Introduction
The ECM is composed of proteins, proteoglycans, and glycosaminoglycans (GAGs). ECM-stable microenvironments favor cell homeostasis, whereas ECM structural abnormalities can cause functional deficits and fibrogenesis [
1‐
3].
ECM is required for ovarian development and participates in folliculogenesis and the generation of ovarian epithelial cells. There are many primordial follicles in the ovarian epithelium, and collagen provides structural support for the primordial follicle pool [
4]. ECM provides suitable rigidity for dormant primordial follicles in the ovarian cortical tissue [
5‐
7]. When follicle activation occurs, the activated follicles migrate to ovarian medulla tissue, which is softer than the cortex tissue and beneficial for follicular development [
8]. Starting from adolescence, a few follicles are activated and can develop into mature follicles and become ovulated during every menstrual cycle [
9].
To date, there have been many studies on follicular development and ovulation and the associated regulation of hormone signaling and molecular mechanisms. However, few studies have focused on the effects of ECM on follicular development. The syntheses of ECM proteins in adult ovaries were significantly greater than those in perimenopausal and menopausal ovaries [
10]. The content of ECM proteins, such as collagen [
11], laminin [
11] and elastin [
12] all increased in the early embryonic developmental stage, providing appropriate stiffness for growth and development. Stable ECM expression supports ovary morphology and promotes ovulation. Abnormal ECM may alter ovarian elasticity and cause ovarian stiffness or fibrosis, affecting ovarian follicle development and ovulation [
13].
The primordial follicles go through the primary follicle, secondary follicle, preantral follicle and antral follicle stages until ovulation. After the luteinizing hormone (LH) surge, antral follicles with a certain diameter enlarge, bulge out the ovarian surface epithelium and ovulate [
14]. Briefly, the top of the preovulatory follicle is constructed with a layer of epithelial cells and collagen. The ECM not only provides structural support for the follicle but also maintains cellular connectivity and regulates a wide variety of biological processes including cell proliferation, development, differentiation, migration and metabolism [
15].
Collagens are major constituents of the ECM in ovarian tissue [
16]. Ouni et al. further studied the expression of elastic matrisome components, such as collagen, elastin, fibrillin-1 and GAGs, in the ovaries of prepubescent girls, reproductive age women and postmenopausal women and further analysed the components of the ECM surrounding preantral follicles at pre- and late-pubertal stages. They concluded that ECM is essential the follicular development and homeostasis [
10]. With respect to follicular growth, studies have demonstrated increased laminin expression and have shown that laminin and fibronectin harbor binding sites for integrin, which can initiate intracellular signaling cascades leading to enhanced cell proliferation and differentiation [
17].
In growing follicles, the communications between granulosa cells (GCs) and theca cells (TCs), GCs and GCs, and GCs and oocytes provide pathways for the exchange of small molecules and nutrients. TCs are critical for follicular growth, and can produce sexual steroid hormones at antral follicles in vitro, indicating crucial roles particularly in gonadotropin independent phases [
18,
19]. Collagen IV is localized at the ovarian TC region and stroma in mice and its expression is initiated in primary follicles and is increased in antral follicles. Laminin is localized the at ovarian TC region, stroma and ovarian surface epithelium, especially around TCs. The above two proteins’ conspicuous staining encircles follicles [
20]. Studies have also shown that the expression of Fbn1 (fibrillin 1) decreases from adolescence to menopause and that Fbn3 (fibrillin 3) is highly expressed during stromal expansion before the follicle formation stage, suggesting that Fbns are essential for ovarian and follicular development [
10,
21].
ECM homeostasis plays a critical role in maintaining normal follicular development, and ECM component expression that is either too high or too low can have adverse effects on ovarian function. Genetic, autoimmune iatrogenic, and environmental factors, etc., can trigger ovarian hypofunction and deplete the primordial follicle pool. In addition, the use of chemotherapeutic drugs in cancer patients may induce damage to ovarian function, such as follicular atresia [
22], cortical fibrosis [
23], oocyte apoptosis [
24], impaired GC function [
25] and abnormal hormone secretion [
26]. MSCs exist in various tissues, such as adipose, bone marrow, placenta and umbilical cord, and they can secrete a wide range of antioxidant, anti-fibrosis, anti-apoptotic and anti-inflammatory cytokines [
27]. It has been reported that MSCs have beneficial effects on ovarian damage repair and they can also regulate hormone secretion, decrease GC apoptosis, increase follicle quantity and quality, improve ovarian reserve, and play other roles in improving ovarian function, thereby alleviating the side effects of chemotherapy [
26,
28,
29].
