IntroductionMammalian spermatogenesis is a complex and sequential biological process involving the transformation and differentiation of various germ cell types1. Deviations from the normal progression of these stages can lead to abnormal spermatogenesis, obstruction, and ultimately male sterility. The regulation of spermatogenesis is controlled not only by numerous coding genes, but also by the involvement of various regulatory small non-coding RNAs. Among these, miRNAs are well-documented regulators of male sterility across germ cell development stages2. In contrast to miRNAs, PIWI-interacting RNAs (piRNAs) are specifically expressed in germ cells in substantial quantities, influencing processes such as de novo DNA methylation, meiosis, sperm morphogenesis, and playing a critical role in germ cell maturation and male fertility3,4,5. Mature piRNAs, typically ranging from approximately 26 to 34 nucleotides in length, exhibit a strong preference for uridine at the 5’ end and 2’-O methylation at the 3’ end. These piRNAs are primarily derived from elongated single-stranded transcripts spanning repetitive sequences, transposable elements, and intergenic regions. Through sequence complementarity, piRNAs guide PIWI proteins to cleave target RNAs, enabling their roles in transposon silencing, germline stem cell maintenance, and post-transcriptional regulation4,6,7,8,9.PIWI proteins, a specialized clade of Argonaute (Ago) proteins, are indispensable for piRNAs synthesis and function2,10,11. In mice, three PIWI subfamily members exhibit stage-specific expression: PIWIL1 is active from pachytene spermatocytes to spermatids, PIWIL2 in embryonic testes, and PIWIL4 in embryonic spermatogonia. Genetic ablation of PIWIL2 or PIWIL4 triggers cell death prior to pachytene spermatocyte formation, while PIWIL1 loss disrupts post-meiotic spermatid development5. In addition to the PIWI proteins, many factors have been identified as crucial mediators for piRNA biogenesis and function, including scaffolding proteins containing the Tudor structural domain protein TDRDs12, mitochondrial-associated proteins13, RNA helicase DDX414, PNLDC115, MOV10L116, FKBP617, among others. Although the functional mechanism of these proteins in piRNA production is well understood, a complete understanding of their expression regulation during spermatogenesis remains elusive. In-depth investigations into the regulation of their expression are necessary to integrate these proteins into a comprehensive regulatory pathway, thereby systematically elucidating the mechanisms by which the PIWI/piRNA complex facilitates piRNA processing. Such research will ultimately enhance our understanding of the piRNA pathway, as well as its roles and mechanisms of action in spermatogenesis.Heterogeneous nuclear ribonucleoprotein K (hnRNPK), a multifunctional RNA-binding protein, plays a crucial role in animal embryonic development, cell proliferation, and cell differentiation through post-transcriptional regulation18,19. The gene encoding mouse Hnrnpk is notably expressed at elevated levels in the testis20,21,22. Initially, hnRNPK was studied as a significant regulator of tumorigenesis23,24,25,26. Subsequent research has provided substantial evidence underscoring the extensive involvement of hnRNPK in various physiological and pathophysiological processes, such as embryogenesis27, axonal development and regeneration28,29, oogenesis30, skeletal muscle development31,32, and virus replication33. Furthermore, our previous study has demonstrated that male mice with a specific knockout of hnRNPK in pre-meiotic male germ cells (referred to as Hnrnpk cKO) exhibited male reproductive defects and impaired spermatogenesis, leading to infertility22. Notably, hnRNPK deficiency disrupts spermatogenesis at the pachytene stage—a phase marked by robust piRNA activity. This temporal overlap suggests hnRNPK may orchestrate piRNA pathway components, though the mechanistic basis remains unknown.In this study, we aimed to investigate the underlying molecular mechanisms responsible for the meiotic cell cycle arrest in Hnrnpk-deficient spermatogonia, utilizing tandem mass tag (TMT) labeling quantitative proteomics technology. Further analyses revealed that perturbed expression of the piRNA pathway protein played a significant role in Hnrnpk-deficient mice. Notably, we have successfully identified the mRNA of piRNA pathway-related genes, including Piwi1, Tdrd7, Ddx4, Asz1, and Mael, as interaction partners