian cells and tissues, including oocytes and preantral follicles.

Trial registration number

Not applicable

Validating Reference Gene Expression Stability in Human Ovarian Follicles, Oocytes, Cumulus Cells, Ovarian Medulla, and Ovarian Cortex Tissue

Human ovarian cells are phenotypically very different and are often only available in limited amounts. Despite the fact that reference gene (RG) expression stability has been validated in oocytes and other ovarian cells from several animal species, the suitability of a single universal RG in the different human ovarian cells and tissues has not been determined. The present study aimed to validate the expression stability of five of the most used RGs in human oocytes, cumulus cells, preantral follicles, ovarian medulla, and ovarian cortex tissue. The selected genes were glyceraldehyde 3-phosphate dehydrogenase (GAPDH), beta-2-microglobulin (B2M), large ribosomal protein P0 (RPLP0), beta-actin (ACTB), and peptidylprolyl isomerase A (PPIA). Overall, the stability of all RGs differed among ovarian cell types and tissues. NormFinder identified ACTB as the best RG for oocytes and cumulus cells, and B2M for medulla tissue and isolated follicles. The combination of two RGs only marginally increased the stability, indicating that using a single validated RG would be sufficient when the available testing material is limited. For the ovarian cortex, depending on culture conditions, GAPDH or ACTB were found to be the most stable genes. Our results highlight the importance of assessing RGs for each cell type or tissue when performing RT-qPCR analysis.