Introduction

Spermatogonial stem cells (SSCs) are responsible for the continuous production of sperm throughout a man’s life, ensuring fertility1. SSCs possess self-renewal and differentiation capacity and are regulated by orderly intrinsic gene expression2,3. RNA-binding proteins (RBPs) are a class of proteins that interact with RNA molecules, and play pivotal roles in post-transcriptional gene regulation4. The RBPs recognize and bind specific RNA sequences or structures, modulating RNA stability, splicing, localization, translation, and degradation5. Several studies have showed the pivotal functions of RBPs in regulating the maintenance and differentiation of SSCs. For instance, NANOS2 is an RBP involved in maintaining the primitive state of mouse SSCs by mediating direct translational repression of differentiation-related transcripts and inhibiting the activity of mammalian target of rapamycin complex 1 (mTORC1)6. The deleted in azoospermia-like (DAZL) protein regulates the expansion and differentiation of progenitor spermatogonia by orchestrating a broad translational program that regulates the expression levels of key proteins, including Lin-28 homologue A (Lin28a), forkhead box O1 (Foxo1) and Sal-like protein 4 (Sall4)7. These results indicate that RBPs are essential for the maintenance of SSCs and expansion of progenitor spermatogonia.

Insulin-like growth factor 2 mRNA-binding protein 1 (IGF2BP1) is a critical post-transcriptional regulator that exerts significant control over gene expression by binding with specific mRNA targets and contains four hnRNP-K homology domains and two RNA recognition motifs domains8. IGF2BP1 belongs to the IGF2BP family and plays an essential role in the regulation of target mRNAs related to metabolism, impacting diverse cellular processes such as stem cell maintenance and tumorigenesis9,10. It has been shown that IGF2BP1 regulates target mRNAs through distinct mechanisms, one mechanism involves IGF2BP1 binding to target mRNA, protecting it from degradation by endonucleases11. In another mechanism, IGF2BP1 associates with target mRNAs in RNA-induced silencing complex-free complexes, inhibiting miRNA-mediated degradation of target mRNAs12. Recent studies revealed that IGF2BP1 promotes the expression of SRF by impairing the miRNA-directed decay of SRF mRNA and promoting the G1/S cell cycle transition13,14. IGF2BP1 also bound to the 3′-untranslated region (UTR) of CD44, CTNNB1 and MAPK4 mRNAs, inhibiting their degradation and stabilizing the mRNAs8,15,16. Interestingly, it has been reported that Imp, an ortholog of Igf2bp1 in Drosophila, maintains germline stem cells (GSCs) by stabilizing upd mRNA in the Drosophila testis stem cell niche17. In humans, IGF2BP1 is highly expressed in the male reproductive system18. However, the in vivo function of IGF2BP1 in mammalian spermatogenesis, particularly regarding its age-related actions, remains inadequately characterized.

The current study integrated multiple approaches to query the biological and molecular roles of IGF2BP1 in spermatogenesis. We found that IGF2BP1 is specifically expressed in human and mouse spermatogonia. It was shown that Igf2bp1 deficiency impaired spermatogenesis in an age-dependent manner, especially in 12-month-old conditional knockout (cKO) mice. Immunoprecipitation and mass spectrometric analysis revealed that ELAVL1, a well-recognized RNA stabilizer, is a potent co-factor of IGF2BP1. Furthermore, single-cell RNA-sequencing (scRNA-seq) revealed that Igf2bp1 deficiency in mice resulted in the downregulation of the expression of key regulators of SSC maintenance. Coupling with RNA immunoprecipitation-sequencing (RIP-Seq), we identified target mRNAs of IGF2BP1 that are enriched in “gene silencing”, “RNA localization” and “regulation of gene expression, epigenetic”, including Lin28a, Dnd1, Gfra1 and Mov10. In addition, immunofluorescence staining showed that IGF2BP1 and LIN28A exhibit simultaneous high expression in human seminoma tissues. Collectively, this study suggests that IGF2BP1 is a key RBP for the maintenance of spermatogonia in mammals and overexpression of IGF2BP1 is associated with human seminoma.

Results

IGF2BP1 is enriched in spermatogonia during human and mouse testicular development

Immunofluorescence staining was used to determine the expression of IGF2BP1 in both human and mouse testes. Human testes were derived from adult obstructive azoospermia (OA)-affected patients with normal spermatogenesis, and pre-pubertal (2- and 5-years of age) and peri-pubertal (8-year-old) children. In OA testes, IGF2BP1 was expressed in glial cell line-derived neurotrophic factor receptor alpha 1-positive (GFRA1+) undifferentiated SSCs and ubiquitin C-terminal hydrolase L1-positive (UCHL1+) spermatogonia, with high expression in cytoplasm (Supplementary Fig. 1a, b). However, the expression of IGF2BP1 rapidly decreased in c-KIT+ differentiating spermatogonia (Supplementary Fig. 1c). The expression pattern was consistent during pre- and peri-puberty. In testes from 2-, 5- and 8-year-old donors, IGF2BP1 also co-localized with UCHL1 (Supplementary Fig. 2a–c).

In mouse testes, IGF2BP1 co-stained with mouse spermatogonia markers cadherin 1 (CDH1) and UCHL1, however it was undetectable in synaptonemal complex 3-positive (SYCP3+) spermatocytes (Supplementary Fig. 3a–c). We also explored IGF2BP1 expression in mouse primary cultured GSCs and observed high expression of IGF2BP1 in the cytoplasm of cultured GSCs (Supplementary Fig. 3d). The conserved expression pattern suggested that IGF2BP1 is enriched in spermatogonia during human and mouse testicular development.

Igf2bp1 deficiency resulted in spermatogenic defects in an age-dependent manner in male mice

According to the UCSC genome browser and the UniProt server, IGF2BP1 was highly conserved in humans and mice (Supplementary Fig. 4a, b). To examine the function of IGF2BP1 during spermatogenesis in vivo, conditional knock-out (cKO, Nanos3-Cre, Igf2bp1delete/flox) mice were generated by crossing Igf2bp1delete/flox mice with Nanos3-Cre recombinase transgenic mice (Fig. 1a). Western blotting and immunofluorescence demonstrated the complete depletion of IGF2BP1 in the Igf2bp1 cKO mouse testes (Fig. 1b; Supplementary Fig. 5a).

