ham is expressed and required for Drosophila male RS development

(A). The adult male RS. TE, testis; SV, seminal vesicle; AG, accessory gland; EJD, ejaculatory duct. (B). White-field images of adult male RS, showing defective morphology in the ham mutants. (C). Quantification of male fertility. Ham_sfGFP with the ham genomic region was used to rescue the sterility. Statistical significance was calculated using unpaired t-test (***, p<0.001; **, p< 0.01; *, p<0.05; n.s., non-significant). (D). Images of the adult RS stained with Ham_V5 in red. DNA is stained with DAPI in blue. The blue and white dashed lines highlight the TE and SV respectively. (E). Illustration of the developing male RS at the larval stage. The testis (TE) terminal epithelial cells localize at the opposite side of the germline stem cell (GSC) niche. The genital disc (GD) contains mesenchymal cells (MC) that will develop into the SV and accessory gland (ag). (F). Images of the larval TE (top) and GD (bottom) stained with Ham_V5 in red or magenta, the reporter in lime-green (btlGal4 driving UAS-GFP) and DAPI in blue. AEL, after egg laying. (G). Illustration of the TE and SV fusion process. APF, after pupal formation. (H). Images of the pupal GD stained with Ham_V5 in magenta, the reporter in lime-green (btlGal4 driving UAS-GFP) and DAPI in blue. (I) Images of the pupal RS with Ham_V5 in red and DAPI in blue. The blue and white dashed lines highlight the TE and SV respectively. (J). Higher magnification image of the joint site between the SV and TE, with myoblasts and myotubes labeled by mef2Gal4>GFP in lime-green.

Ham promotes differentiation of the TE terminal and SV epithelial cells

(A). Images of the male RS at the pupal stage, showing morphological defects of the TE and SV in the ham mutants. Epithelial cells are labeled with Coracle (Cor). The phenotype penetrance (%) and number of animals (n) counted are indicated as % (n). (B). High magnification images of the connection site of the TE and SV, showing that the ham mutant RS failed to form a continuous tube. Ham-sfGFP restored the normal morphology. Epithelial cells are marked by E-Cad in red, Cor in yellow, and the epithelium on the SV side also marked by btlGal4 driving UAS-mCD8GFP. DNA is labeled by DAPI in blue. (C). Images and quantification of the E-Cad signal in the wt and hamPRΔ mutant SV. (D). Images and quantification of the Crumbs signal in the wt and hamPRΔ /Df mutant SV. (E). Images and quantification of the length of Cor signal in the wt and hamPRΔ mutant SV. (F) Images of ham mutant mosaic clones in the TE terminal epithelial cells. Note, the mutant cells are GFP negative and within the area surrounded by the white dashed line. The white arrows indicate junctions between mutant cells, while the yellow arrows highlight the junctions between wild-type cells. (G). Images of pupal male GD, showing that hamRNAi-mediated knockdown reproduces ham mutant phenotypes. (H). Images of pupal male RS, stained with E-Cad in red, Ham in yellow, GFP in green and DAPI in blue. UAS-dcr2 enhanced knockdown efficiency and led to more severe morphological defects. (I). Images and quantification of E-Cad signal in red and Cor length in blue in control and ham RNAi SV. (J) Images and quantification of E-Cad signal in control and ham RNAi TE terminal epithelium. Statistical significance was calculated using unpaired t-test except for TE mosaic clones which used paired t-test. (***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant).

Identification of Ham-regulated genes in the developing TE and GD

(A). Illustration of the sample types and antibodies used in the RNA-seq and CUT&TAG experiments. (B) Venn diagram of dysregulated genes between the two mutant conditions in the GD (left) and TE (right) samples. (C) Heatmaps of Ham CUT&TAG peak density and IgG control, ranked from high to low and centered at peak summits. (D) Venn diagrams of CUT&TAG peaks from the two types of samples. (E-F). Volcano plots of dysregulated genes in GD (E) and TE (F) in the hamSK1/Df mutant, and the ones with at least one Ham CUT&TAG peak are highlighted.

Ham activates expression of epithelial differentiation genes

(A) Ham gene network analysis from STRING. The position of MCL Clusters were manually adjusted to place genes with similar function in close proximity. The functional categories were determined by gene function tools in STRING. (B). A bar plot of the Log2 fold-change value from RNA-seq for wnt2, shot and shg, showing downregulation of these genes in the hamSK1/Df mutant. (C). Schemes of the Ham_FL and other truncation proteins, and illustration of the luciferase reporter assay. (D-E). A genome browser snapshot of the shg and wnt2 loci, showing the CUT&TAG tracks from the TE and GD (TG) mixed samples and the enhancer fragments in the luciferase assays. (F-G). Normalized luciferase activity to the empty reporter and then to the control vector that expresses GFP and RFP proteins. Statistical significance was calculated using unpaired t-test (****, p<0.0001; ***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant).

