Introduction

Follicle-stimulating hormone (FSH), secreted by anterior pituitary gonadotroph cells, is recognized as a crucial regulator of male and female gonadal function; indeed, reproductive biology textbooks strongly emphasize its role in normal physiology. ¹ FSH exerts its effects via a specific receptor, follicle-stimulating hormone receptor (FSHR), which belongs to the highly conserved family of G-protein-coupled receptors. ² Traditional views hold that in females, FSHR is expressed in granulosa cells and controls the maturation of Graafian follicles, granulosa cell proliferation, and estrogen¹, while in males, FSHR is expressed in testicular Sertoli cells and regulates their metabolic functions, which are essential for proper spermatogenesis and germ cell survival. ³ The large FSHR gene is located on chromosome 2p21 and comprises 10 exons. ⁴

However, accumulating evidence demonstrates that FSHR is also expressed in extragonadal tissues, such as endothelium, ⁵ monocytes, ⁶ developing placenta, ⁷ endometrium, ⁸ malignant tissues, ⁹ bone, adipose and neural cells. ¹⁰ Intriguingly, blocking the interaction of Fsh with Fshr mitigates some degenerative disorders in mice, such as low bone mass, ¹¹ obesity, ¹² and neurocognitive decline.¹³ Fshr was reported to be expressed in osteoclasts in vitro, and global Fsh or Fshr knockout resulted in an increase in bone mass by inhibiting bone resorption. ¹¹ Blocking the Fsh-Fshr interaction with either Fsh polyclonal or monoclonal specific antibodies triggered thermogenesis in adipose tissues, significantly reduced body weights, ¹², and reduced Alzheimer’s symptoms in mice. ¹³, Fshr expression is also found in the vasculature of tissues ⁷ and is particularly high in solid tumors ¹⁴. Nevertheless, the relevance of its extragonadal expression and functions have been debated ^2,15 Therefore, precisely defining the localization of Fshr expression remains an imperative challenge in FSH-FSHR biology due to the concern of the ‘non-specificity’ of available antibodies used to localize FSHR expression and the ‘quick turnover’ and ‘rapid degradation’ of Fshr transcripts. ^2,10,16 Thus, a suitable genetic approach to resolve this issue is warranted. ¹⁷

To address this challenge, researchers can utilize a GFP reporter driven by the endogenous promoter of Fshr, which enables the visualization of the expression of low-abundance transcripts in more accurate and context-specific ways, ^18,19 when other approaches, e.g., antibodies and RT-PCR, are limited in their ability to detect Fshr expression. Therefore, in this study, we employed CRISPR/Cas9-mediated technology to create a novel Fshr-ZsGreen reporter mouse model for precisely defining Fsh expression. We believe that this powerful approach, as a reliable readout of Fshr expression at the single-cell level, should allow us to accurately capture Fshr expression in real time in a tissue-specific manner. We also used other techniques to ensure the accuracy and reliability of the results. Our results challenge the current understanding of Fshr expression and clarify that Fshr has a wider profile of expression than previously thought. The findings of Fshr in reproductive and nonreproductive tissues/organs will provide significant insights into new roles of Fsh and Fshr in physiology and pathology and have implications for the development of new therapies for reproductive and nonreproductive disorders, particularly metabolic diseases and degenerative diseases.

Methods

1. Generation of the CRISPR/Cas9-mediated Fshr-ZsGreen knockin reporter line

Fshr-P2A-ZsGreen knockin reporter mice were generated by a CRISPR/Cas9-based approach. Briefly, one sgRNA was designed by the CRISPR design tool (http://www.sanger.ac.uk/) to target the region of the stop codon in the transcript NM_013523.3 exon 10 of mouse Fshr, ²⁰ and then was screened for on-target activity using a Universal CRISPR Activity Assay [UCATM, Biocytogen Pharmaceuticals (Beijing) Co., Ltd]. The targeting vector containing P2A-ZsGreen and 2 homology arms of left (1378 bp) and right (1493 bp) each was used as a template to repair the DSBs generated by Cas9/sgRNA. P2A-ZsGreen was precisely inserted before the stop codon of the Fshr locus. The T7 promoter sequence was added to the Cas9 or sgRNA template by PCR amplification in vitro. Cas9 mRNA, sgRNA and the targeting vector were co-injected into the cytoplasm of one-cell-stage fertilized C57BL/6J eggs. The injected zygotes were transferred into oviducts of Kunming pseudopregnant females to generate F0 mice. F0 mice with the expected genotype as confirmed by tail genomic DNA PCR, DNA sequencing and Southern blotting were mated with C57BL/6J mice to establish germline transmitted F1 heterozygous mice. F1 heterozygous mice were further genotyped by tail genomic PCR, DNA sequencing and Southern blotting. Primer sequences for genotyping F0 and F1 are described in Supplementary Material 1.

The produced Fshr-ZsGreen knockin mice were maintained as heterozygotes, and homozygotes were used for experiments. For genotyping, genomic DNA was extracted from tail tips and assayed using polymerase chain reaction (PCR) primer sets for the Fshr-ZsGreen allele. The primer sequences for genotyping are described in Supplementary Materials 1. All mice were maintained on a 12 h light/dark cycle with food and water ad libitum. The care and treatment of animals in all procedures strictly followed the NIH Guide for the Care and Use of Laboratory Animals. The animal protocols used in this study were approved by the Shanxi Medical University IACUC committee.