Fig. 1: Igf2bp1 depletion resulted in spermatogenic defects in an age-dependent manner in mice.
figure 1

a Schematic strategy for constructing the Igf2bp1 conditional knock-out (cKO) mice. b Western blotting analysis of IGF2BP1 expression in the testes of wildtype (WT) and Igf2bp1 cKO mice. GAPDH was used as a loading control. c The morphology of testes from 8-month-old WT and Igf2bp1 cKO mice. Scale bar: 1 cm. d The testis-to-body weight ratios of 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice (n = 3). Bar graphs represent the means ± standard deviations. ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001. e, f Representative images of periodic acid-Schiff-stained testicular sections from 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice (n = 3). Scale bar: 100 µm. The black asterisks indicate abnormal seminiferous tubules. Bar graphs represent the means ± standard deviations. ns not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. g Representative images of hematoxylin-eosin stained cauda epididymal sections from 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice. Scale bar: 100 µm. h The sperm concentration per cauda epididymis from 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice (n = 3). Bar graphs represent the means ± standard deviations. ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Full size image

At 8 months of age, Igf2bp1 cKO mice exhibited significantly reduced testes volumes compared to wild type (WT) (Fig. 1c). The testis-to-body weight ratio showed no statistical difference at 2 months of age, began to decrease at 4 months, and followed an age-related pattern [12-month-old mice: 0.00086 ± 0.00017 (cKO) vs 0.00284 ± 0.00022 (WT)] (Fig. 1d). To test male fertility, 4-month-old WT and Igf2bp1 cKO male mice were cohabited with WT female mice for 8 months, and the numbers of pups per litter generated by mating of 4- to 8-month-old and 8- to 12-month-old male mice were recorded. We found that the Igf2bp1 cKO mice were fertile during 4- to 8-month-old, with no statistically significant difference in number of pups per litter compared to WT mice (Supplementary Fig. 5b). And Igf2bp1 cKO male mice displayed reduced fertility compared with WT mice from 8 to 12 months of age. Among four Igf2bp1 cKO male mice, two fathered no offspring from the age of 8 to 12 months (Supplementary Fig. 5b). Testicular histology analysis showed 4-month-old Igf2bp1 cKO mice exhibited a higher proportion of abnormal seminiferous tubules with reduced germ cells versus WT mice (Fig. 1e). Furthermore, compared with 4-month-old Igf2bp1 cKO mice, 8- and 12-month-old Igf2bp1 cKO mice had more abnormal seminiferous tubules. Notably, up to 70% of seminiferous tubules were abnormal in 12-month-old Igf2bp1 cKO mice [73.0% ± 9.8% (cKO) vs 2.3% ± 1.5% (WT)] (Fig. 1e, f). A significant reduction in sperm concentration was consistently observed from the caudal epididymis of Igf2bp1 cKO compared with WT mice, accompanied by an extremely low sperm concentration in 12-month-old Igf2bp1 cKO mice [3.9 ± 1.1 M/ml (cKO) vs 24.7 ± 1.4 M/ml (WT)] (Fig. 1g, h). There were no significant differences in serum testosterone levels between 12-month-old Igf2bp1 cKO and WT mice (Supplementary Fig. 5c). These results revealed that loss of Igf2bp1 in male mice resulted in spermatogenic defects in an age-dependent manner, indicating an indispensable role of IGF2BP1 in the maintenance of the male germ line with time.

Igf2bp1 deficiency reduced the pool size of undifferentiated spermatogonia in mice

To elucidate the testicular function of IGF2BP1, we analyzed spermatogenesis in 2-, 4-, 8- and 12-month-old Igf2bp1 cKO and WT mice by immunofluorescence staining. Immunofluorescence staining for Sertoli cell marker sex determining region Y-Box 9 (SOX9) in testes showed that SOX9-positive cell numbers per tubule were similar in Igf2bp1 cKO compared with WT mice of all ages (Fig. 2a; Supplementary Fig. 5d). Using the mouse testicular germ cell-specific antigen antibody TRA98, 8-month-old Igf2bp1 cKO mice exhibited a higher proportion of abnormal testicular tubules, including tubules with massive loss of germ cells, as well as Sertoli cell only (SCO) tubules (Fig. 2a). The percent of SCO tubules increased significantly compared with WT mice in an age-dependent manner (Fig. 2b).

Fig. 2: Igf2bp1 depletion reduced the undifferentiated spermatogonia pool size in mice.
figure 2

a Immunostaining for SOX9 (Sertoli cell marker, green) and TRA98 (germ cell marker, red) in testes from 2-, 4-, 8- and 12-month-old wildtype (WT) and Igf2bp1 conditional knock-out (cKO) mice. The white asterisks indicate Sertoli cell only (SCO) tubules. Scale bar: 20 μm. b Percentage of SCO tubules in 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice. Bar graphs represent the means ± standard deviations. ns not significant, **P < 0.01, ***P < 0.001, ****P < 0.0001. c Immunostaining for PLZF (undifferentiated spermatogonia marker, green) and c-KIT (differentiating spermatogonia marker, red) in 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice. The white arrows indicate PLZF-positive cells. Scale bar: 20 μm. d–f The PLZF-positive cells (d), c-KIT-positive cells (e) and ratio of PLZF/c-KIT positive cells (f) per seminiferous tubule in 2-, 4-, 8- and 12-month-old WT and Igf2bp1 cKO mice (n = 3; tubules per mouse, n = 40). Bar graphs represent the means ± standard deviations. ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001. g Immunostaining for SYCP3 (green) and PNA (red) in 8-month-old WT and Igf2bp1 cKO mice. Scale bar: 20 μm. h Images of TUNEL staining of testicular sections from 8-month-old WT and Igf2bp1 cKO mice. The black asterisks indicate TUNEL-positive tubules. Scale bar: 100 μm. i Percentage of TUNEL-positive tubules in 8-month-old WT and Igf2bp1 cKO mice (mice, n = 3; tubules per mouse, n = 40). Bar graphs represent the means ± standard deviations. **P < 0.01.

