Evolutionary Dampening of Wnt Signaling May Contribute to Naturally Ascrotal Testes in Mammals Authors: Simin Chai, Cuijuan Zhong, Zhenhua Zhang, Daiqing Yin, Shixia Xu, Wenhua Ren, Guang Yang, Xuming Zhou Published: 2025 ABSTRACT AbstractThe testicular descent leading to the exteriorization of male gonads is a complex sexual dimorphic process observed in numerous mammals, whereas some species, such as elephants and cetaceans, retain the lifelong ascrotal testes. Despite the implication of Wnt signaling in testicular development and descent, the evolutionary mechanisms underlying ascrotal testes have not been adequately addressed. Here, we examined selection signatures and unique amino acid substitutions in genes of the Wnt signaling pathway. We identified an ascrotal mammal-specific substitution (S406G) in zinc and ring finger 3 (ZNRF3) located within the disheveled (DVL)-interaction region. Functional assays showed that this substitution enhances the affinity of ascrotal ZNRF3 for DVLs and suppresses Wnt3a-induced Wnt signaling. Comparative transcriptomics of the gubernaculum between male fetal rats from wild-type and spontaneous cryptorchid (orl) strains revealed upregulation of Ctnnb1 and Gsk3b in orl rats, along with ascrotal-specific evolutionary changes in regulatory elements. Collectively, these findings suggest that Wnt signaling may be dampened in ascrotal mammals. This study provides insights into the pathogenesis of cryptorchidism in humans and domestic animals.Keywords: Wnt signaling pathway, evolutionary dampening, cryptorchidism, testicular descent, gubernaculum Introduction The occurrence of testicular descent combined with subsequent exteriorization of testes and epididymides is one of the most complicated sexual dimorphic events and phenotypes in the large number of modern mammals (Reny et al. 2023). For example, a pair of testes descends into a definite scrotum in primates, most of rodents, terrestrial artiodactyls and carnivores, while several lineages such as monotremes, elephants, dolphins and eulipotyphlans have testes and epididymides still in high abdomen or inguinal region, namely the ascrotal testes (Hutson et al. 2016). Testicular descent occurs through a two-stage process: (i) The transabdominal phase, characterized by descent from a high position in abdomen near the kidneys into the lower abdominal position. This stage begins between about gestational week (GW) 10 to 15 in human or gestational day (GD) 16 to 21 in rats. Followed by (ii) the inguinoscrotal phase, where testes move through the inguinal canal and into the scrotum around GW 25 to 28 in human or GD 22 to postnatal day 1.5 in rats (Hutson et al. 1997; Fiegel et al. 2010; Hutson et al. 2016). Descent of the testis is a multifactorial process regulated by the development of the gubernaculum. The gubernaculum is considered one of the primary mechanical elements driving this process (Hutson et al. 2016). Embryonically derived from metanephric mesenchyme, the gubernaculum undergoes progressive differentiation and migration into an abdominal wall appendage, containing striated muscle within an extracellular matrix-enriched mesenchymal core (Radhakrishnan et al. 1979; Hutson et al. 2010). Transgenic experiments and gene expression demonstrated that many genes regulate testicular descent via the development of gubernaculum, such as the hormone insulin-like 3 (INSL3), and its receptor, relaxin family peptide receptor 2 (RXFP2) (Kaftanovskaya et al. 2013). However, limited cases of the undescended testes can be associated with defects in INSL3 and RXFP2, indicating that other additionally essential genes and signaling pathways that regulate testicular descent remain to be elucidated. Wnt signaling is essential for the development of gonad and embryonic muscle (Dong et al. 2015). By activating the canonical and non-canonical pathways, Wnt signaling regulates cell proliferation, orientation, and migration (Logan and Nusse 2004). Importantly, recent studies demonstrated that mutation or altered expression of Wnt signaling pathway genes are related to the maldescent of testes. For example, mutations in both Wnt-5a and β-catenin result in markedly underdeveloped gubernaculum in mice, highlighting the necessity of these factors during the “swelling reaction” of gubernaculum; moreover, mutations in human WNT-5A or ROR2 may impede early gubernacular growth, indicated by undescended testes in 70% of individuals with Robinow syndrome, a condition frequently characterized in males by a hypoplastic scrotum and cryptorchidism (Harisis et al. 2013; Zhu et al. 2022). During the development of gubernaculum and formation of the cremaster muscle, conditional knockout of β-catenin gene (Ctnnb1), one of the effectors of Wnt signaling, causes abdominal cryptorchidism and a failure of cremaster muscle myogenesis (Kaftanovskaya et al. 2011). Similarly, gubernaculum-specific knockout of Wnt4 in mice causes unilateral cryptorchidism and fertility defects (Seth et al. 2022). It has been also suggested that INSL3 acts in gubernaculum via Wnt signaling pathway (Bay and Andersson 2011; von Maltzahn et al. 2012). Nevertheless, the overall evolutionary patterns and the potential molecular mechanisms under the diversity of testicular positions in scrotal and ascrotal mammals based on Wnt signaling pathway remain unclear. Focusing on the protein-coding sequences (CDSs) and the potential regulatory sequences (RESs) of Wnt signaling pathway genes, our study attempted to explore the genetic mechanisms under the evolution of testicular descent by using evolutionary analyses, tissue-specific transcriptomic data and in vitro