2. Tissue harvest and preparation

Mice were terminated by CO₂, and organs were harvested and fixed with 4% paraformaldehyde (PFA) for 12 to 24 hours at 4°C. For bone samples, decalcification was performed with daily changes of EDTA solution (0.5 M, pH 8) for 7 days. After fixation and decalcification processes, samples were transferred to 15% and 30% sucrose overnight, respectively. The tissues were then embedded with optimal cutting temperature (OCT) compound and stored at –80°C. The embedded samples were sectioned into 5– to 25-μm-thick sections using a Leica cryostat CM1950.

3. Immunofluorescence staining

Immunofluorescence staining was performed as described previously. ²¹ Briefly, air-dried 7– to 25-μm-thick frozen sections were washed with 1×PBS for 10 minutes three to five times, followed by blocking with 10% BSA for 20 minutes at room temperature, and the samples were permeabilized with 0.5% Triton X-100 in 1×PBS for 10 minutes. Then, diluted primary antibodies were added to the slides. After overnight incubation with antibodies at 4°C, the slides were washed with 1×PBS for 15 minutes three to five times and then incubated with secondary antibody conjugated with fluorescence for 30 minutes at room temperature. The slides were washed again thoroughly with 1×PBS for 10 minutes three times. The slides were then stained with DAPI, rinsed with 1×PBS three times for 5 minutes each, and mounted with 50% glycerol for confocal imaging. Rabbit IgG was used as the negative control. Primary and secondary antibodies were purchased from Servicebio (Wuhan Servicebio Technology Co., Wuhan, China), except when indicated otherwise. The primary antibodies are listed in Supplementary Materials 2. In addition, sections of different tissues/organs derived from C57BL/6J (B6) mice were used as negative controls for Fshr-ZsGreen expression. A donkey anti-rabbit IgG conjugated with Cy3 was employed at a 1:400 dilution as the secondary antibody (Servicebio, Wuhan, China).

4. Single RNA-fluorescence in situ hybridization (RNA-smFISH)

RNA in situ hybridization, the gold standard for visualizing RNA expression and localization in cells, tissue sections, and whole organs, was performed on tissue sections as described previously. ²² Briefly, frozen tissue sections of the testis from Fshr-ZsGreen mice were washed with 1XPBS three times. The sections were then treated with proteinase K (20 μg/ml) for 5 minutes at 37°C to permeabilize the cells and allow probe penetration, followed by three washes with 1XPBS. Specific oligonucleotide sense and antisense DNA probes for Fshr were designed and synthesized by the manufacturer (GeneCreate Biological Engineering Co., Ltd., Wuhan, China). The probes were labeled with a fluorescent dye (e.g. Cy3) for visualization. RNA-smFISH was performed on tissue sections using a commercially available kit (e.g., SureFISH, Agilent Technologies) at 37°C for 2 hours. After hybridization, the sections were washed with SSC solution (2xSSC, 37°C for 10 mins, 1xSSC, 37°C for 5 mins twice, 0.5xSSC for 10 mins) to remove unbound probes, and counterstained with a nuclear stain (e.g., DAPI) to visualize the tissue architecture. The sections were imaged using a fluorescence microscope equipped with appropriate filter sets. A sense probe was used as a negative control to ensure specificity and sensitivity. The sequences of sense and antisense probes are described in Supplementary Materials 2.

6. Imaging

Imaging of the slides was carried out as described before. ²¹ Briefly, the fluorescence images of the frozen sections were obtained using a Nikon A1 HD25 confocal microscope with a DUVB detector and plan Apo λ 4×, plan Apo VC 20× DIC N2, plan Apo λ 40×, and plan Apo λ 100×C oil objectives, illuminated with a wavelength of 405, 488, or 561 nm to excite DAPI, GFP, or Cy3, respectively; detection was performed with a 425–475, 500–550, or 570–620 nm bandpass filter. To assess the number of cells in each field of view, tissue-cleared images were converted from 3D to MAIP (maximum projection of the Z-stack across the whole section). Data were acquired with NIS-Elements AR 5.20.00 64-bit software.

7. Droplet digital RT‒PCR (ddRT-PCR)

Tissues were harvested from 10-week-old B6 mice. Samples were dissected free of connective tissue and homogenized with a fast multi-sample tissue cryogenic grinder (LC-FG-96, Lichen Instrument Technology Co., Ltd., Shanghai, China), and total RNA was extracted using NucleoZOL (NucleoZOL; Macherey-Nagel GmbH & Co., KG, Dylan, Germany). mRNA was reverse transcribed using an M5 Super qPCR RT kit with gDNA remover (MF012-T, Mei5bio, Beijing). Droplet digital PCR was performed as described previously. ²¹ The primers used for ddRT-PCR were as follows: mFshr Fwd-5’ ccgcagggacttcttcgtcc-3’; mFshr Rev-5’-ttggtgactctgggagccga-3’.

Results

To precisely define Fshr expression in mice, we utilized several complementary strategies, including Fshr endogenous promoter driving ZsGreen knockin reporter mice, immunofluorescence staining with antibodies against tissue/cell type-specific markers and Fshr, ddRT-PCR and RNA-smFISH, to comprehensively examine Fshr expression.

