Abstract

   This study aimed to investigate the influence of sperm miRNAs on
   fertilization rates (FR) in in vitro fertilization (IVF) and to explore
   potential regulatory mechanisms in sperm-mediated fertilization and
   embryo development. Through high-throughput sequencing, we identified
   differentially expressed miRNAs in sperm, with miR-133a-3p
   significantly upregulated in samples associated with low FR and
   available embryo rate (AER). Key regulatory circRNAs and mRNAs were
   further identified via the Starbase database, intersected with
   differentially expressed RNA, and analyzed through GO, KEGG, and PPI
   analyses. The circMYH9/miR-133a-3p/CXCR4 axis emerged as a critical
   regulatory network. In vitro assays using the GC-2 spd mouse
   spermatogenic cell line revealed that miR-133a-3p inhibited cell growth
   and proliferation while promoting apoptosis. circMYH9, acting as a
   competing endogenous RNA (ceRNA) for miR-133a-3p, modulated CXCR4
   expression, enhancing GC-2 spd cell growth and inhibiting apoptosis
   through the miR-133a-3p/CXCR4 axis. In vivo experiments using a mouse
   model confirmed that circMYH9 overexpression increased IVF success
   rates and promoted embryo development via this axis. Mechanistically,
   miR-133a-3p suppresses sperm fertilization and embryo development by
   targeting the circMYH9/miR-133a-3p/CXCR4 axis. These findings suggest
   that this regulatory network could serve as a novel biomarker for
   assessing fertilization potential and embryo quality in clinical
   settings and as a potential therapeutic target to improve IVF outcomes
   and address infertility. This study provides valuable insights into the
   molecular mechanisms governing sperm function and early embryonic
   development.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s43556-024-00236-5.

   Keywords: Sperm, CircMYH9, miR-133a-3p, CXCR4, In vitro fertilization
   rate, Embryo development

Introduction

   In vitro fertilization (IVF) is a widely used technology in
   reproductive medicine that has helped many couples struggling with
   infertility achieve their dream of parenthood [[40]1, [41]2]. The core
   of IVF involves combining sperm and egg in a laboratory setting to form
   a fertilized embryo, which is then implanted into the uterus. Despite
   the success of IVF, its fertilization rate (FR) and embryo development
   success rate remain influenced by various factors [[42]3]. In recent
   years, researchers have increasingly focused on small RNAs in sperm,
   especially microRNAs (miRNAs), which may play key roles in regulating
   sperm function and enhancing IVF success rates [[43]4].

   MiRNAs are a class of non-coding RNA molecules that regulate gene
   expression post-transcriptionally by targeting specific mRNAs [[44]5],
   affecting various biological processes such as cell proliferation,
   differentiation, and apoptosis [[45]6]. Studies have shown that miRNAs
   play essential roles in sperm maturation, motility, and fertilization
   [[46]7]. Additionally, changes in the expression of certain miRNAs have
   been closely associated with male infertility [[47]8], suggesting that
   these miRNAs are crucial to fertility. This has prompted scientists to
   explore their potential regulatory roles in IVF, with the aim of
   developing new strategies to improve fertilization and embryo
   development success rates [[48]9].

   Among these, miR-133a-3p is known for its regulatory role across
   various biological processes and systems, including muscle development,
   cell proliferation and apoptosis, cancer, cardiovascular function, and
   the reproductive system [[49]10, [50]11]. Previous studies have
   indicated that miR-133a-3p is highly expressed in sperm with low FR,
   suggesting it may have an inhibitory effect on IVF [[51]12]. However,
   the regulatory mechanisms of miR-133a-3p in fertilization and embryo
   development remain unclear and warrant further investigation. In
   addition, circMYH9, a circular RNA (circRNA) formed from the MYH9 gene
   via back-splicing, possesses high stability and specificity due to its
   circular structure and plays significant roles in cell proliferation,
   migration, and apoptosis [[52]13, [53]14]. Concurrently, CXCR4, a
   chemokine receptor, is involved in regulating cell apoptosis and
   migration and is considered a potential biomarker for predicting embryo
   implantation success in assisted reproductive technology [[54]15].
   Based on this background, we hypothesized that circMYH9 and CXCR4 might
   be key regulatory factors associated with miR-133a-3p.

   In summary, this study utilized high-throughput sequencing and
   bioinformatics analysis to construct the circMYH9/miR-133a-3p/CXCR4
   regulatory network, validating its roles in sperm growth,
   proliferation, and apoptosis, and further confirming its regulatory
   influence on IVF FR and embryo development. These findings not only
   deepen our understanding of the role of sperm miRNAs in the IVF process
   but also provide new potential targets for improving IVF success rates
   in future clinical applications. This research offers novel therapeutic
   approaches for treating infertility and establishes an important
   theoretical foundation for future studies.

Results

Key role of miR-133a-3p in sperm-mediated IVF and embryo development

   In order to investigate the impact of miRNA and its regulatory network
   on the IVF process, we conducted a differential analysis of miRNA based
   on sequenced clinical samples, identifying differentially expressed
   miRNAs through high-throughput sequencing analysis and determining key
   regulatory miRNAs. Differential analysis of miRNA was performed based
   on FR grouping, resulting in 26 significantly upregulated genes and 31
   significantly downregulated genes (Fig. [55]1a-b). To further pinpoint
   key miRNA in sperm IVF, we conducted miRNA differential analysis based
   on the clinical trait of AER, identifying 3 significantly upregulated
   genes and 31 significantly downregulated genes (Fig. [56]1c-d).
   Intersection of the above differential miRNAs revealed key miRNAs,
   including miR-133a-3p (Fig. [57]1e). Subsequently, we compared the
   miR-133a-3p expression statuses in the low FR group and the low AER
   group from the sequencing data, observing an upregulation in both
   groups (Fig. [58]1f). Therefore, we hypothesize that miR-133a-3p may
   impact IVF and embryo development by influencing sperm growth,
   apoptosis, and other physiological processes.

Fig. 1.