validating assays. We hypothesized and identified the distinct evolutionary trajectories and expression patterns of Wnt signaling pathway genes in gubernaculum of ascrotal and scrotal groups; further we revealed the significant differences in expression regulation capacity of RESs and significant functional changes of Wnt signaling pathway proteins from ascrotal mammals. Our study indicated that Wnt signaling plays a role in mammalian testicular descent and have more insights for the divergent testicular positions in mammals and genetic causes of cryptorchidism. Materials and Methods Sequences Collection and Preparation All of the 167 Wnt signaling pathway genes were gathered from in KEGG PATHWAY Database (https://www.kegg.jp/kegg/pathway.html) under map04310 (Table S1). There are 62 representative mammal species collected from 50 families in 20 orders. Among them, 37 species are scrotal (i.e. the completely descended testes, CDT) and other 25 are ascrotal (including 17 IDT [the incompletely descended testes] and 8 UDT [the undescended testes]) (Table S2). The categories of testicular position were referred to the description in previous studies (Peng et al. 1991; Williams and Hutson 1991; Kleisner et al. 2010; Lovegrove 2014; Hutson et al. 2016). We collected CDSs of Wnt signaling pathway genes by searching gene symbols in OrthoMaM v10 (https://orthomam.mbb.cnrs.fr/) (Scornavacca et al. 2019), Ensembl (https://asia.ensembl.org/index.html) (Cunningham et al. 2022), and NCBI nucleotide database (https://www.ncbi.nlm.nih.gov/nucleotide/). We aligned and trimmed these sequences using MATTF (Katoh and Standley 2013) and Gblocks (Talavera and Castresana 2007) with the default parameters. CDSs were translated into the amino acid sequences by Seaview (Gouy et al. 2010). For the predicted regulatory RESs, we first queried each regulatory elements by looking up 167 Wnt signaling pathway genes with the “Gene Query” function in EpiRegio database (https://epiregio.de/) (Baumgarten et al. 2020). In EpiRegio, the Model score represents the normalized absolute binary logarithm of the P-value of the regression coefficient for the association between a REM and its target gene. The score ranges from 0 to 1, with higher values indicating a stronger predicted contribution of the REM to gene expression. We applied a cutoff of Model score ≥ 0.1 to exclude REMs with only minimal predicted regulatory impact while retaining biologically informative associations. Next, multiple alignments of the specific RESs were retrieved and evaluated the conservation using the UCSC Table Browser (Karolchik et al. 2004) and R package of phastCons100way.UCSC.hg38 which stores phastCons conservation scores from UCSC for the human genome and other 99 vertebrate species (Siepel et al. 2005). Finally, 902 RESs meeting these criteria (Model score ≥ 0.1, length ≥ 40 bp, phastCons ≥ 0.3) were retained for downstream analyses (Table S3). Identification of the Fixed Ascrotal Mammal-Specific Variations FasParser v2.10.0.0 (Sun 2017) and in house Perl scripts were employed to identify the fixed specific amino acid, nucleotide substitutions and indels in ascrotal mammals from multiple alignments of CDSs and RESs. To test whether the identified substitutions were more likely than expected by chance, we applied hypergeometric tests in CAAStools (Barteri et al. 2023). Sites with P-values < 0.05 were considered significant. Evaluation of the Evolutionary Rate and Natural Selection The ratio of dN (nonsynonymous substitution rate) and dS (synonymous substitution rate) is an indicator of selective pressure and evolutionary rate for protein-coding genes, with dN/dS = 1 indicating neutral evolution, dN/dS < 1 purifying selection, and dN/dS > 1 positive selection (Goldman and Yang 1994). The dN/dS value was generated based on the working tree from TimeTree (http://www.timetree.org/) (Kumar et al. 2022) and alignments we obtained before by using the application codeml implemented in PAML (Yang 2007). To identify signals of positive selection and rapid evolution in ascrotal lineages, we specified them as the foreground branches and performed likelihood ratio tests (LRT) comparing the null model with the alternative model. For each gene, the branch-site model provides the overall dN/dS estimate, the log-likelihood values (lnL) under both the null (ma0) and alternative (ma) models, and the 2ΔlnL statistic used in the LRT. In addition, Bayes Empirical Bayes analysis was performed to calculate the posterior probability (PP) of positive selection for each site. In the present study, only sites with PP = 1 were considered to represent robust signals of positive selection. A gene with a dN/dS value greater than one and with the identified positively selected sites is considered as a positively selected gene (PSG). Additionally, the nested branch model detects rapidly evolving genes (REGs) with greater dN/dS values in ascrotal than in scrotal mammals. Multiple hypothesis testing controlling for the false discovery rate was performed, using adjusted P-value < 0.05 as a cutoff. BUSTED in HyPhy was also employed to test for gene-wide episodic positive selection (Murrell et al. 2015). To account for multiple hypothesis testing, the Holm–Bonferroni correction was applied, with adjusted-P < 0.05 considered significant. To identify the accelerated RESs exclusively in ascrotal mammals, we used PhyloAcc (Hu et al. 2019) which restricts the possible substitution rate shift patterns in three nested models—the null model (M0), where the substitution rate in any branch is not allowed to shift to an accelerated rate; the lineage-specific model (M1), where