RUNX proteins and CBFβ create an interconnected network of transcriptional regulation in the male germline

Abstract

The foundation for lifelong spermatogenesis depends on a highly coordinated program of prepubertal germline development, during which a precise balance between spermatogonial proliferation and differentiation is established to ensure sustained spermatogenesis. Any disruptions to this balance can impair germ cell maturation and overall fertility. However, factors critical in maintaining this balance remain incompletely understood. Our previous studies revealed that core-binding factor subunit-β (CBFβ) regulates both proliferation and differentiation during the onset of spermatogenesis. Canonically, CBFβ functions as a co-factor for the Runt-related transcription factor (RUNX) family by forming heterodimeric complexes that can act either as transcriptional activators or repressors. Here, we reveal interactions between CBFβ and RUNX proteins within the male germline and highlight distinct expression patterns of RUNX1 and RUNX3, particularly differential temporal expression during discrete cell cycle phases within spermatogonia. Moreover, Cleavage Under Targets & Release Using Nuclease (CUT&RUN) analyses revealed both overlapping and distinct genomic localization of RUNX1 and RUNX3. Surprisingly, knockdown studies determined that RUNX1 and RUNX3 act in opposition as either transcriptional activators or repressors within overlapping genomic targets. By contrast, genomic regions with differential RUNX1 or RUNX3 localization suggest distinct regulation of proliferation or differentiation, respectively. Furthermore, motif analysis revealed enrichment of disparate transcription factor motifs, including canonical regulators of the germline. Collectively, our findings suggest that CBFβ, RUNX1, and RUNX3 participate in a network to precisely coordinate proliferation and differentiation during prepubertal germline development, thus ensuring continuous spermatogenesis and male fertility.

Postnatal male germline development is modulated by the CBFβ-RUNX network, in which RUNX1 and RUNX3 exhibit distinct expression patterns and transcriptional functions to support spermatogenesis.

Graphical Abstract

Graphical Abstract.

Introduction

Spermatogenesis is a tightly coordinated process by which diploid spermatogonia undergo terminal differentiation with two rounds of meiotic division to produce haploid spermatozoa. During neonatal germline development, establishment of the spermatogonial population is essential for initiating and sustaining spermatogenesis throughout life. The spermatogonia population is heterogeneous and composed of undifferentiated spermatogonia, which includes spermatogonial stem cells (SSCs), and differentiating spermatogonia. The undifferentiated population is uniquely tasked with supporting two essential functions: proliferative self-renewal of SSCs to maintain the germline and the differentiation of spermatogonia on a path to eventually form spermatozoa [1–4]. The balance of self-renewal and differentiation therefore ensures both maintenance of the undifferentiated spermatogonia pool and continuous production of sperm [4–6]. If the balance shifts towards excessive self-renewal without differentiation, undifferentiated spermatogonia accumulate to the point of increased cell death [7–13]. Conversely, if the balance biases differentiation without replenishing the stem cell pool, the ability to sustain spermatogenesis is exhausted over time and results in infertility [14–20]. Therefore, establishing a framework that balances self-renewal and differentiation is essential to ensure sustained spermatogenesis.

The factors that dictate the balance between self-renewal and differentiation remain incompletely understood, but likely include an intricate network of both intrinsic factors and extrinsic factors [2, 21–23]. For example, a number of extrinsic factors have been identified that regulate proliferation, and self-renewal in some instances, including glial cell-derived neurotrophic factor (GDNF), colony stimulating factor 1 (CSF1), wingless-type MMTV integration site family members (WNTs), bone morphogenic proteins (BMPs), and fibroblast growth factors (FGFs) [24–35]. In contrast, retinoic acid (RA) is the essential extrinsic factor that triggers the irreversible differentiation of undifferentiated spermatogonia into differentiating spermatogonia. Once on this differentiation path, germ cells will undergo the timely completion of meiosis and formation of sperm via intrinsic regulators, including, but not limited to, STRA8, MEIOSIN, and DMRT1 [36–40]. While significant progress has been made in understanding these processes, the precise interplay among critical pathways and factors regulating the balance between proliferation and differentiation remains incompletely understood.

In our previous studies, we discovered that CBFβ is necessary for both proliferation within undifferentiated spermatogonia, including self-renewal of the SSC population, and differentiation in spermatogenesis such that spermatid and sperm formation are impaired [41]. Thus, CBFβ somehow intersects both self-renewal and differentiation. Interestingly, transcriptomic analysis of Cbfb conditional knockout (cKO) males revealed that CBFβ regulates cell cycle checkpoint genes for mitosis and meiosis [41]. However, it remains unclear how CBFβ can intersect both self-renewal and differentiation functions. Importantly, using a pair of Cbfb cKO approaches in the male germline, our previous studies demonstrated that the ability of CBFβ to regulate both self-renewal and differentiation is likely confined to the undifferentiated spermatogonia population, even with distal aspects of differentiation impacted by Cbfb cKO [41].

Canonically, CBFβ is a common co-factor for the RUNX transcription factor family, which is composed of three familial constituents that share a conserved RUNT DNA-binding domain in the N-terminal region [42, 43]. This domain also enables RUNX interactions with CBFβ, at which point CBFβ allosterically enhances and stabilizes RUNX protein associations with DNA [42–45].

The mammalian RUNX transcription factor family consists of RUNX1, RUNX2, and RUNX3, which serve as key regulators in development, cell proliferation, differentiation, and lineage specification [46–57]. However, the role of RUNX proteins in male reproduction and spermatogenesis remains limited. In the developing gonads, prenatal RUNX1 expression can be detected in both male and female gonads at the onset of sex determination, but RUNX1 expression becomes enriched within ovarian somatic cells and acts in concert with other factors to ensure ovarian specification [58]. In the postnatal male germline, transcripts of Runx2 variants have been reported in spermatocytes and spermatids in the testis [59]. Finally, studies suggest that RUNX1 and RUNX2 are important for epididymal epithelial differentiation and thus, may be involved in sperm maturation after the testis [60]. Collectively, however, limited studies have illustrated the role of the RUNX family within male reproductive systems.

For the studies herein, we turned our attention to individual RUNX proteins in the male germline because our previous studies of CBFβ utilized a cKO strategy that ablated the N-terminal region of CBFβ responsible for CBFβ-RUNX protein interactions. Thus, we broadly asked if individual RUNX proteins may separately underlie the capacity of CBFβ to regulate either self-renewal or differentiation in the male germline. Outcomes of expression profiling indicated that RUNX1 and RUNX3, but not RUNX2, are expressed in spermatogonia. Furthermore, immunoprecipitation of CBFβ revealed interactions with both RUNX1 and RUNX3 in spermatogonia. Immunofluorescence analysis further determined that RUNX1 and RUNX3 are expressed in subtly different patterns across spermatogonial subtypes. Interestingly, cell cycle analysis of undifferentiated spermatogonia indicated that while CBFβ and RUNX3 are expressed through all phases of the cell cycle, RUNX1 is limited within the late stages of mitosis (G₂/M phases), suggesting a unique role in proliferation. Finally, we performed CUT&RUN and motif enrichment analyses to investigate whether RUNX1 and RUNX3, in relation to CBFβ, bind to unique or shared targets in spermatogonia during germline development. Our findings revealed that RUNX1 and RUNX3 have both distinct and overlapping targets with CBFβ, but disparate targets linked each RUNX protein to either proliferation or terminal differentiation programs in spermatogonia. Collectively, our studies not only demonstrate that CBFβ interacts with RUNX proteins in the male germline, but also that RUNX transcription factors may be integral for CBFβ to balance between proliferation and differentiation.

