Sine oculis homeobox 1 (Six1) was first recognized as a homologues homeobox gene sine oculis (so) in Drosophila and responsible for compound eyes formation (1). Former research focused on the Six1 function as a transcriptional factor to regulate gene expression, involving in development and metastasis (2). SIX1 expression can be detected in multiple human organs and play a vital role for organ development as well as embryonic development (3). Mice depleted with Six1 showed reduced metanephric mesenchyme during embryonic development and leading to hypoplastic kidneys formation (4). Beside the role in normal tissue development, various cancer cells showed the upregulation of Six1. Overexpression of Six1 increases metastasis and proliferation of cancer cells (5), and knockdown Six1 can impede these functions (6).

Although there have been extensive studies on the function of Six1 in organogenesis and cancer cell proliferation and malignancy, the expression and localization of Six1 in male reproductive organs are still not well-understood. Previous studies suggested that Six1 is localized to the nucleus in various cells (7, 8). However, it is important to note that the size of Six1 protein can affect their localization in the same cell type, as reported by the recent study (9). Further research is needed to fully understand the expression and localization of Six1 in male reproductive organs and its potential role in reproductive biology.

With the development of bulk RNA sequencing and single cell RNA-sequencing (scRNA-Seq) technology, it is possible to explore gene expression profile of different organs and different cells in one organ (10-12). The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases show the significant upregulation of Six1 in 13 cancer samples when compared with testicular germ cell tumor (TGCT) (13). We were intrigued by the localization of Six1 in human and mouse stem cells. As mouse is a widely used animal model for gene investigation, we then further explored Six1 localization in mouse testis and mouse spermatogonial stem cells (mSSCs).

mSSCs are a type of stem cell found in the epithelium of the seminal tubules that play a crucial role in supporting spermatogenesis in mammals. These cells hold great promise for applications in regenerative medicine and stem-cell genomic therapies (14). Researchers have achieved significant progress in the study of germ-line stem (Gs) cells in vitro. In particular, they have been able to purify and culture Gs cells from mouse pups for up to 5 months, and also obtain functional sperm via recipient mouse transplantation with Gs cells derived from adult mice (15, 16). The expression of testis-specific genes was studied using scRNA-Seq data from mouse testis (GSE148032). Specifically, mSSCs derived from mice aged 3 or 7 years old were analyzed to identify differentially expressed genes using the scRNA-Seq database. This analysis helped to identify core factors that are involved in germ cell development (10, 11).

In our study, we investigated the subcellular localization of Six1 in male reproductive organs and mSSCs. Our goal is to reveal the functional of Six1 in reproductive stem cells. Our results showed that Six1 was co-localized with germ cell marker Vasa (DEAD-box helicase 4 or an RNA binding protein with an ATP-dependent RNA helicase) as well as glial cell line-derived neurotrophic factor receptor 1 (Gfra1) in the cytoplasm of mSSCs, but not in the nuc-leus. These finding provides important insight into the role of Six1 in the regulation of male reproductive stem cells.

TCGA and GTEx bioinformatics

Of all the 33 cancer cell lines based on TCGA data, the expression of SIX1 was examined via GEPIA2 (http://gepia2.cancer-pku.cn/) by comparing the RNA-Seq data based on TCGA and GTEx database. After log₂(TPM＋1) scaling, the differential expression was significantly determined between tumor and normal tissues by “limma” with absolute value of log₂FC≥1 and p-value≤0.01. Total 6 reproductive related carcinomas, listed as breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), ovarian serous cystadenocarcinoma (OV), TGCTs, uterine corpus endometrial carcinoma (UCEC), and uterine carcinosarcoma (UCS), were analyzed for SIX1 expression compared with normal controls.

Human Protein Atlas database

The data of protein localization and abundance in human normal and tumor tissues were provided based on IHC examination (Human Protein Atlas, HPA; https://www.proteinatlas.org/) online (17). The protein localization of SIX1 was analyzed in TGCT, BRCA, CESC, OV, UCS vs. their correspondence normal tissues.

