Extraction, Labeling, and Purification of Lineage-Specific Cells from Human Antral Follicles

Summary

Here, we present protocols for the identification and purification of ovarian cells from antral follicles. We elaborate on methods for processing whole ovaries for the cryopreservation of cortical strips while also harvesting intact antral follicles that are treated enzymatically to liberate multiple follicle resident cell types, including granulosa, theca, endothelial, hematopoietic, and stromal cells.

Abstract

The activation, growth, development, and maturation of oocytes is a complex process that is coordinated not just between multiple cell types of the ovary but also across multiple points of control within the hypothalamic/pituitary/ovarian circuit. Within the ovary, multiple specialized cell types grow in close association with the oocyte within the ovarian follicles. The biology of these cells has been well described at the later stages, when they are easily recovered as byproducts of assisted reproductive treatments. However, the in-depth analysis of small antral follicles isolated directly from the ovary is not commonly carried out due to the scarcity of human ovarian tissue and the limited access to the ovary in patients undergoing assisted reproductive treatments.

These methods for processing whole ovaries for the cryopreservation of cortical strips with the concurrent identification/isolation of ovary resident cells enable the high-resolution analysis of the early stages of antral follicle development. We demonstrate protocols for isolating discrete cell types by treating antral follicles enzymatically and separating the granulosa, theca, endothelial, hematopoietic, and stromal cells. The isolation of cells from the antral follicles at various sizes and developmental stages enables the comprehensive analysis of the cellular and molecular mechanisms that drive follicle growth and ovarian physiology and provides a source of viable cells that can be cultured in vitro to recapitulate the follicle microenvironment.

Introduction

The primary functional elements of the human ovary are the follicles, which govern the growth and development of oocytes. Protocols for the isolation of follicular cells have been well established in the context of in vitro fertilization, but these are appropriate only for the collection of cells from luteinized follicles at the point of oocyte retrieval¹. We have developed a protocol that enables the isolation of discrete cell populations from antral follicles at different developmental stages that arise from native ovaries or xenotransplanted ovarian tissue². Although there is consensus that the contributions of follicle resident cells to the cultivation of the oocyte are highly important, few studies have prospectively identified and extracted the unique phenotypic subtypes present in antral-stage follicles. A deeper understanding of the differentiation hierarchy and signal transduction between specialized cells during the different developmental stages could broaden our understanding of ovarian physiology under homeostatic and pathological conditions. Moreover, the discrimination of discrete cellular subtypes and their molecular contributions to follicle growth/maturation may provide a means of generating ex vivo surrogates that reconstruct ovarian function to foster oocyte maturation and/or treat endocrine dysfunction.

Each unique cell type within the ovary contributes to the complex function of the follicle, which effectively functions as a discrete mini-organ to foster the growth and maturation of the oocyte it contains. The oocyte, the centerpiece of the follicle, is directly enveloped by a continuous layer of granulosa cells (GCs), with the theca cells (TCs) forming a secondary layer of cells that combine with the oocyte and GCs to compose the follicular unit. Although classified into two groups, GCs and TCs contain numerous subtypes. GCs are classified according to their position within the follicle; GCs that surround the oocyte versus those that are adjacent to the basement membrane are designated as oophorous and mural GCs, respectively, and these subtypes display unique transcriptomic signatures. TCs have numerous subtypes that function to provide steroidogenic, metabolic, and structural support. Endothelial, perivascular, and immune cells play a central role in maintaining normal ovarian physiology. The ovarian stroma serves not only as a substrate for follicle growth but also likely provides a source of progenitors that give rise to TCs. This multilayered complex of cellular subtypes within the ovary is what enables its function as both an endocrine and a reproductive organ.

This paper presents a protocol for the identification and purification of granulosa, theca, stromal, endothelial, and hematopoietic cells from antral follicles. We have utilized this protocol to isolate these ovarian cells and analyze them using single-cell sequencing, followed by specific staining in follicles of different developmental stages. The protocol provides a straightforward methodology that is replicable and will enable the high-resolution analysis of both the physiology and pathology of the ovary.

