C18-4 cells were established by stably transfecting type A spermatogonia from 6-day-old mice with the large T antigen gene under the control of a ponasterone A-driven promoter (Hofmann et al., 2005; He et al., 2009), and they were cultured in DMEM/F12 supplemented with 10% FBS and maintained in a 5% CO₂ atmosphere.

TM3 cells were purchased from the American Type Tissue Culture Collection (ATCC, Manassas, VA, United States) and maintained in a 5% CO₂ atmosphere in DMEM/F12 (HyClone, United States) supplemented with 2.5% FBS (HyClone, United States), 5% HS (Solarbio Beijing, China) and 1% P/S solution (Solarbio Beijing, China).

C18-4 cells were cultured in 96-well plates and treated with different concentrations of CGA (Cat. No. YZ-110753, HPLC ≥96.1%), quercetin (Cat. No. SQ8030, HPLC ≥98%), kaempferol (Cat. No. SK8030, HPLC ≥98%) and rutin (Cat. No. SR8250, HPLC ≥98%) (0–100 μM) (Solarbio Beijing, China). Thirty-six hours after treatment, 10 μl of WST-1 (Beyotime Biotechnology, Shanghai, China) solution per well was added and the plate was incubated for 1 h at 37°C. The absorbance of each well was measured at 450 nm by a microplate spectrophotometer. The cell viability was calculated using the following formula: Cell viability= (OD_{treated group}−OD_{blank group})/(OD_{control group}−OD_{blank group}).

EF (voucher specimen number: TCVM-15082501) was collected from Lueyang, Shaanxi Province (33°07′55″N, 105°42′31″E) and stored in the specimen room of Traditional Chinese Veterinary Medicine (TCVM) of the College of Veterinary Medicine. EF powder (160.0851 g) was refluxed for 30 min, at 97°C, with 1600 ml 50% ethanol solution (v/v). After repeated extraction for four times and filtration, the filtrate was concentrated at 90°C and dried to a constant weight by vacuum freeze dryer. The extract was 66.0413 g and the yield of the extract was 41.25% (w/w) (Hou et al., 2016). The content of CGA in EFEE was analysis by HPLC (Supplementary Material).

All experiments were performed on 5-week-old healthy male KM mice. These mice were purchased from Chengdu DOSSY Experimental Animals Co., Ltd. The mice were housed in wire cages at 25°C under a 12 h light-dark cycle with 70% humidity.

EFEE spermatogenic activity assay. Fifty mice were divided into five groups: a negative control group, three experimental groups and a positive control group. The negative control group was intragastrically administered with water, the experimental groups were intragastrically administered with 0.4, 0.8 and 1.2 g/kg EFEE, and the positive control group was intraperitoneally injected with 5 mg/kg testosterone propionate. Each group of mice was given intragastric administration or injection once a day for 10 days.

CGA spermatogenic activity assay. The experimental method was the same as above. The experimental groups were intragastrically administered with 20, 40 and 80 mg/kg CGA.

After 10 days of administration, the mice were anaesthetized, the serum isolated from blood that was collected through heart punctures, and the testes and epididymides were collected. Testosterone and FSH were detected by ELISA kits (Shanghai Enzyme-Linked Biology Company Shanghai, China) according to the instructions. The testes were carefully removed and fixed with 10% formalin solution. After dehydration procedures, the sections were embedded in paraffin. The tissues were after that cut (with a microtome) to produce 4–5 μm sections, transferred to slides and subsequently stained using the traditional hematoxylin and eosin (H&E) stain. The testicular organ coefficient was calculated using the following formula: Testicular organ coefficient = Testicular weight/Body weight. The sperm were collected from the cauda epididymides. The cauda epididymides were minced with scissors and placed in 5 ml of 0.9% saline at 37°C for 30 min. After filtration, 10 μl aliquots of these mixtures were placed on a hemocytometer, and sperm numbers were determined by counting under an optical microscope (Park and Han, 2013; El-Khadragy et al., 2021).