Full size image

Promyelocytic leukemia zinc-finger (PLZF) is a common marker for undifferentiated spermatogonia, which comprises SSCs and progenitor germ cells, and KIT proto-oncogene receptor tyrosine kinase (c-KIT) is a common marker for differentiating spermatogonia. We counted the average number of PLZF-positive cells per seminiferous tubule from male mice of different ages. Notably, the average number of PLZF-positive cells per tubule significantly decreased in Igf2bp1 cKO mice from 4-, 8- and 12-month-old compared with WT mice and showed a decreasing trend with age (Fig. 2c, d).

The average number of c-KIT-positive cells per tubule in 2-month-old Igf2bp1 cKO mice showed no statistical difference, but it decreased in 4-, 8-, and 12-month-old Igf2bp1 cKO mice and showed a decreasing trend with age (Fig. 2e). Also, we quantified the ratio of PLZF to c-KIT per tubule. The results showed that at 4-, 8- and 12-month-old, the PLZF/c-KIT ratio in cKO mice was lower compared to WT mice (Fig. 2f). Meanwhile, immunofluorescence staining of SYCP3 (components of the axial/lateral element) and PNA (a marker for the acrosome of sperm cells) showed that Igf2bp1 cKO testes have normal meiosis and spermiogenesis (Fig. 2g). In addition, TUNEL staining showed significantly increased germ cell apoptosis in 8-month-old Igf2bp1 cKO mice compared with WT mice (Fig. 2h, i). These results suggested that IGF2BP1 is required to maintain the undifferentiated spermatogonia pool over time, and depletion of Igf2bp1 destabilized the quantity of undifferentiated spermatogonia in mice.

IGF2BP1 interacted with mRNA stabilizer ELAVL1 in human and mouse testes

Immunoprecipitation and mass spectrometric analysis were used to investigate the partners of IGF2BP1 in human and mouse testes. After quality control, we identified 78 proteins in human and 73 proteins in mouse testes that potentially interacted with IGF2BP1, and 22 of them overlapped in both human and mouse testes (Fig. 3a). Gene Ontology (GO) analysis showed that these 22 proteins were enriched in the terms “regulation of mRNA stability”, “mRNA binging” and “cytoplasmic ribonucleoprotein granule” (Fig. 3b). The protein-protein interaction network was constructed using the STRING database to find the hub gene. Interestingly, the results showed that ELAVL1 is the central protein in the protein-protein interaction network and has an extremely high combined score with IGF2BP1 (Fig. 3c; Supplementary Data 2).

Fig. 3: IGF2BP1 interacted with ELAVL1 in human and mouse spermatogonia.
figure 3

a Venn diagram showing the proteins enriched in IGF2BP1-immunoprecipitation (-IP) complexes in mouse and human testes, identified by mass spectrometry analysis. b Gene Ontogeny (GO) analysis of the overlapping targets of IGF2BP1. c STRING analysis of the proteins overlapping in the IGF2BP1-IP complex in mouse and human obstructive azoospermia testes. d Co-IP assay by anti-IGF2BP1 antibody showing the interaction between IGF2BP1 and ELAVL1 in both mouse and obstructive azoospermia testes. e Co-IP assay by anti-Myc or anti-Flag antibody showing the interaction between transfected 3× Flag-IGF2BP1 and Myc-ELAVL1 fusion proteins with or without RNase A treatment in HEK293T cells. f Co-localization of IGF2BP1 and ELAVL1 in human testis samples revealed by immunofluorescence staining (white arrows). Scale bar: 20 µm.

Full size image

Previous studies revealed that ELAVL1 is a highly conserved mRNA stabilizer involved in post-transcriptional gene regulation19,20,21,22,23. To study the interaction between IGF2BP1 and ELAVL1, we performed co-immunoprecipitation (co-IP) of IGF2BP1 and ELAVL1 using testis samples from patients with OA and mice. IGF2BP1 was able to efficiently pull down ELAVL1, indicating interaction between these two proteins in both human and mouse testes (Fig. 3d). Using HEK-293T cells transfected with plasmids containing 3× Flag-tagged IGF2BP1 and Myc-tagged ELAVL1, co-IP showed that IGF2BP1 and ELAVL1 reciprocally interacted with each other with or without RNase A, establishing an in vitro interaction between the two proteins independent of RNAs (Fig. 3e). Furthermore, immunofluorescent staining of IGF2BP1 and ELAVL1 revealed their co-localization in the cytoplasm of human spermatogonia (Fig. 3f). These results suggested that ELAVL1 is a key interaction protein of IGF2BP1 involved in the regulation of target mRNA stability.

scRNA-seq demonstrated that IGF2BP1 deficiency altered mRNA profiles of spermatogonia