Protocol

All procedures involving mice were approved by the Institutional Animal Care and Use Committee (IACUC) at Weill Cornell Medicine. All xenotransplantation experiments using ovarian tissue were performed in accordance with relevant guidelines and regulations. Both ovaries were isolated from a 14-year-old brain-dead organ donor with no history of radio/chemotherapy and no documented history of endocrine or reproductive conditions. The institutional review board (IRB) Committee of Weill Cornell Medicine approved the collection of tissue, and approval from the family of the organ donor was obtained following informed consent regarding the use of tissue.

1. Ovarian tissue collection and handling

1. Collect whole ovaries and the associated tissue, and rinse in sterile saline solution.
2. Using sterile tweezers, place the ovaries in a sterile container. Pour sterile, buffered, cold solution (e.g., Leibovitz's L-15 medium) into the container. Ensure that the tissue is completely covered with the liquid.
3. Close the container and double-bag it using sterile plastic bags. Transport the container on ice within an isolated box (e.g., Styrofoam). Transfer the specimens to the lab that will process the ovaries as soon as possible, preferably in a time not exceeding 5 h^3,4.

2. Processing the ovarian tissue

NOTE: Make sure all the reagents and tools are prepared before the tissue arrives in the lab to reduce the ischemic interval as much as possible. The buffers can be prepared up to 1 week prior to freezing and refrigerated until use.

1. Preparations for the ovarian processing and freezing
   1. Set sterilized surgical tools in the biosafety cabinet in preparation for tissue processing: one number 21 scalpel; one number 11 scalpel; sharp fine curved scissors; one long forceps (~150 mm length); and two medium forceps (~110 mm length).
   2. Prepare 100 mL of tissue processing medium (see the Table of Materials): Leibovitz's L-15 medium containing antibiotic-antimycotic solution and filtered using a 0.2 μm filter. Keep the medium chilled.
   3. Label the cryovials, add 1.5 mL of freezing solution (see the Table of Materials) to each vial, and store it at 4 °C.
   4. Have available a 6-well plate on hand for use.
2. Upon arrival of the tissue
   1. Spray the container with 70% ethanol solution, wipe thoroughly, and place it within the biosafety cabinet.
   2. Wearing sterile gloves, open the container. Place the ovary in a sterile Petri dish, and pour chilled medium (from step 2.1.2) to hydrate the tissue. Ensure that the ovary is half-submerged in the medium.
   3. Remove any sutures or other material and dissect the residual surrounding tissue away from the ovary.

3. Isolation of antral follicles

1. Identify individual follicles for isolation. In choosing healthy follicles, exclude any blood-filled, dark, or non-symmetric follicles, as those are likely to be atretic.
2. Isolate the chosen antral follicles from the whole ovary using scalpels, keeping the antrum intact by excising the cortical tissue encompassing the follicle.
3. Place each excised intact antral follicle in one well of the 6-well plate. If needed for descriptive purposes, use a ruler placed underneath the dish to determine the follicle diameter.
4. Using a scalpel, bisect the intact antral follicle to gain access to the antral cavity, and add 4 mL of enzymatic cell detachment medium. Incubate the plate in a humidified incubator at 5% CO₂ and 37 °C for 10 min.
5. Repeat the extraction of additional antral follicles as needed.

4. Isolation of follicle resident cells

1. Following incubation, add 4 mL of available medium containing 20% FBS, and flush by repeated, vigorous pipetting of the medium over the bisected follicles with a P1,000 micropipette.
2. Collect the medium, and pass it through a 100 μm filter.
3. Centrifuge the filtered supernatant at 300 × g and 4 °C for 3 min. Aspirate the supernatant, and reserve the cell pellet (enriched with GCs) on ice for labeling and fluorescence-activated cell sorting (FACS).
4. Place the remaining tissue in a mixture of collagenase (100 U/mL) and dispase (1 U/mL) diluted in a balanced saline solution (e.g., Hanks), and incubate for 30 min at 5% CO₂ and 37 °C, mechanically disrupting the tissue by trituration and vigorous pipetting (i.e., 10-15 passes through the pipet tip) twice after 10 min and 20 min.
5. Recover media containing the tissue, flush via pipetting through a P1,000 micropipette tip, and strain through a 100 μm filter.
6. Centrifuge at 300 × g and 4 °C for 3 min. Aspirate the supernatant, and put aside the cell pellet (enriched with TCs and stroma cells) on ice for labeling and FACS.

