Here, we present a protocol to quantify follicles in cultured ovaries of young mice without serial sectioning. Using whole organ immunofluorescence and tissue clearing, physical sectioning is replaced with optical sectioning. This method of sample preparation and visualization maintains organ integrity and facilitates automated quantification of specific cells.
Research in the field of mammalian reproductive biology often involves evaluating the overall health of ovaries and testes. Specifically, in females, ovarian fitness is often assessed by visualizing and quantifying follicles and oocytes. Because the ovary is an opaque three-dimensional tissue, traditional approaches require laboriously slicing the tissue into numerous serial sections in order to visualize cells throughout the entire organ. Furthermore, because quantification by this method typically entails scoring only a subset of the sections separated by the approximate diameter of an oocyte, it is prone to inaccuracy. Here, a protocol is described that instead utilizes whole organ tissue clearing and immunofluorescence staining of mouse ovaries to visualize follicles and oocytes. Compared to more traditional approaches, this protocol is advantageous for visualizing cells within the ovary for numerous reasons: 1) the ovary remains intact throughout sample preparation and processing; 2) small ovaries, which are difficult to section, can be examined with ease; 3) cellular quantification is more readily and accurately achieved; and 4) the whole organ imaged.
In order to study the cellular composition and morphological features of mammalian ovaries, scientists often rely on in vivo experiments followed by immunohistological staining of paraffin embedded ovaries. More recently though, whole ovary organ culture has proven to be an effective alternative to study ovarian function1,2,3,4 because the technique can be coupled with better visualization and quantification tools. Traditionally, analysis of ovarian morphology depends on reconstructing three-dimensional ovarian architecture from paraffin embedded serial sections, but, in addition to being laborious and time consuming, serially sectioning paraffin embedded tissue does not guarantee proper reconstruction of the organ, and sections are often lost or mis-ordered in the process.
In addition to the technical challenges associated with serial sectioning, there are also variations in the methods routinely used to quantify follicle numbers per ovary5,6. The methodological variability currently used impairs meta-analysis of ovarian reserve across studies5,7. For example, follicle numbers from different research articles can vary by 10-fold or more between similar developmental ages within a specific strain6. These large differences in reported follicle quantification can lead to confusion and have hindered cross-study comparisons. Experimentally, traditional approaches to follicle quantification from serial sections are performed by counting follicles of a pre-defined number of sections (e.g., every fifth, tenth, or other section). Variability in follicle counts using this approach arises not only from the periodicity in which sections are counted but also from variations in section thickness, and technical experience in generating serial sections5,6. In addition to its variability, another disadvantage of traditional tissue sectioning is that the sectioning of small ovaries from young animals is especially challenging and highly dependent on tissue orientation8.
The protocol below describes a routinely used ovary culture technique1 but greatly improves upon traditional follicle quantification by substituting physical sectioning with tissue clearing and optical sectioning using confocal microscopy8,9. Clearing using tissue immersion (without the need for transcranial perfusion or electrophoresis) in a urea- and sorbitol-based solution (e.g., ScaleS(0)10) proved compatible with the immunostaining and allowed for the reduction of clearing time without compromising depth of imaging. Other reported methods (e.g., ScaleA28,10, SeeDB11, ClearT12, and ClearT212) are either more time consuming or do not allow in-depth optical resolution of the sample. Optical sectioning is advantageous because it is less labor intensive and maintains the organ's three-dimensional architecture7,8. Another benefit of this approach is that preparation of the samples does not require costly reagents to clear the tissue and can be conducted with relative ease.
Specifically, the protocol described has been optimized for cultured mouse ovaries at postnatal day five but has been conducted on ovaries ranging from postnatal day 0 – 10. The method makes use of an ovary culture system in which the tissue naturally attaches to the membrane on which it is cultured, facilitating organ handling and manipulation. The culture system described can be used to maintain explanted ovaries for up to 10 days and to assess how different experimental conditions may interfere with oocyte survival13. The quantification procedure described is performed using the non-commercial image processing package FIJI-ImageJ14 and can be conducted on most personal computers. Furthermore, images used for quantification can be made widely available for the scientific community, thus allowing for future meta-analysis.
Cornell University's Institutional Animal Care and Use Committee (IACUC) has approved all the methods described here, under protocol 2004-0038 to JCS.
1. Preparation of Instruments and Culture Media
2. Ovary Dissection and Organ Culture
NOTE: Ovaries can be dissected at room temperature as long as the exposure to the non-ideal temperature is minimal.
3. Tissue Fixation
4. Whole Mount Immunofluorescence
5. Ovary Clearing and Imaging
6. Oocyte Quantification
NOTE: There are many different computer programs that are able to quantify cells. Described below is a protocol for oocyte quantification using the non-commercial image processing package, FIJI-ImageJ14.