the substitution rates of ascrotal mammals are allowed to shift to an accelerated rate; and the full model (M2), where the substitution rate of any branch is allowed to shift. As recommended by Hu et al. (2019), the present study considered RESs with logBF1 > 1 and logBF2 > 1, and the PP for accelerated evolution under M2 > 0.6 in at least 60% of ascrotal species. Functional Changes Verification by Dual-Luciferase Reporter Assay and Overlaps With the Known Enhancers To verify the potential regulatory function of RESs with the ascrotal mammal-specific changes, dual-luciferase reporter assay was employed. The human kidney epithelial cell line 293T is used as the tool cell. First, 30% of the cells was planted in each well of a 24-well plate and grow overnight. Next, cells were transfected with plasmids according to the ratio of plasmid (mass/μg): PEI (volume/μL) of 1:3. Since the bio-fluorescence intensity of sea cucumber luciferase catalyzed by the substrate is stronger than that of firefly luciferase, the ratio of sea cucumber luciferase plasmid and firefly luciferase plasmid is 1:10. Medium was replaced after 6 to 8 h, followed by 30 to 48-h culture. The lysate for lysing the cells and luciferase substrate is from the Dual-Luciferase Reporter Assay Kit (Cat. No.: DL101) of Nanjing Novozan Company. Subsequently, we use the BIOTEK SYNERGY/H1 microplate reader for detection, setting the parameters as: vibration plate 5 s, gain 120, and detection time 3 s. We identified the overlap between RESs and known enhancer by querying Human Enhancer Disease Database (Wang et al. 2018) and WashU Epigenome Browser (http://epigenomegateway.wustl.edu/) (Li et al. 2022). The peak information is from (i) mesenchymal stem cell DNase-seq signal (ENCFF387STL), (ii) mesenchymal stem cell DNase-seq signal (ENCFF594UFP), (iii) testis CTCF (ENCFF686GDH), (iv) testis CTCF (ENCFF245HIM), v) E021 iPS DF 6.9 Cells AuxiliaryHMM (E021_18_core_K27ac), respectively. Molecular Docking Simulation Canonical protein sequences of zinc and ring finger 3 (ZNRF3), DVL1, DVL2, and DVL3 were obtained from NCBI and Uniprot. AlphaFold was used to predict the three-dimensional structure of proteins (https://golgi.sandbox.google.com/). GRAMM was used to simulate molecular docking (https://gramm.compbio.ku.edu/request). Finally, PDBePISA was used to analyze protein binding energy (ΔG, kcal/mol) (https://www.ebi.ac.uk/pdbe/pisa/). Transcriptomics Analysis The gene transcription data of gubernaculum from wild type and cryptorchid male rats in transabdominal phase (GD18-20) were retrieved from published transcriptome data (GSE7755) (Barthold et al. 2008). The transcripts per million reads normalized expression counts representing the expression measurement were calculated using GEO2R with the default parameters (Love et al. 2014). Heatmap and Veen diagrams were drawn using TBtools (Chen et al. 2020) and online tool from https://bioinformatics.psb.ugent.be/webtools/Venn/. Expression Constructs The plasmids, named pIRES2-EGFP (using XhoI and EcoRI restriction sites), used in this study were synthesized by Gentlegen Co. Ltd. (Suzhou, China), containing CDSs of gene ZNRF3 from human, dolphin and elephant (XM_049866947, XM_019950445, and NM_001206998). Codon optimization was performed on the plasmids, and a 3× Flag epitope tag was added to the C-terminal. Mutated plasmids were created using QuickMutationPlus Site-Directed Mutagenesis KitM (Beyotime Biotechnology) (Table S4). After the mutagenesis, the entire coding region was sequenced to confirm the presence of the desired mutation and the absence of additional mutations. Details of plasmids are available on request. Cell Culture and Transfection HEK293T cells were cultured in DMEM (Gibco) media supplemented with 1% penicillin/streptomycin and 10% FBS at 37 °C and 5% CO2. Transfections were performed using the jetPRIME transfection reagent (Polyplus-transfection) according to the manufacturer's protocol. HeLa cells were cultured under the same conditions as HEK293T cells. For transfection, Lipofectamine 3000 (Lipo3000) reagent (Thermo Fisher Scientific) was used following the manufacturer's protocol. Immunofluorescence Cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100, at 48 h post-transfection. Cells were blocked with 5% bovine serum albumin to reduce nonspecific antibody binding, and the primary antibody against the DYKDDDDK (Flag) tag (Proteintech) was diluted 1:200 and incubated with the cells for 1 h at room temperature. Detection was performed using an Alexa Fluor 488-conjugated secondary antibody, and DAPI staining was used to show the cell nuclei. Confocal microscopy was used to capture the localization images of ZNRF3 within 293T cells. Images were taken using specific filters for Alexa Fluor 488 (ZNRF3) and DAPI (nuclei). Western Blots For protein half-life assay, after 24 h of transfection, cycloheximide (CHX) was added to the culture medium at a final concentration of 100 µg/mL to inhibit protein synthesis. Cells were harvested at 0, 3, 6, and 9 h post-CHX treatment. Total cellular proteins were extracted using RIPA lysis buffer containing protease inhibitors. Equal amounts of protein (20 µg) were separated on 10% SDS-PAGE gels and transferred to PVDF membranes. Membranes were blocked with 5% nonfat milk in TBST (Tris-buffered saline with 0.1% Tween-20) for 2 h at room temperature. The membranes were then incubated with the following primary antibodies: Anti-FLAG (1:1,000, Sigma) for ZNRF3 detection. Anti-β-actin (1:10,000, Proteintech) as a loading control. After primary antibody incubation, membranes were washed with TBST and incubated with the appropriate