To explore the impact of IGF2BP1 on gene expression at single-cell resolution, we performed 10× Genomics scRNA-seq using 8-month-old Igf2bp1 cKO and WT testes and obtained 6021 quality-controlled single cells. Dimensionality reduction and clustering analysis, combined with previously described known cell-type markers identified major germ cell populations covering the whole process of spermatogenesis, including A1: mainly undifferentiated spermatogonia, In-Bspg: differentiated spermatogonia, PL: pre-leptotene spermatocytes, L: leptotene spermatocytes, Z: zygotene spermatocytes, P: pachytene spermatocytes, D: diplotene spermatocytes, M1-M2: metaphase I and II spermatocytes, RS2-RS8: round spermatids and elongating spermatids (Fig. 4a, b). We utilized single-cell data to quantify the number and proportion of various germ cell types. Our analysis revealed that the number of all germ cell types significantly decreased in Igf2bp1 cKO mice compared to WT (Fig. 4c). In terms of proportion, the In-Bspg and RS8 cell types appeared to have a higher proportion in Igf2bp1 cKO mouse compared to WT (Fig. 4c). However, due to the uncertainties associated with single-cell enzymatic digestion and capture, accurately determining the proportions of different germ cell types using single-cell data may be challenging. To explore the transcriptional changes of Igf2bp1 depletion in the A1 cluster (mainly undifferentiated spermatogonia), we conducted differentially expressed genes (DEGs) analysis of the transcriptome of A1 spermatogonia between WT and Igf2bp1 cKO samples. Notably, 1627 genes were downregulated and 1622 genes were upregulated in A1 spermatogonia from Igf2bp1 cKO versus WT mice (Fig. 4d). Previous studies have shown that IGF2BP1 can stabilize target mRNAs, such as c-Myc mRNA, through binding mRNA stabilizer ELAVL121. Analysis of GO terms of downregulated genes in the Igf2bp1 cKO compared with WT A1 spermatogonia groups revealed terms including “ribonucleoprotein complex biogenesis”, “mRNA processing” and “gene silencing” (Fig. 4e). Meanwhile, we analyzed the upregulated genes in A1 and found that they are mainly associated with “establishment of protein localization to organelle”, “ncRNA processing” and “RNA splicing” (Fig. 4f). Interestingly, key mRNAs that encode essential regulators of SSC proliferation and maintenance (Lin28a, Dnd1, Gfra1, and Mov10) were significantly decreased in the Igf2bp1 cKO compared with WT A1 spermatogonia clusters (Fig. 4g). Notably, these transcripts were also enriched in Zbtb16 (Plzf)+ A1 cluster spermatogonia, consistent with the expression of Igf2bp1 (Fig. 4h). These results indicated that the depletion of Igf2bp1 significantly altered the transcriptome of spermatogonia and downregulated key factors of SSC maintenance.

Fig. 4: Overview of single-cell transcriptome profiling on Igf2bp1 conditional knock-out (cKO) testes.
figure 4

a Uniform manifold approximation and projection (UMAP) plot of combined single-cell transcriptome from 8-month-old wild-type (WT) and Igf2bp1 cKO mice testes. b UMAP plots of single-cell transcriptomes of 8-month-old WT or Igf2bp1 cKO mice testes. c The stacked bar charts showing cell counts and proportions for different cell types in WT and Igf2bp1 cKO mice. d Heatmap showing the differentially expressed genes (DEGs) for WT versus Igf2bp1 cKO A1 cluster groups. Columns represent individual cells and rows represent DEGs. The gradient of two colors (red and blue) indicates low to high expression levels. e, f Gene Ontogeny (GO) analysis of the downregulated DEGs (e) and upregulated DEGs (f) in the Igf2bp1 cKO versus WT A1 spermatogonia cluster groups. g Bubble diagram showing the expression levels of key mRNAs in WT and Igf2bp1 cKO A1 clusters. The gradient of bubbles sizes indicates low to high percentage of expression in each cluster, and the gradient of two colors (blue and red) indicates low to high expression levels. h Bubble diagram showing the expression levels of key mRNAs in all clusters. The gradient of bubbles sizes indicates low to high percentage expression in each cluster, and the gradient of two colors (blue and red) indicates low to high expression levels.

Full size image

IGF2BP1 interacted with key mRNAs that encode essential regulators of SSC maintenance

To elucidate the function of IGF2BP1 in germ cells, we conducted sequencing of RNAs isolated by RNA immunoprecipitation sequencing (RIP-Seq) using IGF2BP1 antibody. Western blotting analysis confirmed successful immunoprecipitation of IGF2BP1 (Fig. 5a). Cultured mouse GSCs (~4 ×107) were used to pull-down RNAs, and 3578 genes were identified as putative targets, with 445 RNAs overlapping with downregulated RNAs found in the A1 spermatogonia cluster of Igf2bp1 cKO mice identified by scRNA-Seq (Fig. 5b). GO analysis of biological processes of these target RNAs showed that key RNAs are clustered in terms of “gene silencing”, “RNA localization” and “regulation of gene expression, epigenetic” (Fig. 5c). Intriguingly, many target RNAs involved in these GO terms have been widely recognized as key regulators for the proliferation and maintenance of SSCs, including Lin28a, Dnd1, Gfra1 and Mov10. RIP-qPCR validated the results of RIP-Seq of IGF2BP1 in mouse P10 testes (Fig. 5d). Additionally, to verify the interaction between ELAVL1 and these target mRNAs, we conducted RIP-qPCR for ELAVL1 in mouse P10 testes. The results demonstrated binding relationships between ELAVL1 and Lin28a, Dnd1, and Mov10 (Fig. 5e).

Fig. 5: IGF2BP1 associated with mRNAs that encode essential regulators of spermatogonial stem cell (SSC) maintenance.
figure 5

a Immunoblot analysis of IGF2BP1 immunoprecipitation using cultured germline stem cells (GSCs). b Venn diagram showing the overlapping transcripts identified by RNA immunoprecipitation-sequencing (RIP-Seq) of GSCs and downregulated differentially expressed genes (DEGs) in the conditional knock-out (cKO) A1 spermatogonia cluster by single cell RNA-sequencing (scRNA-Seq). c Gene Ontology enrichment analysis of the overlapping transcripts identified by RIP-Seq of GSCs and downregulated DEGs in the cKO A1 cluster by scRNA-Seq. d, e RIP-qPCR analysis of IGF2BP1 (d) and ELAVL1 (e) potential targets in P10 mouse testes, with Gapdh as a negative control. Relative enrichment in IGF2BP1-IP and ELAVL1-IP was compared with input (n = 3). Bar graphs represent the means ± standard deviations. ns, not significant, ****P < 0.0001. f Luciferase reporter analysis demonstrating reduced activity of indicated reporters in Igf2bp1-knockdown (shRNA, red) compared with control HEK293T cells (shNC, blue). ns, not significant, ***P < 0.001. g Reducing LIN28A mRNA half-life by silencing IGF2BP1 in Huh-6 cells. Bar graphs represent the means ± standard deviations. h Western blotting assay to compare the protein expressions of LIN28A between 8-month-old wildtype (WT) and Igf2bp1 cKO mice. i The band intensity of LIN28A was normalized to GAPDH. ***P < 0.001. j, k Immunostaining of LIN28A in testes from 8-month-old WT and Igf2bp1 cKO mice (mice n = 3; tubules per mouse, n = 40). Cell nuclei were stained with DAPI. The white arrows indicate LIN28A-positive cells. Fluorescence intensity of LIN28A was normalized to DAPI intensity. Bar graphs represent the means ± standard deviations. **P < 0.01.