of hnRNPK in the testis. This identification was achieved through the utilization of RNA immunoprecipitation (RIP), dual-luciferase reporter system, and Fluorescent in situ hybridization coupled with immunofluorescence (FISH/IF) staining techniques. Consequently, it can be inferred that hnRNPK directly binds to the mRNA of these piRNA metabolic pathway-related genes, thereby regulating the translation of multiple transcripts. Importantly, these genes collectively exert a significant regulatory influence on piRNA production during the pachytene stage, thereby playing a crucial role in the process of spermatogenesis. Collectively, our work uncovers hnRNPK as a critical upstream regulator of the piRNA pathway, offering new insights into the molecular basis of pachytene-stage spermatogenesis.Materials and methodsAnimal UsedMice were maintained on a standard 12 h light/12 h dark cycle (8:00 AM to 8:00 PM) at the humidity of 40 ~ 70% and ambient temperature of 22 ± 2 ℃ with free access to food and water. The Hnrnpk cKO strain was obtained by constitutive defloxing of the Floxed Hnrnpk transgenic mice using the Stra8-Cre line (Jax No. 008208, obtained from The Jackson Laboratory) as previously described22. Hnrnpk flox/flox mice were generated with the aid of Cyagen Biosciences Inc. (Suzhou, Jiangsu, China) in a C57BL/6 genetic background using TurboKnockout® technology. Wide-type (WT) and Hnrnpk cKO mice were sacrificed by cervical dislocation at P28 and the testes were harvested to be measured and weighed.Protein extraction and digestion and TMT labelingFrozen testis samples were ground and lysed in STD buffer containing a 1% protease inhibitor cocktail (Beyotime). Then the sample was homogenized on ice by sonication with the ultrasonic cell crusher at 80 W for 3 min. The homogenate was centrifuged at 16,000 g for 45 min at 4 °C. Supernatants were collected and then quantified by the BCA assay. Protein digestion was performed as described in our previous publications34. For proteome analysis, the peptide was labeled using TMT 6-plex reagents (Thermo Fisher Scientific) according to the manufacturer’s recommendations. In detail, each TMT reagent was dissolved in 88 µL anhydrous acetonitrile and added to different peptide mixtures. Three independent biological repeats were performed. WT samples were labeled with TMT 126, 127, and 128, while Hnrnpk cKO samples were labeled with TMT 129, 130, and 131. After labeling, peptides from each TMT sample were mixed in a vacuum concentrator for further analysis by liquid chromatography-tandem mass spectrometry (LC-MS/MS).LC-MS/MS analysis based on Q-Exactive mass spectrometryLC-MS/MS analyses were applied on a Q-Exactive mass spectrometer coupled to an Easy nLC1000 nano-HPLC system (Proxeon Biosystems). Briefly, the vacuum-dried peptides were reconstituted in 20 µL Buffer I (0.1% formic acid), loaded on a Thermos Scientific EASY columm equilibrated with 95% Buffer I, and then the peptides were separated on a C18 column at a flow rate of 250 nl/min. Peptide mixtures were eluted using buffer I, and buffer II (100% acetonitrile and 0.1% formic acid) under a 170 min gradient with a flow rate of 250 nl/min. For MS data acquisition, the eluted peptides were analyzed in positive ion mode on a Q-Exactive MS with MS survey scan of 300–1800 m/z range and MS/MS spectra (15,000 resolution at 200 m/z, high-energy collision dissociation, 2 m/z isolation window, 27 normalized collision energy, 2 × 105 AGC, 100 ms maximal ion time, 30 s dynamic exclusion, and top number 20). The mass spectrometry raw data have been deposited to the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProXpartner repository35,36 with the dataset identifier PXD056125.MS data analysisThe Raw MS/MS files were interpreted with Proteome Discoverer 2.4 using the search engine Mascot (Version 2.2, Matrix Science) against the Mus musculus UniProt protein reference database (comprising 17,167 entries, Jun 2022). For proteome analysis, the parameters for database searching were set as previously described34. All reported data were based on 99% confidence for protein and peptide identifications as determined by an FDR of ≤ 0.01. The quantitative analysis was performed on the log2-values of the measured reporter intensities using Perseus software (version 1.5.5). A two-sample t-test was carried out within SPSS18.0. Proteins with q-value ≤ 0.05 and fold-change ratios ≥ 1.5 or ≤ 0.67 were considered as differentially expressed