   [59]Fig. 1
   [60]Open in a new tab

   Key miRNAs identified through bioinformatics analysis in the IVF
   process. Note: a-b Volcano plots and heatmaps showing differentially
   expressed miRNAs grouped according to FR. Red represents upregulated
   genes; blue represents downregulated genes. The control group consists
   of 3 samples, and the treatment group consists of 3. c-d Volcano plots
   and heatmaps showing differentially expressed miRNAs grouped according
   to AER. Red represents upregulated genes; blue represents downregulated
   genes. The control group consists of 3 samples, and the treatment group
   consists of 3. e Venn diagram representing the intersection of
   differentially expressed miRNAs: miR-133a-3p. f Upregulation of
   miR-133a-3p expression in the FR and AER experimental groups compared
   to the control group (*P < 0.05, **P < 0.01)

miR-133a-3p modulates growth, proliferation, and apoptosis in GC-2spd germ
cells

   The research demonstrates that miR-133a-3p can impact cell survival
   [[61]16]. We investigated the influence of miR-133a-3p on GC-2spd
   spermatogonia cells and validated its effects through overexpression
   and knockdown experiments. RT-qPCR confirmed significant upregulation
   of miR-133a-3p in the mimic group and downregulation in the inhibitor
   group compared to their respective controls (Fig. [62]2a). As a result,
   we successfully established GC-2spd cell models with miR-133a-3p
   overexpression or silencing for further experiments.

Fig. 2.

   [63]Fig. 2
   [64]Open in a new tab

   Biological effects of miR-133a-3p on GC-2spd cells. Note: a RT-qPCR
   analysis of miR-133a-3p expression levels in each group. b MTT assay to
   measure cell viability in each group. c EdU experiment to assess cell
   proliferation capacity (scale bar: 50 μm). d Statistical analysis of
   EdU experiment results. e TUNEL assay to determine apoptosis rate in
   each group (Scale bar = 50 μm). f Statistical analysis of TUNEL
   experiment results. g Western blot analysis of Caspase-3, cleaved
   Caspase-3, Bax, and Bcl-2 protein expression levels in each group.
   Comparisons were made to the mimic-NC group (*P < 0.05, **P < 0.01,
   ***P < 0.001, ##P < 0.01). Cell experiments were performed in
   triplicate

   In a study of miR-133a-3p effects on GC-2spd cells, MTT assay showed
   reduced cell viability, EdU experiment indicated decreased cell
   proliferation, and TUNEL staining demonstrated elevated apoptosis in
   the miR-133a-3p mimic group compared to mimic-NC (Fig. [65]2b-g);
   whereas compared to the inhibitor-NC group, the miR-133a-3p inhibitor
   group showed notably strengthened cell viability and proliferation,
   with reduced levels of cell apoptosis (Fig. [66]2b-g). It is indicated
   by the results that miR-133a-3p can impede the proliferation and growth
   of GC-2spd cells and stimulate cell apoptosis.

Construction and analysis of the ceRNA network: implicating circMYH9,
miR-133a-3p, and CXCR4 in IVF and embryo development

   To construct a comprehensive ceRNA network, differential analysis was
   performed on circRNAs and mRNAs. The results revealed 57 significantly
   upregulated genes and 82 significantly downregulated genes in circRNA
   analysis (Figure S[67]1a-c), 167 significantly upregulated genes and
   502 significantly downregulated genes in mRNA analysis (Figure
   S[68]1b-d). Utilizing the identified disease-associated miRNA,
   miR-133a-3p, predictions of circRNAs interacting with it were made
   using the StarBase database (Fig. [69]3a). The intersection of these
   predictions with differentially expressed circRNAs led to the
   identification of key circRNAs: circMYH9 and EZH2, where circMYH9 was
   downregulated and EZH2 was upregulated (Fig. [70]3b-c). This suggests
   circMYH9 as a potential ceRNA for miR-133a-3p.

Fig. 3.

   [71]Fig. 3
   [72]Open in a new tab

   Identification of Key mRNA and circRNA in the IVF Process through
   Bioinformatics Screening. Note: a miRNA-circRNA regulatory network,
   where yellow represents miRNA and green represents circRNA. b Venn
   diagram representing the intersection of predicted miRNA target
   circRNAs from the Starbase database and differentially expressed
   circRNAs. c Relative expression levels of circRNA circMYH9 and EZH2 in
   the experimental and control groups (n = 2). *P < 0.05, **P < 0.01,
   ***P < 0.001 compared to the control group. d Enrichment analysis of
   differentially expressed mRNAs using GO. e Enrichment analysis of
   differentially expressed mRNAs using KEGG. f miRNA-mRNA regulatory
   network, where yellow represents miRNA and blue represents mRNA. g Venn
   diagram representing the intersection of predicted miRNA target mRNAs
   from the Starbase database and differentially expressed mRNAs analyzed
   by PPI. h Relative mRNA CD44, FOS, CXCR4, and MSN expression levels in
   the experimental and control groups (n = 3). i
   circMYH9/miR-133a-3p/CXCR4 regulatory network, where yellow represents
   miRNA, blue represents mRNA, and green represents circRNA

   More in-depth analysis was executed by carrying out GO and KEGG
   enrichment analysis on differentially expressed mRNAs. The GO analysis
   indicated that the BP these mRNAs may be involved in include signal
   transduction, immune response, and cytokine-mediated signaling
   pathways. Their potential roles encompass cellular components such as
   the cell membrane, membrane, and membrane components, and their
   molecular functions could include protein binding, metal ion binding,
   and hydrolase activity (Fig. [73]3d). KEGG analysis revealed potential
   regulatory pathways that the target genes may participate in, such as
   Fluid shear stress and atherosclerosis, Hematopoietic cell lineage, and
   Mineral absorption (Fig. [74]3e). Additionally, a PPI network analysis
   using the STRING database on differentially expressed mRNAs unveiled
   core regulatory genes (Figure S2).

   Furthermore, predictions of miRNAs binding to mRNAs were made using
   StarBase (Fig. [75]3f), and an intersection was taken with the core
   regulatory genes identified in the PPI analysis, resulting in key
   downregulated mRNAs: CD44, FOS, CXCR4, and MSN (Fig. [76]3g-h). Based
   on the differential genes and enriched gene analysis from the
   databases, it is hypothesized that miR-133a-3p may regulate sperm IVF
   and embryo development processes through the circMYH9/miR-133a-3p/CXCR4
   network (Fig. [77]3i).