Animal use and care

The wild-type mouse of the C57BL/6 strain and DBA/2N strain were purchased from Beijing Charles River Experimental Animal Technology Co., Ltd., and maintained and raised according to the Animal Protection Act of School of Basic Medical Sciences, Southern Medical University. Mouse testis from C57BL/6 and B6D2F1 (C57BL/6N×DBA/2N) F1 (10 weeks, 10w) were dissected, and their testis were fixed for section examination. All the research work were performed according to the Ethics Committee of the Southern Medical University (no. L2016149).

Cell culture of mSSCs

The mSSCs were kindly provided by Prof. Xiaoyang Zhao lab and stored in the Department of Developmental Biology, School of Basic Medical Sciences, Southern Medical University. They were generated from male mice pups (C57BL/6N×DBA/2N) F1 (B6D2F1) at 7∼14 postpartum (dpp) mating female C57BL/6 with male DBA/2 mice, harvested and cultured as previous studies (15, 18).

The mSSCs were seeded in plates with feeder cells generating from mouse embryonic fibroblasts treated with mitomycin C (M5353; Sigma-Aldrich), with 2∼3×10⁵ cells per well in 12-well plates. The mSSC colonies were harvested or passaged every 6 days and culture medium was prepared as described previously (19). The cells are free from mycoplasma contamination.

Immunofluorescence staining

Triple immunostaining was performed as described below. Briefly, testicular paraffin sections were subjected to sequential dehydration, antigen retrieval, then blocked with Goat serum. The sections were then incubated with Goat anti-Gfra1 (AF560; Novus Biologicals) primary antibody overnight at 4℃, and subsequently incubated sections with the second antibody rabbit anti-Goat IgG (HRP) (A21030; Abbkine), washed and treated with FITC (B40953; Thermo Fisher Scientific). The sections were washed again and subsequentially stained with mouse anti-Vasa (AF2030; Novus Biologicals) and rabbit anti-Six1 (10709-1-AP; Proteintech Group) antibodies overnight. The samples were washed and incubated with secondary antibodies Alexa 594-donkey anti-mouse lgG (H＋L) (A21203; Life Technologies) and Alexa 647 donkey anti-rabbit IgG (H＋L) (ab150075; Abcam). Finally, all sections were counterstained the nuclei with DAPI (C0060; Solarbio), washed and sealed and then observed with the ZEISS LSM 880 confocal microscope for image capture.

Immunocytochemistry

The mSSCs growing on glass coverslips in 4-well plate were washed and then fixed with 4% paraformaldehyde for 10 minutes. After washing, the cells were treated with 0.1% Triton X-100, washed with TBS, and incubated with donkey serum. Next, the coverslips were incubated overnight at 4℃ with primary antibodies, either mouse anti-Vasa (ab27591; Abcam) and rabbit anti-Gfra1 (ab186855; Abcam) or mouse anti-Vasa (ab27591) and rabbit anti-Six1 (TD4129; Abmart). In next day, the coverslips were washed with Tris-buffered saline with 0.1% Tween^Ⓡ 20 detergent (TBST; MERCK) and incubated with secondary antibodies, Alexa Fluor^Ⓡ 488 donkey anti-rabbit IgG (H＋L) (ab21206; Invitrogen) and Alexa Fluor^Ⓡ 594 donkey anti-mouse IgG (H＋L) (A21203). The cells were then washed, stained with DAPI for the nucleolus (C0060), and sealed with an anti-fluorescence quencher. Finally, the coverslips were stored at 4℃ for further microscopic examination.