secondary antibodies: HRP-conjugated mouse secondary antibody (1:10,000, Proteintech) for anti-FLAG detection. β-Actin detection did not require a secondary antibody. Membranes were incubated at room temperature for 1.5 h, followed by washing with TBST. Protein bands were visualized using an enhanced chemiluminescence detection kit and imaged using a ChemiDoc Imaging System. TOPFlash Reporter Assay To measure Wnt signaling activity, we used the SuperTOPFlash-CP luciferase reporter plasmid (Beyotime Biotechnology), which contains eight TCF/LEF binding sites driving luciferase expression. As a negative control, the SuperFOPFlash-CP plasmid containing mutant TCF/LEF binding sites, was used. Cells were co-transfected with Renilla luciferase plasmid to normalize transfection efficiency. After 24 to 48 h of transfection, the cells were treated with Wnt3a Surrogate Protein (MedChemExpress) or control treatments. Luciferase activity was measured using a dual-luciferase reporter assay system, and the TOPFlash activity was normalized to Renilla luciferase. The relative luciferase activity between TOPFlash and FOPFlash was calculated to determine the Wnt signaling activity. Cell Scratch Assay To evaluate the effect of ZNRF3 on cell migration, a cell scratch assay was performed using HeLa cells transfected with plasmids expressing ZNRF3. The transfected cells were cultured in a 12-well plate until confluent, and a scratch was made using a pipette tip. After scratching, the cells were treated with Wnt3a (100 ng/mL) to activate the Wnt signaling pathway. The wound area was imaged at 0 and 48 h post-scratch. Cell migration was quantified as the percentage of the wound area covered by cells over time. Data Analysis Statistical calculations were performed using the GraphPad Prism. The D’Agostino and Pearson normality and Shapiro–Wilk tests were used to check the data distribution. If data were normally distributed, N-way ANOVA (α = 0.05) was used with correction for multiple comparisons using Holm–Sidak multiple post hoc test. If nonparametric, the Kruskal–Wallis test was used with correction for multiple comparisons using Dunn–Sidak (α = 0.05). All experiments were independently repeated at least three times. All data were expressed as the mean ± SD. The labels “ns,” “*,” “**,” “***,” and “****” indicate not significant, P < 0.05, P < 0.01, P < 0.001, and P < 0.0001, respectively. Results Evolutionary Pattern of Wnt Signaling Pathway in Ascrotal Mammals Unique Amino Acid Substitutions The alignment and comparison of 82,588 amino acid residues in Wnt signaling pathway proteins identified one ascrotal mammal change in ZNRF3, GPC4, and APC, respectively. Specifically, most ascrotal mammals have a glycine (G) at position 406 of ZNRF3 (P = 7.16 × 10−15), a glutamic acid (E) at position 328 of GPC4 (P = 1.56 × 10−16) and a valine (V) at position 2,209 of APC (P = 3.24 × 10−16), whereas the scrotal mammals typically have a serine (S), an aspartic acid (D) and an isoleucine (I) at these positions, separately (Fig. 1a). All three affected residues are intriguingly situated within crucial domains that could potentially impact protein function, yet they do not coincide with any known pathogenic mutations in humans. The ascrotal mammal-specific evolutionary changes. a) The graph shows three unique amino acid substitutions of ZNRF3 S406G, GPC4 D328E, and APC I2209V in ascrotal mammals, with changes in residue volume are suited in important functional domains in each protein. The ascrotal species (UDT and IDT) were highlighted. SP, signal peptide; TM, transmembrane; DIR, Dvl-interaction region; R, ring E3 ligase domain. b) Segment deletion shared by ascrotal mammals in chr10:113164550-113166049 (hg38), which is predicted to regulate the expression of TCF7L2. c) and d) shows RESs with specific changes in ascrotal mammal identified in the present study were overlapped with the known regulatory elements. Positive Selected and REGs Of the tested Wnt signaling pathway genes, 48 genes showed significantly accelerated evolutionary rate in ascrotal species compared to the background of the scrotal counterparts (branch model, adjusted-P < 0.05, Table S5). Further, by employing the BUSTED method in HyPhy, we found 53 genes were significantly positively selected (adjusted-P < 0.05) in ascrotal mammals (Table S6). Besides, 27 positively selected genes and a number of specific codons were also detected with episodes of positive selection by PAML (branch-site model, adjusted-P < 0.05, Table S7 and Fig. S1). RESs With Ascrotal Mammal-Specific Changes Based on Bayes factor thresholds, we identified eight and four RESs showing significantly accelerated rates of evolution in ascrotal mammals and in UDT mammals, respectively (Figs. S4 and S5, Tables S8 and S9). Additionally, several nucleotide substitutions and segment indels were exclusively shared in RESs of several ascrotal branches (Figs. S2 and S3), leading to the speculation that the expression of predicted target Wnt signaling pathway genes would be affected convergently. Thereinto, some of them are overlapped with the known regulatory elements related to Wnt signaling genes, such as TCF7L2 (Fig. 1b-1d). Ascrotal Mammal-Specific Changes in Wnt Signaling Pathway Alter Protein Function and Regulatory Activities Scrotal ZNRF3 G406S Prolongs Protein Half-Life We selected three species to represent three different testicular positions, i.e. UDT elephant, IDT dolphin and CDT human. Immunofluorescence results indicated predominant cell membrane localization of ZNRF3 proteins (Fig. 2a), implying unaffected subcellular distribution of ZNRF3 across species with varying