Full size image

LIN28A is an RNA-binding protein involved in pluripotency and stem cell maintenance and is strongly expressed in mouse and primate SSCs24,25. Testicular conditional depletion of Lin28a in mice resulted in impaired clonal expansion of undifferentiated spermatogonia26. To investigate whether Lin28a mRNA could be bound by IGF2BP1 via the motif (GGAC), WT and mutant type Lin28a 3′-UTR containing luciferase reporters were constructed and shRNA were used to efficiently knock down IGF2BP1 expression in the HEK293T cell line (Supplementary Fig. 5e, f). IGF2BP1 knock-down resulted in downregulation of luciferase activity of the WT Lin28a 3′-UTR containing reporter (Fig. 5f). This downregulation was eliminated when the recognition motif GGAC was mutated to CCTG (Fig. 5f). Meanwhile, accelerated mRNA decay of LIN28A upon knockdown of IGF2BP1 was confirmed in Huh-6 cells (Fig. 5g, Supplementary Fig. 5g, h).

Furthermore, we examined the protein expression levels of LIN28A between WT and Igf2bp1 cKO mouse testes using WB and immunofluorescence. The results showed that the expression of LIN28A in the testes of 8-month-old Igf2bp1 cKO mice is significantly reduced compared to WT (Fig. 5h, i). Based on fluorescence intensity relative to DAPI, LIN28A expression was significantly reduced in 8-month-old cKO compared with WT mice (Fig. 5j, k). These results suggested that IGF2BP1 is associated with the maintenance of spermatogonia, potentially through the stabilization of target mRNAs, such as Lin28a mRNA.

IGF2BP1 and LIN28A are highly expressed in human seminoma

Immunofluorescence staining on human seminoma sections was used to investigate the expression pattern and correlation of IGF2BP1 and LIN28A in germ cell tumors. Human seminoma tissues were confirmed by histological analysis. Co-staining of IGF2BP1 with the seminoma cell marker octamer-binding transcription factor 4 (OCT4) showed that IGF2BP1 expression is significantly increased in OCT4+ seminoma cells compared with OCT4- spermatogonia in human seminoma and paracancer testicular tissues (Fig. 6a). Furthermore, immunofluorescence staining showed that human seminoma tissues simultaneously express high levels of IGF2BP1 and LIN28A (Fig. 6b). These expression patterns suggested that the expression of IGF2BP1 and LIN28A may be correlated in human seminoma.

Fig. 6: IGF2BP1 and LIN28A are highly expressed in human seminoma.
figure 6

a Co-staining of IGF2BP1 with the seminoma cell marker OCT4 in human seminoma and paracancer testicular tissues. The white hollow arrows indicate OCT4+ cells and the white solid arrows indicate OCT4− cells. The white dashed-line oval indicates paracancer testicular tissues (seminiferous tubule). Scale bar: 20 µm. b Co-staining of IGF2BP1 with LIN28A in human seminoma. Scale bar: 20 µm.

Full size image

Discussion

Our research provided compelling evidence that IGF2BP1 has a critical role in mammalian spermatogenesis, particularly in the maintenance of SSCs. A novel mouse model showed that conditional depletion of Igf2bp1 in germ cells lead to disruption of spermatogenesis in an age-dependent manner, with most seminiferous tubules becoming pathologically SCO in 12-month-old mice. Furthermore, we identified that ELAVL1 is a crucial co-factor of IGF2BP1 in testes, implicating a potential cooperative role in the regulation of mRNA stability during spermatogenesis. ScRNA-Seq and RIP-Seq analysis revealed that IGF2BP1 is largely associated with the maintenance of spermatogonia, potentially by regulating mRNA stabilization of essential regulators of SSC proliferation and maintenance, such as Lin28a mRNA. In addition, immunofluorescence staining showed that IGF2BP1 and LIN28A are synchronously expressed at high levels in human seminoma tissue.

The RBP IGF2BP1 plays a crucial role in recognizing and regulating the expression of key mRNAs. In lower organisms, such as Drosophila, the homolog of IGF2BP1, known as IMP, plays a critical role in the long-term maintenance of male GSCs, by counteracting endogenous small interfering RNAs to stabilize upd RNA in the stem-cell niche17. The current research showed aberrant spermatogenesis is increased in an age-dependent manner in Igf2bp1 cKO mice, accompanied by a gradually declining testis-to-body weight ratio and an increasing number of SCO tubules from 4 to 12 months of age. Furthermore, PLZF-positive undifferentiated spermatogonia were markedly decreased in 4-, 8-, and 12-month-old Igf2bp1 cKO mice. Meanwhile, TUNEL staining showed significantly increased germ cell apoptosis in 8-month-old Igf2bp1 cKO mice compared with WT mice. We found that while most apoptotic cells are in the basal region of seminiferous tubules, specifically in spermatogonia, there are indeed instances of apoptosis observed in other germ cells. This observation appears inconsistent with the expected expression pattern of IGF2BP1. We hypothesize that this discrepancy may be related to the early transcription of key mRNAs of other germ cells in spermatogonia, which are subsequently regulated by IGF2BP1. RIP-seq of IGF2BP1 revealed the presence of several crucial mRNAs, such as Cdc45 and Ccnd2. However, additional experiments are required to validate and elucidate the molecular mechanisms underlying these observations. Subsequent IP-MS, STRING and co-IP assay analysis showed that ELAVL1 is a crucial co-factor for IGF2BP1 in the testis.