proteins (DEPs). The biological functions of identified DEPs were annotated using Metascape bioinformatics analysis website tools (https://metascape.org)37.RT-qPCR analysisTotal RNA was isolated from testis samples using Trizol (Takara) according to the manufacturer’s protocol. The first strand cDNA was synthesized by reverse transcription of 1.0 µg total RNA using M-MLV Reverse Transcriptase. RT-qPCR was performed on the Bio-rad C1000 Touch PCR System using the iTaq Univer SYBR Green Supermix (Bio-rad). In the experiment, three parallel repeating wells were designed, and all specimens were repeatedly tested for 3 times. The gene expression was quantified relative to the internal parameters Gapdh and 18 S, and 2(−ΔΔCt) was used to analyze the data. For piRNA expression analysis, total small RNA was isolated from testis samples using miRcute miRNA extraction Kit (DP501, TIANGEN, USA), and reverse transcription reactions of miRNAs were performed with the miRcute miRNA First-Strand cDNA Synthesis Kit (KR211, TIANGEN, China) using the poly(A) method. U6 was used as the internal reference gene for piRNAs. All the primer sequences are listed in Table S1.Western blotting analysisThe testis samples were cracked with RIPA buffer containing a 1% protease inhibitor cocktail (Beyotime), and protein concentration and quantification were determined with BCA assay. Then, proteins were denatured by heating at 100 °C, separated on 12.0% SDS-PAGE gels, and transferred to PVDF membranes. After blocking with 5% defatted milk for 1 h, the membranes were incubated in the diluted primary antibodies (Table S2) overnight at 4 °C, and followed with HRP-labeled Goat Anti-Mouse IgG or HRP-labeled Goat Anti-Rabbit IgG as secondary antibodies at a dilution of 1:1000 and detected by the enhanced chemiluminescence system. Finally, the blot was imaged using a FluorChem M multicolor fluorescence imaging system (ProteinSimple). The experiment was repeated three times. The obtained protein band intensities were analyzed by Image J Software.Tissue collection and histological examinationP28 mouse testes were collected and fixed in 4% paraformaldehyde and then embedded in paraffin. For immunostaining, cross-sections were deparaffinized in xylene for 10 min and rehydrated in a descending alcohol series (100%-90%-80%-70%). Tissue sections were then performed antigen retrieval step by keeping them in boiling Tris-EDTA buffer for 10 min. Then sections were blocked in 5% BSA and incubated with diluted primary antibodies (Table S2) at 4 ℃ overnight. For immunofluorescence (IF) staining, sections were washed and incubated with Alexa Fluor 488 and 555-conjugated secondary antibodies for 2 h at room temperature and DAPI Staining Solution (Beyotime) were used to visualize the nucleus. For immunohistochemistry, sections were washed and incubated with HRP-labeled Goat Anti-Mouse IgG or HRP-labeled Goat Anti-Rabbit IgG as secondary antibodies, and then stained the tissues using the DAB Horseradish Peroxidase Color Development Kit (Beyotime). The images were visualized using a Nikon 80i with NIS-Elements system (Nikon).Small RNA sequencingTotal RNA was extracted from P28 WT and Hnrnpk cKO testes as described above. For each sample, 2 ug total RNA was used as the starting material for the construction of the Small RNA sequencing library. Small RNA (18–40 nt) was isolated from total RNA samples using 15% PAGE gel and recovered by gel cutting. Then the isolated small RNA was enriched by ethanol precipitation and centrifugation. After enrichment, the small RNA samples were subjected to library preparation according to the manufacturer’s protocol of TruSeq Small RNA Sample Preparation Kit (Illumina). The library preparations were sequenced on an Illumina Novaseq6000 platform and paired-end reads were generated. Mm9 was used as the mouse genome reference sequence. The lengths of miRNAs and piRNAs were defined as 20–24 and 24–45 nt, respectively.Transmission electron microscopy (TEM)For TEM analysis, testes collected from P18 WT and Hnrnpk cKO were cut into 1–2 mm3 pieces and fixed in cacodylate buffer at 4 °C overnight. After three washes in 0.1 M cacodylate buffer, the samples were incubated in 1% OsO4 for 1 h at room temperature, and stained in 2% uranyl acetate for 30 min. Subsequently, dehydration was done through consecutive incubation in sequentially ethanol solutions (30, 50, 70, 90, and 100%) and an acetone bath. Then samples were sequentially embedded in an Eponate mixture