For all experiments, logarithmic growth phase C18-4 cells were plated in six-well cell culture plates and treated with 0, 10, 50 and 100 μM CGA for 36 h. Logarithmic growth phase TM3 cells were plated in six-well cell culture plates and treated with 0, 0.5, 1 and 10 μM of CGA for 36 h. Then, the cells were resuspended as single cells, washed in precooled PBS and incubated using a Cell Cycle Kit (LianKeBiology, Hangzhou, China) with 1 ml DNA staining solution and 10 μl permeabilization solution for 30 min. Cell cycle analysis was performed with a flow cytometer (Cao et al., 2012).

Total RNA was reverse-transcribed to cDNA using the HiScript III RT SuperMix reverse transcriptase reagent kit according to the reagent manual (DiNing, Beijing, China). qRT–PCR was performed on a CFX96 real-time PCR detection system (Bio–Rad, CA 94547) according to the manual for the ChamQTM Universal SYBR qPCR Master Mix kit (DiNing, Beijing, China). For more details on the qRT–PCR protocol, please refer to Wu et al. (2013). The relative expression levels of target genes and differentially expressed miRNAs were normalized to Gapdh expression for each sample respectively. The relative expression levels were calculated using 2^−ΔΔCt (Wu et al., 2014). The verified primers of mRNAs are listed in Supplementary Table S1.

TM3 and C18-4 cells were transfected with shRNA (Shp2-mus-405: 5ʹ-GCT​GAA​CTG​GTT​CAG​TAT​TAC​TTC​A AGA​GAG​TAA​TAC​TGA​ACC​AGT​TCA​GCT​T-3ʹ) and NC (5ʹ-TTC​TCC​GAA​CGT​GTC​ACG​TTT​CAA​GAG​AAC GTG​ACA​CGT​TCG​GAG​AAT​T-3ʹ) (Genepharma Co., Shanghai, China) in a 24-well plate. After 24 h, two kinds of transfected cell were respectively treated with G418 (400 μg/ml) and hygromycin (100 μg/ml). The stably transfected cell clusters were screened by drugs and cultured. These cells were detected by qRT–PCR and Western blot.

The protocol of immunofluorescence staining was performed according to Yu et al. (2014). Briefly, C18-4 cells and Shp2 knockdown cells, which were cultured in a 48-well plate after treatment with CGA, were fixed with 4% formaldehyde for 15 min and washed with PBS three times for 3 min each. The cells were permeabilized with by 0.1% Triton X-100 (Solarbio, Beijing, China) for 15 min and blocked for 30 min with 1% BSA at 37°C. Then, the cells were incubated with primary antibodies specific against PLZF (1:200, Santa Cruz Biotechnology, California), SHP2 (1:100, Abways, Shanghai, China) and C-KIT (1:300, Biolegend, San Diego) for 12 h at 4°C. Alexa-488 (1:500, Beyotime Biotechnology, Shanghai, China) secondary antibodies were used to incubate cells for 1 h at 37°C. The negative control was stained with conjugated secondary antibodies alone: goat anti-rabbit IgG and goat anti-mouse IgG. The nuclei of cells were stained with DAPI (Beyotime Biotechnology, Shanghai, China) (Niu et al., 2016).

The cells were treated with different doses of CGA for 36 h. The proteins were extracted using a protein extraction reagent kit (Solarbio, Beijing, China). Protein quantification was performed using a BCA quantification kit (Solarbio Beijing, China). 10% separation gel was prepared. Each well was loaded with 30 μg of the protein. The denatured proteins were separated by SDS–PAGE (80 V, 30 min; 120 V, 60 min) and transferred onto PVDF membranes (80 V, 135 min). The membranes were blocked for 2 h at 37°C in TBST containing 5% skim milk and then incubated for 12 h at 4°C in TBST containing specific primary antibodies (SHP2, ERK1/2, pERK1/2, StAR, or β-actin, 1:1000) (Abways, Shanghai, China). After four washes with TBST, the membranes were incubated with HRP-conjugated secondary antibody (Solarbio Beijing, China) for 1.5 h at 37°C. After washing the membranes with TBST three times, the signals were visualized using ECL (DiNing, Beijing, China), and the membranes were exposed on X-ray films. The quantification of protein bands were analyzed by ImageJ version 1.51.