1. Generation of Fshr-ZsGreen knockin reporter mice

CRISPR/Cas9-mediated homologous recombination was used to generate embryonic stem cell (ESC) clones, in which a P2A-ZsGreen cassette was precisely inserted before the stop codon of the Fshr gene followed by the 3′ UTR of the Fshr allele, as described in the Methods (Fig. 1A). This P2A-ZsGreen expression cassette under the control of endogenous Fshr regulatory elements ultimately generates ZsGreen proteins without disrupting Fshr expression. The injected zygotes were transferred into oviducts of Kunming pseudopregnant females to generate F0 mice. F0 mice with the expected genotype were confirmed by tail genomic DNA PCR, DNA sequencing and Southern blotting (Fig. 1B and C) and then mated with C57BL/6J mice to establish F1 heterozygous mice with the germline-transmitted transgene. F1 heterozygous mice were further genotyped by tail genomic PCR, DNA sequencing and Southern blotting (Fig. 1B and C). The results from these tail genomic PCR, DNA sequencing and Southern blots in both F0 and F1 pups demonstrated that the targeted P2A-ZsGreen cassette was accurately inserted into the designed site between exon 10 and the stop codon of the Fshr gene before the 3’ UTR. We maintained heterogeneous Fshr-ZsGreen mice and used homogeneous mice for experiments. These mice were genotyped using primer sets specific to P2A-ZsGreen, as shown in Figure 1D. Fshr and GFP are transcribed in a bicistronic mRNA but subsequently translated into two independent proteins rather than as a fusion protein. This design ensures unaffected Gαi3 transcription and function and simultaneous GFP expression as a reporter protein. All animals were fertile and showed normal behavior and no obvious abnormal phenotypes.

Figure 1.

2. Examination of Fshr expression in Fshr-ZsGreen reporter mice

With confirmation that the Fshr-P2A-ZsGreen targeting vector was successfully inserted into the Fshr locus, we investigated Fshr expression by confocal microscopy to detect ZsGreen and immunostaining for tissue/cell markers in frozen sections of tissues/organs of Fshr-ZsGreen reporter mice. To ensure that there was no nonspecific expression of Fshr-ZsGreen in the examined tissues/organs, we took frozen sections derived from wild-type mice (B6) as negative controls. The negative controls were imaged under the conditions used for examining Fshr-ZsGreen expression. The representative results are shown in Supplementary Data 1, showing no nonspecific expression of Fshr-ZsGreen in the negative controls. On this basis, we performed the following imaging to examine Fshr-ZsGreen expression in the major organs and tissues.

1) Reproductive organs

As the reproductive system is well known to express Fshr, we first tested Fshr-ZsGreen expression in the ovary and testis to ensure ZsGreen expression driven by the endogenous Fshr promoter. In the ovary, we observed Fshr-ZsGreen expression in the different stages of follicles from primordial follicles to primary follicles, secondary follicles, and the corpus luteum (Fig. 2A-b and e). In the ovarian/Graafian follicles, expression was observed in the oocytes, granulosa cells/follicle cells and theca (interna and externa) (Fig. 2A-b, c, e and f). We also found Fshr-ZsGreen expression in the ciliated epithelial cells in the oviduct (Fig. 2A-d and h). Furthermore, we employed an antibody against Stra8 ²³ to perform immunofluorescence staining to identify reproductive cells and observed the colocalization of Fshr-ZsGreen with Stra8 staining as a marker for reproductive cells (Fig. 2A).

Figure 2.

In the testis, we found Fshr-ZsGreen expression and its colocalization with Stra8 ²⁴ staining in the cells of seminiferous tubules (STs), including primary spermatocytes, Sertoli cells and spermatids, and particularly in interstitial Leydig cells (LCs) between STs (Fig. 2B). Figure 2B-a shows an image of the whole sectioned testis. A representative ST is displayed at two magnifications (Fig. 2B-b and c), demonstrating strong expression of Fshr-ZsGreen in Leydig cells, as indicated by empty white arrows. In addition, we also applied an antibody against Set to identify testis cells, whose expression was reported in multiple cell types of the mouse testis at different developmental stages ²⁵. In Figure 2C, a representative image of ST with Leydig cells is shown at a lower magnification (400X) and a higher magnification (1000X) (Fig. 2C-a and b), showing colocalization of Fshr-ZsGreen and Set staining in testis cells. We found strong Set signal in Fshr-ZsGreen-positive spermatogonia, as indicated by empty white arrowheads (Fig. 2C-b). Fshr-ZsGreen was also observed in the arterioles of the testis (Fig. 2C-c).

To confirm Fshr expression in the LCs, we further carried out single RNA molecule-fluorescence in situ hybridization using antisense probes for identifying Fshr expression in the sections of Fshr-ZsGreen mice, whereas its sense probe was used as a negative control (Fig. 2D and E). This finding validated that Fshr was not only expressed in the cells in the ST but also strongly expressed in the LCs, verifying the expression of Fshr in Leydig cells.

Overall, we observed Fshr-ZsGreen expression in the reproductive system, demonstrating that ZsGreen is a readout for Fshr expression. In addition to its sole expression in granulosa cells and Sertoli cells, as reported previously, our findings clearly reveal that Fshr is also expressed in other cell types in the reproductive system, particularly in Leydig cells.

2) Skeletal tissues

Fsh has been thought to have a direct role in bone, ¹¹ therefore we next examined the expression pattern of Fshr in femoral sections, as shown in Figure 3A, under confocal fluorescence microscopy. The representative areas are presented at two magnifications of 400X and 1000X. In the epiphyseal growth plate, we observed lower expression of Fshr in chondrocytes, as indicated by two dotted lines, compared to its expression in cells located in the sponge area above or under the dotted lines (Fig. 3A-a). At higher magnification, Fshr was expressed in the columns of chondrocytes from the resting zone to the transformation zone (Fig. 3A-e).