Western blotting analysis

The mSSCs were harvested and suspended in RIPA (P0013B; Beyotime) containing protease inhibitor cocktail (10×) (Roche) and supplemented with 1 mM phenylmethylsulfonyl fluoride (10×) and then resolved on sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The homogenates were then transferred to nitrocellulose membranes. Blots were blocked in TBST containing 5% milk for one hour and incubated overnight with primary antibody at 4℃. Specific proteins were analyzed using commercial antibodies as followings: rabbit anti-Gfra1 (ab186855), rabbit anti-Six1 (10709-1-AP), and rabbit anti-actin (KM9007; Sungene Biotech) antibodies. Blots were then washed and incubated with appropriate secondary antibodies coupled to horseradish peroxidase. Enhanced chemiluminescence peroxidase-labeled anti-mouse or rabbit antibodies (ZSGB-BIO) were used. Blots were again washed and developed with Super ECL Detection Reagent (Gbcbio), then viewed with a Smart Chemi imager (Beijing Sage Creation Science Co., Ltd.).

Protein-protein interactions

Six1 protein-protein network from human and mouse was analyzed by STRING (https://cn.string-db.org/) (20). Total 50 proteins were selected for protein-protein interactions (PPIs) network construction based on minimum required interaction score. PPIs from the previous examination were analyzed and visualized by Cytoscape software (21).

Gene ontology and Kyoto Encyclopedia of Genes and Genomes analysis

The 50 genes that potentially interacted with Six1 via PPI analysis were submitted to Enrichr (https://maayanlab.cloud/Enrichr/). To examine the related biological processes, the gene ontology (GO) term cell component application of ontologies database was utilized. Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were examined for both human and mouse species.

SIX1 expressed in human testicular tissue rather than TGCT samples

To illustrate the expression pattern of SIX1, we consulted various database. According to TCGA data profile ana-lysis, the expression of SIX1 was significantly different in 15 out of 33 human cancer samples (Supplementary Fig. S1). We found that SIX1 was only downregulated in TGCT samples among the 6 reproductive-related cancer samples, whereas the expression of SIX1 was significantly upregulated in BRCA, CESC, OV, UCEC, and UCS samples (Fig. 1A). Furthermore, on HPA database we found that in normal testis SIX1 protein was not detected in the somatic cells, such as peritubular cells, Leydig cells, and Sertoli cells, but was localized in the cytoplasmic and membra-nous of the pachytene spermatocytes and round or early spermatids with moderate intensity under the HPA001893 staining (Fig. 1B).

Figure 1. SIX1 expression in human cancers and localization in testicular germ cell tumor (TGCT) samples. (A) The expression levels of SIX1 in 33 human cancer samples based on GEPIA2 online examination compare with normal tissues. TGCT, breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), ovarian serous cystadenocarcinoma (OV), uterine corpus endometrial carcinoma (UCEC), and uterine carcinosarcoma (UCS). T: tumor, N: normal. Red bar represents tumor, blue bar represents normal. *p<0.05. (B) SIX1 localization in testis and tumor tissues by HPA001893 staining. Black arrows indicated the positive signal for staining and scale bar=10 μm.

Cytoplasmic localization of SIX1 in testis compare with nucleus localization in other tissues

SIX1 expression and localization were examined and collected in the HPA database. We examined reproductive organs and explored the expression of SIX1 in those sections via IHC examination. The results of studying ovary tissue showed that SIX1 could not be detected in the follicle and ovarian stroma cells but was detected in the nucleus of ovarian cancer cells when stained by HPA001893 (Fig. 2A). In the endometrium, SIX1 was undetectable in the stroma and glandular cells, but was moderately expressed in urothelial cancer cells (Fig. 2B). In breast tissue, SIX1 was not detected in adipocytes, glandular cells, or myoepithelial cells, but SIX1 expression in tumor cells was strong and localized in the nucleus (Fig. 2C). However, in cervix samples, SIX1 was highly expressed in the glandular cells and localized in both the cytoplasmic/membranous and nuclear compartments. SIX1 was also highly expressed in the tumor cells of cervical cancer (Fig. 2D).