Validating the targeted association: miR-133a-3p Binds with circMYH9 and
CXCR4, establishing circMYH9 as a ceRNA influencing CXCR4 function

   Based on the predictions from Starbase databases, we identified the
   binding sites of miR-133a-3p with circMYH9/CXCR4 and designed mutant
   sequences for circMYH9 and CXCR4 (Fig. [78]4a-b). Initially, we
   executed experiments using the dual-luciferase reporter gene system.
   The results revealed a marked decrease in luciferase activity in the
   miR-133a-3p mimic group, contrasting with the mimic-NC group in
   circMYH9-wt (Fig. [79]4c), indicating the putative interaction of
   miR-133a-3p with circMYH9. Similarly, in CXCR4-wt, the luciferase
   activity in the miR-133a-3p mimic group significantly decreased when
   juxtaposed with the mimic-NC group (Fig. [80]4d), suggesting the
   interaction between miR-133a-3p and CXCR4. Furthermore, it was
   evidenced by the results of the RNA pull-down study that the enrichment
   levels of circMYH9 and CXCR4 exhibited a notable escalation in the
   Bio-miR-133a-3p-wt group compared to the Bio-probe-NC and
   Bio-miR-133a-3p-mut groups (Fig. [81]4e-f), validating the connection
   between miR-133a-3p and circMYH9/CXCR4. To sum up, based on these
   discoveries, we have identified that miR-133a-3p indeed targets
   circMYH9/CXCR4, where circMYH9 serves as a competitive endogenous RNA
   for miR-133a-3p, thereby influencing the functionality of CXCR4.

Fig. 4.

   [82]Fig. 4
   [83]Open in a new tab

   Validation of the interaction among circMYH9/miR-133a-3p/CXCR4. Note: a
   Prediction of the binding site between miR-133a-3p and circMYH9(wt),
   where circMYH9(mut) denotes the mutated sequence. b Prediction of the
   binding site between miR-133a-3p and CXCR4(wt), where CXCR4(mut)
   denotes the mutated sequence. c Dual-luciferase reporter assay to
   detect the targeting relationship between miR-133a-3p and circMYH9
   (experiment repeated three times, *P < 0.05 compared to the mimic-NC
   group). d Dual-luciferase reporter assay to detect the targeting
   relationship between miR-133a-3p and CXCR4 (experiment repeated three
   times, *P < 0.05 compared to the mimic-NC group). e RNA pull-down assay
   to investigate the interaction between miR-133a-3p and circMYH9
   (experiment repeated three times, *P < 0.05 compared to Bioprobe NC or
   Bio-miR-133a-3p-mut). f RNA pull-down assay to investigate the
   interaction between miR-133a-3p and CXCR4 (experiment repeated three
   times, *P < 0.05 compared to Bioprobe NC or Bio-miR-133a-3p-mut)

miR-133a-3p modulates growth, proliferation, and apoptosis in GC-2spd Cells
via regulation of CXCR4 expression

   We overexpressed miR-133a-3p alone or with CXCR4 in GC-2spd cells to
   study their effects. Results showed elevated miR-133a-3p and decreased
   CXCR4 in mimic-miR-133a-3p + oe-NC group. In
   mimic-miR-133a-3p + oe-CXCR4 group, miR-133a-3p levels were unchanged,
   but CXCR4 expression recovered (Fig. [84]5a). Subsequent MTT assays,
   EdU experiments, TUNEL staining, and Western blot analysis revealed
   that the mimic-miR-133a-3p + oe-NC group displayed decreased cellular
   viability, weakened cell proliferation, and increased apoptosis when
   contrasted with the mimic-NC + oe-NC group. However, these effects were
   reversed upon the overexpression of CXCR4 on this basis
   (Fig. [85]5b-g).

Fig. 5.

   [86]Fig. 5
   [87]Open in a new tab

   The Impact of miR-133a-3p regulation on CXCR4 in GC-2spd cell
   biological functions. Note: a Expression levels of miR-133a-3p and
   CXCR4 were measured using RT-qPCR; b Cell viability in each group was
   assessed using the MTT assay; c Proliferation ability of cells in each
   group was evaluated through the EdU assay (scale bar: 50 μm); d
   Statistical analysis of the EdU assay results; e Apoptosis rates of
   cells in each group were determined using the TUNEL assay (scale bar:
   50 μm); f Statistical analysis of the TUNEL assay results; g Protein
   expression levels of Caspase-3, cleaved Caspase-3, Bax, and Bcl-2 in
   each group were detected by Western Blot; * denotes a difference
   compared to the mimic-NC + oe-NC group (P < 0.05), # denotes a
   difference compared to the mimic-miR-133a-3p + oe-NC group (P < 0.05),
   all cell experiments were performed in triplicate

   These outcomes illustrate that miR-133a-3p suppresses the growth and
   proliferation of GC-2spd cells and stimulates cell apoptosis by
   regulating CXCR4 expression.

MYH9 enhances GC-2spd cell proliferation and inhibits apoptosis by modulating
the miR-133a-3p/CXCR4 regulatory axis

   To investigate the effects of circMYH9 on GC-2spd cells through the
   miR-133a-3p/CXCR4 axis, circMYH9 and miR-133a-3p were overexpressed
   using lentivirus transfection. RT-qPCR revealed that in the
   oe-MYH9 + mimic-NC group, circMYH9 and CXCR4 expression increased while
   miR-133a-3p expression decreased. Overexpression of miR-133a-3p
   counteracted the circMYH9-induced suppression of miR-133a-3p, resulting
   in reduced CXCR4 expression (Fig. [88]6a). The outcomes indicated that
   the oe-MYH9 + mimic-NC group demonstrated higher cell viability and
   proliferation capacity than the oe-NC + mimic-NC group. Additionally,
   this group exhibited a decrease in the level of cellular apoptosis. The
   cell viability and proliferation ability of the
   oe-MYH9 + mimic-miR-133a-3p group exhibited a reduction in contrast to
   the oe-MYH9 + mimic-NC group. Additionally, elevated levels of cellular
   apoptosis were detected in the oe-MYH9 + mimic-miR-133a-3p group
   (Fig. [89]6b-g).