testicular positions. In our analysis of the degradation rates of ZNRF3 proteins treated with CHX, we noted a significant acceleration in the degradation of UDT elephant ZNRF3 compared to that of dolphin and human ZNRF3. Furthermore, the degradation rate of elephant and dolphin ZNRF3 showed a marked decrease after the G406S mutation (Fig. 2b-2c). The results suggest that ZNRF3 stability is the least in UDT elephant, and scrotal G406S has the ability to restore the stability of the ZNRF3 protein itself. Under identical experimental conditions, human ZNRF3 S406G mutant was not detectable in Western Blot assay. We speculated that the S406G mutation may alter or disrupt the antibody recognition epitope, thereby preventing antibody binding and resulting in the absence of a detectable protein band. Subcellular localization and stability of ZNRF3. a) Subcellular localization of ZNRF3 from UDT elephant, IDT dolphin, and CDT human was on the cell membrane. Scale = 10 μm. b) ZNRF3 protein stability in Hela cells under CHX treatment. ZNRF3 constructs from UDT elephant, IDT dolphin and CDT human, and mutant G406S were analyzed via Western Blot at 0, 3, 6, and 9 h post-CHX treatment. β-Actin served as a loading control. c) The line graph shows relative ZNRF3 levels (100% of 0 h). Data are displayed as the mean ± SD (n = 3). Enhanced Affinity of Ascrotal ZNRF3 Along With DVLs To explore the molecular mechanisms underlying the ascrotal mammal-specific substitution of ZNRF3 S406G in the disheveled (DVL)-interaction regions (DIRs), we initially compared the interactions between wild type and mutant ZNRF3 with DVL1-3. These adaptors bind to the DIR of ZNRF3 and are crucial for its inhibitory function. It was found that the total binding energy of the docking complex between ascrotal ZNRF3 and DVLs (i.e. IDT dolphin or/and UDT elephant) was observed to be lower than that of scrotal ZNRF3 and DVLs, suggesting a higher affinity (Fig. 3a-3c). For example, the affinity of human ZNRF3 and DVL2 (−31.2 kcal/mol) is lower than the affinity of ZNRF3 and DVL2 from dolphin and elephant, respectively (−39.3 kcal/mol and −43.4 kcal/mol). Further, we observed that the ascrotal mutation of S406G introduced a lower affinity, representing higher binding energies, between ZNRF3 and DVLs to some extent. On the contrary, the scrotal mutation of G406S showed an opposite trend. Functional modifications in ZNRF3 induced by species with different testicular positions and amino acid changes at residue 406. a) to c) Presents the total biding energy (ΔG) of molecular docking between ZNRF3- and DVL1, DVL2, and DVL3, with lower ΔG values indicating stronger binding affinity. d) S406G mutation enhances the inhibitory effect of ZNRF3 on Wnt signaling pathway. Relative luciferase activity was measured using the TOPFlash reporter assay. The EV group represents cells transfected with an empty vector without Wnt3a treatment, while the Control group refers to empty vector-transfected cells treated with Wnt3a (60 ng/mL). All other groups were treated with Wnt3a. Data are displayed as mean ± SD (n > 3). (****P < 0.0001; ***P < 0.001; ns, not significant). e) Representative images of scratch assays performed on Hela cells expressing ZNRF3 wild-type and mutant variants from human (S406G), dolphin (G406S), and elephant (G406S) at 0 and 48 h. All groups, except the empty vector (EV) group, were treated with 100 ng/mL Wnt3a. Dashed lines indicate the wound edges, and cell migration was quantified by measuring wound closure at 48 h. Scale bar = 100 μm. f) Quantification of cell migration percentage (mean ± SD) across groups (one-way ANOVA, n = 3; *P < 0.05, **P < 0.01, ****P < 0.0001, ns: not significant). Ascrotal ZNRF3 S406G Increases Potency in Inhibiting Wnt Signaling Pathway and Cell Migration In order to compare regulatory effects of ZNRF3 from different representative species on Wnt signaling pathway, TOPFlash reporter assays were employed and showed that the expression of ZNRF3 significantly inhibited Wnt signaling which was induced by Wnt3a; Importantly, the UDT elephant ZNRF3 exhibits the most potent inhibitory effect on the Wnt signaling pathway, with humans showing intermediate inhibition and dolphins displaying the weakest effect (Fig. 3d). Further, it was showed that the inhibitory activity of human ZNRF3 was significantly enhanced after mutation of site 406 from S-to-G, while no significant change was observed based on the G-to-S mutation. These results indicated that ZNRF3 of UDT elephant showed the most inhibition, and ZNRF3 S406G strengthens the inhibitory effect on Wnt signaling pathway. Given the facts that Wnt pathway plays an important role in regulating cell motility and adhesion, and the observation that gubernaculum swells and migrates during the testicular descent, we next investigated if and how ascrotal ZNRF3 and S406G would affect cell migration by regulating the Wnt signaling pathway. Wnt3a treatment significantly enhanced cell migration in control groups, validating Wnt pathway activation (Fig. 3e-3f). Furthermore, cell migration was significantly reduced in the ZNRF3 overexpression groups compared to the control group, irrespective of the specific type of ZNRF3. In particular, the migration inhibition in human ZNRF3 was particularly robust, and this effect was amplified following the S406G mutation. Combined with the results of TOPFlash reporter assays, it was revealed that ZNRF3 had a significant effect on cell migration by inhibiting Wnt signaling pathway. Additionally, the ascrotal mammal-specific S406G mutation markedly increased ZNRF3 activity, leading to the suppression of cell migration through