Figure 3.

In contrast, we found that Fshr was brightly expressed in osteoblasts and osteoclasts in the metaphyseal trabeculae, which were recognized based on nuclear DAPI staining-osteoblasts were stained with a single DAPI nucleus, while osteoclasts contained more than two DAPI-stained nuclei (Fig. 3A-c and d). Similarly, we further observed that these cells on the surfaces of trabeculae in bone marrow and cortical bone also clearly expressed Fshr-ZsGreen, as shown in Fig. 3A-e and f (trabeculae) and-g and h (cortex). Importantly, we noted Fshr-ZsGreen expression in osteocytes (indicated by empty white arrowheads, Fig. 3A-d, g and h), the most abundant cell in the skeleton. These were embedded in trabecular and cortical bone, as well as in the periosteum (P) (Fig. 3A-h).

To confirm the identification of osteoblasts/osteocytes and osteoclasts that expressed Fshr, we performed immunofluorescence (IF) staining using an antibody against osteocalcin, a marker of osteoblasts. ²⁶ As shown in Fig. 1B, we observed colocalization of Fshr expression with osteocalcin staining in chondrocytes in the transforming zone, as indicated by white arrows (Fig. 3B-a), in cuboid and nucleated osteoblasts and bone lining cells on the trabecular and endosteal surfaces (Fig. 3B-b and c), cortical bone, and osteocytes within the mineralized matrix (Fig. 3B-b, c and e).

To identify osteoclasts, we performed fluorescence immunostaining with an antibody against Trap, an osteoclast marker. ²⁷ Osteoclasts were recognized by positive staining for TRAP with more than two DAPI-stained nuclei. Multinucleated Fshr-ZsGreen-positive cells were stained positively for Trap located on the surface of the resorptive bays or areas adjacent to the trabecular bone (Fig. 3B-e and f).

To examine whether skeletal stem/progenitor cells express Fshr, we performed IF staining for stem markers with antibodies against CD34 or CD133. ^28–32 Using these well-known stem markers, we identified Fshr-ZsGreen-positive cells as stem/progenitor cells located in the bone marrow, growth plate/articular cartilage, and periosteum, as shown in Figure 2C-a, b, c, e and d, respectively. These cells also featured an increased nuclear-cytoplasmic ratio, except for these cells on the trabecular surface and in the periosteum.

3) Adipose tissues

Because our previous works and others have provided evidence on the role of Fsh in adipose tissues, we then examined Fshr-ZsGreen expression in adipose tissues. As expected, we found Fshr-ZsGreen expression in adipocytes of the frozen sectioned inguinal WAT, as shown in the left panel of Figure 4A at a low magnification (40X), which was further confirmed under a higher magnification (400X) in the three representative areas, demonstrating that Fshr-ZsGreen was expressed in the cellular membranes of individual adipocytes (Fig. 4A-a, b and c).

Figure 4.

To confirm adipocyte identification, we performed IF with an antibody against mouse Ucp1 ¹² to recognize adipocytes, showing that the majority of Fshr-ZsGreen-positive cells were stained positively for Ucp1, and found two types of Fshr⁺ adipocyte populations: one with colocalization of the two markers only in the membrane (indicated by empty white arrowheads) and another with two markers in both the membrane and cytoplasm (indicated by white arrows) (Fig. 4B-a and b). In addition, we observed that Fshr-ZsGreen was expressed in arterioles in adipose tissue, denoted by white empty arrowheads in Figure 4B-c and a dotted circle in Figure 4B-d.

We further examined Fshr-ZsGreen expression in BAT. Fshr-ZsGreen was observed across the whole section of examined BAT at a lower magnification (Fig. 3A-a). At a higher magnification (400X), three representative areas are presented, as shown in Figure 3A-b, c and d, in which Fshr-ZsGreen was expressed not only in the cellular membranes but also in areas of cytoplasm close to the membranes, as indicated by white arrows (Fig. 4C-b, c, and d).

Furthermore, we also performed IF with three antibodies against Ucp1, Th and Peri to identify brown cells and peripheral fibers in the Fshr-positive section. We found that several areas in the section were strongly stained for Ucp1 at low magnification (40X), as shown in Figure 4D-a. Under higher magnifications of 400X and 1000X, we colocalized Fshr-ZsGreen with Ucp1 in the three representative areas, in which some locations had higher Fshr-ZsGreen expression and others had higher Ucp1 expression (Fig. 4D-a, b and d).

To examine whether peripheral neural fibers express Fshr, we used antibodies against Th and Peri that can recognize peripheral neural fibers. We found that neural fibers stained positively and surrounded Fshr-ZsGreen-positive arterioles in BAT (Fig. 4E-a and b). In addition, Fshr-ZsGreen was expressed in the nodes of Ranvier of TH-stained small neural fibers (indicated by empty red arrows, Fig. 4E-e). We further confirmed Fshr-ZsGreen expression in peripheral neural fibers by IF with an antibody against peripherin (Peri), which is a type III intermediate filament protein found predominantly in peripheral nerves, specifically in sensory and autonomic neurons (Fig. 4F). We noted colocalization of Fshr-ZsGreen and Peri staining in large neural fibers (Fig. 4F-b, d and e). Interestingly, we observed that both markers for peripheral neural fibers were also expressed in the cytoplasm of BAT cells, in which Fshr-ZsGreen was strongly expressed (Fig. 4E-b, c and f; Fig. 4Fe, c and f).