Results

CBFβ interacts with the RUNX family in the male germline

To begin exploring the role of the RUNX proteins in the male germline, we first examined the expression of each RUNX family member across different germ cell types. To do this, we mined published single-cell RNA-sequencing (scRNA-seq) data of the postnatal male germline during development [2, 21, 61–63]. Consistent with previous findings, Cbfb transcripts were heterogeneously expressed across many adult (P56) germ cell subtypes, which include both undifferentiated and differentiating spermatogonia (marked by Lin28a, Gfra1, Stra8, Kit), spermatocytes (marked by Sycp3, Meiob), and round spermatid (marked by Prm1, Prm2), but Cbfb was most abundant in undifferentiated and differentiating spermatogonia (Figure 1A–C and Supplementary Figure 1A–E). Runx1 transcripts were predominantly found in the undifferentiated spermatogonia population, albeit at low levels and with sparse detection (Figure 1A, D). Runx2 transcripts were sparsely detected in germ cells (Figure 1A, E), and Runx3 transcripts were primarily detected in differentiating spermatogonia population (Figure 1A, F). Next, we evaluated expression across development in spermatogonia, which revealed relatively constant abundance of Cbfb, Runx1, and Runx3 transcripts through postnatal development (Figure 1G). Western blot analysis of whole-testis lysates using validated antibodies (see Methods) indicated the presence of CBFβ, RUNX1, and RUNX3 and the absence of RUNX2 (Supplemental Figure 2A–E). Furthermore, a doublet banding pattern from both RUNX1 and RUNX3 in testis lysates suggests that more than one isoform of each protein is expressed in the testis, consistent with other tissues [49].

Figure 1.

Expression of Runx family and their co-factor, Cbfb, in postnatal male germline development. (A–G) Analysis of published scRNA-seq data from wild-type testes. (A) Violin plots of Cbfb, Runx1, Runx2, and Rux3 transcript abundance across germ cell subtypes at P56. Uniform manifold approximation projection (UMAP) representations of germ cell subtypes (B) alongside Cbfb (C), Runx1 (D), Runx2 (E), and Runx3 transcripts (F) in P56 (adult) germ cells. Additional violin plots illustrate Cbfb, Runx1, Runx2, and Runx3 transcript in spermatogonia during postnatal development. (G) Violin plots of Cbfb, Runx1, Runx2, and Rux3 transcript abundance in spermatogonia population across ages. (H-I) Qualitative analysis of RUNX protein expression in male germline via western blot. RUNX1 and CBFβ detection (H), and RUNX3 and CBFβ detection (I) in immunoprecipitation (IP) and flow-through (FT) fractions from CBFβ and IgG control IP assays.

In many biological contexts, the ability of RUNX family proteins to interact with DNA is dependent on allosteric interactions with the co-factor CBFβ [64]. Our recent studies utilized a Cbfb cKO strategy, which disrupts RUNX-CBFβ interactions through ablation of the N-terminal domain of CBFβ [41]. However, this interaction between RUNX proteins and CBFβ has not been examined in the male germline. Therefore, we performed a co-immunoprecipitation of CBFβ using internally validated antibodies to determine if RUNX1 and RUNX3 interact with CBFβ in undifferentiated spermatogonia. With co-immunoprecipitation of CBFβ, bands corresponding to RUNX1 were detected at ~52 kDa (Figure 1H) and bands corresponding to RUNX3 were detected at ~43 kDa and ~45 kDa (Figure 1I). Thus, these results indicate the association of RUNX1 and RUNX3 with CBFβ, suggesting a conserved CBFβ-RUNX interaction within the male germline.

Subtle, yet divergent expression of RUNX1 and RUNX3 in spermatogonia subtypes

To further investigate RUNX expression within the heterogeneous spermatogonia population, we performed immunofluorescent staining of testis cross-sections across male germline development. First, we used LIN28A as a marker enriched within undifferentiated spermatogonia [65]. RUNX1 immunofluorescence co-localized with LIN28A, consistent with our scRNA-seq findings that RUNX1 is expressed in undifferentiated spermatogonia (Figure 2A, I – white and pink arrows). Interestingly, while Runx1 mRNA transcripts were sparse in other germ cell types, RUNX1 protein was detected in spermatocytes and spermatid. On the other hand, immunofluorescence analyses also revealed co-localization of RUNX3 with LIN28A+ germ cells, indicating protein expression in undifferentiated spermatogonia throughout postnatal germline development despite weak transcript representation by scRNA-seq (Figure 2B, I – white arrow).

Figure 2.

Expression profile of RUNX1 and RUNX3 within the heterogeneous spermatogonia population. (A–H) Representative immunofluorescence images of tissue cross-sections from P6, P21, and P56 testes stained with antibodies recognizing RUNX1 or RUNX3 (green), DAPI (grey), TRA98 (not shown), and selected spermatogonia markers (red), including LIN28A (A and B; undifferentiated spermatogonia), GFR⍺1 (C and D; SSC-enriched spermatogonia), SOX3 (E and F; progenitor-enriched spermatogonia), and cKIT (G and H; differentiating spermatogonia). White arrowheads mark RUNX1+ or RUNX3+ germ cells co-expressing the spermatogonia marker of interest. Yellow arrowheads indicate examples of RUNX1- and RUNX3- spermatogonia within each sub-type. Pink arrowheads highlight proliferative RUNX1+ germ cells, and blue arrowheads identify RUNX1 expression in advanced germ cells, such as spermatocytes and spermatids. Images are representative of n = 3 biological replicates for each immunostain at each developmental age. Scale bar 50 μm for images in A–J. (I) Schematic representation of marker gene expression within germ cell sub-types alongside RUNX1 and RUNX3 expression patterns based on co-immunofluorescence analyses. (J) Representative image of testis cross-sections at P6 stained with antibodies recognizing RUNX1+ (green) and DAPI (grey) in germ cell (TRA98; not shown). (K–M) Quantifications of CBFβ+ (K), RUNX1+ (L), and RUNX3+ (M) cells within the LIN28A+ TRA98+ (undifferentiated spermatogonia) population across development are plotted. Quantifications in K–M are presented as means ± SEM for n = 3 biologically independent animals and age point. Different letters indicate significant differences (P < 0.05) as determined by the Tukey HSD test.

To further characterize RUNX expression within the undifferentiated spermatogonia, we performed additional RUNX immunostaining with GFRα1 and SOX3, markers of stem-cell-enriched and progenitor-enriched spermatogonia, respectively [66, 67]. RUNX1+ germ cells co-localized with both GFRα1+ and SOX3+ cells across all developmental stages (Figure 2C, E, I – white and pink arrows). Meanwhile, RUNX3 showed weak and rare co-localization with GFRα1+ cells but moderate co-localization with SOX3+ cells (Figure 2D, F, I – white arrow). Thus, RUNX1 and RUNX3 appeared to have subtle differences in expression within undifferentiated spermatogonia subtypes. Furthermore, RUNX1 protein was also detected at low levels in cKIT+ differentiating spermatogonia (Figure 2G, I – white and pink arrows), while RUNX3 was highly expressed in cKIT+ differentiating spermatogonia population (Figure 2H, I – white arrow).

Further analysis also revealed a high occurrence of RUNX1+ germ cells that appeared to be undergoing active proliferation, as evidenced by metaphase alignment and condensed chromosomes (Figure 2A, C, E, G, J – pink arrow). The same expression pattern was not observed with RUNX3 expression.

Immunofluorescence staining of embryonic day (E) 18.5 and newborn (P0) testes under the same conditions did not resolve detectable RUNX1 and RUNX3 expression in the TRA98+ germ cells (Supplementary Figure 3A and B), suggesting primary activation of RUNX1 and RUNX3 protein expression in germ cells during postnatal development.