This protocol includes 6 major steps following dissection of the ovaries, as outlined in Figure 1. Figure 2, Figure 3, Figure 4 highlight the most novel features of this protocol, which include optimization of tissue clearing for ovaries and whole tissue oocyte quantification using FIJI-ImageJ. Figure 2A shows images of an uncultured 5-day postnatal fixed ovary before (left) and 1 h after (right) adding clearing solution to the ovary. The ovary will begin to become translucent within minutes of being submerged in clearing solution. Once cleared, small ovaries like the one imaged in Figure 2A become difficult to handle without damaging. Therefore, working with cultured explanted ovaries is advantageous because they become attached to the porous surface of the insert membrane on which they are cultured. Attachment of the ovary to the insert membrane allows the experimenter to handle the membrane insert rather than the ovary itself (Figure 2B and Figure 3). Also, ovaries cultured in close proximity will fuse and Figure 2B shows attached fused ovaries before (left) and after clearing (right) on hematoxylin stained samples.
Performing the optical sectioning of cultured ovaries without clearing is possible; however, cells deep within the tissue have a signal that is difficult to differentiate from the background and this lack of clear signal impedes proper oocyte quantification (Figure 3A). In contrast to Figure 3A, Figure 3B is a representative image of a cleared sample in which oocyte quantification throughout the entire organ is possible. Figure 3B demonstrates that whole organ imaging of cleared cultured ovaries can be conducted without significant loss of signal deep within the tissue and images such as the one shown (Figure 3B) can be readily quantified. One approach for quantifying oocytes in samples such as these is by converting the Z-stack series of optical sections into a Maximum Intensity Projection and then further processing the file with an image processing package. Figure 4 and Supplemental Figure 1 highlight the steps used to quantify the number of oocytes in Figure 3B using FIJI-ImageJ. Supplemental Table 1 includes FIJI-ImageJ's particle (oocyte) quantification. Quantification of particles using this method also allows for the analysis of different follicles based on oocyte size, because information on both the number of particles and the corresponding area of each particle is calculated by the software and provided to the user.
Figure 1: Schematic representation of the entire protocol. Visual summary of the protocol beginning with dissection of the ovary (1), followed by ovary culture (2), tissue fixation with 4%PFA (3), immunostaining using specific antibodies (4), clearing tissue for deep imaging (5), obtaining the optical section using a confocal microscope (6) and ending with oocyte quantification (7). Please click here to view a larger version of this figure.
Figure 2: Difference in tissue opacity between un-cleared and cleared ovaries. (A) Ovary from a postnatal day five pup without clearing (left) and after mild clearing of about 1 h (right). (B) Multiple ovaries cultured in close proximity to each other for seven days. Two different magnifications of each sample are shown. Leftmost images are without clearing and rightmost images are with clearing. Ovaries will attach to the membrane of the culture insert and can be better handled if kept attached throughout the protocol. In order to improve the contrast of tissue for imaging using white light, ovaries were submerged in hematoxylin for 5 min after fixation. Please click here to view a larger version of this figure.
Figure 3: Immunofluorescence imaging of un-cleared and cleared ovaries. Ovaries from postnatal day 5 pups were cultured for 7 days and stained according to the whole mount immunofluorescence staining protocol described. Shown in red are cells labeled with mouse vasa homolog (MVH) to identify germ cells and in green are cells labeled with the nuclear oocyte-specific marker, p63. DAPI, in blue, was used to label all cell nuclei. (A) Immunofluorescence image of ovaries without clearing. (B) Immunofluorescence image of ovaries with clearing. Please click here to view a larger version of this figure.
Figure 4: Visual workflow for how to quantify oocytes using FIJI-ImageJ software. In order to facilitate data analysis, images derived from the confocal planes can be reduced to a maximum intensity projection instead of a 3D image. With the maximum intensity projection, the user can define image threshold parameters in such a way that the target particles become evident. Once the simplified image is obtained, the software is used to count the oocytes. Critically examining images while setting the parameters is crucial. This figure shows an example of how the particle/oocyte threshold can be set. The text above each image highlights the parameter used to obtain that image in FIJI-ImageJ. The software generates a table with the area measurement of the particles (see Supplemental Table 1). Please click here to view a larger version of this figure.
Supplemental Figure 1: Visual workflow for how to quantify oocytes using FIJI-ImageJ software (lower magnification). This figure shows oocyte quantification for two ovaries in close proximity. The steps performed are the same as in Figure 4. 2,436 particles/follicles were counted by the software. Quantification data obtained from the software can be found in Supplemental Table 1. Please click here to download this file.
Supplemental Table 1: Follicle quantification computed by FIJI-ImageJ. This table contains the list of counted follicles and the corresponding area generated from the image in Supplemental Figure 1. The particle size used for quantification was set from 10 µm2-infinity. Please click here to download this file.
The study of mammalian reproduction requires using and quantifying specialized cells that are not routinely amenable to cell culture. However, ex vivo culture systems are effective at maintaining ovary and follicle viability1,15. During ovary culture, the tissue requires a larger surface area for exchange of nutrients through diffusion. Therefore, 5-day old mouse ovaries are ideal in size and shape for organ culture. This protocol was optimized for ovaries maintained for seven days in culture, but ovary culture length can be adjusted according to the specific scientific question1,4. Comparable results have been achieved in ovaries cultured for as little as three days and as long as ten days (data not shown), but ovaries from animals older than ten days are large and culture may not be as effective, thus highlighting a limitation of the protocol.