the Wnt signaling pathway. RESs With Ascrotal Mammal-Specific Changes Show Distinct Regulatory Activities To verify the expression regulation potential induced by the ascrotal species-specific RES variations, the dual-luciferase reporter gene assays were performed. First, compared to the blank control group, all the five RESs in ascrotal species significantly promoted the luciferase activity of the reporter gene, suggesting potentials for their transcriptional regulation activities (Fig. 4). Additionally, the constructs containing the ascrotal species-specific RESs showed higher transcriptional enhancer activities for the target gene WNT3, WNT9B, and WNT10B than did the other constructs containing the mouse RESs, while it showed lower transcriptional enhancer activities for FZD5 and TCF7. Dual-luciferase activity assay revealed potential expression effects of RESs with ascrotal mammal-specific changes. Data are mean ± standard deviation (mean ± SD) of three independent experiments analyzed by a Student's t-test across groups (*P < 0.05, **P < 0.01, ****P < 0.0001, ns: not significant). Different Gene Expression Patterns of Gubernaculum in Cryptorchid and Scrotal Rats To investigate mechanisms by which Wnt signaling pathway contributed to the testicular descent at the transcript level, we first extracted and compared the gene expression of the gubernaculum between the male fetal rats of wild type (wt) and spontaneous cryptorchid (orl) strains (Barthold et al. 2008). We detected differential expression in 30, 31, and 50 Wnt signaling pathway genes during the transabdominal phase of rat testicular descent (i.e. at GD18-20; Fig. 5a). Among them, nine upregulated (Tbl1xr1, Sfrp2, Ppp3ca, Btrc, Prickle2, Gsk3b, Ctnnb1, Frzb, and Smad3) and five downregulated (Camk2d, Csnk2b, Prkacb, Skp1, and Myc) genes were shared throughout the whole transabdominal phase in cryptorchid orl rats (Fig. 5b). Interestingly, seven out of the nine upregulated genes could serve as the negative regulators of Wnt signaling pathway, while all of the five downregulated genes usually function as the positive regulators within Wnt signaling pathway (Fig. 5b). In addition, several muscle development-related genes, such as Bmp2, Spag9, Paxbp1, and Chd2, were also identified as the differentially expressed genes (Fig. S6). Together, these findings indicate that both Wnt signaling and muscle developmental processes undergo substantial transcriptional reorganization during testicular descent. Different expression patterns of Wnt signaling pathway genes between wt scrotal and orl cryptorchid rats. a) Differential expressed Wnt signaling pathway genes at GD18-20 between wt scrotal and orl cryptorchid rats. b) Differential expressed Wnt signaling pathway genes throughout the transabdominal phase. Importantly, a subset of the differentially expressed genes also overlapped with gene sets showing distinct evolutionary signatures (e.g. positive selection, rapid evolution, or regulatory sequence divergence) in ascrotal lineages (Fig. S7). For example, Ctnnb1 and Gsk3b were upregulated in orl rats and were also associated with ascrotal-specific evolutionary changes in regulatory elements. This suggests a potential mechanistic link in which evolutionary modifications may drive altered gene expression patterns, contributing to the development of ascrotal testes. Moreover, in some cases, both coding and non-coding regions of the same gene exhibited evolutionary signals. For instance, Csnk2b, a gene with accelerated evolution in regulatory elements and rapid evolution in coding regions, was also differentially expressed. Such convergence of transcriptional and evolutionary evidence highlights these genes as strong candidates for functional involvement in the regulation of testicular descent (Fig. S7). Nonetheless, it should be acknowledged that further experimental validation will be essential to verify the causal impact of these evolutionary and expression changes of the candidate genes on the process of testicular descent. Discussion Sexually dimorphic position of gonads is exclusive to mammals and is present in the majority of them. Male individuals experience the externalization of testes through a developmental process of testicular descent (Hutson et al. 2009). Descent of the testis depends on a complex interaction between the gubernaculum and inherited genetic factors (Hutson et al. 2016). Formation and development of the gubernaculum has been suggested to depend on the conserved Wnt signaling (Zimmermann et al. 1999; Harisis et al. 2013), while the overall evolutionary impacts and molecular mechanisms of Wnt signaling pathway on the evolution and divergence of testicular positions in mammals are unclear. Examining the selection, fixed specific variants of CDSs and RESs, and the gene expression of the Wnt signaling pathway, this study identified significantly different evolutionary patterns between ascrotal and scrotal species. These results underscore the importance of Wnt signaling in the divergence of testicular positions in mammals. Our findings establish that suppression of core Wnt pathway components through mutation- or dosage-dependent mechanisms drives ascrotal testes evolution. In this study, we identified several Wnt signaling pathway genes, known negative regulators of this pathway, that exhibit unique evolutionary signatures in natural ascrotal mammal lineages, including specific amino acid substitutions and evidence of accelerated evolution. Among these, three fixed amino acid substitutions in Wnt signaling proteins distinguished ascrotal from scrotal mammals (Figs 1a and 6). Statistical support