Taken together, the above-described findings on Fshr-ZsGreen expression in reproductive, skeletal, and adipose tissues convincingly demonstrate that Fshr-ZsGreen is a reliable readout of Fshr expression. Furthermore, we identified Fshr expression in Leydig cells and follicles at different developmental stages, cells of osteoblast lineage and peripheral nerve fibers. With confidence that Fshr-ZsGreen is a reliable readout, we used this powerful tool to further examine Fshr expression in other tissues and organs.

4) Heart and aorta

To examine Fshr expression in the cardiovascular system, we used the heart and aorta as key organs/tissues to detect Fshr-ZsGreen expression. As expected, we observed strong Fshr-ZsGreen expression in the myocardium (Fig. 5A-a) and large muscular arteries (a representative is shown in Fig. 5A-b). Then, we further confirmed Fshr-ZsGreen expression by IFs with two antibodies against α-SMA and EMCN ²¹ that recognize alpha smooth muscle actin of smooth muscle and endomucin of endothelial cells at higher magnifications. With IF staining for α-SMA, we imaged several areas of both heart and blood vessels, as shown in the left image of the upper panel (40X). In the heart, we observed that Fshr-ZsGreen was highly expressed in cardiomyocytes in longitudinal and transverse orientations of the myocardium, which were positively stained for α-SMA (Fig. 6B-a, b and c). At a magnification of 1000X, it was also expressed in the endothelial layer of arterioles between muscle fibers (Fig. 5B-i).

Figure 5.

Figure 6.

In addition to the cardiomyocytes, we also observed Fshr-ZsGreen in α-SMA-stained smooth muscles and endothelial cells of large blood vessels above the heart (Fig. 7B-d). Interestingly, we found Fshr-ZsGreen in adipose tissue around the blood vessels, and the adipocytes morphologically appeared to be brown adipose cells, as the majority of these brown-like cells were full of ZsGreen-positive cytoplasm, instead of single large fat droplets with Fshr-ZsGreen expression in cellular membranes (Fig. 5B-c, d, e and j). The adipose tissue was stained positively for α-SMA, suggesting that the ZsGreen⁺ structures costained for α-SMA are blood vessels within the beige tissues, which are indicated by white empty arrowheads (Fig. 5B-e and j).

Figure 7.

Furthermore, we observed bright Fshr-ZsGreen expression with slightly weak staining for α-SMA in the layer of endothelial cells of the large vein in Figure 5B-f and h. Under the endothelial cells, a cluster of smooth muscles showed colocalization of Fshr-ZsGreen with positive staining for α-SMA (Fig. 5B-f and h). In contrast to the vein, we observed a thin layer of endothelial cells (tunica intima) and stronger Fshr-ZsGreen expression with positive staining for α-SMA in the large artery (tunica media); Fshr-ZsGreen was also present in the cells in the tunica adventitia (Fig. 5B-k). In frozen sections of hearts, we observed strong Fshr-ZsGreen expression in neural fiber clusters that were in adipose tissue, as shown in Figure 5B-g and m.

Using an antibody against EMCN, a marker for endothelial cells, we further confirmed Fshr-ZsGreen expression in the layer of endothelium (tunica intima), which showed visible positive staining for EMCN when imaged at 400X and 1000X (Fig. 5C-a, b and d). In addition, Fshr-ZsGreen⁺ cardiomyocytes and smooth muscles under the endothelial layer were positive for EMCN staining (Fig. 5C-a, b and c).

In addition to large blood vessels on the heart, we also took a close look at the ascending aorta. We obtained two types of sections with longitudinal and transverse orientations for IFs with the two antibodies as above. In both sections, we found that Fshr-ZsGreen⁺ smooth muscle fibers were strongly stained for α-SMA in the first layer of SM close to the endothelium, where the second layer of smooth muscle was relatively weak for staining with α-SMA (Figs. 5D-a and 6B-e). Fshr-ZsGreen was also present at the endothelium, which was costained positively for α-SMA (Fig. 6A-a and 6B-e). Using an anti-EMCN antibody, we noticed that positive staining was in the upper part of Fshr-ZsGreen⁺ endothelial cells facing the lumen of the examined blood vessels (Fig. 5A-c and 5B-f). In addition to Fshr-ZsGreen expression in the endothelium of the tunica intima, it was also observed in the areas of subendothelial connective tissue and tunica media (smooth muscle), but these areas were not stained for EMCN (Fig. 5A-c and 5B-f).

5) Lung and kidney

To identify whether Fshr is expressed in epithelial cells, we first detected Fshr-ZsGreen in the lung. Unexpectedly, we observed high Fshr-ZsGreen expression in the lung (Fig. 6A). Fshr-ZsGreen was brightly expressed in the columnar epithelium of the respiratory conducting zone/tract, including the trachea, bronchus, bronchi and bronchiole, at low magnification (Fig. 6A-left panel). At higher magnifications, it was clearly shown that Fshr-ZsGreen was expressed not only in the columnar epithelium but also in the bronchial gland and alveoli (Fig. 6A-a, b, d, e, f and g). In the alveoli, Fshr-ZsGreen was observed in the respiratory portion of both type I and II cells, as indicated by empty arrowheads and white arrows, respectively (Fig. 6A-d and h). We confirmed the identification of respiratory cells by IF with an anti-PD-L1 antibody, which showed the colocalization of Fshr-GFP with PD-L1 ³³ staining (Fig. 6A).