To quantify these protein expression observations, we used flow cytometry to determine the distribution of CBFβ, RUNX1, and RUNX3 within the undifferentiated spermatogonia population (LIN28A+ TRA98+) through postnatal development. Notably, CBFβ was detected within most LIN28A+ TRA98+ cells across all developmental ages, suggesting stable expression in the undifferentiated spermatogonia (Figure 2K). Intriguingly, RUNX1 expression in the undifferentiated spermatogonia was rare (2%–23% depending on developmental age), with peak representation at P21 before declining in adulthood, (Figure 2L), consistent with immunofluorescent staining. In contrast, RUNX3 expression increased from P6 to P21 and remained upregulated through adulthood (Figure 2M).

While stark differences in expression profiles for RUNX1 and RUNX3 were not observed, these findings revealed subtle, yet recognizable expression patterns of RUNX1 and RUNX3 in spermatogonia. Furthermore, these findings indicate that RUNX1+ spermatogonia are quite rare.

RUNX1+ and RUNX3+ spermatogonia have different cell cycle profiles

Our previous Cbfb cKO studies demonstrated that CBFβ regulates proliferation in the undifferentiated spermatogonia population [41]. Additionally, an elevated number of RUNX1+ germ cells were observed undergoing active proliferation (Figure 2J). Therefore, to explore the connection between cell cycle progression and CBFβ, RUNX1, and RUNX3, we performed multi-parameter flow cytometry analyses of each protein in undifferentiated spermatogonia (LIN28A+ TRA98+). We then used DNA staining to determine if CBFβ+, RUNX1+, and RUNX3+ undifferentiated spermatogonia exhibit disparate cell cycle profiles throughout postnatal germline development.

Flow cytometry of P6, P21, and P56 testis suspensions revealed that CBFβ+ LIN28A+ TRA98+ cells were within all cell cycle phases, with the highest proportion in G₁/G₀ and the remainder in S and G₂/M phases (Figure 3A and B, Supplementary Figure 4A and B). In contrast, RUNX1+ LIN28A+ TRA98+ cells were predominantly within G₂/M phases. Interestingly, not all LIN28A+ undifferentiated spermatogonia within G₂/M phases were RUNX1+, indicating that RUNX1 is primarily expressed within a subset of undifferentiated spermatogonia in the late stages of the cell cycle (Figure 3C and D, Supplementary Figure 4C and D). RUNX3 expression during cell cycle progression resembled that of CBFβ, with RUNX3+ LIN28A+ TRA98+ cells distributed across all cell cycle phases (Figure 3E and F, Supplementary Figure 4E and F). However, when comparing RUNX1+ and RUNX3+ populations in G₂/M, we found that RUNX3+ LIN28A+ TRA98+ cells were approximately half as abundant as RUNX1+ cells, yet both maintained consistent proportions across all ages (~92.6% and ~42.4%, respectively; Figure 3G).

Figure 3.

Proliferative profiles of CBFβ, RUNX1, and RUNX3 undifferentiated spermatogonia through postnatal germline development. (A–F) Multiparameter flow cytometric analysis of P6 undifferentiated spermatogonia assessing CBFβ, RUNX1, and RUNX3 protein expression. Cells were co-stained with FxCycle (DNA marker), LIN28A, and TRA98. Representative Density plots show LIN28A+ TRA98+ cells, with corresponding histograms of DNA staining for CBFβ+ (blue) or CBFβ- (grey) cells (A), RUNX1+ (green) or RUNX1- (grey) cells (C), and RUNX3+ (red) or RUNX3- (grey) cells (E) at P6. Quantifications of CBFβ+ (B), RUNX1+ (D), and RUNX3+ (F) cells within the LIN28A+ TRA98+ population across cell cycle stages are plotted. (G) Percentage of LIN28A+ TRA98+ cells that are RUNX1+ (green) and RUNX3+ (red) enriched during the G₂/M phase across developmental time points. (H–I) Representative immunofluorescence images (H) and quantification (I) of P6 testis cross-sections stained for CBFβ, RUNX1, and RUNX3 (green), KI67 (magenta), DAPI (grey) and LIN28A (red). Scale bar is 50 μm. Quantifications in B, D, F, G, and I are presented as means ± SEM for n = 3 biologically independent animals and age point. Statistical differences in G are denoted as ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. Different letters in I indicate significant differences (P < 0.05) based on one-way ANOVA and Tukey HSD test.

To further validate the expression of CBFβ, RUNX1, and RUNX3 in proliferative undifferentiated spermatogonia, we performed immunofluorescent staining of testis cross-sections using LIN28A and the proliferation marker KI67. We found that CBFβ, RUNX1, and RUNX3 co-localized with KI67+ LIN28A+ germ cells at all developmental ages (Figure 3H, Supplementary Figure 4G–I). Quantification of CBFβ+, RUNX1+, and RUNX3+ populations with KI67 showed distinct co-localization levels: ~62.1% for CBFβ, ~93.7% for RUNX1, and ~50.9% for RUNX3 (Figure 3I). Together, these findings indicate that CBFβ, RUNX1, and RUNX3 are expressed at distinct stages of the cell cycle, with RUNX1 specifically enriched in late-stage cell cycle progression.

CBFβ, RUNX1, and RUNX3 regulate overlapping and unique downstream targets in spermatogonia

Allosteric binding of CBFβ to RUNX protein induces a large conformation change in the complex to expose the Runt domain and enable DNA binding for downstream transcriptional regulation [64]. Canonically, RUNX proteins associate with DNA sequences in a CBFβ-dependent manner to act as either transcriptional activators or repressors [49, 68, 69]. Therefore, to identify genomic targets of CBFβ, RUNX1, and RUNX3, we performed CUT&RUN. Initial analysis identified 232 peaks for CBFβ, 2591 peaks for RUNX1, and 503 peaks for RUNX3 (Figure 4A). Several peaks overlapped between CBFβ and each of RUNX1 or RUNX3, and only 20 that overlapped among all three factors.

Figure 4.

Genomic localization of CBFβ, RUNX1, and RUNX3 identify regions of both overlapping and disparate localization. (A) Venn diagram showing the number of differential peaks associated with CBFβ (blue), RUNX1 (green), and RUNX3 (red), along with overlapping peaks in CUT&RUN analyses of spermatogonia. (B–E) Track plots of representative differential peaks with co-enrichment of CBFβ, RUNX1, and RUNX3 (B); co-enrichment of CBFβ and RUNX1 (C); co-enrichment of CBFβ and RUNX3 (D); and enrichment for only CBFβ (E). (F–I) HOMER motif analysis of transcription factor motifs enriched withing peaks sets co-enriched for CBFβ, RUNX1, and RUNX3 (F), co-enriched for CBFβ and RUNX1 (G), co-enriched for CBFβ and RUNX3 (H), or unique to CBFβ only (I). Data represent n = 2 biological replicates per CUT&RUN.

To further investigate whether RUNX1 and RUNX3 shared downstream targets in relation to CBFβ, we categorized the CUT&RUN peaks into four groups: (1) co-enrichment among CBFβ, RUNX1, and RUNX3, (2) co-enrichment with both CBFβ and RUNX1, (3) co-enrichment with both CBFβ and RUNX3, and (4) enrichment for only CBFβ.

First, among the 20 peaks co-enriched for CBFβ, RUNX1, and RUNX3, notable genes included Meiob, Pea15a, Kdm3a, Csnk1g2, Tubb5, and Mad2l1, which are associated with meiotic progression, proliferation, and histone modifications (Figure 4A and B, Supplementary Figure 5A). Interestingly, most of these peaks were found near transcription start sites (TSS), with the exception of Pea15a.