ELAVL1 is a highly conserved RBP involved in post-transcriptional gene regulation, significantly contributing to RNA stability19. ELAVL1 can bind to 3′-UTR AREs (AU-rich elements) and stabilize mRNAs by preventing their degradation20,22,23. Recent advancements have revealed that IGF2BP1 recruits ELAVL1 as co-factor to protect mRNAs from degradation21. Intriguingly, the present study also found that IGF2BP1 interacts with ELAVL1 in human and mouse testes and co-locates with ELAVL1 in the cytoplasm of spermatogonia. Therefore, these results indicate that IGF2BP1 may stabilize target mRNAs by recruiting ELAVL1 in mammalian spermatogenesis. In addition, spermatogenesis in Elavl1-deficient mice is arrested during meiosis and is characterized by a failure of spermatid elongation27. The differences in phenotypes between Igf2bp1-deficient mice and Elavl1-deficient mice are as follows. First, in Igf2bp1-deficient mice, abnormalities, including SCO tubules and a reduction in sperm count, began to appear as early as 4 months of age and worsen significantly with advancing age. Elavl1-deficient mice exhibit defects in delayed elongation of spermatids as early as 4 weeks of age, with compromised meiotic division observed from 9 weeks of age. Second, Igf2bp1 depletion appeared to affect the long-term maintenance of spermatogonia, leading to age-dependent depletion of germ cells and eventual SCO tubules. On the other hand, Elavl1 deficiency primarily caused extensive death of spermatocytes at meiotic divisions and failure of spermatid elongation. The phenotypic differences observed between IGF2BP1 and ELAVL1 deficiencies in mice can be attributed to several factors. One primary reason could be the broader expression range of ELAVL1 compared to IGF2BP1. ELAVL1 is expressed at higher levels in both spermatocytes and round spermatids, suggesting it plays a crucial role across various stages of spermatogenesis. Furthermore, the differences in observed phenotypes may also relate to the timing of phenotypic assessment in the mice. In Igf2bp1-deficient mice, abnormalities such as SCO tubules and reduced sperm count become evident as early as 4 months of age and worsen with advancing age. On the other hand, Elavl1-deficient mice show defects in spermatid elongation as early as 4 weeks, with compromised meiotic division observable from 9 weeks of age. However, the phenotypes of Elavl1-deficient mice have not been assessed in older mice, leaving a gap in understanding how these deficiencies manifest over a longer period. Thus, the distinct expression profiles and the timing of phenotypic observations likely contribute to the different impacts of IGF2BP1 and ELAVL1 deficiencies.

The RBPs fulfil their functions through regulation of target mRNAs in a post-transcriptional manner, and via synergistic action with other proteins. In the current study, scRNA-Seq identified 1627 downregulated RNAs in spermatogonia from 8-month-old testes Igf2bp1 cKO versus WT testes. The application of RIP-Seq with GSCs identified more than 3000 potential targets of IGF2BP1. Because IGF2BP1 may regulate the stability of target mRNAs by recruiting ELAVL1, we examined the potential targets identified by RIP-Seq and the downregulated DEGs found in the spermatogonia A1 cluster of Igf2bp1 cKO mice and identified 445 overlapping transcripts. Bioinformatics analysis indicated that these targets are enriched in the terms “gene silencing”, “RNA localization” and “regulation of gene expression, epigenetic”. Notably, transcripts regulating SSC maintenance were identified as the target mRNAs, including Lin28a, Dnd1, Gfra1, and Mov10. LIN28A is a developmental regulator known for its involvement in pluripotency and stem cell maintenance via repressed let-7 microRNAs24,25. Conditional depletion of Lin28a in the mouse male germline resulted in impaired clonal expansion of undifferentiated spermatogonia26. Dead end 1 (DND1) is a vertebrate-specific RNA-binding protein implicated in the survival of primordial germ cells and maintenance of spermatogonia. Murine Dnd1 deficiency resulted in the depletion of both differentiating and undifferentiated spermatogonia28. DND1 binds the 3′-UTR region of mRNAs and destabilizes target mRNAs through direct recruitment of the CCR4-NOT deadenylase complex29. GFRA1 is a member of the glial cell line-derived neurotrophic factor receptor family and a common cell surface marker of undifferentiated spermatogonia1. Gdnf+/− mice showed depletion of the stem cell pool, resulting in SCO seminiferous tubules in older mice30. These findings showed that IGF2BP1 regulates the maintenance of spermatogonia through targeting these key genes.

Aberrant development in SSCs can lead to the formation of testicular germ cell tumors, specifically seminomas31. Seminomas resulting from the uncontrolled proliferation of spermatogonia highlighting the complex relationship between SSCs and testicular germ cell tumors. Interestingly, the present research showed that IGF2BP1 and LIN28A are highly expressed in seminomas. This suggests a potential balance in the expression of IGF2BP1 and LIN28A for spermatogenesis or seminomas. Deficiency of Igf2bp1 may lead to decreased expression of LIN28A, consistent with the largely reduced SSC pool characterized by an abundance of SCO tubules in 12-month-old Igf2bp1 cKO testes. Conversely, overexpression of IGF2BP1 may disrupt the normal regulation of LIN28A, leading to abnormal proliferation of SSCs and thereby the development of germ cell tumors.

In conclusion, our study provides compelling evidence for the critical role of IGF2BP1 in mammalian spermatogenesis, particularly in the maintenance of SSCs. The conditional depletion of Igf2bp1 in germ cells resulted age-dependent and severe spermatogenic defects. Furthermore, co-IP revealed that ELAVL1 is a crucial co-factor of IGF2BP1 in the testis, implicating a potential cooperative role in the regulation of mRNA stability during spermatogenesis. Additionally, this study identified Lin28a, Dnd1, Gfra1 and Mov10 as crucial target mRNAs of IGF2BP1 in the regulation of spermatogonia (summarized in Fig. 7). Thus, our research expands the function of IGF2BP1 in regulating the long-term maintenance of mammalian spermatogenesis and provides new potential diagnostic and therapeutic targets for severe oligozoospermia, azoospermia and human seminoma in the future.