Fig. 6.

   [90]Fig. 6
   [91]Open in a new tab

   The Impact of circMYH9/miR-133a-3p/CXCR4 Axis on GC-2spd cell
   biological functions. Note: a Expression levels of circMYH9,
   miR-133a-3p, and CXCR4 were measured using RT-qPCR; b Cell viability in
   each group was assessed using the MTT assay; c Proliferation ability of
   cells in each group was evaluated through the EdU assay (scale bar:
   50 μm); d Statistical analysis of the EdU assay results; e Apoptosis
   rates of cells in each group were determined using the TUNEL assay
   (scale bar: 50 μm); f Statistical analysis of the TUNEL assay results;
   g Protein expression levels of Caspase-3, cleaved Caspase-3, Bax, and
   Bcl-2 in each group were detected by Western Blot; * denotes a
   difference compared to the oe-NC + mimic-NC group (P < 0.05), # denotes
   a difference compared to the oe-MYH9 + mimic-NC group (P < 0.05), all
   cell experiments were performed in triplicate

   We overexpressed circMYH9 or co-overexpressed circMYH9 and silenced
   CXCR4 in GC-2spd cells. Results showed circMYH9 upregulation,
   miR-133a-3p downregulation, and CXCR4 elevation in oe-circMYH9 + sh-NC
   group. Silencing CXCR4 reversed circMYH9's effect on CXCR4 without
   affecting miR-133a-3p expression (Fig. [92]7a). The MTT, EdU, TUNEL,
   and Western blot results showed that compared to the oe-NC + sh-NC
   group, the oe-circMYH9 + sh-NC group exhibited significantly enhanced
   cell viability and proliferation, with reduced apoptosis levels. In
   contrast, the oe-circMYH9 + sh-CXCR4 group showed significantly reduced
   cell viability and proliferation, with increased apoptosis levels
   (Fig. [93]7b-g).

Fig. 7.

   [94]Fig. 7
   [95]Open in a new tab

   Effects of the circMYH9/miR-133a-3p/CXCR4 axis on the biological
   functions of GC-2spd cells. Note: a RT-qPCR analysis of circMYH9,
   miR-133a-3p, and CXCR4 expression levels; b MTT assay to assess cell
   viability in each group; c EdU assay to evaluate cell proliferation in
   each group (scale bar: 50 μm); d Statistical analysis of the EdU assay
   results; e TUNEL assay to detect apoptosis rates in each group (scale
   bar: 50 μm); f Statistical analysis of TUNEL assay results; g Western
   blot analysis of Caspase-3, cleaved Caspase-3, Bax, and Bcl-2 protein
   expression levels in each group. * indicates P < 0.05 compared to the
   oe-NC + sh-NC group, # indicates P < 0.05 compared to the
   oe-circMYH9 + sh-NC group. All cell experiments were repeated three
   times

   The outcomes described above imply that circMYH9 is capable of boosting
   the growth and proliferation of GC-2spd cells while also preventing
   cell apoptosis through the regulation of the miR-133a-3p/CXCR4
   signaling pathway.

MYH9 overexpression enhances IVF and early embryo development via the
miR-133a-3p/CXCR4 axis in a mouse model

   To evaluate the impact of circMYH9/miR-133a-3p/CXCR4 on embryo
   development, lentiviral transfections overexpressing circMYH9 and
   co-overexpressing circMYH9 and miR-133a-3p were constructed in a mouse
   model. RT-qPCR results showed increased circMYH9 and CXCR4 expression
   and decreased miR-133a-3p in the oe-MYH9 + mimic-NC group, compared to
   the oe-NC + mimic-NC group. However, miR-133a-3p expression was
   restored, and CXCR4 expression was decreased in the
   oe-MYH9 + mimic-miR-133a-3p group (Fig. [96]8a). The results affirm the
   efficient creation of a mouse model that could be utilised for future
   research.

Fig. 8.

   [97]Fig. 8
   [98]Open in a new tab

   The Impact of circMYH9/miR-133a-3p/CXCR4 Axis on Mouse FR and Early
   embryo development. Note: a RT-qPCR analysis of circMYH9, miR-133a-3p,
   and CXCR4 expression levels; b Analysis of sperm motility parameters in
   each group using the CASA system, including progressive motility rate
   (PR) and total motility rate (PR + NP), where NP represents the
   non-progressive motility rate; c Analysis of sperm motility parameters
   in each group using the CASA system, including curvilinear velocity
   (VCL), straight-line velocity (VSL), average path velocity (VAP), and
   beat cross frequency (BCF); d Statistical data on the proportion of
   2-cell stage embryos after sperm-oocyte binding in each group; e
   Observation of sperm attachment (white arrows) 4 h after fertilization
   using DAPI staining, with red arrows indicating cumulus cells (scale
   bar: 50 μm); f Statistical chart of the number of sperm attached to
   each fertilized egg; g Observation of the percentage of embryos with 1,
   2, or 3 pronuclei (PN) during early fertilization using TLI. * denotes
   a difference compared to the oe-NC + mimic-NC group (P < 0.05), #
   denotes a difference compared to the oe-MYH9 + mimic-NC group
   (P < 0.05), each group consisted of 6 mice

   Sperm motility analysis revealed significant increases in PR, PR + NP,
   VCL, VSL, VAP, and BCF in the oe-circMYH9 + mimic-NC group compared to
   the control, but these parameters were reduced in the
   oe-circMYH9 + mimic-miR-133a-3p group (Fig. [99]8b-c).

   Co-culture experiments showed increased 2-cell embryos in the
   oe-MYH9 + mimic-NC group and a decrease in the
   oe-MYH9 + mimic-miR-133a-3p group (Fig. [100]8d).

   Sperm binding to the zona pellucida and the development ratio of
   fertilized embryos were higher in the oe-MYH9 + mimic-NC group compared
   to the control, while both were reduced in the
   oe-MYH9 + mimic-miR-133a-3p group (Fig. [101]8e-g). The results suggest
   that circMYH9 overexpression promotes sperm IVF rates and early embryo
   development, potentially through the miR-133a-3p/CXCR4 axis.