Among peaks co-enriched by CBFβ and RUNX1, we identified 42 shared peaks, many of which are involved in proliferation and apoptosis, including Ralbp1 (Rip1), Trp53, Dgat2, Wrap53 and Dffa (Figure 4A, C, Supplementary Figure 5B). Among peaks co-enriched with CBFβ and RUNX3 were genes primarily associated with differentiation pathways, including Deaf1, Sox6, H2ac12, and Rad23a (Figure 4A, D, Supplementary Figure 5C). Most peaks co-enriched for CBFβ-RUNX1 or CBFβ-RUNX3 were found near or upstream of TSSs.

Finally, we identified peaks enriched for only CBFβ, suggesting that CBFβ may regulate additional pathways independent of RUNX proteins. These targets included Ptma, Wdr77 (p44), Nsun, Fubp1, Rraga, Tmem18, Psmd5, Med1, and Cdk12, which were largely associated with cell cycle regulation and apoptosis (Figure 4A, E, Supplementary Figure 5D).

To gain deeper insights into the biological roles of the identified targets, we performed gene ontology (GO) analysis on all targets discovered through CUT&RUN. GO analysis revealed that CBFβ targets were enriched in developmental processes across multiple organs (Supplementary Figure 6A). Meanwhile, RUNX1-associated peaks were linked to cell fate commitment, development, and differentiation (Supplementary Figure 6B). By contrast, RUNX3 targets were enriched in response to steroid hormones and other developmental pathways (Supplementary Figure 6C). Furthermore, GO analysis of shared peaks among CBFβ, RUNX1, and RUNX3 corroborated enrichment in differentiation, regulatory, and developmental processes (Supplementary Figure 6D). Interestingly, KEGG pathway analysis revealed that many CBFβ-RUNX3 target genes are associated with oocyte meiotic signaling, a pathway that includes key processes that are also essential for terminal differentiation in spermatogenesis (Supplementary Figure 6E).

As many transcription factors act in concert together, we conducted HOMER motif analyses on CUT&RUN peaks to identify potential co-enrichment of other transcription factors. Outcomes of motif analysis revealed that, as a positive control, the Runt motif was enriched within all peak sets that we analyzed (Figure 4F and Supplementary Figure 6F–H). Interestingly, among peaks shared by CBFβ, RUNX1, and RUNX3, we identified enriched motifs for transcription factors with known roles in spermatogenesis, such as Dmrtb1 (Dmrt6), Dmrt1, Rev-erb-α (Nr1d1), Foxp1, and Eomes (Figure 4F). Previous functional studies on DMRTB1 (DMRT6), DMRT1, and NR1D1 indicated critical roles of these transcription factors in facets of the differentiation process within the germline [37, 38, 70–72]. Notably, previous studies indicate that DMRT1 is a critical factor in the switch from mitotic amplification of undifferentiated spermatogonia to initiating the program leading to meiosis [37].

Within peaks co-enriched for CBFβ and RUNX1, motif analysis primarily identified co-enrichment of Gli2 and Gli3 motifs, which are components of the Hedgehog signaling pathway (Figure 4G). Interestingly, analysis of CBFβ and RUNX3 co-enrichment peaks revealed motifs linked to differentiation-related factors, including Mre, Rbpj, and Rarg (Figure 4H). Notably, RARγ plays a critical role in terminal differentiation within the male germline, with Rarg^−/− germ cells stalling at the onset the of differentiation [73]. Finally, motif enrichment among peaks for CBFβ alone identified motifs of cell-cycle regulators, such as Myc (c-Myc), Max, Mycn (n-Myc), Pax7, E2f (E2f3) (Figure 4I). Taken together, motif analysis suggests that while CBFβ and RUNX1/3 locate to common regions enriched for other notable transcription factors, CBFβ and RUNX1 or CBFβ and RUNX3 also locate to genomic regions enriched for motifs of disparate transcription factors.

CBFβ, RUNX1, and RUNX3 can act as transcriptional activators or repressors in undifferentiated spermatogonia

To validate CBFβ- and RUNX-dependent expression of target genes linked to differential peaks found in CUT&RUN, we performed siRNA knockdown of Cbfb, Runx1, or Runx3 and performed qPCR analysis of selected targets at 24 and 48 h post-knockdown. Using this approach, siRNA knockdown efficiencies at 24 h for Cbfb, Runx1, and Runx3 were 64.0%, 82.6%, and 83.0%, respectively and 64.5%, 66.8%, and 88.3% at 48 h, respectively (Figure 5A). These levels are consistent with previous studies [74, 75]. Notably, peak protein knockdown of CBFβ, RUNX1, and RUNX3 were observed at 48 h post-knockdown (Supplementary Figure 2). First, we examined Meiob, Pea15a, and Kdm3a, which were common targets of CBFβ, RUNX1, and RUNX3. Interestingly, Meiob and Pea15a were downregulated with Cbfb and Runx1 siRNA knockdowns but were significantly upregulated with Runx3 siRNA knockdowns (Figure 5B and C and Supplementary Figure 7A). This suggests that the interaction between CBFβ and RUNX1 promotes Meiob and Pea15a activation, while CBFβ and RUNX3 together may suppress their activation. By contrast, Kdm3a was downregulated in all siRNA knockdowns, suggesting overlapping regulation by CBFβ, RUNX1, and RUNX3 (Figure 5C).

Figure 5.

Validation of CUT&RUN target genes of CBFβ, RUNX1, and RUNX3. (A–F) Quantitative PCR validation of CUT&RUN target genes following siRNA knockdown. (A) Expression of Cbfb, Runx1, and Runx3 in their respective siRNA knockdowns compared to siRNA controls. (B and C) Expression of Meiob (B), Kdm3a (C) among Cbfb, Runx1, and Runx3 siRNA knockdowns relative to controls. (D) Expression of co-enriched CBFβ and RUNX1 targets, including Ralbp1 (Rip1), Dffa, and Trp53 (p53). (E) Expression of co-enriched targets within CBFβ and RUNX3, including Deaf1 and Sox6. (F) Expression of CBFβ-specific targets, such as Ptma, Cdk12, Med1, and Wdr77 (p44) in Cbfb siRNA knockdown compared to control. Quantifications in A-F are presented as means ± SEM for n = 3 biologically independent primary undifferentiated spermatogonia cultures. Statistical differences in A-F were indicated as ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. (G) Heatmap representation of select gene targets from previously published Blimp1-Cbfb cKO spatial transcriptomics data, comparing Blimp1-Cbfb cKO and heterozygous control testes at P6 and P21 within the undifferentiated spermatogonia segment.

Next, we analyzed potential target genes with peaks of CBFβ and RUNX1 in CUT&RUN analyses, such as Ralbp1 (Rip1), Dffa, and Trp53 (p53). Interestingly, Rablp1 and Dffa were significantly downregulated with siRNA knockdown of Runx1, whereas Trp53 (p53) was upregulated (Figure 5D and Supplementary Figure 7B). Notably, genes co-enriched within CBFβ and RUNX3 peaks, such as Sox6 and Rad23a, were downregulated, whereas others, including Deaf1 and H2ac12 were upregulated following Runx3 siRNA knockdown (Figure 5E and Supplementary Figure 7C). Finally, we assessed genes with enrichment of only CBFβ peaks in CUT&RUN analysis, including Ptma, Wdr77 (p44), Cdk12, and Med1. All four targets were significantly downregulated with Cbfb siRNA knockdown but remained unchanged with Runx1 and Runx3 knockdowns (Figure 5F and Supplementary Figure 6E and F), suggesting that CBFβ may somehow regulate target genes independent of RUNX proteins. Collectively, these validation studies demonstrate that CBFβ, RUNX1, and RUNX3 regulate both overlapping and distinct targets, with the potential to act as either transcriptional activators or repressors.