5. Fluorescence-activated cell sorting

1. Resuspend the cell pellets in blocking solution (PBS + 0.1% FBS), and incubate for 10 min at 4 °C.
2. Centrifuge at 300 × g and 4 °C for 3 min, and aspirate the supernatant.
3. Resuspend the cell pellets in blocking solution containing directly conjugated antibodies (see Table 1 for appropriate antibody combinations to identify GCs and TCs), and incubate on ice for 10 min.
4. Wash and centrifuge the cells at 300 × g and 4 °C for 3 min. Resuspend the cells in FACS buffer containing 4′,6-diamidino-2-phenylindole (DAPI), and run the cell suspension through a FACS instrument to capture a small population of the cells (500-1,000) using the fluorescence intensity of fluorophore-conjugated antibodies for gating.
5. For the identification of unique subpopulations, utilize the following gating strategies:
   1. For the purification of GCs, draw a gate around the CD45⁺ cells (Figure 1B), and exclude this population from the CD99⁺ population (Figure 1A and Figure 1C); then, further subdivide the remaining CD99⁺ fraction into PVRL^+/- fractions (Figure 1A and Figure 1C).
   2. For the purification of TCs and stroma, gate the total ENG⁺ (Figure 1A) or CD55⁺ (Figure 1D) population to exclude the CD34⁺ endothelial fraction (Figure 1E) and distinguish the steroidogenic (ANPEP⁺) and non-steroidogenic (ANPEP^–) cells. Be careful to compensate for bleed-through between fluorophores that are close in the emission spectra.
6. To validate the purity of the sorted cell fraction, run 5% of the total captured volume through the FACS instrument, and ensure that cells populate the expected gates and are present at >95% purity.

6. Processing of the ovarian cortical tissue

1. Bisect the remainder of the ovary, and excise the medulla using curved fine scissors and scalpels, being careful to avoid damage to the ovarian cortex.
2. Continue processing and thinning the ovarian cortex by gentle scraping until minimal medulla remains. Ensure to use the minimum force needed.
   NOTE: The cortical tissue thickness should be between 1 mm and 1.5 mm.
3. Divide the cortical tissue into 2-3 mm wide strips along the length of the ovary.

7. Ovarian tissue slow freezing

1. Preparation for freezing
   1. Place one cortical strip into each chilled cryogenic vial containing freezing solution.
   2. Shake the cryogenic vials containing the cortical strips on a rotating plate for 20 min at 4 °C.
2. Slow freezing of the ovarian tissue⁵
   1. Place the cryovials containing ovarian tissue into a programmable freezer set to start at 0 °C.
   2. Cool at a rate of 2 °C/min until −7 °C.
   3. Maintain the cryovials at this temperature for 10 min.
   4. Nucleate ice crystal formation by touching each cryovial with a cotton tip that has been briefly immersed in liquid nitrogen (LN₂).
   5. Cool at a rate of 0.3 °C/min until the sample temperature reaches −40°C.
   6. Increase the rate of cooling to 10 °C/min until the sample temperature reaches −140 °C.
   7. Transfer the cryovials for storage in LN₂.

Representative Results

We isolated follicles from the surface of the ovary and enzymatically treated them to isolate the GCs as well as theca and stroma cells surrounding the antral cavity. The cells were collected, and the cell fractions were sorted from the antral follicles (diameters ranging between 0.5 mm and 4 mm) by FACS to >95% purity (Figure 1).

To label and purify unique cellular fractions within the human antral follicles, we combined enzymatic digestion with flow cytometric sorting. We previously identified surface proteins that specifically mark follicle-resident fractions². We have shown that CD99 (red in Figure 1A) specifically labels GCs, and PVRL1 (yellow in Figure 1A) is increasingly localized to the oophorous GC compartment as the antral follicles increase in size². We have also shown that antibodies against endoglin (ENG, green in Figure 1A) or CD55 (Figure 1D) specifically demarcate all stroma and TCs and that alanine aminopeptidase (ANPEP, Figure 1E) is specifically expressed by androgenic TCs^2,6.