Discussion

   Research indicates that miRNAs are vital in reproduction, such as in
   regulating sperm formation, maturation, and function [[102]11]. Studies
   have also found that miR-133a-3p is significantly upregulated in low FR
   and low AER groups [[103]11]. However, the operational mechanism of
   miR-133a-3p in IVF remains unclear [[104]17]. The utilization of
   high-throughput sequencing techniques was pivotal in the detection of
   varied expression levels of miRNAs linked to IVF FR and the
   developmental stages of embryos, particularly miR-133a-3p, further
   validating the key role of miRNAs in reproduction. This aligns with
   existing research and deepens our understanding of specific miRNAs in
   the IVF process. Our study fills this gap by clarifying the regulatory
   roles in IVF and embryo development.

   Existing studies often utilize GO and KEGG functional and pathway
   enrichment analysis to reveal the biological functions and involved
   signaling pathways of differentially expressed mRNAs [[105]18–[106]20].
   Similar analyses in our study identified key mRNAs related to IVF FR
   and embryo development. Together with miRNA and circRNA regulatory
   networks, this further unveils the specific regulatory mechanisms of
   sperm miRNAs in reproduction. This consistency with existing research
   methods and results enhances the scientific rigor and credibility of
   our study.

   In recent years, the role of circRNAs as ceRNAs in modulating gene
   expression has garnered attention. Studies indicate that the regulation
   of miRNA target genes can be modulated by circRNAs through the
   sequestration of miRNAs [[107]13]. While studies on sperm IVF FR exist,
   few have addressed early embryo development [[108]21]. By predicting
   miRNA-binding circRNAs and combining them with differentially expressed
   circRNAs, we constructed the circMYH9/miR-133a-3p/CXCR4 regulatory
   network, validating the regulatory role of circRNAs in IVF. This
   finding corresponds with existing research, further confirming the
   importance of circRNAs in reproduction.

   Our in vitro cellular experiments validated the specific roles of
   miR-133a-3p, circMYH9, and CXCR4 in sperm growth, proliferation, and
   apoptosis, with in vivo animal experiments further confirming their
   regulatory roles in IVF FR and embryo development. This experimental
   design and validation method align highly with existing research,
   ensuring the reliability and repeatability of our results. Existing
   research has shown that sperm quality and function directly impact IVF
   success rates [[109]22]. Our study reveals how sperm miRNAs and their
   regulatory networks affect IVF FR and embryo development, offering new
   potential targets to improve IVF success rates. This finding aligns
   with current research objectives, further enriching the studies aimed
   at enhancing IVF success rates. This provides a new perspective in the
   field of reproduction, allowing researchers to delve deeper into the
   role of miRNAs in sperm fertilization and embryo development.

   The circMYH9/miR-133a-3p/CXCR4 axis has been identified as a crucial
   regulatory network. Research on this network not only provides a deeper
   understanding of gene regulation mechanisms but also offers new
   strategies and targets for treating infertility and embryo development
   disorders. This study demonstrates that miR-133a-3p, via the
   circMYH9/miR-133a-3p/CXCR4 axis, inhibits sperm fertilization and
   embryo development in vitro. This provides clinicians with new
   biomarkers that may hold significant value in assessing fertilization
   potential and embryo quality. Moreover, modulation of this regulatory
   axis could present new treatment opportunities for infertility
   patients.

   However, our study has several limitations. Firstly, this study
   primarily utilized the mouse spermatogonial cell line GC-2spd as the
   experimental model. Yet, differences exist between the reproductive
   systems of mice and humans, necessitating further validation of whether
   these findings can be directly applied to humans. Secondly, although
   this study identified the circMYH9/miR-133a-3p/CXCR4 regulatory
   network, it may only represent a fraction of the regulatory processes
   during reproduction. Hence, further research is required to unveil
   other potential regulatory molecules or pathways. While numerous in
   vitro and in vivo experiments have confirmed the function of
   miR-133a-3p, additional mechanistic studies and in-depth functional
   validations are still necessary. Future research could delve deeper
   into the circMYH9/miR-133a-3p/CXCR4 regulatory network to uncover more
   critical molecules and pathways. Expanding the scope of the research to
   include other animal models and humans could validate the universality
   and applicability of these findings. Exploring how to translate the
   discovery of this regulatory network into clinical applications to
   provide more effective diagnosis and treatment methods for infertility
   patients is crucial (Fig. [110]9).

Fig. 9.

   [111]Fig. 9
   [112]Open in a new tab

   Molecular mechanism diagram of circMYH9/miR-133a-3p/CXCR4 regulatory
   network IVF of sperm

Materials and methods

Sample collection

   Patients treated at Xuzhou Central Hospital and Shuyang Hospital from
   August 2022 to May 2023 were included in this study. Couples with
   female infertility caused by tubal factors or unexplained reasons were
   divided into experimental and control groups. Females aged 20–44 were
   included, while males with abnormal semen parameters, smoking, alcohol
   abuse, chronic medication, infections, or tumors were excluded.
   Clinical data information was recorded (Table S[113]1).

   Semen was collected after 3 days of abstinence, processed through
   density gradient centrifugation, and stored for transcriptome
   sequencing. Patients with FR below 50% were placed in Group A, and
   those with FR above 50% in Group B. The study was approved by the
   Clinical Ethics Committee of Xuzhou Central Hospital
   (XZXY-LK-20230822–0147), and patients were informed and agreed to
   participate in adherence to the Helsinki Declaration guidelines
   [[114]23].