The main components and their related genes in EF were collected by summarizing the research work on chemical components of EF (He et al., 2014; Wang C.Y. et al., 2019) and searching TCMSP (Ru et al., 2014) (http://tcmspnw.com), Pharm Mapper (Wang et al., 2017) (http://www.lilab-ecust.cn/pharmmapper/) and other traditional Chinese medicine components databases. The selected genes were analysed by literature research (Paz et al., 2016; Puri and Walker, 2016; Zhou et al., 2016), Kegg (Kanehisa et al., 2017) (https://www.kegg.jp/kegg/) and KOBAS (http://kobas.cbi.pku.edu.cn/) enrichment to investigate the possible biological functions of the potential targets and the biological pathways involved in spermatogenesis. Cytoscape version 3.5.1 was used to draw the network diagram of “component-target-pathway-function” (Shannon et al., 2003; Killcoyne et al., 2009) (Supplementary Material).

Data analysis was performed using IBM SPSS statistical software (version 23.0). One-way analysis of variance with a post hoc test was used for multiple comparisons. The histograms were drawn using GraphPad Prism 7. The results were considered significant or extremely significant at a level of p < 0.05 or p < 0.01.

As shown in Figures 1A,B, compared with the control group, the serum testosterone and FSH levels of 0.8 and 1.2 g/kg EFEE groups were significantly increased (p < 0.01, p < 0.05). Similarly, the testicular tissue coefficient of mice in the medium-dose groups was significantly higher (p < 0.05) than that in the control group (Figure 1C). The sperm count of the 0.8 and 1.2 g/kg EFEE groups were higher (p < 0.05) than that in the control group (Figures 1D,E). Therefore, EFEE could promote the secretion of spermatogenic hormones and sperm production in mice.

Through network pharmacological analysis, chemical compounds in EF related to spermatogenesis mainly involved MAPK, RAS, PI3K-Akt and regulating the pluripotency of the stem cell signaling pathway. These signaling pathways were involved in the biological processes of SSC proliferation, self-renewal and Leydig cell testosterone secretion. The corresponding compounds were screened to obtain quercetin, kaempferol, CGA and rutin. The targets of these compounds were mainly concentrated in four signaling pathways (Figure 2A). These results suggested that quercetin, kaempferol, CGA and rutin were the main active components in EF that may be involved in spermatogenesis.

As shown in Figure 2B, the four compounds screened above were tested for their biological activities in promoting SSC proliferation and testosterone secretion. Compared with the control group (0 μM), C18-4 cell proliferation in the CGA groups (10 and 50 μM) below their safe concentrations was significantly increased (p < 0.05, p < 0.01). In addition, quercetin treatment (10 μM) also significantly promoted cell proliferation (p < 0.05).

As shown in Figure 2C, the CGA (0.5 and 1 μM) and quercetin (10 μM) treatment groups also significantly enhanced (p < 0.01, p < 0.05) testosterone secretion in TM3 cells, and the CGA groups exhibited stronger activity at low concentrations. The content of CGA in EFEE was 2.58% (w/w) by HPLC (Supplementary Figure S1). These results indicated that CGA was the main compound of EFEE and involved in spermatogenesis. Therefore, CGA was selected for further experiments.

In order to determine the effect of CGA on spermatogenesis in vivo, the mice were intragastrically administered with 20, 40 and 80 mg/kg CGA. As shown in Figures 3A,B, compared with the control group, the serum testosterone and FSH levels in groups treated with 20 and 40 mg/kg of CGA were significantly increased (p < 0.05). Similarly, the testicular tissue coefficient of mice in the medium-dose groups was significantly higher (p < 0.05) than that in the control group (Figure 3C). The sperm count of 20 and 40 mg/kg CGA groups were higher (p < 0.01, p < 0.05) than that in the control group (Figures 3D,E). The results suggested that CGA could promote the secretion of spermatogenic hormones and sperm production in mice.