The cells were labeled with an antibody directed against CD99 to identify GCs, and antibodies against CD45 were used to exclude hematopoietic cells (Figure 1B,C). Antibodies against PVRL1 enabled the separation of the GC population into oophorous and mural compartments (Figure 1C). Following enzymatic digestion with collagenase/dispase, the cell preparations labeled with antibodies against CD55 (or alternatively ENG), CD34, and ANPEP enabled the capture of all the stroma/TCs (CD55⁺) and/or androgenic TCs (ANPEP⁺), with the exclusion of endothelial cells (CD34⁺) from the TC populations (Figure 1D,E). Using this panel and a stepwise enzymatic dissociation protocol, we labeled and purified hematopoietic, endothelial, granulosa (oophorous and mural), and theca (stromal and androgenic) cells from the antral follicles of freshly resected ovaries.

Figure 1: Labeling and purification of lineage-specific cells from human antral follicles. (A) Representative micrograph showing GC (CD99⁺ PVRL^+/-) and TC (ENG⁺) compartments within human follicles at the antral stage; white stroke boxes are magnified to the right. (B–E) Representative sort plots identifying GCs as CD45^– CD99⁺ PVRL1^+/- cells (a and b) as well as generic (CD34^– CD55⁺) and androgenic (CD34^– CD55⁺ ANPEP⁺) TCs (C,D). Gated populations are shared between (B,C) and (D,E); unstained controls are shown to the left in (B–E). Scale bar = (A) 100 µm. Abbreviations: GC = granulosa cell; TC = theca cell; PVRL = nectin; ANPEP = alanine aminopeptidase; Nuc = nuclei; SSC = side scatter; PE = phycoerythrin; APC = allophycocyanin. Please click here to view a larger version of this figure.

Cell Type	Inclusion Antibodies	Exclusion Antibodies
Granulosa cells	CD99 (PVRL+/-)	CD45
All Theca Cells	CD55 or ENG	CD34 & CD45
Steroidogenic Theca Cells	ANPEP	CD34 & CD45

Table 1: List of antibodies that can be used to identify GCs, TCs, and ovarian stroma. Abbreviations: GC = granulosa cell; TC = theca cell.

Discussion

Better resolution of the cellular diversity within the ovarian follicles is clinically important for several reasons. In applying the above protocol to the isolation of the unique phenotypic subtypes that reside within antral stage follicles, several factors should be considered. First, the health and viability of the ovarian tissue from which the antral follicle is derived is critical in determining the quality of the cells and the success of downstream applications. This can be optimized by minimizing the ischemic interval, working quickly (preferably in pairs), and preparing the laboratory spaces that will be used for the tissue processing and FACS. Second, these protocols are well-suited to antral-stage follicles precisely because they can be easily identified and extracted from the ovary to efficiently access the antral cavity; a similar approach for the study of early-stage and pre-antral follicles would yield a cell population with much greater heterogeneity. Finally, the number and quality of cells obtained from the follicles will vary widely depending on how developmentally advanced a follicle is and whether or not the follicle has entered a state of atresia. While these limitations make the study of the reconstruction of folliculogenesis a challenge, the efficient and robust isolation of follicle-derived cells will enable further interrogation of the mechanisms underlying human ovarian biology and infertility.

The purification of GCs from the antral follicles has been utilized previously by our group to demonstrate the deleterious influence of super physiologic levels of anti-Mullerian hormone (AMH) on follicle development⁶. In that study, cryopreserved and thawed ovarian tissue from both fertility preservation patients and organ donors was transplanted into immunocompromised mice in different experimental arms: control or AMH-expressing endothelial cells. Chronic exposure to AMH prompted a phenotype of accelerated luteinization that was reflected in increased follicle size, but this was only clearly elucidated in the molecular signature of follicle-resident GCs.