Finally, we leveraged data from our previous transcriptomic analysis of undifferentiated spermatogonia from P6 and P21 Cbfb cKO animals to further evaluate whether targets identified in CUT&RUN analyses and validated by qPCR were also differentially expressed in the undifferentiated spermatogonia of Cbfb cKO. Several proliferation and anti-apoptotic genes, including Dffa, Ralbp1 (Rip1), Cdc20, Gli3, and Wdr77 (p44), were downregulated with Cbfb cKO, whereas pro-apoptotic genes, such as Trp53, and Wrap53, were upregulated (Figure 5G). Moreover, transcription factors critical for differentiation identified from motif analysis, including Dmrtb1 (Dmrt6), Dmrt1, Eomes, and Nr1d1, were differentially regulated with Cbfb cKO (Figure 5G). Similarly, differentiation-associated genes like Meiob, Sox6, Rad23a, and Rarg were also dysregulated in Cbfb cKO (Figure 5G). Lastly, dysregulation of cell cycle-related genes, such as Cdk12, Fubp1, E2f1, Ef23, Myc, Max, and Mycn, was observed, suggesting the role of CBFβ in overall cell cycle regulation (Figure 5G). These findings provide in vivo context and validation for the regulation of downstream targets by CBFβ, RUNX1, and RUNX3 identified through CUT&RUN and qPCR analyses.

Differential roles of CBFβ, RUNX1, and RUNX3 in regulating cell cycle progression of undifferentiated spermatogonia

CUT&RUN studies above suggested that the RUNX family may be involved in differentiation. While evaluating more advanced stages of differentiation, like meiosis or spermiogenesis, are not tractable using cultures of undifferentiated spermatogonia, the first step in differentiation, which is the transition from undifferentiated to differentiating spermatogonia, can be readily assessed. To evaluate this, we treated cultures of spermatogonia with RA and utilized flow cytometric analysis to quantify the percentage of cells that induce the canonical differentiating marker cKIT. Outcomes revealed that siRNA knockdown of Cbfb, Runx1, and Runx3 did not significantly change the percentage of cKIT+ spermatogonia with RA induction (Figure 6A–C).

Figure 6.

Roles of CBFβ, RUNX1, and RUNX3 in proliferation of spermatogonia. (A–C) Flow cytometric analysis of cKIT induction following 48 h siRNA knockdown of Cbfb, Runx1, or Runx3 and subsequent treatment with either DMSO control or RA in primary cultures of spermatogonia. (D–F) Flow cytometric analysis of EdU incorporation following 48 h siRNA knockdown of Cbfb, Runx1, or Runx3. (G) The percent difference in EdU+ cells between siRNA-treated and control conditions were plotted. Quantifications in A-G are presented as means ± SEM for n = 3 biologically independent cultures. Statistical differences are denoted as ^*P < 0.05 and not significant (ns) based on two-way ANOVA. Different letters in G indicate significant differences (P < 0.05) based on one-way ANOVA and Tukey HSD test. (H) Schematic representation depicting the predicted functions of CBFβ, RUNX1, and RUNX3 in undifferentiated spermatogonia.

Previously, Cbfb cKO studies indicated that CBFβ is important for the proliferation of undifferentiated spermatogonia [41]. Moreover, our CUT&RUN analysis further revealed that gene targets of the RUNX family including those associated with proliferation. Therefore, we next sought to determine whether CBFβ, RUNX1, and/or RUNX3 regulate cell cycle progression in the undifferentiated spermatogonia population. To test this, we first performed and EdU incorporation assay with siRNA knockdown of Cbfb, Runx1, and Runx3. Knockdown of Cbfb and Runx1 reduced the proportion of EdU-positive undifferentiated spermatogonia by ~20.2% and ~29.0%, respectively (Figure 6D, E, G). In contrast, Runx3 knockdown did not significantly alter EdU incorporation (Figure 6F, G). Together, these results point to the roles of CBFβ and RUNX1, but not RUNX3, in proliferation of undifferentiated spermatogonia population.

Discussion

Our previous studies demonstrated that CBFβ regulates both proliferation, including self-renewal, and differentiation, which are two cellular processes essential for sustaining spermatogenesis. The approach in these previous studies included the generation of germline Cbfb cKO in which the N-terminal region of CBFβ, which is responsible for associations with RUNX proteins, is ablated. Thus, we hypothesized that the dual regulation of proliferation/self-renewal and differentiation may be attributed to CBFβ-RUNX protein associations. Furthermore, we hypothesized that disparate functions of the RUNX proteins may be responsible to intersecting each of self-renewal or differentiation.

Studies of other tissues illustrate that despite commonality in the necessary co-factor CBFβ, each RUNX protein can perform vastly disparate roles within a common cell lineage [49, 76]. For example, in the mammary epithelium, RUNX1 is required for differentiation of luminal cells, while RUNX2 promotes proliferation and self-renewal [76]. Furthermore, while RUNX2 is a master regulator of bone development, studies indicate that RUNX2 promotes differentiation and inhibits chondrocyte proliferation, while RUNX3 promotes proliferation [77–79]. Finally, in the peripheral nervous system, the expression of RUNX1 or RUNX3 define distinct neural circuits [80]. Thus, even though all three RUNX transcription factors share a common consensus motif and predominantly depend on CBFβ to regulate gene expression, each factor can regulate distinct programs even from within a common lineage.

Here, we characterized the expression profiles of CBFβ, RUNX1, and RUNX3 during germline development, particularly focusing within the undifferentiated spermatogonia population because our previous studies indicated the ability of CBFβ to regulate both self-renewal and differentiation was likely confined to the undifferentiated population. Co-immunoprecipitation assays confirmed that CBFβ interacts with both RUNX1 and RUNX3 in the male germline. Co-immunofluorescence staining demonstrated that CBFβ is broadly expressed across most undifferentiated spermatogonia (95%–98%), while RUNX1 and RUNX3 were expressed in subsets of undifferentiated spermatogonia (2%–23% and 14%–66%, respectively). Whether RUNX1+ and RUNX3+ populations overlap is not clear, as our efforts to double-immunostain were inconclusive. Unfortunately, the only commercially available antibodies that presented target specificity by immunostaining in our hands were all raised in rabbit hosts, and direct conjugation of these antibodies with fluorophores did not generate sufficient signal to discern co-localized expression with spermatogonial subtype markers. Interestingly, however, RUNX1 expression was specifically enriched in the G₂/M phases of proliferating undifferentiated spermatogonia, while RUNX3 mirrored the broad expression pattern of CBFβ across all cell cycle stages. It is important to note that these observations are dependent upon antibody-based detection. While we internally validated antibody specificity using targeted siRNA knockdown, technical limitations with primary antibodies are important to consider. Collectively, however, the findings from our studies suggest that RUNX1 and RUNX3 may indicate discrete states within the complexity of the undifferentiated population.

RUNX proteins share a conserved RUNT binding domain to facilitate interactions with CBFβ and allosterically enable transcriptional regulation via DNA binding [42, 43, 45, 48, 49]. Each RUNX protein also possesses a nuclear localization signal critical for nuclear recruitment to then function as transcription factors [76, 81]. Through our CUT&RUN analyses and qPCR validation, we revealed that CBFβ, RUNX1, and RUNX3 shared the regulation of multiple downstream targets. Interestingly, among some of these shared target genes, particularly Meiob which is essential for meiotic progression [82], Runx1 knockdown reduced expression while Runx3 knockdown drastically increased expression (see Figure 5B). Thus, these data suggest that within the same target gene, RUNX1 and RUNX3 can differentially act as either transcriptional activators or repressors, which highlights their potential for disparate functions. The ability of each RUNX protein to modulate activating or repressive roles may be influenced by additional transcriptional regulators that form complexes with each RUNX protein. Indeed, in our studies, we identified clear differences between RUNX1 and RUNX3 in terms of motif enrichment for other transcription factors. Collectively, our CUT&RUN and qPCR validation demonstrate that the differences in the transcriptional regulation by RUNX1 or RUNX3 can be distinct at certain, often critical, loci.