for polymerization about 24 h at 60 °C. Ultrathin sections (~ 70 nm) were cut on an ultra-microtome, and these sections were subsequently re-stained with uranyl acetate and lead citrate. Finally, the sections were photographed using an H7000 electron microscope (Hitachi) at 120 kV.RNA Immunocoprecipitation (RIP)Imprint® RNA Immunoprecipitation Kit (Sigma) was applied for RIP assay, and the operation steps were carried out according to the instructions. Briefly, after removing the tunica albuginea of the testis, the seminiferous tubule samples were cracked in the RIP lysis buffer. Then cell lysate, RIP buffer including magnetic beads with hnRNPK antibody and normal mouse IgG were incubated at 4 °C for 4 h. Finally, immunoprecipitation RNA was extracted and RT-qPCR analysis was performed.Double luciferase analysisThe partial 3’UTR sequences of Piwil1, Piwil2, Tdrd7, Ddx4, Aszl, and Mael containing hnRNPK binding motif ‘UCCC’ and corresponding mutation binding site were amplified by PCR and cloned into the downstream of the Renilla luciferase gene in psi-CHECK2 vector to construct wild-type 3’UTR and mutants reporter vectors of Piwil1, Piwil2, Tdrd7, Ddx4, Aszl and Mael (The 3’UTR sequence information see in Supplemental materials 1). The pcDNA3.1-Hnrnpk and the above-mentioned reporter vector were co-transfected into HEK-293T cells using Lipofectamine 3000. Fireflies and renin luciferase activities in cell lysates were detected using a dual luciferase reporting kit after 48 h of transfection. For analysis, R-Luc activity normalized with F-Luc activity was set to 1.RNA fluorescence in situ hybridization (FISH) and immunofluorescenceRNA FISH and antibody staining were performed on dissected P28 testes. For RNA FISH experiments, the procedure was carried out according to the instructions of Fluorescent In Situ Hybridization Kit (RiboBio). The Piwil1 probe labeled with FITC dye (125 nM) was used in this study and sequences of custom probe sets are listed in Table S2. All hybridizations were done overnight in the dark at 37 °C in a humidifying chamber.Statistical analysisAll the experiments were carried out at least in triplicate and the values were shown as mean ± SEM. Statistical significances were measured based on a t-test using the SPSS18.0 software. P ≤ 0.05 (*) was considered significant difference, and p ≤ 0.01 (**) was considered a very significant difference between the two groups.ResultsDifferential proteomic analysis of testes between WT and Hnrnpk cKO miceBefore performing proteomics analysis to identify changes associated with Hnrnpk deletion, testis tissues were harvested from P28 WT and Hnrnpk cKO mice. Six samples were subsequently lysed, digested, labeled with multiplex TMT, and finally pooled and analyzed using LC-MS/MS. This analysis identified 62,371 unique peptides corresponding to 8,058 proteins, with 7,219 proteins quantified for relative comparison (refer to Table S3). Principal component analysis (PCA) revealed a distinct separation between the proteomes of the two groups, with the first principal component accounting for approximately 70.68% of the variance and the second component accounting for 13.57% (Fig. 1a). The analysis indicated significant differences in the proteomes of cKO testes compared to WT, identifying 791 DEPs, with 256 upregulated and 535 downregulated (Fig. 1b and c, Table S4). Several DEPs associated with spermatogenesis, including SYCP3, SYCP1, RPA32, AKAP3, AKAP4, and PHKG2, were selected for validation through Western Blot analysis. The validation results demonstrated that the levels of SYCP3, SYCP1, RPA32, AKAP3 and AKAP4 were significantly reduced in the Hnrnpk cKO testes compared to WT, whereas PHKG2 was more abundant in cKO testes (Fig. 1d, e, and Fig. S3). The tendency of expression changes of these Six proteins was consistent with the TMT quantitative analysis, confirming the reliability of our data.Fig. 1Whole proteome analysis of Hnrnpk cKO mice testis. (a) PCA analysis of TMT-proteoome in each group. (b) Volcano plot of the differentially expressed proteins in Hnrnpk cKO testes compared with the WT control at P28. Green dots represent significantly downregulated proteins, red dots represent significantly upregulated proteins (p-adj value 2), and gray dots represent unchanged proteins. (c) The heatmap of differentially expressed proteins. (d) Validation of proteins expression by western blot analysis. (E) Statistical analysis of the results shown in D (n = 3). *p