Fig. 7: Diagram of proposed functions and mechanisms of IGF2BP1.
figure 7

IGF2BP1 is essential for the long-term maintenance of spermatogonia. Further experiments showed that mRNA stabilizer ELAVL1 is a crucial co-factor of IGF2BP1 in the testis and Lin28a, Dnd1, Gfra1 and Mov10 are crucial target mRNAs of IGF2BP1 in the regulation of spermatogonia.

Full size image

Materials and methods

Antibodies

The antibodies used for immunofluorescence and western blotting have been listed in Supplementary Data 3.

Human samples

Testicular tissue samples were obtained from three patients diagnosed with OA who had normal karyotypes, genotypes, and sex hormone levels, and exhibited normal spermatogenesis. Postoperative pathological examination showed that their Johnsen scores are not less than 9. Testicular tissues were obtained from children who underwent testicular biopsies or partial excision for the following indications: benign testicle mass (2 and 5 years of age) and contralateral testis to cryptorchidism (8-year-old), and children patient’s source in accordance with the previously described32. All children had normal karyotypes, genotypes, and sex hormone levels. Seminoma samples were confirmed by histological analysis. All participants in this study provided informed consent. The clinical data of the donors are summarized in Supplementary Data 4. This study was approved by the Ethical Review Committee of Shanghai General Hospital (Permit Number: 2020SQ199). All ethical regulations relevant to human research participants were followed.

Mouse model and cell line

Nanos3-Cre mice were maintained and used as a tool to conditionally induce gene depletion33. Igf2bp1delete/flox mice were used to delete Igf2bp1 specifically in germ cells. All animals were maintained in a C57BL/6/MCH background. All mice were housed in a specific pathogen free animal facility under standard conditions by Gempharmatech Co., Ltd. We have complied with all relevant ethical regulations for animal use. All animal care and experiments adhered to the guidelines of the National Institutes of Health and were approved by the Animal Care Committee of Shanghai General Hospital. Supplementary Data 5 presents a list of the primers employed for genotyping Igf2bp1 cKO mice. The HEK-293T and Huh-6 cells were cultured with DMEM medium containing 10% FBS at 37 °C and 5% CO2 incubation. Primary GS cell lines were established and cultured as previously described34. Briefly, mouse embryonic fibroblasts (MEFs) derived from day 13.5 post-coitum mouse embryos were used as feeders. Postnatal day 5 C57BL/6 male mouse testes were aseptically removed and digested using 1 mg/ml collagenase IV (17104019, Thermo Fisher Scientific) and 1 mg/ml DNase I (90083, Thermo Fisher Scientific) to dissociate the tissue. The digested seminiferous tubules were mechanically dissociated into fragments using pipette tips. These fragments were then seeded into 12-well plates coated with 0.2% gelatin and cultured with 20 ng/ml rat glial cell line-derived neurotrophic factor (GDNF, 512-GF, R&D systems) for 18 h to promote initial colony formation. Each germline stem cell (GSC) colony was carefully blown off using a pipette and transferred to new plates coated with 0.2% gelatin. For the P1 generation, the cells were cultured in MEM-α/StemPro-34 medium (12561, GIBCO) supplemented with 20 ng/ml GDNF and 1 ng/ml mouse fibroblast growth factor 2 (FGF2; 450–33, Peprotech). The GSC colonies were then cultured in a feeder-free environment for 3–4 passages. During this period, somatic cell contamination decreased significantly, leaving only a thin layer of somatic cells. Once the somatic cells were sufficiently reduced, the GSCs were transferred to plates containing mitomycin C-treated MEFs. At this stage, the concentration of FGF2 was increased to the normal level of 10 ng/ml to support optimal GSC proliferation and maintenance. After the establishment of the GSC line, the cells can be passaged using 0.25% trypsin digestion, and the GDNF concentration can be reduced to 10 ng/ml.

Histological analysis

The tissues were fixed in 4% paraformaldehyde solution then dehydrated through a gradient of alcohol solutions, followed by infiltration with molten paraffin wax for subsequent sectioning. Cut 5-μm-thick sections were used for further histological staining. After deparaffinization and hydration, the testicular sections were subjected to periodic acid-Schiff staining, dehydration, blocking and scanning analysis. Similarly, the epididymis sections were subjected to hematoxylin-eosin staining according to the manufacturer’s instructions.

Immunostaining of testicular tissues

Immunofluorescence staining was performed on 5 μm sections of human and mouse testicular samples. The sections were subjected to deparaffinization, rehydration and heat-mediated antigen retrieval in a 10 mM sodium citrate buffer solution (pH 6.0) at 105 °C for 15 min. After treatment with 5% donkey serum for 1 h at room temperature, individual sections were incubated overnight at 4 °C with a mix of diluted primary antibodies. Antigen detection was conducted using the appropriate combination of Alexa Fluor 488 and 594 secondary antibodies (Supplementary Data 3) for 1 h at room temperature in the dark. DAPI was used to label nuclei. Images were captured with an OLYMPUS confocal microscope (FV3000).

Plasmids construction

Plasmids pCMV-Myc-ELAVL1 and pCMV-3×Flag-IGF2BP1 were synthesized by amplifying specific coding sequences using PCR and then recombining into pCMV-Myc or pCMV-3×Flag vectors via the ClonExpress One Step Cloning Kit (C112, Vazyme). For knockdown of IGF2BP1, a DNA fragment encoding shRNA was synthesized in vitro and incorporated into the pGMLV-SC5 vector to produce the pGMLV-SC5-IGF2BP1 shRNA plasmid, which was purchased from Genomeditech Co., Ltd. The 3′-UTR of Lin28a containing the wild-type (GGAC) motif was amplified from mouse testis cDNA using PCR and then recombining into pmirGLO vector via the ClonExpress One Step Cloning Kit (C112, Vazyme). The 3′-UTR of Lin28a containing the mutant (CCTG) motif was synthesized by TsingkeBiotechnology Co., Ltd to ensure precise mutation and incorporated into the pmirGLO vector.