Motif analysis of our CUT&RUN studies provided significant insights into the function of RUNX proteins in the germline. For instance, enrichment of DMRT1 and DMRTB1 (DMRT6) motifs within regions of CBFβ, RUNX1, and RUNX3 co-enrichment suggests that RUNX proteins may co-regulate critical germline genes alongside DMRT1 and DMRTB1. Specifically, DMRT1 is a necessary regulator of SSC self-renewal and also suppresses the meiotic program in progenitor spermatogonia, thus acting as a key modulator in the balance between self-renewal or differentiation [37, 38]. Likewise DMRTB1 functions primarily within progenitor cells and differentiating spermatogonia to ensure the timely initiation of meiosis [70]. Thus, RUNX proteins potentially work in concert with DMRT1 and DMRTB1 to balance fate in the germline. Additionally, our Cbfb cKO data indicate that these transcriptional regulators are CBFβ-dependent. Taken together, our findings suggest that CBFβ, through interactions with the RUNX family, may contribute to maintaining a balance between proliferation and differentiation through both interactions with and modulating the expression of these transcriptional regulators (Figure 6H).

Collectively, our studies revealed that CBFβ, RUNX1, and RUNX3 participate in interwoven functions within the male germline. From within the spermatogonia population, RUNX1 and RUNX3 exhibit a mixture of subtle differences, including expression patterns, genomic localization, and transcriptional regulation. Thus, further functional studies to explore the borders of action for each factor are warranted. However, our findings suggest that the interplay between RUNX1 and RUNX3 may contribute to the capacity of CBFβ to regulate multiple essential functions in the germline.

Methods and materials

Animals

All procedures for the care and use of animals were approved by the Washington State University Animal Care and Use Committee. To isolate germ cells using Fluorescently Activated Cell Sorted (FACS) for CUT&RUN analysis (discussed below), Nanos2-Cre⁺; Rosa26-tdTomato^fl/+ animals were generated by mating Nanos2-Cre⁺ [83] and Rosa26-tdTomato^fl/fl (Jackson Laboratories, stock no. 007909) transgenic mice. Nanos2-Cre mice were backcrossed onto the C57BL/6 J background for a minimum of nine generations prior to mating with other transgenic mice. Wild-type, C57BL/6 J (Jackson Laboratories, stock no. 000664) males were used for other analyses in this study.

Preparation of single-cell suspensions

Cell suspensions were prepared from P6 and P21 testes using trypsin/EDTA digestion. Briefly, detunicated testes were incubated at 37°C for 5–10 min in a solution containing 0.25% trypsin/EDTA (Thermo Fisher Scientific, 25–200-056) and 7 mg/mL deoxyribonuclease I (Sigma-Aldrich, DN25) with gentle agitation.

For adult mice, detunicated testes were incubated in 1 mg/ml of collagenase type IV (Worthington Biochemical, LS004189) and 1 mg/ml of deoxyribonuclease I per testis at 37°C for 10 min with gentle agitation to break up seminiferous tubules. Tubules were then rinsed three times with HBSS on ice to remove interstitial cells. Following washing, single cell suspensions were generated from tubules through a second digestion step with 0.25% trypsin/EDTA and 7 mg/mL deoxyribonuclease I under gentle agitation.

Trypsin digestions were quenched by adding 10% fetal bovine serum (FBS; Life Technologies, 10,438,034), followed by filtration through a 40 μm filter. Cells were then centrifuged at 600 × g for 7 min at 4°C and washed once with HBSS (Thermo Fisher Scientific, 14175–079) prior to proceeding to other analyses.

STO feeder preparation

Mitotically inactivated STO (mSTO) (SNL76/7) cells (ATCC, SCRC-1049) were used as a feeder layer for primary cultures of spermatogonia. STO cells were mitotically inactivated using mitomycin C (Sigma-Aldrich, M4287) at 0.1 mg for 4–5 h. Following treatment, mSTOs cryopreserved according to the manufacturer’s recommendations. To support cultures of spermatogonia, mSTO cells were plated in media containing DMEM, 2 mM glutamine (Fisher Scientific, 25030–081), 50 units/mL penicillin, 50 μg/mL streptomycin, 7% FBS, and 100 μM 2-mercaptoethanol (Sigma-Aldrich, M3148) at a density of 1 × 10⁵ per well in 24-well plates coated with 0.1% gelatin. Feeder STOs were incubated at 37°C and 5% CO₂ at least 24 h prior to use in spermatogonia cultures.

Primary cultures of undifferentiated Spermatogonia

Primary cultures of spermatogonia were established as previous [84]. Briefly, single-cell suspensions were prepared from pooled neonatal (P6–P8) testes. To enrich for germ cells, single-cell suspensions were first processed by gradient separation using a Percoll solution (30% Percoll [Sigma-Aldrich, P1644], 1% FBS, 10 units/mL penicillin, 10 μg/mL streptomycin [Fisher Scientific, 15–140-122], and 1x DPBS [Fisher Scientific, 14–200-075]) and centrifugation (600xg, 4°C, 8 min). Isolated cells were washed twice with HBSS and resuspended in a mouse serum-free medium (mSFM) containing MEMα (Fisher Scientific, 12561–056), 50 units/mL penicillin, 50 μg/mL streptomycin, 0.2% BSA (MP Biomedicals, 08810661), 0.01 mg/mL transferrin (Sigma-Aldrich, T1283), 0.03 μM sodium selenite (Sigma-Aldrich, S5261), 2 mM glutamine (Fisher Scientific, 25030–081), 100 μM 2-mercaptoethanol (Sigma-Aldrich, M3148), 5 μg/mL insulin (Fisher Scientific, 12585–014), 10 mM HEPES, and 60 μM putrescine (Sigma-Aldrich, P5780). Cells were then plated at a density of 2-3 × 10⁶ cells per well onto 0.1% gelatin-coated 6-well plates and incubated at 37°C with 5% CO₂ and 10% O₂ for 18–24 h. After incubation, non-adherent cells were removed and adherent cells were washed once with HBSS and transferred to 6-well plates coated with 23.75 μg/mL laminin (Thermo Fisher Scientific, 23017015) at a density of 1x10⁶ cells per well. After a 45 min incubation, laminin-bound cells were collected by applying 0.5% BSA, centrifuged, and resuspended in mSFM with 20 ng/ml GDNF (Fisher Scientific, 450–10) and 2 ng/ml FGF2 (Thermo Fisher, 100-18B). Finally, cells were plated at 1 × 10⁵ cells per well onto gelatin-coated, 24-well plates seeded with mSTO feeder cells. Spermatogonia are loosely adherent to mSTOs in culture. Therefore, spermatogonia were isolated from mSTOs for analysis by gentle mechanical pipetting.

scRNA-seq analysis

Single-cell transcriptomic data of the adult mouse male germline was obtained from the Gene Expression Omnibus (GEO) under accession numbers GSE124904, GSE121904, GSE109049, and GSE109033. The aligned and demultiplexed transcriptomes were integrated and processed in R using the Seurat package. To ensure data quality, transcriptomes with more than 25% mitochondrial reads and over 500 detected genes per cell were excluded. The combined dataset underwent normalization, scaling, and dimensionality reduction using principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) with default settings and 50 principal components. Germ cells were isolated based on clustering patterns and Ddx4 transcript expression and reassembled for downstream analysis. Germ cell classification was inferred based on marker genes previously characterized by Hermann et al., 2018, with a subset of these markers shown in Supplementary Figure 1A–F.