RNA sequencing

   RNA was extracted from sperm samples using Trizol reagent (Thermo
   Fisher, 16096020). RNA quality was assessed via the Qubit®2.0
   Fluorometer® (Life Technologies, [115]Q33216), the Qubit® RNA assay kit
   (Shanghai Baoji Biotechnology, HKR2106-01), and the Bioanalyzer 2100
   system (Agilent, 5067–1511) [[116]24, [117]25]. For mRNA sequencing,
   the NEBNext® UltraTM RNA Library Prep Kit (NEB, E7435L) was used, and
   libraries were generated on the Illumina-HiSeq 2000 platform.
   Paired-end reads were processed using FastQC software, followed by
   adapter trimming with Cutadapt, and sequence filtering using FASTX
   Toolkit [[118]24–[119]26]. HISAT2 (v0.7.12) was used to align reads to
   the human genome. CircRNA sequencing involved enrichment and RNase A
   digestion, followed by cRNA generation and quantification via a circRNA
   microarray (Arraystar). For miRNA sequencing, the samples were
   processed with the SOLiD Small RNA Expression Kit (Applied Biosystems),
   followed by purification, adapter ligation, reverse transcription, and
   PCR amplification. The products were sequenced using the SOLiD 4 system
   and analyzed against miRNA databases [[120]27, [121]28].

Disease characteristic gene screening

   The "edgeR" package in R software was used to filter circRNA, miRNA,
   and mRNA data to identify differentially expressed genes between the
   experimental (A) and control (B) groups. The criteria were |logFC|> 1
   and P-value < 0.05. R's "heatmap" and "ggplot2" packages were used to
   visualize differential expression. Differential miRNAs based on
   available embryo rate (AER) were identified and intersected with those
   based on FR. The jvenn database was employed to generate Venn diagrams
   of shared disease-associated miRNAs. GO and KEGG enrichment analysis
   was performed on mRNAs using the "clusterProfiler" package, and
   protein–protein interaction (PPI) analysis was conducted using the
   STRING database to identify key regulatory genes [[122]29, [123]30].

Construction of competing endogenous RNA (ceRNA) network

   A comprehensive ceRNA network was constructed using Starbase to predict
   mRNA interactions with disease-specific miRNAs. These mRNAs were
   intersected with core genes from PPI analysis. Additionally,
   miRNA-circRNA binding predictions were made, and key circRNAs were
   identified. The ceRNA network was built using Cytoscape software to
   illustrate the interactions between mRNA, miRNA, and circRNA [[124]31,
   [125]32].

Formation and nurturing of cellular models in vitro

   The GC-2spd cell line (CRL-2196, ATCC) was cultured in DMEM medium
   containing high glucose (11965092, Gibco) and supplemented with 10%
   fetal bovine serum (FBS, 10099141C, Gibco) at 37°C with 5% CO[2] in a
   humidified incubator (Heracell™ Vios 160i CR CO[2] incubator, Thermo
   Scientific, catalogue number 51033770). Cell passaging was performed
   when the cells reached 80%-90% confluence. After three passages, the
   cells were assessed for optimal conditions. Once 80%-90% confluence was
   reached, the cells were washed with phosphate-buffered saline (PBS) and
   digested with trypsin. The digestion process was halted by adding a
   complete medium. After centrifugation, the supernatant was discarded,
   and 1 mL of freezing solution was added. The cell mixture was then
   transferred to cryogenic vials and stored at -80 °C or in liquid
   nitrogen for long-term preservation.

   The 293 T cell line (CRL-3216, ATCC) was cultured in DMEM with 10% FBS,
   10 μg/mL streptomycin, and 100 U/mL penicillin at 37 °C with 5% CO[2].
   Cell passaging was performed when cells reached 80%-90% confluence
   [[126]33–[127]35].

Construction of lentiviral vectors

   Lentiviruses overexpressing circMYH9 and CXCR4 were assembled by
   Genechem Company (Shanghai), using the LV-PDGFRA vector. The
   miR-133a-3p overexpression and silencing lentiviruses were created with
   a mimic and inhibitor from Guangdong RayBiotech Co., Ltd. The packaged
   virus and target vector were transfected into 293 T cells using
   Lipofectamine 2000 when cells reached 80–90% confluence. After 48 h of
   incubation, the viral supernatant was collected, and viral titers were
   measured. Lentivirus was added to 293 T cells at a density of 5 × 10^4
   cells/mL in 6-well plates, with stable strains selected using 2 μg/mL
   puromycin (Sigma-Aldrich) over 14 days [[128]36–[129]39].

Cell transfection

   Cell culture was maintained until 80–90% confluence was reached. Cells
   were washed 2–3 times with PBS and digested with trypsin, after which
   the digestion was stopped by adding a complete culture medium
   containing 10% FBS. The cell suspension was transferred to a 15 mL
   centrifuge tube and centrifuged at 1000 rpm for 5 min. After
   centrifugation, the supernatant was removed, and 3 mL of complete
   culture medium was added to the cell sediment. Cells were thoroughly
   resuspended by gentle pipetting, followed by cell counting. Based on
   the results, the cell density was adjusted, and the cells were seeded
   evenly into 6-well or 12-well plates. Cells were placed in an incubator
   for cultivation. In RT-qPCR experiments, transfection efficiency was
   validated using 12-well plates, while 6-well plates were used for
   Western blot experiments.

   The groups for transfection were as follows: (1) mimic-NC group (miRNA
   mimic negative control); (2) inhibitor-NC group (miRNA inhibitor
   negative control); (3) miR-133a-3p mimic group (transfected with
   miR-133a-3p mimic); (4) miR-133a-3p inhibitor group (transfected with
   miR-133a-3p inhibitor); (5) oe-NC + mimic-NC group (lentiviral empty
   vector + miRNA mimic negative control); (6) oe-MYH9 + mimic-NC group
   (oe-MYH9 plasmid + miRNA mimic negative control); (7)
   oe-MYH9 + mimic-miR-133a-3p group (oe-MYH9 plasmid + miR-133a-3p
   mimic); (8) mimic-miR-133a-3p + oe-NC group (miR-133a-3p mimic + empty
   vector); (9) mimic-miR-133a-3p + oe-CXCR4 group (miR-133a-3p
   mimic + oe-CXCR4 plasmid); (10) oe-NC + sh-NC group (overexpressed and
   silenced non-coding MYH9); (11) oe-circMYH9 + sh-NC group (oe-circMYH9
   plasmid + silenced non-coding MYH9); (12) oe-circMYH9 + sh-CXCR4 group
   (oe-circMYH9 plasmid + sh-CXCR4 plasmid). The effectiveness of each
   transfection was verified after 48 h via RT-qPCR [[130]27].