from hypergeometric tests (ZNRF3 S406G, P = 7.16 × 10−15; GPC4 D328E, P = 1.56 × 10−16; APC I2209 V, P = 3.24 × 10−16) argues against their occurrence by chance alone. Nevertheless, their evolutionary interpretations require caution. The patterns observed in APC and GPC4 may reflect lineage-specific or partially convergent processes with limited relevance to testicular descent, given their incomplete overlap with ascrotal evolution. By contrast, ZNRF3 S406G shows a more compelling convergent distribution across mammals, suggesting that its evolutionary trajectory may be functionally linked to ascrotal conditions in a subset of mammalian lineages. The evolution of Wnt signaling pathway in the ascrotal mammals. Genes with positive selection (identified by PAML) and differential expression, proteins with the specific amino acid substitution, and RESs with specific genetic changes were shown. Together with ring finger protein 43 (RNF43), zinc and ring finger 3(ZNRF3 ) plays as a negative feedback regulator of Wnt signaling by promoting the turnover of Wnt coreceptors frizzled and LDL receptor related protein 6 (LRP6) to suppress the β-catenin signaling initiated by endogenous Wnt proteins (Hao et al. 2012). Physical interaction between ZNRF3 and a positive Wnt regulator DVL through DVL-interaction regions (DIRs) is essential for the Wnt inhibitory activity of ZNRF3/RNF43 (Jiang et al. 2015). Notably, S406G is located in DIRs (from amino acid 346 to 528) of ZNRF3 (Fig. 1a). Thus, these findings suggested that the substitution of ZNRF3 S406G in ascrotal mammals might affect Wnt signaling and subsequently introduce a negative influence on the evolution of testicular position. To further verify this hypothesis, our in vitro functional assays revealed that ZNRF3 and the cryptorchid-specific amino acid substitution of S406G had a significant effect on cell migration by inhibiting Wnt signaling pathway (Fig. 3). Although HEK293T cells are not gubernaculum-derived, transcriptomic data from wild type and cryptorchid rat gubernaculum indicate that core Wnt signaling pathway components are expressed and active during testicular descent (Barthold et al. 2008), suggesting that the effects observed in vitro likely reflect meaningful trends in gubernaculum development. Besides, the expression of FZD5, one of the positive regulators of Wnt signaling, is suggested to be negatively influenced by the unique sequence changes in ascrotal mammals including the elephant and the manatee (Fig. 4), indicating an evolutionarily dampening Wnt signaling shared by them. TCF7 (Transcription Factor 7) and TCF7L2 (Transcription Factor 7 Like 2) are members of the T cell factor/lymphoid enhancer-binding factor (TCF/LEF) family of transcriptional effectors playing bimodal functions (Tang et al. 2008). It has been observed that the expression of TCF7L2 in gubernaculum upregulated significantly during the exposure to dihydrotestosterone or INSL3 (Barthold et al. 2013), suggesting TCF7L2 is a common target responding for the key stimulations in the process of the testicular descent. Through the present study, it was discovered that TCF7L2 is a rapidly evolving and PSG (Tables S5 and S7). Furthermore, the analysis revealed specific changes in the regulatory sequence of TCF7 unique to ascrotal mammals, resulting in a significant reduction in reporter gene expression according to dual-luciferase reporter assay (Fig. 4 and Fig. S2). Thus, the decreased expression of TCF7 might reflect a slow reaction to respond for upstream stimulations such as INSL3 in ascrotal mammals. These lines of evidence provide support for a dosage-dependent manner, at least, of Wnt signaling pathway being involved in the evolution of divergent testicular positions. Besides, it has been suggested that the overexpression of Wnt signaling effectors led to varying degrees of genital malformation including cryptorchidism (Jordan et al. 2003; Basu et al. 2018). Here, we identified a number of fixed variants and indels in the RESs specifically in ascrotal mammals, and verified the significant expression impacts on their potential target genes, such as WNT9B (Fig. 4). As a result, we speculate that members of Wnt protein family in ascrotal mammals might antagonize the signaling pathway and the development of the scrotal testes to some extent by the subtle modulation of expression. We also noticed that the β-catenin destruction complex might play an important role in the development of the ascrotal testes. APC, AXIN1, AXIN2, CKIα and GSK3β are involved in the destruction complex to cooperatively trigger the degradation of β-catenin. The mutation or knockout of their conserved coding genes (e.g. APC and AXIN1) resulted in the perturbation of Wnt signaling and subsequent congenital anomalies related to the genital and urinary systems (Katoh and Katoh 2007; Zhou et al. 2015). We previously reported a 36-amino acid fragment deletion of AXIN1 shared by UDT mammals, being suggested to be associated with the testicular descent (Chai et al. 2021); here we identified more specific evolutionary footprints from other scaffold proteins of β-catenin destruction complex in ascrotal species. For instance, APC was detected to experience the rapid evolution and possess a specific amino acid substitution (I2209V) in ascrotal mammals (Fig. 1a and Table S5); Gsk3b was differentially expressed in the gubernaculum of the transabdominal phase between cryptorchid orl and scrotal wild-type rats and one of the possible RESs of it was recognized rapid evolution signal (Fig. S4). These genetic innovations might introduce changes of protein function, interaction or the altered