The aim of this study was to decipher the therapeutic role of hUMSCs transplantation on damaged ECM components in mouse ovaries. We focused on ECM organization through transcriptome sequencing and performed both transcriptional and protein analyses. IHC results demonstrated the abnormal expression and localization of collagen IV and laminin gamma 3 in the damaged ovaries, whereas hUMSCs treatment rescued their expressions. This study illustrated that ECM-stable microenvironments were critical for follicular development, and hUMSCs transplantation was effective in restoring chemotherapy-induced ovarian damage.
Materials and methods
Animals
Six- to eight-week-old ICR female mice were purchased from the Jinan Pengyue Experimental Animal Breeding, Co., Ltd. (Shandong, China). The mice were housed in a temperature-controlled room with a 12 h/12 hr light-dark cycle, and fed a regular diet. All procedures with mice were conducted according to the rules stipulated by the Animal Care and Use Committee of Yantai Yuhuangding Hospital.
hUMSC culture and characterization
The hUMSCs were kindly provided by Shandong Qilu Stem Cell Engineering Co., Ltd. The cells were cultured in serum-free medium according to standard experimental protocols approved by the provider. Passage 4 hUMSCs were collected for tail vein injection.
Establishment of the primary ovarian insufficiency (POI) model
Female ICR mice were treated with CTX (120 mg/kg, dissolved in saline) and BUS (20 mg/kg, dissolved in DMSO) to generate the POI model. Total 64 mice were randomly divided into four groups (n = 16/group): control, POI, POI + hUMSCs and POI + PBS. In the POI group, the mice were injected intraperitoneally with a 120 mg/kg CTX and 20 mg/kg BUS mixture. Controls were injected intraperitoneally with an equal volume of solvent at the same time. In the POI + hUMSCs group, 7 days after the injection of CTX and BUS, 1 × 106 hUMSCs diluted in 100 µl PBS were injected into the tail vein. For the POI + PBS group, POI mice were injected with 100 µl of PBS via the tail vein. Another 7 days after the injection of hUMSCs or PBS, the mice were sacrificed for the following studies.
Hematoxylin and eosin (HE) staining and ovarian follicle counting
The ovarian tissues were collected and washed at least 3 times with cold PBS and fixed in 4% paraformaldehyde (PFA) for at least 24 h. After dehydration and paraffin embedding, the samples were processed by sectioning (5 μm) and HE staining. To analyse the numbers of ovarian follicles, primordial, primary, secondary and antral follicles were classified and counted as described previously [
30].
RNA-seq data processing
RNA-seq raw reads were trimmed to remove adapters and low-quality reads using TrimGalore (version 0.6.6) with the parameters “-q 25 --phred33 --stringency 3 --length 36 -e 0.1”. These processed reads were then aligned to the mouse reference genome (Ensemble mm10) using STAR (version 2.7.3) with the default parameters. Mapped reads with high confidence were kept for further analysis using SAMtools (version 1.9). Expression levels for all Refseq genes were quantified to transcripts per kilobase of exon model per million mapped reads (TPM)using StringTie (version 2.1.4), and TPM values of replicates were averaged. To perform differential gene expression analysis, we first calculated the read counts of each RNA-seq sample using FeatureCounts (version 2.0.0). Then, we used Deseq2 in R to perform differential analysis. Genes with p ≤ 0.05 and FC > 1.2 or FC < -1.2 were defined as differentially expressed. DEG functional enrichment was analysed using online tools in Metascape (
https://metascape.org/). The most significant GO terms and KEGG pathways were selected to visualize in R. To evaluate potential PPIs, DEGs were mapped in R using the function of STRINGdb (version 2.6.5) for the Retrieval of Interacting Genes under the default settings; node connections less than 5 were excluded. The PPI network of 23 downregulated ECM genes was visualized in R.
Real-time reverse transcription polymerase chain reaction
Reverse transcription was conducted using HiScript III RT SuperMix for qPCR (+ gDNA wiper) (Vazyme, R323-01) following the manufacturer’s instructions. Real-time PCR was performed using 2 × SYBR Green qPCR Mix (with ROX) (SparkJade, AH0104). Data are shown as the fold change = 2
-ΔΔCt mean ± s.d. The primers are listed in Table
S1.