RT-qPCR

   RNA was extracted using Trizol reagent (Invitrogen), and its
   concentration and purity were assessed with NanoDrop LITE (Thermo
   Scientific). The PrimeScript RT reagent Kit (TaKaRa) was used to
   synthesize cDNA. Total RNA was also isolated using the miRNeasy micro
   kit (QIAGEN) for miR-133a-3p expression analysis, followed by cDNA
   synthesis with the miRNA First Strand cDNA Synthesis Kit (Shanghai
   Sangon Biotechnology). RT-qPCR detection used SYBR Green PCR Master Mix
   (Applied Biosystems) and the ABI PRISM 7500 system. Primers were
   synthesized by Shenggong Biotechnology based on Table S2. miR-133a-3p
   expression was normalized using U6, while other genes were normalized
   to GAPDH. The relative expression levels of DLEU2 and other genes were
   analyzed using the 2^−ΔΔCt method, calculated as ΔΔCt = (target gene Ct
   in the experimental group—housekeeping gene Ct in the experimental
   group)—(target gene Ct in the control group—housekeeping gene Ct in the
   control group). All RT-qPCR experiments were repeated thrice
   [[131]40–[132]43].

Dual-luciferase reporter gene experiment

   The binding sites for miR-133a-3p and circMYH9/CXCR4 were retrieved
   from bioinformatics databases. Wild-type (wt) and mutant (mut)
   sequences for the 3'-UTR regions of circMYH9 and CXCR4 were designed.
   These sequences were inserted into the pmiR-GLO-ReportTM vector
   (Biosage). The plasmids containing wt and mut sequences, along with
   miR-133a-3p mimic and mimic-NC plasmids, were transfected into 293 T
   cells using Lipofectamine 2000. After 48 h, the cells were lysed, and
   the supernatant was collected. The Dual-Luciferase Reporter Assay
   System (Promega) was used to analyze relative luciferase activity. The
   experiment was performed in triplicate for reliability [[133]44].

RNA pull-down

   Cells transfected with Bio-miR-133a-3p-wt and Bio-miR-133a-3p-mut RNA
   (50 nM) were collected after 48 h. After washing with PBS, the cells
   were lysed in a buffer containing protease and RNase inhibitors. The
   lysate was incubated overnight at 4℃ with streptavidin-coated magnetic
   beads pre-coated with RNase-free BSA and yeast tRNA. After incubation,
   the expression levels of circMYH9 and CXCR4 were detected using RT-qPCR
   analysis. The experiment was repeated three times for accuracy
   [[134]45].

MTT assay

   The test cells were inoculated onto a 96-well cell culture plate at a
   density of 3–5 × 10^4 cells/mL, followed by a 48-h incubation period.
   Add MTT solution (10 mg/mL, Cat. No. ST316, Beyotime Biotechnology Co.,
   Ltd, Shanghai) to the cell suspension, incubate for 4 h, then add
   dimethyl sulfoxide (DMSO) and mix for 10 min. The spectrophotometer
   from Laspec, China, was employed for measuring absorbance at 490 nm
   [[135]46].

EdU experiment

   Cells were seeded in a 24-well plate, and EdU (10 µmol/L, Guangzhou
   RiboBio Co.) was added to the medium for a 2-h incubation in a CO[2]
   incubator. After medium removal, cells were fixed with 4% formaldehyde
   for 30 min at room temperature. The cells were washed with PBS
   containing 3% BSA, followed by treatment with 0.5% Triton-100 for
   20 min to enhance permeability. After washing, 100 µL of staining
   solution was added to each well, incubating in the dark for 30 min at
   room temperature. DAPI was added for nuclear staining and incubated for
   5 min. Randomly selected stained cells were analyzed using a
   fluorescence microscope (IX73, Olympus), and the number of EdU-positive
   cells was assessed using Image-Pro Plus 6.0 software. The EdU labeling
   rate was calculated as the percentage of positive cells. The experiment
   was performed in triplicate [[136]47].

TUNEL staining

   GC-2spd cells were fixed with 4% paraformaldehyde (Shanghai YESBIOTECH)
   for 15 min and permeabilized with 0.25% Triton X-100 for 20 min at room
   temperature. After blocking with 5% BSA (Shanghai Yisheng Biological
   Technology), the cells were stained with TUNEL reagent (Beyotime
   Biotechnology) and DAPI. Apoptotic cells were imaged using a confocal
   microscope (LSM 880, Carl Zeiss). TUNEL-positive cells emitted green
   fluorescence (apoptotic), while DAPI-stained nuclei appeared blue. The
   apoptosis rate was calculated by dividing the number of TUNEL-positive
   cells by the total number of cells, based on five different fields
   [[137]48].

Western blot

   Cells were lysed with RIPA buffer (Beyotime) containing protease
   inhibitors for total protein extraction. Protein concentrations were
   determined using the BCA protein quantification kit (Beyotime). After
   denaturation, proteins were separated by 10% SDS-PAGE electrophoresis
   and transferred to a PVDF membrane. The membrane was blocked with 5%
   BSA for 2 h at room temperature to prevent nonspecific binding.
   Afterward, primary antibodies (details in Table S3) were incubated for
   1 h, followed by washing and incubation with goat anti-rabbit secondary
   antibody (HRP-conjugated, Abcam). The membrane was exposed using
   Pierce™ ECL Western Blot Substrate (Thermo Scientific) and imaged using
   a Bio-Rad imaging system. Grayscale values of the Western blot bands
   were analyzed with ImageJ software, with GAPDH used as the internal
   reference [[138]41]. Each experiment is repeated 3 times.