Figure 2. Sub-localization of SIX1 in cancer samples vs. the controls by HPA001893 staining. (A) SIX1 expression is undetectable in normal ovary. The localization of SIX1 in ovarian serous cystadenocarcinoma is in the nucleus. (B) SIX1 expression is undetectable in normal endometrium, but highly expressed in the uterine corpus endometrial carcinoma and localized in the nucleus. (C) SIX1 expression is undetectable in normal breast tissue, but highly expressed in the breast invasive carcinoma and localized in the nucleus. (D) SIX1 highly expressed in glandular cells of cervix samples, and localized in both cytoplasmic/membranous and nucleus, while SIX1 in cervical squamous cell carcinoma and endocervical adenocarcinoma was high expression in the nucleus. All scale bars=200 μm.

Six1 localization in cytoplasm in mSSCs and expression in germ cells

scRNA-Seq data of mouse testis (GSE148032) showed that the Six1 was highly expressed in SSCs vs. Leydig and Sertoli cells. We found that the expression level of Six1 was elevated in the differentiating spermatogonia and was reduced severely in the differentiated spermatogonia, suggesting that it may play potential role in mSSC differentiation process. Other genes, such as Gfra1, FGFR3, Etv5, and Id4, were also highly expressed in mSSCs according to our scRNA-Seq mining. The expression level of Six1 is comparable with other stem cell markers, such as Sall4, Sohlh2 in mSSCs, and Fgfr3, Sall4, and Nr6a1 in differentiating spermatogonia (Fig. 3A).

Figure 3. Six1 expression and localization in mouse testis. (A) Dotplot illustration for gene expression based on single cell RNA-sequencing data analysis of mouse testis. L: leptotene, Z: zygotene, P: pachytene, D: diplotene, MII: meiosis II, RS: round sperm. (B) Triple immunofluorescent staining of 10-week adult mouse testis by Vasa (red), Gfra1 (green), and Six1 (purple). The nucleus was stained by DAPI (blue). Scale bar=10 μm.

Vasa is a specific marker of germ cells, and Gfra1 is a specific membrane receptor of SSCs (22, 23). Triple immunofluorescent staining in mouse adult testis showed that Gfra1 was detected in Asingle, Apair, and Aalign spermatogonia cells and partially colocalized in Vasa positive C57 and F1 testis tissues. Six1 expression was detected in mSSCs marked with Gfra1 and also partially colocalize with Vasa positive cells in both species (Fig. 3B).

Cytoplasmic localization of Six1 in mSSCs

To guarantee the purity of mSSCs, EGFP positive cells derived from the Oct4-EGFP transgenic mouse were used for cell culture for morphology confirmation (24), indicating the pluripotency state of mSSCs (Fig. 4A). Western blot examination showed that Six1 was detected in cultured mSSCs, and Gfra1 was also detected in the same sample (Fig. 4B). Immunofluorescent staining indicated that the Six1 was localized in the cytoplasm, but not in the nucleus of mSSCs (Fig. 4C).

Figure 4. Six1 expression in mouse spermatogonial stem cells (mSSCs). (A) EGFP positive mSSCs were cultured and photographed with fluorescent microscope. Scale bar=20 μm. (B) Western blot analysis by Tubulin (55 kDa), Gfra1 (55 kDa), and Six1 (37 kDa) in mSSCs. (C) Immunofluore-scent examination of Vasa (red)/Gfra1 (green) and Vasa (red)/Six1 (green) colocalization in mSSCs. The nuclei were stained by DAPI (blue). All pictures were taken with 63× magni-fication. Scale bar=20 μm.

Six1 interacted proteins in human or mouse samples via bioinformatic examination

Analysis of protein-protein network provides the clue to reveal the potential interacting factors. The human SIX1 PPI network showed that proteins of EYA1, SOX2, BMP4, PAX2, FGF8, and MYOD1 showed strong connection with SIX1 (Fig. 5A); while in mouse, Eya1, Sox2, Bmp4, Pax2, Fgf8, and Shh proteins are highly related with Six1 (Fig. 5B).