Then, we aimed to examine Fshr expression in epithelial cells of the kidney. In the frozen section of the kidney, it was astonishing to observe high expression of Fshr-ZsGreen in the proximal and distal convoluted tubules (Fig. 6B-e and f), whereas relatively weak expression was observed in the glomerular capillaries at different magnifications (Fig. 6B-e and f at 400X and g at 1000X). Again, we observed Fshr-ZsGreen expression in the endothelial layer of the arteriole in the kidney tissue (Fig. 6B-h). We also observed colocalization of Fshr-ZsGreen with positive staining for Col1a1 ³⁴ in the kidney (Fig. 6B).

6). Other key tissues and organs (liver, pancreas, thyroid, skin and skeletal muscle, spleen, bone marrow, and brain)

With the power of Fshr-ZsGreen, we characterized Fshr-ZsGreen expression in the liver. We observed Fshr-ZsGreen expression in the hepatocytes and arteries inside the hepatocytes, which were positively stained for Col1a1 or CD31 ^35,36, respectively (Fig. 7-A and B). Although it is weakly expressed inside large nerve fibers, Fshr-ZsGreen is strongly expressed in small neural fibers and shows a costaining pattern with KCBMA1, ³⁷ a marker used for the detection of peripheral nerve fibers (Fig. 7-C, indicated by white arrowhead, whereas large nerve fibers are indicated by white arrows).

We then examined Fshr-ZsGreen expression in the pancreas and found that it was expressed not only in acinar cells but also in islets of Langerhans at low and high magnifications (Fig. 8-A-a and b). We then confirmed the identification of α and β-cells by IFs with antibodies against NG3, insulin or glucagon. The imaging results clearly demonstrated that Fshr-ZsGreen was expressed in both α and β-cells as well as acinar cells at 400X and 1000X magnifications, respectively (Fig. 8-A, –B and –C).

Figure 8.

Regarding Fshr-ZsGreen expression in the thyroid, we found Fshr-ZsGreen in both follicular cells and parafollicular cells (C-cells) at low and high magnifications (400X and 1000X) (Fig. 9-A-a, b, c, d and e). We further used an anti-Tsh receptor antibody to confirm Fshr-ZsGreen-positive cells, which showed that follicular cells were stained positively for Tshr in the nuclei, as indicated by dotted circles and white arrowheads, and C-cells were indicated by white arrows (Fig. 9-b to e).

Figure 9.

Next, to examine Fshr-ZsGreen expression in the skin, we used two types of skin sections: thick skin from the tail and thin skin from the abdomen (Fig. 10-A-a and b). Although a dermis layer was not included in the image taken for a thick sample, we observed that Fshr-ZsGreen was expressed in hair follicle (HF) and sweat gland cells and fibroblasts in the dermis and hypodermis (Fig. 10-A-a). Similarly, Fshr-ZsGreen was present in HFs and fibroblasts in the dermis and keratinocytes in the epidermis of the thin skin (Fig. 10-A-b). As we detected Fshr-ZsGreen expression in hair follicles, we wondered whether stem cells in HFs express Fshr-ZsGreen. To address this question, we carried out IF with an antibody against CD34, a stem cell marker. Not surprisingly, we found that CD34 staining was colocalized with cells with Fshr-ZsGreen in Bulge as quiescent stem cells and in the dermal papilla (DP) and epidermis as active stem/progenitor cells, as shown in Figure S4-B-a and b. Therefore, these results indicate Fshr-ZsGreen expression in stem/progenitor cells in the skin (Fig. 10B-a and d).

Figure 10.

We then detected Fshr expression in skeletal muscle (gastrocnemius), followed by IF staining using three antibodies against αSMA, PAX7 and TH for the identification of satellite cells and peripheral nerve fibrils. We observed Fshr-ZsGreen across the muscle sections at a lower magnification of 40X (Fig. 11a). At higher magnifications, we found that Fshr-ZsGreen was present in the muscle fibers, in which one type had higher Fshr-ZsGreen expression (indicated by white arrowheads) and another type had lower Fshr-ZsGreen expression (Fig. 11b to h). Fshr-ZsGreen was also highly expressed in satellite cells stained positively for either αSMA (Fig. 11-b to e) or PAX7 (Fig. 11-f and g) in both longitudinal and transverse sections, as indicated by white arrows. In addition, we detected Fshr-ZsGreen expression in the peripheral neural fibers identified by positive staining for TH, which were around vascular structures in the muscle tissue, as indicated by red arrowheads (Fig. 11-h and i).

Figure 11.

To detect Fshr expression in immune cells, we examined Fshr-ZsGreen expression in the spleen and bone marrow using antibodies against CD11b and CD3, as integrin αM (CD11b) is expressed in myeloid-lineage cells such as monocytes/macrophages, neutrophils, eosinophils, and basophils and in lymphoid cells such as NK cells and B-1 cells, ³⁸ while CD3 marks T cells. ³⁹ We imaged frozen sections of the spleen under magnifications of 40X, 400X and 1000X and observed that Fshr-ZsGreen was highly expressed in trabeculae and cells in both red and white pulps (RP and WP). The Fshr-ZsGreen-positive cells were further identified by IFs with antibodies against CD11B or CD3, indicating that Fshr is expressed in immune cells, such as monocytes/macrophages, neutrophils, eosinophils, basophils, NK cells, B cells and T cells (Fig. 12-A and B). In addition, we further confirmed Fshr-ZsGreen expression in macrophages of bone marrow, which were identified with anti-CD4/80 antibody, as shown in two representative images of BM-one from an area close to cortical bone (Fig. 12-C-a) and another located in the center of BM (Fig. 12C-b) at 1000X magnification (empty white arrows indicate the colocalization of Fshr-ZsGreen and CD11B, CD3 or CD4/80 in the RPs, while white arrows point to their colocalization in the WPs).