IP-mass spectrometry and co-IP

Human and mouse testes were lysed in Pierce IP Lysis Buffer (88804, Thermo Fisher Scientific) containing complete EDTA-free protease inhibitor cocktail, then immunoprecipitated with anti-IGF2BP1 antibody (22803-1-AP, Proteintech) by a Pierce Crosslink IP Kit (88804, Thermo Fisher Scientific). The immunoprecipitated protein complexes were detected using western blotting. The IGF2BP1 protein complexes pulled down by anti-IGF2BP1 antibody from human and mouse testes were analyzed by the mass spectrometry center of ShanghaiTech University using the entire elutes. HEK-293T cells transfected with pCMV-Myc, pCMV-Myc-ELAVL1, pCMV-3×Flag or pCMV-3×Flag-IGF2BP1 were also lysed with Pierce IP Lysis Buffer and precipitated with anti-Myc or anti-Flag antibodies in the presence or absence of 5 U/μL RNase A. The immunoprecipitated protein complexes were detected using western blotting.

scRNA-seq analysis

Samples from 8-month-old Igf2bp1 cKO and WT mice were used for scRNA-seq. Detailed scRNA-seq analysis of testicular samples has been described in a previous study32. Briefly, individual mouse testicular cells were obtained through enzymatic digestion and loaded on a 10× Genomics chip to form gel beads in emulsion, which were subjected to reverse transcription and PCR amplification, and then sequenced on a Nove-seq 6000 platform (Illumina). A cell-gene expression matrix was obtained with Cell Ranger software (10× Genomics). After quality control, standardization and dimensionality reduction, cell identification and clustering analysis were performed in the R environment. The Seurat FindMarkers function was used for identification of differentially expressed genes in the Igf2bp1 cKO versus WT A1 cluster groups. GO analysis was performed using WebGestalt 2019 (http://www.webgestalt.org/#).

RIP and sequencing

The process of RIP was carried out using the EZ-Magna RIP kit (17-701, Millipore) in accordance with the manufacturer’s guidelines. Briefly, cultured GSCs (~4 ×107) or P10 mouse testes (400 mg) were subjected to RIP lysis buffer. The resulting lysates were then centrifuged at 20,000 g for 15 min and the supernatant collected. The supernatant was subsequently used for immunoprecipitation using either anti-IGF2BP1 beads or IgG beads. Both input and co-immunoprecipitated RNAs were isolated using RNA Clean & Concentrator-5 (R1013, Zymo), and were subjected to sequencing (Cloud-Seq Biotech Co., Ltd.) or real-time qPCR analysis. The primers for target mRNAs are listed in Supplementary Data 5.

Western blotting

Both tissues and cultured cells were added to RIPA lysis buffer (P0013B, BiotechWell) for a duration of 30 min on ice. Following centrifugation at 12,000 g for 20 min at 4 °C, the protein concentrations of lysates were measured using a BCA kit (P0012, BiotechWell). SDS-PAGE was performed with 20 μg of lysate from each sample, and western blottings were carried out in accordance with the previously described protocol35. After conducting SDS-PAGE on a 10% gel to load and separate proteins (95 V, 30 min and 120 V, 80 min for loading and separation, respectively), the samples were transferred to 0.45 µm polyvinylidene membranes. The membranes were then blocked with 5% non-fat dried milk, followed by incubation overnight at 4 °C with primary antibodies (Supplementary Data 3) specific for the target proteins. The next day, the membranes were washed with TBS containing 0.1% Tween and incubated with HRP-conjugated secondary antibodies (Supplementary Data 3) for 1 h at room temperature. After washes with TBS 0.1% Tween, the proteins were detected by chemiluminescence (34080, Thermo Fisher Scientific).

RNA extraction and real-time qPCR

The TRIzol method or Total RNA Isolation Kit (RC112-01, Vazyme) were used to extract total RNA from cultured cells or tissues. The quality and concentrations of RNA were assessed using a NanoDrop (840-317400, Thermo Fisher Scientific). Then, 1 µg of extracted RNA was reverse transcribed using a First Strand cDNA Synthesis Kit (K16215, Thermo Fisher Scientific). The cDNAs were used for qRT-PCR in a final 20-µL volume with Power SYBR Green PCR Master Mix in a 96-well thermal cycler. Gapdh mRNA expression was used to standardize the expression of target mRNA. The comparative CT (∆∆CT) method was used to quantify the PCR results. Primers for target mRNAs are listed in Supplementary Data 5.

mRNA stability assay

Huh-6 cells, stably expressing shRNAs targeting IGF2BP1 or a non-specific control (shNS), were seeded into 6-well plates to achieve 50% confluency after 24 h. The cells were then treated with 5 μg/ml actinomycin D (A9415, Sigma-Aldrich) and harvested at specific time points. Total RNA was extracted using the Total RNA Isolation Kit (RC112-01, Vazyme) and analyzed by RT-qPCR.

Luciferase reporter assay

Luciferase plasmids were transfected into cells using Lipofectamine 3000 (L3000001, Thermo Fisher Scientific) according to the manufacturer’s protocol. Control and Igf2bp1 knockdown HEK-293T cells were seeded to reach 70–80% confluency and incubated with the DNA-transfection reagent complexes for 6 h. After incubation, the medium was replaced with fresh growth medium, and cells were further incubated for 48 h. Luciferase activity was determined by a Dual-Glo Luciferase Assay System (Promega, USA) and was shown relative to the control shRNA group.

Statistics and reproducibility

The mean ± standard deviation was used to represent the data unless otherwise specified. For experiments involving multiple treatments, a one-way analysis of variance was conducted to determine significance by GraphPad software. Alternatively, for experiments with only two treatments, Student’s t test was used to examine statistical significance. P < 0.05 was statistically significant. The sample sizes (n numbers) are detailed in the figure legends. Except for scRNA-seq and RIP-seq data, all experiments were independently repeated at least three times. No specific methods were applied for sample allocation into groups. No data were excluded, as all collected data were included in the analysis. Data collection and analysis were performed without blinding to the experimental conditions. For the representative images shown in the paper, each experiment was repeated independently at least three times with consistent results.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.