As the oocyte exits dormancy, GCs acquire a cuboidal morphology⁷, begin to proliferate, and together with the oocyte, secrete signals that recruit the surrounding stroma cells and promote their differentiation toward a TC fate⁸. Although the proper differentiation of stroma cells into specialized TCs is known to be essential for fertility, the mechanisms that drive TC specification and govern their interaction with follicle-resident cells during oocyte maturation are not well-defined. In humans, the cellular heterogeneity that exists in the TCs surrounding the follicle remains particularly elusive because, unlike GCs, which are readily accessible in an IVF setting, the isolation and culture of TCs require a direct biopsy of preovulatory follicles. In addition, most studies have relied on mechanical dissociation of the theca layer from isolated antral follicles^9,10, yielding a heterogeneous mixture of stromal, theca, vascular, and immune populations.

Single-cell multi-omics approaches have revealed the cellular heterogeneity of numerous tissues and the diverse cell types that foster oocyte development within the ovary exist within follicles across a wide range of sizes and maturation stages. The methods described above complement existing protocols for the isolation and freezing of ovarian tissue for the purpose of fertility preservation and can be implemented in parallel for the procurement of follicle-resident cells for clinical or experimental purposes. The application of these methods will be important for deciphering the molecular mechanisms that govern human folliculogenesis and are a critical prerequisite to approaches aiming to foster oocyte development/maturation from primordial follicles in vitro.

Materials

Chemicals, reagents
Antibiotic-Antimycotic 100x	Thermo Fisher Scientific	15240062	Anti-Anti
Antifade Mountant solution	Thermo Fisher Scientific	P36930	ProLong Gold
Collagenase from Clostridium histolyticum	Millipore Sigma	C 2674
DAPI	Thermo Fisher Scientific	D1306
Dispase II, powder	Thermo Fisher Scientific	17105041
DMSO	Millipore Sigma	D 2650	Dimethyl sulfoxide
DPBS, no calcium, no magnesium	Thermo Fisher Scientific	14190144
Enzyme Cell Detachment Medium	Thermo Fisher Scientific	00-4555-56	Accutase
Fetal Bovine Serum, heat-inactivated	Thermo Fisher Scientific	10438026
Hanks′ Balanced Salt solution	Thermo Fisher Scientific	14175079	no calcium, no magnesium, no phenol red
Leibovitz’s L-15 medium	Thermo Fisher Scientific	11415064
Normal Saline	Quality Biological	114-055-101
Sucrose	Millipore Sigma	S 1888
Freezing Medium (100 mL, filtered through a 0.2 micron filter)
– 69.64 mL of Leibovitz's L-15
– 17.66 mL of fetal bovine serum
– 3.42 g of sucrose
– 10.65 mL of DMSO
– 1 mL of antibiotic-antimycotic
Lab Plasticware and Supplies
6-well Clear Flat Bottom Not Treated	Corning	351146	Falcon
Cell Strainer 100 µm	Fisher scientific	352360	Corning, Falcon
Cryovials	Thermo Fisher Scientific	377267	CryoTube 1.8 mL
Petri dish, D x H 150 mm x 25 mm	Millipore Sigma	CLS430599	60EA
Round-Bottom Polystyrene Test Tubes with Cell Strainer Snap Cap, 5 mL	Fisher scientific	352235	Corning, Falcon
Vacuum Filter/Storage Bottle System, 0.22 µm	Corning	431154
Antibodies
ANPEP	BioLegend	301703
CD34	R&D Systems	FAB7227A
CD45	BioLegend	304019
CD55	BioLegend	311306
CD 99	BioLegend	371308
PVRL	BioLegend	340404
Surgical tools
long forceps (~150 mm length)	Fisherbrand	12-000-128	Fisher Scientific
medium forceps (~110 mm length)	Fisherbrand	12-000-157	Fisher Scientific
number 21 scalpel	Andwin Scientific	EF7281H	Fisher Scientific
number 11 scalpel	Andwin Scientific	FH/CX7281A	Fisher Scientific
sharp fine curved scissors	Roboz Surgical	RS-5881
Instruments
FACSJazz Flourescence activated cell sorter	BD
LSM 710 META Confocal microscope	Zeiss