In vivo animal experiments

   Eight-week-old male C57BL/6 J mice (n = 18) were sourced from Beijing
   Vital River Laboratory Animal Technology Co., Ltd. (Catalog#: 219,
   Beijing, China) and placed in a specific-pathogen-free (SPF) animal
   facility under controlled conditions (humidity: 60%-65%, temperature:
   22–25 °C) after a one-week acclimatization period. Observations on the
   health status of the mice were made prior to the commencement of the
   experiments. These experiments were performed with the necessary
   approval from the Institutional Animal Care and Use Committee.
   Anesthesia was initiated for the mice by intraperitoneal injection of
   copper fluoride (0.4 ml/kg, Sigma, Catalog#: SML1875) and diazepam
   (5 mg/kg, Catalog#: 439–14-5), followed by shaving and disinfection of
   the scrotal region. Subsequently, the scrotal skin was exposed, and the
   epididymides were carefully dissected and isolated. A precise volume of
   lentiviral vector (200 μL per side) was injected into both epididymides
   using a micropipette, followed by suturing and disinfection. After a
   14-day period of postoperative care in a sterile animal facility, the
   mice were utilized for subsequent experiments [[139]49].

   The experimental groups were as follows: oe-NC + mimic-NC,
   oe-circMYH9 + mimic-NC, and oe-circMYH9 + mimic-miR-133a-3p, each
   consisting of 6 mice per group [[140]49]. The expression levels of the
   corresponding proteins in mouse sperm and sperm viability were
   determined by qRT-PCR and IVF assays, respectively.

Separating sperm

   The epididymis and vas deferens were extracted from male mice and
   transferred into a culture dish. The dish was coated with liquid
   paraffin and contained 500 μL of GIVF Plus medium (Vitrolife AB,
   Sweden). The tissue was surgically cut, and the sperm were extracted
   and placed into the culture medium. The sperm were then incubated at
   37 °C and used for IVF experiments following energy acquisition
   [[141]50].

Sperm motility analysis

   Male mice were anesthetized with 1% sodium pentobarbital and euthanized
   by decapitation. The bilateral epididymides were removed and placed in
   pre-warmed human tubal fluid (HTF) at 37 °C. The caudal epididymis was
   cut, and mature sperm were released into the HTF. A 10 μL sample
   containing sperm was analyzed using a CASA microscope system to measure
   sperm motility, including progressive motility (PR), non-progressive
   motility (NP), immotile sperm, total motility rate (PR + NP),
   curvilinear velocity (VCL), straight-line velocity (VSL), average path
   velocity (VAP), and beat cross frequency (BCF) [[142]51].

IVF

   To retrieve oocytes, 6-week-old female C57BL/6 J mice were injected
   with human chorionic gonadotropin (hCG) for superovulation and
   euthanized 15–18 h later. After exposing the uterus, fallopian tubes,
   and ovaries, the fallopian tubes were cleaned and placed in paraffin
   oil. The egg mass was gently expressed from the enlarged fallopian tube
   section into the oil. The dish containing the ovum clumps was incubated
   at 37 °C with 5% CO[2] for 20–30 min. Capacitated sperm (1–2 μL) was
   added to the droplet containing the oocytes, and Fertilization occurred
   over 3 h. Post-incubation, the oocytes were washed three times with
   GIVF Plus medium. The IVF FR was calculated as the ratio of fertilized
   embryos to total eggs, while the 2-cell development rate was determined
   as the ratio of embryos developing into 2-cell stages to the total
   number of eggs [[143]52].

Delayed imaging analysis of embryo development

   After IVF of mouse embryos obtained from IVF 4 h post-fertilization,
   the sperm attached to the zona pellucida was removed using a pipette.
   Fertilized eggs should exhibit at least one polar body as an indication
   of MII oocyte. Primovision (EmbryoScope monitoring system) was used to
   capture time-lapse images of the development of fertilized eggs prior
   to implantation. The fertilized eggs were cultured in G-TL™ medium
   (Vitrolife, [144]https://www.vitrolife.com) for 96 to 100 h, with
   photographs taken every 10 min. Data collection was performed using
   Primovision Analyzer software [[145]52].

Statistical analysis

   Expression data for miRNA, mRNA, and circRNA were obtained through
   high-throughput sequencing. Differentially expressed genes between the
   experimental and control groups were identified using the "edgeR"
   package in R software, with selection criteria of |logFC|> 1 and
   P-value < 0.05. The selected differentially expressed genes were
   further analyzed using GO and KEGG enrichment to reveal potential
   biological processes and signaling pathways associated with IVF and
   embryo development. Protein–protein interaction (PPI) analysis was
   performed using the STRING database to identify key regulatory genes.
   Additionally, disease-specific miRNA binding sites with mRNA and
   circRNA were predicted using the StarBase database, and a competing
   endogenous RNA (ceRNA) network was constructed using Cytoscape software
   to illustrate the interactions between mRNA, miRNA, and circRNA.

   The experimental data were statistically analyzed using SPSS 26.0
   software, and the generation of bar graphs was accomplished using
   Graphpad Prism software. All measurement data were represented as
   mean ± standard deviation. Before conducting between-group comparisons,
   data normality and homogeneity of variance were verified to ensure
   normal distribution and equal variances. Independent samples were
   analyzed using an independent t-test. In cases of multiple-group
   comparisons, a one-way analysis of variance (One-Way ANOVA) was used to
   determine intergroup differences, followed by Tukey’s post hoc test.
   Statistical significance was inferred at a P-value < 0.05.

Supplementary Information

   [146]43556_2024_236_MOESM1_ESM.docx^ (732.7KB, docx)

   Supplementary Material 1: Figure S1. Bioinformatics analysis for
   selection of differentially expressed circRNA and mRNA during IVF of
   sperm. Note: (A) Volcano plot showing differentially expressed circRNA,
   with red indicating upregulated genes and blue indicating downregulated
   genes, n=2 for the control group, n=2 for the treated group; (B)
   Volcano plot showing differentially expressed mRNA, with red indicating
   upregulated genes and blue indicating downregulated genes, n=3 for
   control group, n=3 for treated group; (C) Heatmap showing
   differentially expressed circRNA, with red indicating upregulated genes
   and blue indicating downregulated genes, A represents the treated
   group, B represents the control group; (D) Heatmap showing
   differentially expressed mRNA, with red indicating upregulated genes
   and blue indicating downregulated genes, A represents the treated
   group, B represents the control group. Figure S2. PPI analysis of
   differentially expressed mRNA.Note: The figure shows key proteins in
   the PPI network of differentially expressed mRNA.

Acknowledgements