Finally, we examined Fshr-ZsGreen expression and its colocalization with markers for either astrocytes, microglia, or neurons in the brain. As expected, we observed Fshr expression across the brain sections and the three representatives from the olfactory bulbs, pallidum, and hippocampus (CA3) are shown in Figure 12, demonstrating that Fshr is indeed expressed in astrocytes (Fig. 13-a to c), microglia (Fig. 12-d to f) and neurons (Fig. 13-g to j), as well as neuronal fibers (synapses or projections). Other cell types are needed to be further defined.

Figure 13.

3. Confirmation of Fshr-ZsGreen expression with a Fshr-specific antibody and ddRT-PCR

Finally, to confirm the accuracy of the above results obtained from Fshr-ZsGreen mice, we performed fluorescence immunostaining with a specific antibody against mouse Fshr and accurate droplet digital RT‒PCR (ddRT-PCR) with mouse Fshr-specific primers to confirm the above data. An isotype-matched rabbit IgG was used as a negative control for IFs with anti-Fshr antibody using sections from Fshr-ZsGreen mice. Imaging was performed under the same conditions to record each corresponding tissue/organ stained with anti-Fshr antibody. The images of negative controls are shown in Supplementary Data 2, showing specific binding of the secondary antibody to anti-Fshr antibody without any nonspecific binding of the secondary antibody to the examined sections.

As shown in Figure 14A, in all the examined tissues, including bone, BAT, thyroid, cardiac muscle, kidney, liver, lung, aorta, ovary and testis, we observed colocalization of Fshr-ZsGreen with positive staining for Fshr, further confirming the specific expression of Fshr-ZsGreen in the examined tissues/organs.

Figure 14.

Figure 15.

We also obtained total RNA from the following tissues: lung, kidney, thoracic vertebra, calvaria, femur, jejunum, liver, teeth, tibia, skeletal muscle, tails, cartilage, skin, spleen, stomach, heart, bladder, tongue, BAT, WAT, thyroid, brain, pancreas and duodenum. After reversing mRNA from total RNA, we performed ddRT-PCR with mouse-specific primers to check Fshr expression at the transcriptional level as described in the Methods section. The results demonstrated Fshr expression in these tissues, and the expression profile was categorized into three groups: 1) high expression in the lung and kidney tissues; 2) moderate expression in the thoracic (T) vertebrae, calvaria, femur, jejunum, liver, teeth, tibia, and muscle; and 3) low expression in the tail, cartilage, skin, spleen, stomach, heart, bladder, tongue, BAT, WAT, brain, pancreas, and duodenum (Fig. 14B).

Taken together, these data further confirmed Fshr-ZsGreen expression patterns from the Fshr-ZsGreen reporter line, convincingly demonstrating that Fshr is not limited to previously reported cells, tissues, or organs, such as the reproductive system, osteoclasts, adipose, endothelium in tumors, and neurons in the brain, but rather has a wide expression in the cells, tissues, and organs in the body, particularly in the lung, kidney, and heart, as well as Leydig cells in the testis, which has not previously been recognized.

Discussion

Although numerous efforts have been made to characterize Fshr expression in tissues/cells, defining the locations of Fshr expression remains an imperative challenge in Fsh-Fshr biology, primarily because of concerns about the specificity of available antibodies against Fshr. ^2,15,16 In this case, we developed CRISPR/Cas9-mediated Fshr-ZsGreen knockin reporter mice to address this issue. The P2A-ZsGreen construct was precisely inserted into the site between the last exon (exon 10) and the stop codon of the Fshr locus, which was confirmed by integration detection PCR, sequencing of PCR fragments and Southern blotting with Fshr locus-specific enzyme restriction digestions. Successful insertion allows the endogenous promoter of Fshr to drive ZsGreen reporter expression. This approach greatly enhanced our understanding of this critical cellular pathway in several ways. Firstly, employing the native regulatory element responsible for governing the expression of the target gene ensures that the expression pattern of the GFP reporter closely mirrors that of the gene of interest, Fshr, within its natural context. Consequently, this approach provides a more faithful representation of gene expression. Secondly, endogenous promoters often exhibit specific spatial and temporal expression patterns, driving gene expression in specific cell types or developmental stages. This capability facilitates the capture of the dynamic nature of gene regulation and enables the study of gene expression under diverse physiological conditions. Thirdly, utilizing endogenous promoters minimizes potential perturbations to the native gene regulation machinery. It avoids the need to introduce exogenous elements or artificial constructs, reducing the likelihood of altering the gene’s expression behavior or interfering with its regulatory interactions. Unlike biochemical assays or immunostaining, using a tagged protein under endogenous regulation avoids fixation artifacts and allows detection of the target’s activity in live cells. Therefore, these distinct advantages of enhanced physiological relevance, precise spatiotemporal control, preservation of regulatory elements, and minimal perturbation provide us with a powerful tool to understand Fshr expression in more accurate and context-specific ways compared to other methods, such as using antibodies, Northern blotting, RT‒PCR and in situ hybridization.