Immunohistochemistry
The expression of Collagen IV, Laminin gamma3 and Fibrillin 2 within ovarian tissues were detected by immunohistochemistry staining on the paraffin slide respectively. The ovarian sections fixed on the paraffin performed dewaxing and rehydration, antigen retrieval, and incubated in wet box with 3% H2O2 for 10 min sequentially. After blocking with 3% BSA, the slides were then immunostained with primary antibodies of Collagen IV (1:400, Abcam, ab236640), Laminin gamma3 (1:200, Abcam, ab234429) and Fibrillin 2 (1:200, Proteintech, 20252-1-AP) in a humidified box at 4 ℃ overnight. The slides were washed three times, and incubated with secondary antibody anti-rabbit (1:500, Sangon Biotech, D110065) at 37 °C for 30 min. Then the slides were stained with diaminobenzidine (DAB) as chromogen and counterstained with hematoxylin according to manufacturer’s instruction. After dehydration, the coverslips were mounted, and immunohistochemical images were taken using Leica Camera (Leica Microsystem, Buffalo Grove, IL, United States). Protein expressions were quantified with Image J software.
Statistical analysis
Statistical analyses were performed in R (
www.r-project.org/) and GraphPad Prism software. Statistical significance was calculated with One-way ANOVA with Tukey test (
P < 0.05 was considered statistically significant). All experiments were repeated at least three times unless otherwise stated. The results are expressed as the mean ± s.d. Differences are shown with *(
P < 0.05) and **(
P < 0.01).
Discussion
In this study, we successfully induced mouse ovarian damage via intraperitoneal injection of CTX and BUS, and partially recovered this injury through hUMSCs transplantation. The results revealed that chemotherapy exerted negative effects on the weights of ovaries and the number of follicles at different stages, and hUMSCs injection partially rescued these negative effects and restored follicular development and gene expression partially.
Studies have shown that chemotherapy drugs may evoke ovarian damage, including follicular atresia, GC apoptosis, abnormal hormone secretion, oocyte apoptosis and fibrosis [
31,
32]. Hormone replacement therapy (HRT) is one of the most common treatment modalities, but it cannot restore ovarian function from the root. Since MSCs have advantages such as easy isolation and culture and low immunogenicity, MSCs have become a hot topic in clinical and scientific research. Many studies have showed that stem cell therapy can promote follicle growth, regulate hormone secretion, suppress oocyte and GC apoptosis, inhibit cell cycle arrest and improve ovarian function [
33‐
35].
In the study, we found it is hard to obtain mature follicles exposed to chemotherapeutic drugs. However, there exist follicles restored their development and matured finally after MSCs translation. In normal physiological conditions, most primordial follicles are in a dormant state to avoid early depletion of follicle pool. During each menstrual cycle, a small population of primordial follicles gets activated and begins to grow and develop. However, chemotherapy treatment can disturb hormone secretion and result in follicle arrest. Both LH and FSH are required for follicular maturation, it has been shown that MSCs translation could regulate hormone levels, increase the secretion of LH and FSH, and restore follicle development [
36]. Our study also demonstrated that the mature follicles number was increased in POI + hUMSCs group compared with POI + PBS group (Fig.
3E).
To analyse which aspects were affected by CTX and BUS, we performed transcriptome sequencing of mouse ovaries, and further enriched gene functions and signaling pathways, therefore exploring the factors that affect follicle development and maturation. We obtained 2768 differentially expressed genes between the control and POI groups. The upregulated signaling pathways were mainly involved in the inflammatory response, oxidative damage response and other regulatory pathways, and the downregulated genes were mainly involved in signaling pathways such as ECM-receptor interaction, ECM organization and negative regulation of cell population proliferation. Changes in ECM elasticity can lead to variations in mechanical signaling. As a result, it may affect cellular communication and regulate cell physiological processes or pathological fibrosis by altering transcription factors and gene expression [
37,
38]. The ECM mediates oocyte dormancy. ECM components, such as collagen IV and fibronectin, and granular cells act together to maintain the dormant state of oocytes, thus playing a role in oocyte preservation [
39]. In ageing ovaries, the ECM contents were significantly reduced compared with those in younger ovaries, and the synthesis of elastin proteins in adult ovaries is also slower than that in pubertal ovaries [
10,
40]. Hyaluronic acid (HA) is also a glycosaminoglycan in the ECM, and its alteration may change the structure and properties of the ECM, thereby leading to oocyte ageing and meiotic arrest [
41].