KEGG enrichment analysis showed that CGA might affected the SSC proliferation, self-renewal, and leydig cell testosterone secretion by regulating MAPK, RAS, PI3K-Akt and Pluripotency of stem cells signaling pathways (Figure 2D). As shown in Figure 2E, SHP2 was involved in these signaling pathways above. These results indicated that Shp2 is a key gene in CGA regulation of spermatogenesis. Therefore, the subsequent studies mainly focused on the regulation of CGA on the expression of SHP2 and its downstream genes.

SHP2 is a widely expressed protein tyrosine phosphatase that is necessary for signal transduction from multiple cell surface receptors (Puri and Walker, 2016). Compared with the control group (0 μM), CGA (10 and 50 μM) significantly increased (p < 0.05) the S phase and reduced the G1 phase of the C18-4 cell cycle (Figures 4A,B). This indicated that CGA could enhance the C18-4 cells proliferation. The inhibition of SHP2 expression significantly decreased the proportion of cells in S phase and increased the proportion of cells in G1 phase (p < 0.01). Moreover, CGA had no significant effect on the cell cycle of Shp2 knockdown C18-4 cells. These results suggested that the Shp2 gene plays a key role in the promotion of C18-4 cells proliferation by CGA.

The SHP2-MAPK signaling pathway is required for maintaining SSC proliferation and self renewal, and spermatogenesis (Puri et al., 2014). As shown in Figures 5A,B, CGA (1, 10, 20 and 50 μM) treatment induced the SHP2 expression (p < 0.05, p < 0.01) and ERK1/2 phosphorylation (p < 0.05, p < 0.01) in C18-4 cells. After inhibiting Shp2 gene expression, ERK1/2 phosphorylation was significantly reduced compared with that in the control group (0 μM). Meanwhile, CGA could not significantly promote ERK1/2 phosphorylation in Shp2 knockdown cells. These results indicated that CGA could induce SHP2-MAPK signaling pathway activation in SSCs.

SSCs play an essential role in maintaining highly productive spermatogenesis by self-renewal and the continuous generation of daughter spermatogonia that differentiate into spermatozoa, transmitting genetic information to the next generation (Subash and Kumar, 2021). GFRA1 and PLZF are required for the regulation of SSC self-renewal (Buaas et al., 2004; Costoya et al., 2004), and C-KIT is a critical factor in SSC differentiation (Zhang et al., 2014). As shown in Figures 6A,B, the C18-4 cells treated with CGA (10 and 50 μM) induced the expression of SHP2 and PLZF, but decreased C-KIT expression (p < 0.05, p < 0.01). The inhibition of SHP2 expression reduced PLZF expression and promoted C-KIT expression (p < 0.05, p < 0.01). These results suggested that CGA could promote the self-renewal of C18-4 cells and maintain their homeostasis.

Compared with the control group, CGA (0.5 and 1 μM) effectively promoted testosterone secretion in TM3 cells (p < 0.01). Testosterone secretion in TM3 cells was significantly reduced after the Shp2 gene was inhibited (p < 0.05) (Figure 7C). Furthermore, CGA (0.5 and 1 μM) also significantly increased the proportion of TM3 cells in the S phase of the cell cycle (p < 0.01, p < 0.05). The knockdown of Shp2 gene expression significantly reduced the proportion of cells in S phase compared with that in the control group (p < 0.01, p < 0.05) (Figures 7A,B). These results suggested that CGA could promote testosterone secretion and proliferation in TM3 cells.

SHP2 regulates the expression of the downstream Star gene, which is a rate-limiting enzyme involved in testosterone secretion in leydig cells (Cooke et al., 2011). As shown in Figures 8A,B, CGA (0.5 and 1 μM) induced SHP2 expression (p < 0.01) and promoted StAR expression in TM3 cells (p < 0.01). The inhibition of Shp2 gene expression significantly reduced the expression of StAR compared with that in the control group (p < 0.01, p < 0.05), and CGA did not promote Star gene expression in Shp2 knockdown TM3 cells. These results suggested that CGA could promote the testosterone secretion by improving the expression of SHP2 and StAR.