Immunoprecipitation and Western blot

Spermatogonia from primary cultures were isolated from STOs, washed once with HBSS, and lysed in ice-cold IP buffer (25 mM Tris–HCl pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS, and 1X protease inhibitor cocktail; Roche, 1183617000) at a concentration of 3000 cells per μL. The lysates were centrifuged at 15 000 × g for 15 min at 4°C, and total protein concentration was measured from the supernatant. Approximately 500 μg of cell lysate was pre-cleared with 20 μL of Protein A/G PLUS agarose beads (Santa Cruz Biotechnology, sc-2003) for 1 h at 4°C on a rotator, then centrifuged at 1000 × g for 5 min to collect the pre-cleared lysate. Lysates were then incubated on ice for 1–2 h with 2 μg of either rabbit anti-CBFβ [85] or rabbit anti-IgG (Cell Signaling, 2729), followed by the addition of 20 μL of Protein A/G PLUS agarose beads for overnight incubation at 4°C with rotation. The next day, the IP mixture was centrifuged at 1000 × g for 5 min at 4°C, and the flow-through was collected. Meanwhile, beads were washed four times with ice-cold IP buffer, then resuspended in 40 μL sample buffer (2X Laemmli buffer [Bio-Rad, 1610737] containing 5% 2-mercaptoethanol), heat-denatured for 5 min, and loaded onto a 4–20% SDS polyacrylamide gel (Bio-Rad, 5671094).

For protein extraction from P6 wild-type testes (Supplemental Figure 1F), detunicated testes were homogenized in 10% NP-40 lysis buffer, centrifuged at 15 000 × g for 15 min at 4°C, and combined with an equal volume of sample buffer before heat denaturation as described above.

Western blotting was performed, with nitrocellulose membranes blocked in 5% non-fat dry milk, incubated overnight at 4°C with primary antibodies, and treated with secondary antibodies for 2 h at room temperature (see  Table 1 for antibody details). Blots were visualized using Fujifilm LAS-4000 Image Reader (v. 2.1) after incubation with ECL (Thermo Fisher Scientific, 32,109).

Table 1.

Information pertaining to antibodies used in this study

Antibody	Vendor and Cat. #	Application	Concentration
Rabbit anti-CBFβ	gifted from Ichiro Taniuchi, RIKEN Center for Integrative Medical Sciences	IF, Flow cytometry, CUT&RUN, IP, WB,	IF, Flow: 1:200 CUT&RUN: 1 μg IP: 2 μg WB 1:500
Goat anti-LIN28A	R&D Systems, AF3757	IF, Flow cytometry	IF 1:250 Flow cytometry 1:125
Goat anti-cKIT	R&D Systems, AF1356	IF	1:125
Goat anti-SOX3	R&D System, AF2569	IF	1:250
Goat anti-GFR⍺1	Fisher - AF560	IF	1:250
Rat anti-TRA98	Abcam, ab82527	IF	1:500
Rat anti-Ki67, FITC	Invitrogen, 11569882	IF	1:250
Mouse anti-RUNX1	Santa Cruz Biotechnology, sc-365644	WB	1:500
Mouse anti-RUNX3	Cell Signaling Technology, 13089	WB	1:500
Rabbit anti-GAPDH	Cell Signaling Technology, 2118	WB	1:1000
Normal Rabbit IgG	Cell Signaling Technology, 2729	IP	2 μg
Donkey anti Rabbit Alexa Fluor 488	Invitrogen, A21206	IF, Flow cytometry	1:500
Donkey anti Goat Alexa Fluor 555	Invitrogen, A21432	IF, Flow cytometry	1:500
Donkey anti Rat Alexa Fluor 546	Invitrogen, A11081	IF, Flow cytometry	1:500
Donkey anti Goat Alexa Fluor 647	Invitrogen, A21447	IF	1:500
Donkey anti Rabbit Alexa Fluor 555	Invitrogen, A31572	IF	1:500
Donkey Serum	EMD Millipore, S30	IF	-
Goat anti-rabbit IgG, HRP-linked	Cell Signaling Technology, 2118	WB	1:1000/1:2000
Goat anti-mouse IgG, HRP-linked	Cell Signaling Technology, 7074	WB	1:500/1:1000

Immunofluorescence

Immunofluorescence (IF) staining of testis-cross sections were done as previously described [41]. Briefly, paraformaldehyde-fixed paraffin-embedded (FFPE) blocks of tissue were sectioned at 5 μm, deparaffinized, and prepared as testis cross-sections. Following sodium citrate antigen retrieval, slides were washed three times with 1X PBS at RT and blocked for 1 h in a humidity chamber with blocking buffer (1% serum matching the secondary host species, 0.1% fish skin gelatin, and 1X PBST). To reduce autofluorescence, slides were treated with 0.1% Sudan black/EtOH, followed by series of 1X PBS washes, and incubated overnight at 4°C with primary antibodies in blocking buffer (see  Table 1 for antibody details). The next day, slides were washed in 1X PBS and then incubated with secondary antibodies diluted in blocking buffer for 2 h at RT. Slides were then washed again with 1X PBS, and mounted with Vectashield HardSet containing DAPI (Vector Laboratories, H-1200-10). Secondary-only controls (i.e. omission of the primary antibody) were included in all immunofluorescent staining assays as negative controls. Imaging was performed using a Leica DMi8 inverted fluorescent microscope (Leica-Microsystems; LasX v3.7.5) at 20X or 40X magnification.

Flow cytometry

Single-cell suspensions were generated as above and resuspended in DPBS-S solution containing 1% FBS, 10 mM HEPES (Thermo Fisher Scientific, 15630080), 1 mM sodium pyruvate (Fisher Scientific, 11–360-070), 1 mg/mL glucose (Sigma-Aldrich Inc. G8270), Penicillin (100 units/mL) —Streptomycin (100 μg/mL) (Fisher Scientific, 15–140-122), and 1x PBS (VWR, 16750–078). After a wash in DPBS-S, cells were fixed with 4% paraformaldehyde (PFA) at 37°C for 6 min and then centrifuged at 3000xg for 5 min at 4°C. Permeabilization was performed using 90% methanol with gentle agitation, followed by a 30 min incubation on ice. Cells were then washed twice with DPBS-S and incubated overnight at 4°C with primary antibodies on a rotator (see  Table 1 for antibody details). The next day, cells were washed twice, incubated with secondary antibodies for 2 h at room temperature, and washed three times with DPBS-S. For DNA staining, cells were treated with FxCycle Stain (Thermo Fisher Scientific, F10347) at a concentration of 1 ul per 1x10⁶ cells and RNase A for 30 min at room temperature before analysis using the Attune Nxt Flow Cytometer (Thermo Fisher Scientific). Background fluorescence was determined from unstained controls and single-stained fluorescent controls were used to calculate compensation.