Interestingly, we identified a series of ECM-related genes that were downregulated in the chemotherapeutic drug treatment groups, including Ace2, Dpp4, Col26a1, Col8a2, Fbn1, Lamc3, Col4a6, Fbn2, Col9a2, Adamts2 and other genes, and most of the above genes were upregulated after hUMSCs injection. The ECM not only acts as a supporting material, but also contains diverse receptors that mediate a variety of physiologic processes. There have been studies reporting that the ECM plays significant roles in follicular development. Therefore, we performed GSEA analysis and PPI network construction of ECM-related genes, and selected Col4a6, Fbn1, Fbn2 and Lamc3 for qRT-PCR. The results showed expressions of Col4a6, Fbn2 and Lamc3 were rescued in the POI + hUMSCs group compared with POI + PBS group (Fig.
5A-D). However, the qRT-PCR results just indicate global gene expression in the ovary tissues, and we are more interested in their locations and expressions around follicles. So, we performed IHC of Collagen IV, Laminin gamma 3 and Fibrillin 2, and the results showed Collagen IV and Laminin gamma 3 were localized around the follicles, and the immunostaining of the two proteins were decreased in the chemotherapeutic group (Fig.
6A-B). TCs distribute around the follicles [
18], previous studies have shown that collagen IV and Laminin are localized at the TC region [
20]. Therefore, it is reasonable to speculate that the abnormal expressions and localizations of Collagen IV and Laminin gamma 3 led to the suppression of follicular development. However, excessive ECM deposition may also result in fibrosis, and subsequently repress ovarian function. Studies have demonstrated that collage I and collage III can participate in tissue fibrosis [
40,
42,
43]. Mesenchymal stem cell transplantation could reduce fibrosis levels and reverse ovarian function by inhibiting the expression collage I and collage III by decreasing the synthesis of TGFβ1 and SMAD3 [
23].
It has been shown that the ECM plays essential roles in the regulation of follicular development through the Hippo and PI3K-AKT signaling pathways [
13]. The Hippo signaling pathway is important for moderate follicle activation and development. When the growing follicle migrates from the cortex to the medullary regions, the Hippo signaling pathway is activated, which in turn slows the growth of follicles, avoiding early depletion of the primordial follicle pool [
44]. However, overactivation of the Hippo signaling pathway may lead to excessive TCs proliferation and a disproportionate LH/FSH ratio ultimately cause polycystic ovary syndrome (PCOS) [
45]. The PI3K-AKT signaling pathway functions together with the Hippo signaling pathway in follicular development, and the PI3K-AKT signaling pathway regulates primordial follicle activation and follicular development in different species [
46,
47].
Previous study indicated that hUMSCs improved ovarian function by paracrine mechanism [
26], the cell secreted several bioactive molecules, such as VEGF, IGF and HGF, as well as micro-RNAs and extracellular vehicles (EVs) to modified the ovarian micro-circumstances [
48,
49]. Studies have shown that MSCs migrated into multiple organs after transplantation, which depended on damage degrees, blood vascular volume and chemokines regulation [
50]. It has also been shown that MSCs transplantation promoted follicle growth through migrating to the ovarian hilum and medulla. The MSCs exerted their therapeutic function through the features of migration and homing [
51,
52]. Regrettably, we did not perform the tracing experiment, we performed pre-experiment to confirm the influences of CTX and BUS on ovarian weights and developing follicle numbers directly. In the following experiments, we will label the MSCs with a fluorescent dye and re-construct animal models, make clear that the MSCs migrate into ovarian and then exert their effects.
Furthermore, in the present study, we focused on the whole ovary to explore DEGs at the transcriptional level in the different groups. Despite we performed IHC to verify the localizations and expressions of Collage IV and Laminin gamma 3, we did not isolate TCs to knock out col4a6 or lamc3 to explore the molecular mechanisms. We will continue to focus on this field and investigate the deeper molecular mechanisms, and we hope our findings provide a better understanding of the relationship between the ECM and follicular development.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.