If ex vivo culturing of the ovary is not needed, it is possible to fix and stain older ovaries with the protocol described, although modifications may be required to adjust to the larger tissue size9,16. The staining protocol described depends on passive diffusion of the antibodies into the organ, which indicates that it may be possible to stain larger ovaries with either longer antibody incubation steps or by partitioning the tissue into few smaller pieces that can be reconstructed after imaging. The use of urea and sorbitol in the ScaleS(0) clearing agent proved effective for neonatal cultured ovaries because it required a shorter incubation period than those reported for clearing agents with urea and glycerol (ScaleA2)8,10. Furthermore, the use of a low concentration of sodium borohydride (NaBH4) in the permeabilization solution decreased the background noise, and, together with longer incubations of the samples with primary and secondary antibodies, facilitated staining deep within the tissue. Additionally, the use of PVA in the permeabilization and washing solutions prevents spurious particles from sticking to the tissue17,18.
In this protocol, cultured healthy ovaries will attach to the insert membrane, while unhealthy ovaries may detach from it upon fixation and handling. Handling loose tissue with forceps will damage the ovary and likely compromise imaging. Alternatively, loose tissue can be embedded in 5% agarose plugs or handled with transfer pipettes as long as the tip opening is wide enough not to damage it. With regard to immunostaining, primary antibodies other than the ones used in this protocol may perform differently and may require optimization in the permeabilization and incubation steps.
Lastly, the protocol describes an alternative approach towards quantifying follicles from young ovaries as compared to traditional paraffin embedded serial sections or other previously described volumetric quantification of follicles from whole mount images5,9,16. Depending on the researcher's requirements, the procedure for oocyte quantification described in the protocol can also be used the characterize different stages of follicular development19. The particle area generated by the software can be used to determine the cell diameter and thus infer follicular stage.
The authors have nothing to disclose.
We thank Rebecca Williams and Johanna Dela Cruz from the Cornell BRC Imaging and Andrew Recknagel for helpful suggestions and technical assistance. This work was supported to National Institutes of Health grant S10-OD018516 (to Cornell's Imaging Facility), T32HD057854 to J.C.B. and R01-GM45415 to J.C.S.
Micro dissecting scissor 4.5", straight, sharp points | Roboz | RS-5912 | Micro dissecting scissor |
Jewelers style forceps 4-3/8", style 5 | Integra Miltex | 17-305X | Fine tip forceps |
Micro Iris Scissors 4", straight, sharp points | Integra Miltex | 18-1618 | Micro dissecting iris scissor or micro dissecting spring scissor |
1X Minimal Essential Media (MEM) | ThermoFisher Scientific | 11090-081 | |
Fetal Bovine Serium | Corning | 35011CV | |
HEPES (1M) | Gibco | 15630080 | |
Antibiotic-Antimycotic (100X) | Gibco | 15240062 | Used in Step 1.3.1 |
35mm Tissue Culture Treated Dish | Corning | 430165 | |
Nunc™ 24-Well Carrier Plate for Cell Culture Inserts | ThermoFisher Scientific | 141006 | Pore size: 8μm |
Paraformaldehyde (PFA) | Electron Microscopy Sciences | 15710 | 16% solution |
1X Phosphate-Buffered Saline (PBS) | Gibco | 10010023 | |
Nutator mixer GyroTwister™ | Labnet | S1000-A-B | three dimensional shaker |
Normal Goat Serum | VWR | 103219-578 | |
Bovine Serum Albumen (BSA) | VWR | 97061-416 | |
Sodium Azide (NaN3) | Sigma-Aldrich | S2002-25G | |
Sodium borohydride solution (NaBH4) | Sigma-Aldrich | 452904-25ML | Use the solution, rather than the tablet or powder form |
Polyvinyl Alcohol (PVA) | Sigma-Aldrich | P8136-250G | Cold water soluble |
Triton™ X-100 | Sigma-Aldrich | 93443-100ML | polyethylene glycol tert-octylphenyl ether |
Syringe filters | ThermoFisher Scientific | 725-2520 | 25mm |
10mL Syringes | BD | 309695 | |
Mouse anti-p63 antibody (4A4) | Biocare Medical | CM 163A | Dilution 1:500 |
Rabbit anti-MVH antibody | Abcam | ab13840 | Dilution 1:600 |
Alexa Fluor® goat anti-mouse 594 | ThermoFisher Scientific | A-11032 | Dilution 1:1000 |
Alexa Fluor® goat anti-rabbit 488 | ThermoFisher Scientific | A-11034 | Dilution 1:1000 |
4,6-Diamidino-2-phenylindole (DAPI) | ThermoFisher Scientific | 62248 | |
D-(-)sorbitol | Sigma-Aldrich | 240850-100G | |
Glycerol | Sigma-Aldrich | G9012-100ML | |
Urea | Sigma-Aldrich | U5378-500G | |
Dimethyl sulfoxide (DMSO) | Sigma-Aldrich | D2650-5X10ML | |
FIJI-ImageJ | Image processing software | ||
Disposable plastic transfer pipettes | VWR | 414044-034 |