In this study, we systemically investigated Fshr expression at the single-cell level in this reporter line with confocal fluorescence microscopy and further confirmed the location of Fshr-ZsGreen expression by IF staining, in situ hybridization and ddRT‒PCR. The results from this work demonstrate that as a receptor for Fsh, Fshr is widely expressed in virtually every cell at variable levels in the examined tissues/organs of mice. As expected in the examined testis, we noticed Fshr expression in Sertoli cells in the testis and granular cells in the ovary. Surprisingly, we observed that Fshr was also more strongly expressed in Leydig cells than in Sertoli cells. This expression was also detected in spermatocytes, spermatids, spermatozoa and spermatogonia. Although this finding is different from present thoughts that Fshr is only present in Sertoli cells but not in other cell types of the testis, our finding is in line with previous works in fishes (including African catfish, zebrafish, teleosts, and Japanese eels) ^40–45, rats ⁴⁶, dogs ⁴⁷ and humans ⁴⁸ performed by IHC and in situ hybridization. In humans, Fshr was also highly expressed in Leydig cells, although the non-specificity of the anti-Fshr antibodies was noted. However, as shown in this study, this is not the case, because Fshr is widely expressed, as demonstrated in our work. Furthermore, our finding is supported by a report of the failure of normal Leydig cell development resulting from the deficiency of Fshr but not Fsh-beta. ⁴⁹ The Fshr expression in Leydig cells strongly indicates that Fsh-Fshr plays a role in the production of steroids in males. In the ovary, Fshr is highly expressed in follicles at different stages, from primordial cells, primary follicles, and secondary follicles to mature/Graafian follicles and the corpus luteum. In addition to granulosa cells, Fshr expression was observed in oocytes of follicles.

In the skeletal system, we observed Fshr expression in vivo not only in osteoclasts, as reported previously, ^11,50,51 but also, interestingly, in osteoblast lineage cells, such as osteoblasts, bone lining cells, osteocytes and progenitor cells of the periosteum, as well as in chondrocytes. In previous reports, Fshr expression in osteoclasts was detected only by RT‒PCR, Western blot and immunostaining in cultures of primary murine precursors. ^11,50,51 Using the Fshr-ZsGreen reporter line, we visualized Fshr expression in multinucleated osteoclasts in frozen sections, clearly demonstrating Fshr expression in osteoclasts. Intriguingly, this powerful tool enabled us to examine Fshr expression in other cell types in bone. Surprisingly, we observed Fshr expression in cells of the osteoblast lineage, from osteoprogenitor cells and osteoblasts to osteocytes and bone lining cells. This finding indicates that Fsh may regulate not only osteoclast-mediated bone resorption but also osteoblasts for bone formation. To functionally prove the presence of Fshr in osteoblasts/osteocytes, we also deleted Fshr in osteocytes using an inducible model. The conditional knockout of Fshr triggered a much more profound increase in bone mass and decrease in fat mass than blockade by Fsh antibodies (unpublished data), illuminating Fshr expression in the osteoblast lineage.

In addition to its expression in the reproductive system and skeletal system, we also strikingly identified other cell types that highly express Fshr: endothelial cells in blood vessels and epithelial cells in the lung and kidney. In every examined tissue/organ, we found that endothelial cells stained positively for CD34 lining on the arterioles had a higher expression of Fshr-ZsGreen than other cell types. This bright Fshr-ZsGreen is more obviously seen in large arteries, such as the ascending aorta and others located in the heart. Similarly, Fshr was detected in vessels in solid malignant tumors by immunohistochemistry or RT‒PCR. ⁵² However, it was not seen in normal tissues or organs by these methods, possibly due to its rapid turnover, fast degradation, or selected antibodies. Strikingly, we detected the highest Fshr-ZsGreen expression in bronchial and bronchiole ciliated epithelial cells by both fluorescence microscopy and ddRT‒PCR. It was also more highly expressed in other cell types in the lung, such as type I pneumocytes (alveolar lining cells), type II pneumocytes (great alveolar or septal cells), and gland cells. However, the role of this unexpectedly high Fshr expression in the lung remains unknown. Similarly, we also found the second highest expression of Fshr-ZsGreen in renal epithelial cells in proximal and distal convoluted tubules but weak expression in renal corpuscles.

In summary, we established and validated an Fshr-ZsGreen protein reporter in vivo that faithfully recapitulates endogenous Fshr expression at single-cell resolution. Our compelling findings reveal that in addition to gonadal tissues, Fshr is also highly expressed in extragonadal systems, such as the lung, kidney, heart, and pancreas. This will provide insight to better understand the biology of the Fsh-Fshr axis and its roles in the physiology and pathology of these tissues/organs. In addition to the above described, we detected Fshr expression in cells of the teeth and brain; those findings are not presented here because of space limitations and will be published elsewhere, except that three representative areas of the brain are shown in Figure 7-6.

Although the Fshr-ZsGreen reporter line is a powerful tool for detecting the location of Fshr expression, it is limited in the definition of individual isoforms of Fshr transcripts and the detection of their turnover, which should be addressed by specific antibodies and determined by the rates of transcription and RNA degradation, including Xrn1 for 5’-to-3’ degradation, exosomes for 3’-to-5’ degradation and nonsense-mediated decay ⁵³.

Acknowledgements

This study was supported by Shanxi Medical University, Taiyuan, China.

Competing interests

The authors declare that we have no competing interests related to this research.

Supplementary Materials 1

1.

2.

3.

Supplementary materials 2

1.

2.

The Supplementary Data 1

Supplementary Data 2