expression of the core complex elements to facilitate the rapid destabilization and degeneration of β-catenin in time of testicular descent. As a possible downstream pathway of INSL3, the Wnt signaling pathway with changes in gene expression was involved in the differentiation and growth of the gubernaculum, whose development is from a mesenchymal tissue to a muscle-like structure (Seiler 1817; Johnson et al. 2010; Hadziselimovic 2012). Consistently, we found that the altered expression patterns of muscle- and cell cycle-related genes during the first stage of the descent of the testis in the natural cryptorchid rats compared to the scrotal wild-type counterparts (Fig. S6). Among them, the expression of MYC, which could be activated by Wnt signaling and participates in the regulation of cell cycle and the proliferation of mesenchymal cells (Zhang et al. 2019), was significantly downregulated in the cryptorchid rats, suggesting to be possibly related to the interruption of the testicular descent by affecting the development of gubernaculum. Moreover, β-catenin is another signaling required in early myogenic differentiation, the overexpression of it has been considered to be inhibitory for myogenic determination (Tanaka et al. 2014). Consistently, the elevated expression of Ctnnb1 in the whole-course of the first stage of the testicular descent in cryptorchid rat gubernaculum and the specific indel in the regulatory sequence targeting this gene implying the possible inhibiting effect on the gubernaculum development (Fig. 5, Fig. S2). Conclusion The evolutionary and expression changes of the Wnt signaling pathway, involved in developmental processes like organogenesis and cell differentiation, are linked to cryptorchidism in scrotal mammals, including humans. However, the comprehensive understanding of how the evolutionary and expression patterns of this pathway contribute to the evolution of testicular descent remains limited. Here, we revealed the different evolutionary pattern of Wnt signaling pathway genes and their potential regulatory elements between natural ascrotal and scrotal mammals. Specifically, evolutionary signals which are unique to ascrotal mammals can be found in numerous genes that have a negative regulatory role in the pathway. Further, evidenced by the comparison of transcriptome data between spontaneous cryptorchid orl rats and their scrotal wild-type counterparts, we concluded that the dampened Wnt signaling in mutation- or dosage-dependent manners as well as the subsequent cell migration of the gubernaculum might contributed to the evolution of the ascrotal testes in mammals. Evolutionary signatures and functional validation of Wnt pathway alterations in ascrotal species provide mechanistic insights into mammalian sexual dimorphism and cryptorchidism pathogenesis. Supplementary Material Acknowledgments We thank all the members of Center for Marine Evolutionary and Conservation Biology in Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou) for suggestions and support during the project. Contributor Information Simin Chai, Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou, Guangdong, China. Cuijuan Zhong, Jiangsu Key Laboratory for the Biodiversity Conservation and Sustainable Utilization in the Middle and Lower Reaches of the Yangtze River Basin, College of Life Sciences, Nanjing Normal University, Nanjing, Jiangsu, China. Zhenhua Zhang, Jiangsu Key Laboratory for the Biodiversity Conservation and Sustainable Utilization in the Middle and Lower Reaches of the Yangtze River Basin, College of Life Sciences, Nanjing Normal University, Nanjing, Jiangsu, China. Daiqing Yin, Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou, Guangdong, China. Shixia Xu, Jiangsu Key Laboratory for the Biodiversity Conservation and Sustainable Utilization in the Middle and Lower Reaches of the Yangtze River Basin, College of Life Sciences, Nanjing Normal University, Nanjing, Jiangsu, China. Wenhua Ren, Jiangsu Key Laboratory for the Biodiversity Conservation and Sustainable Utilization in the Middle and Lower Reaches of the Yangtze River Basin, College of Life Sciences, Nanjing Normal University, Nanjing, Jiangsu, China. Guang Yang, Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou, Guangdong, China; Jiangsu Key Laboratory for the Biodiversity Conservation and Sustainable Utilization in the Middle and Lower Reaches of the Yangtze River Basin, College of Life Sciences, Nanjing Normal University, Nanjing, Jiangsu, China. Supplementary Material Supplementary material is available at Molecular Biology and Evolution online. Author Contributions Conceptualization: G.Y., S.C. Methodology: S.C., Z.Z., C.Z., D.Y. Investigation: S.C., Z.Z., C.Z. Visualization: S.C., Z.Z., C.Z. Supervision: G.Y., S.X., W.R. Writing—original draft: S.C., C.Z., Z.Z. Writing—review & editing: S.C., G.Y. Funding This work was supported by: Young Scientists Fund of the National Natural Science Foundation of China (32200345). China Postdoctoral Science Foundation (2022M710878). Guangzhou Basic and Applied Basic Research Foundation (2023A04J0770). National Key Research and Development Program of China (2022YFF1301601). PI Project of Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou) (GML2021GD0805). Key Project of the National Natural Science Foundation of China (32030011). Joint Funds of the National Natural Science Foundation of China (U24A20362). Data Availability All data are available in the main text or the Supplementary Materials. References Associated Data Supplementary Materials Data Availability Statement All data are available in the main text or the Supplementary Materials.