CUT&RUN analysis

Single-cell suspensions of P6 Nanos2-Cre and Rosa26-tdTomato^fl/+ testes were prepared and germ cells isolated using FACS with an SH800 machine (Sony Biotechnology) gating for tdTomato+ cells through the service of FACS Core at Washington State University. Next, nuclei were extracted from live, sorted cells following the manufacturer recommendations and CUT&RUN was performed using the CUTANA ChIC/CUT&RUN Kit (EpiCypher 14–1048, v.3.5). Approximately 1 μg of antibody was used for each CUT&RUN reaction, including a rabbit anti- CBFβ [85], rabbit anti-RUNX1 (Abcam, ab23980), rabbit anti-RUNX3 (Cell Signaling Technology, 9647), H3K4me3 (EpiCypher; positive control), and IgG (EpiCypher; negative control). Genomic libraries were generated using CUTANA CUT&RUN Library Prep Kit (EpiCypher, v.1.3) with slight modification, including 50°C for 60 min for enzyme inactivation during end-repair due to lower input (as recommended), using 1.75X and 1.25X SPRIselect in DNA cleanup and PCR cleanup, respectively, and increasing cycles during PCR indexing to 16 cycles to increase yield. Resulting libraries were sequenced at equal read depth on an Illumina NextSeq2000 (provided as a service from the Laboratory of Biotechnology and Bioanalysis Genomics Core at Washington State University). Fastq files were aligned using bowtie2 (v.2.5.3) and converted to sorted bam files and BED files via Samtools (v.1.21). EaSeq software (v.1.2) was used to call differential peaks-finding, IgG normalization, and trackplot visualization [86]. Gene ontology and KEGG pathway analysis was performed using clusterProfiler (v.4.14.6) in R software (v.4.4.2). Motif Analysis was done using HOMER2 analyze enrichment of known motifs (findMotifsGenome.pl) and position weight matrix (pwm) of known motifs were created through JASPAR database (JASPAR2022) in in R software.

siRNA transfection

Primary cultures of germ cells were transfected using with 75 pmol of ON-TARGETplus siRNA targeting Cbfb (Dharmacon, L-062486-00-0005), Runx1 (Dharmacon, L-048982-00-0005), Runx3 (Dharmacon, L-049822-00-0005), or a non-targeting control (Dharmacon, D-001810-10-05) per 2x10⁵ cell. Transfections were carried out for 24 h at 37°C in 5% CO₂ and 10% O₂ utilizing Lipofectamine 3000 (Fisher Scientific, L3000001) in Opti-MEM (Fisher Scientific, 11–058-021) under feeder-free conditions. For long-term culture following transfection (>24 h), spermatogonia were washed with HBSS and plated onto mSTO feeders in mSFM supplemented with 20 ng/ml GDNF and 2 ng/ml FGF2.

Quantitative PCR

Following siRNA transfection and TRIzol isolation of RNA, 1 μg of RNA was reverse transcribed to cDNA using the QuantaBio qScript cDNA Synthesis Kit (VWR, 101414–098) on the same day as the RNA extraction. Once cDNA was generated, we performed qPCR using Applied Biosystems SYBR Green PCR Master Mix (Fisher Scientific, 43–091-55). Gene expression analysis was done on each siRNA transfection (n = 3 primary cultures per siRNA knockdown) using primers listed in Table 2. Rps2 was used as reference internal control, and qPCR reactions were performed using qTOWER³/G (Analytik Jena) and qPCRSoft software (v.1.1.3.0).

Table 2.

qPCR primer sequences used in this study

Gene	Forward	Reverse
Cbfb	TCGAGAACGAGGAGTTCTTCAGGA	AGGCGTTCTGGAAGCGTGTCT
Cdc20	GATGGACGACATCTGGCAAGTG	GTTGCCAGGATATTGGACTGCC
Cdk12	CTGAATAACAGCGGGCAAAT	AGCTCTGGAGGTCGATACCA
Deaf1	CCTTCCCTTGGCCCACTT	AAGCACACAGCCTCGACATCT
Dffa	TTATCTAGTCAGGATTTGGAGTCTG	CTCTGCCTTCCTTATGTCCCA
H2ac12	AACGACGAGGAGCTCAACAAG	TATTGAAACCTGCTTCACTTGGC
Kdm3a	AACTATTGAGCCACACAGACAGG	ACACATACTCCAAACCCACACC
Med1	GGCTCTCCAATCCTTAGAACAG	GTGAGATAACCAACACTTCCATG
Meiob	ACTACATCCGCTCGCTGTCC	TTATCACACACTCAGCAACCCTG
Pea15a	GACCAACAACATCACCCTTGA	TCTCCAGGAAGCTAAACCAGG
Ptma	ACCAGCCTTCAGAGCGTTTT	AGTTACTGGAATGCTCGGAATAAGA
Rad23a	CTCTGATGTAGAGGGGGAGGT	TCTATAGCCTCCTTCTCCTGCG
Ralbp1 (Rip1)	AAGTGTGAAGGCGTCTACCG	GCTGGCTACAGTGTTAGGCT
Rps2	CTGACTCCCGACCTCTGGAAA	GAGCCTGGGTCCTCTGAACA
Runx1	CAGTCGACTCTCAACGGCT	AGGTAGGTGTGGTACCGAGA
Runx3	AGCACCACAAGCCACTTCAG	GGGAAGGAGCGGTCAAACTG
Sox6	CTCCTGCAGCAACAGATCCA	AGAGGAATCCCTGTTGGGCA
Trp53 (p53)	CACAGCACATGACGGAGGTC	TCCTTCCACCCGGATAAGATG
Wdr77 (p44)	CATCACAGATGGGCTGCAAT	ACAAGGGAGACAGATCCATTCT
Wrap53	AGCCCGAAAGGATGCTGAAC	ACCGCTCGGGTCCAGATCAAA

Antibody validation

Antibodies from the experiments herein have been utilized with in previous studies that employed targeted gene disruption approaches such as knockdown or knockout to demonstrate antibody specificity [52, 85, 87–89]. However, we further validated these antibodies in-house using primary cultures of undifferentiated spermatogonia treated with siRNAs targeting Runx1, Runx3, or Cbfb for 24 and 48 h and collected cell lysates for qPCR and Western blot analysis. Under these conditions, transcript depletion of Runx1 was 63.9% at 24 h and 64.5% at 48 h; Runx3 was 82.6% at 24 h and 66.8% at 48 h; and Cbfb was 82.9% at 24 h and 88.3% at 48 h (also refer to Figure 5A). Outcomes of Western blot analysis revealed a corresponding reduction in bands for RUNX1, RUNX3, or CBFβ (refer to Supplementary Figure 2).

EdU incorporation assay

Cultures of undifferentiated spermatogonia were incubated with 10 mM of EdU (Thermo Fisher Scientific, C10337) for 24 h in mSFM with 20 ng/ml GDNF and 2 ng/ml FGF2 media. After incubation period, spermatogonia were isolated from the STO feeder layer using gentle mechanical pipetting and processed for staining according to the manufacturer’s recommendations.

Quantification and statistical analyses

Quantifications in Figure 3I were performed on wild-type testis cross-sections from at least three randomly selected and non-adjacent sections per animal using images in Fiji (ImageJ v1.0). Quantifications in Figure 2K–M and Figure 3B, D, F, G, were done from three biological replicates using the Attune Nxt Flow Cytometer software (Thermo Fisher Scientific; software v2.7.873). Two-tailed t-tests or one-way ANOVAs with Tukey Honest Significant Difference (HSD) tests were performed using in Prism software (GraphPad v10.4.1). All quantitative data are presented as mean ± standard error of the mean (SEM), with each data point representing an individual animal (biological replicate).

Author contributions

MR and NCL conceived and designed the study. MR, CB, AJB, KMS, and TMH conducted the experiments. MR and NCL analyzed and interpreted the data. MR wrote the original draft. MR, MR, CB, AB, KMS, TMH, and NCL reviewed and edited the manuscript. MR and NCL funded the project. All authors contributed to the study and approved of the final manuscript version.

Conflict of Interest: The authors declare no conflict of interest related the studies herein.

Data availability

Raw sequencing files (.fastq format) as well as processed BED files for each CUT&RUN reaction were deposited within the NCBI Gene Expression Omnibus (GEO) database under accession no. GSE292902. Any additional data is available upon request.