Figure 5. Protein-protein interaction (PPI) examination of potential proteins that may interact with Six1. (A) Human SIX1 PPI network was analyzed and visualized by STRING and Cytoscape. (B) Mouse Six1 PPI network was analyzed and visualized by STRING and Cytoscape. (C) Gene ontology (GO) cell component term analysis of human SIX1 interacted proteins based on PPI analysis. (D) GO cell component term analysis of mouse Six1 interacted proteins based on PPI analysis. (E) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of human Six1 interacted proteins based on PPI analysis. (F) KEGG analysis of mouse Six1 interacted proteins based on PPI analysis. Asterisk (*) indicates significant difference with p<0.01, and the blue-colored part indicates p<0.05.

GO analysis showed the close relation of Six1 with nucleus and intracellular membrane-bounded organelles in both human and mouse (GO:0005634, GO:0043231) (Fig. 5C, 5D). Of which CtnnB1 involved in several GO terms in human, such as intracellular membrane-bounded organelle (GO:0043231) according to cellular component profile (Supplementary Table S1-S4). KEGG examination indicated the highly relation of Six1 with cancer development and signaling pathways regulating pluripotency of stem cells in both species (Fig. 5E, 5F).

Sub-localization of SIX1 varies in different tissues and cancer cells

Two conserved protein domains exist in the Six family proteins, the Six domain (SD, 110±115 amino acids) and the Six-type homeodomain (HD, 60 amino acids) (1, 25). Former research about SIX1 localization was mainly in the nucleus (26-28), while other reported that SIX1 in hepato-cellular carcinoma (HepG2 HCC) tissue was subcellularly localized in cytoplasm and a small amount in nucleus (29). Localization of Six1 varies due to different protein domains in mouse skeletal muscle cell line C2C12. Full-length CDS of porcine Six1 and individual HD domain were preferentially distributed to the nucleus in both C2C12 and PK15 cells (epithelial morphology from porcine kidney), but the SD domain diffusely detected in both cytoplasm and the nucleus and biased to cytoplasm in C2C12 cells, indicating the importance of SD domain for protein localization in the cytoplasm (9). Thus, we supposed that localization of Six1 in human and mouse testis and mSSCs is probably due to the SD domain of the protein. SD domain was regarded as the necessary domain for protein-protein inter-action. It verified to bind C-terminal region of EYA, a homolog protein related with eyes absent (30). In our study, immunofluorescence of mouse testis and mSSCs provided us an interesting phenomenon of cytoplasm and membrane localization of Six1, the interaction proteins such as Bmp4 and CtnnB1 may contribute to the cytoplasmic localization of Six1 in the male reproductive cells.

Cytoplasmic localization of Six1 in mSSCs

mSSCs maintains spermatogenesis and Gfra1 expressing spermatogonia showed stem cell character and recognized as stem cells. With the two markers Vasa and Gfra1 for mSSCs, we further verified the cytoplasmic localization of Six1 in mouse testis and mSSCs, which also verified the mRNA expression pattern by scRNA-Seq examination.

Specific expression of Six1 in germ cells but not Sertoli cells, myoid cells, macrophage and Leydig cells with familiar expression pattern of Six4 protein, which hints us the potential regulation role of Six1 and Six4 in male reproduction. Interestingly, Six1 and Six4 double knockout mice showed the lack of germ cells, which further demonstrate the synergistic effect of them (31). The higher level of Six1 in normal testis vs. TGCT tumor cells hints the importance of Six1 in male reproduction and the highly expression of Six1 in differentiating mSSCs provide its potential function for stem cell maintenance.

Six1 involvement in stem cell signal pathway

Five proteins showed high relation with Six1 protein based on PPI network examination. Of them, Sox2, Bmp4, and Fgf8 are well-known factors involved in stem cell proliferation in nucleus as transcription factor (Sox2) or in cytoplasm as growth factors (Bmp4 and Fgf8). BMP4 and FGF8 are regarded to localize in cytoplasm or extracellular space (32, 33) while the others showed nucleoplasm localization in various cell types (7). KEGG examination showed the affected signal pathway by Six1 proteins in both stem cell signal pathway as well as cancer development, demonstrating the potential function of Six1 in stem cell proli-feration and intracellular membrane-bounded organelle, provide us the clue to explore potential function of Six1 in male reproduction, especially in mSSCs.

As for the role of Six1 concerning of BMP4 function, cleavage under targets and tagmentation (CUT&Tag) were performed in the clonal mouse embryo cell line C3H10 T1/2 cells and Six1 binding promoter regions were examined. Result showed the directly binding of Six1 in Bmp4 promoter region. Further quantitative RT-PCR showed the decreased Bmp4 in Six1 knockdown cells (34). Meanwhile, SIX1 expression was absent totally in SMAD4 null Müllerian duct epithelium, showed its downstream role involving in BMP4 signal pathway (35). In human embryonic stem cells, endogenous WNT signals lead to the loss of pluripotency and it was recognized as BMP targets and requi-red for mesoderm induction with inducer BMP4 (36).

CTNNB1 belongs to the intracellular membrane-bounded organelle (GO:0043231), also recognized as β-catenin, is a part of protein complex that constitute adherens junctions, is an important part for the Wnt/β-catenin signaling pathway. SIX1 is a factor which affect Wnt/β-catenin pathway. Former research showed its’ proliferation of SSCs that Wnt6 initiated the Wnt/β-catenin signaling pathway and then promotes proliferation of SSCs (37). Cells undergoing epithelial-mesenchymal transition (EMT) showed stem cell characteristics and multiple signaling pathways including Wnt/β-catenin pathway have been involved in this process. SIX1 overexpression driven a partial EMT and activated Wnt signaling and increased nuclear β-catenin level which strongly associated with epithelial stem cell maintenance (38). Meanwhile, dual-luciferase reporter gene assay and chromatin immunoprecipi-tation indicated the two binding regions of β-catenin in SIX1 promoter (39). Consistent with this, silencing β-catenin significantly blocked Six1-induced nuclear localization of β-catenin in cancer cells (40).

According to our investigation, we explored the colocalization of Six1 with Vasa as well as Gfra1 proteins in mouse testis and mSSCs. The reason of the localization shift from cytoplasm in normal tissue of Six1 to nuclear localization in cancer samples is an intriguing issue to explore. Whether and how the localization of Six1 related with growth factor Bmp4 in the cytoplasm or CtnnB1 in the membrane can be investigated in further functional examination by generating knockdown or knockout mSSCs cell lines in the future. Western blot examination of non-phosphorated β-catenin and Smad1/5/9 level can be examined for the interference of Six1 depletion with WNT and BMP4 signal pathways. IF of Six1 in CtnnB1/Bmp4 depletion cells can also be checked to explore the reason of localization shift. Altogether, our research provides a clue for the importance of Six1 in normal male reproductive system and would provide a new insight to investigate mSSCs differentiation and proliferation.

Supplementary data including four tables and one figure can be found with this article online at https://doi.org/10.15283/ijsc23093

We showed great appreciation for animal maintenance and cell line donation from Prof. Xiaoyang Zhao lab. Meanwhile, we provide our great thanks to Dr. Mengze Du for bioinformatics examination, Dr. Xiaoman Wang, Dr. Lin Tang, Mr. Yi Zheng for kindly help, and other helpful lab members from Prof. Zhao lab.

There is no potential conflict of interest to declare.

Conceptualization: MQ, GL, ZL. Data curation: LM, WD, MQ, DC. Formal analysis: GL, ZL. Funding acquisition: MQ, ZL. Investigation: MQ, LM, WD. Methodology: LM, WD. Project administration: MQ, LM. Resources: GL, WD, DC. Software: MQ. Supervision: GL, ZL. Validation: LM, WD. Visualization: MQ, LM. Writing – original draft: MQ, LM. Writing – review and editing: GL, ZL.

This research was supported by Guangdong Basic and Applied Basic Research Foundation (No. 2023A1515030255, No. 2020A1515110045).