A candidate RNAi screen identified Ham-downstream effectors in RS epithelial tissue fusion

(A). Scheme for the RNAi screen. (B) Summary of positive hits. The extent of sterility is determined by the number of progenies from the F1 male crossing to the wild-type female (see Methods). 100% means no progeny. >70%, the number of progenies being fewer than 30% of the number of progenies produced by the control cross (the same Gal4 driver x UAS-GFPRNAi). 40-70%, the number of progenies being between 30-60% of the number of progenies produced by the control. (C). Images of the control and gene-specific RNAi knockdown RS. The left panels show the fusion site stained with E-Cad in magenta, GFP in lime-green, Ham in grey and DAPI in blue, the middle panels are GD, and the right ones are TE. The orange arrows indicate mobilized epithelial cells in the shg knockdown TE.

Spatial-temporal expression dynamics of Ham-downstream genes

(A). Images from multiplexed in situ hybridization (the SCRINSHOT method), showing mRNA signal of the indicated genes in wt TE samples. DNA is marked by DAPI in grey. (B). Images from SCRINSHOT, showing mRNA signals of indicated genes in wt/Df and hamSK1/Df mutant GD. The last panel is an enlarged view from the sample with Tl, ham and spz probes. The yellow arrowheads indicate myoblast cells with Tl signal. (C). Quantification of signal intensity for genes shown in B in two areas, the SV and EjD. Statistical significance was calculated using unpaired t-test (***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant). (D) Images from SCRINSHOT, showing mRNA signals of indicated genes in GFP and ham RNAi GD samples. (E). Quantification of signal intensity for genes shown in D in two areas, the SV and EjD. (F). A summary of genes expressed in the TE terminal and SV epithelium. Highlighted genes show temporal or spatial dynamic expression in the two tissues. Statistical significance was calculated using unpaired t-test (****, p<0.0001; ***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant).

Wnt2 and its ligand genes are required for male fertility and RS development

(A) Images of wnt2o mutant TE and GD, showing disrupted morphology in these tissues. (B). Images of the TE and SV fusion site from the control and wnt2 overexpression animals. The epithelial cells of the TE and SV are labeled with hamRSGal4 driving UAS-mCD8GFP in lime-green and E-Cad in magenta and Ham in grey. The white arrows indicate mobilized GFP positive cells in the middle of the TE, and these cells have a lower level of E-Cad. The blue and white dashed lines highlight the TE and SV respectively. (C). Images of the TE and SV fusion site from the control and fz2 RNAi animals. The white arrows indicate mobilized GFP positive cells in the middle of the TE, and these cells have a lower level of E-Cad. (D) Graphs of progeny number from male fertility tests for fz and fz2 RNAi lines in comparison to GFP RNAi. Number of tested males in each genotype, 3. (E) Illustration of cell-cell interactions in the developing RS, and a speculated model for Tl and Wnt2 signaling functions. Wnt2 signaling may be involved in epithelial cell interaction between the two ends. Tl signaling and Tl may be involved in the interaction between epithelial cells and myoblasts/myotubes. In three conditions (shgRNAi, wnt2OE and fz2RNAi), some TE terminal epithelial cells (illustrated in orange) move up to the distal end of the testis.

Ham gene locus and mutant alleles.

(A) Illustration of the ham genomic locus and predicted protein isoforms, highlighting the mutation sites for the three mutant alleles and the protein domains of each isoform. All Ham protein isoforms maintain a functionally unknown motif PTZ00121 and the C-terminal zinc finger (ZF) clusters, while the long isoforms B/D and G also comprise the PR domain, the N-terminal ZF clusters. Isoform H has a truncated PR domain, and the shortest isoform E lacks the PR domain and the N-terminal ZF clusters. (B) Flanking sequences of the sgRNA target site from wt and four alleles with indels. Red letters indicate the sgRNA sequence. The corresponding amino acid sequence is shown above. “*” indicates the stop codon. (C). An image from western-blot, showing the two Ham protein bands in the control samples and absence of the Ham band(s) in the indicated mutants. “*” indicates an unspecific band, used as the loading control. (D) A table showing viability and sensory organ (SO) defects in four ham alleles. Note: the - 6nt allele has a 6 nucleotide in-frame deletion. (E). Mechano-sensory organ phenotypes in the three indicated alleles. Red arrowheads indicate double sockets and double shafts phenotype.

Ham expression and function in the female RS

(A). Bright-field images of the adult female RS from the indicated genotypes, showing accumulated eggs in the ham mutant ovaries. (B). Quantification of the number of progenies produced by wild-type and ham mutant females after crossed with wild-type males. Statistical significance was calculated using unpaired t-test (***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant). (C). Images of pupal and adult ovaries and oviducts, showing high expression of the Ham_V5 protein in the connecting sites of the ovaries and oviducts.

Split channel images for images in Fig. 2

Ham controls the formation of TE epithelial cells

(A). Images of the testes from the early L3 larval stage, showing the absence of or mis-polarized TE terminal epithelial cell cluster in the ham mutant. E-Cad marks epithelial cells in green and DAPI marks the nuclei. The white arrows indicate mis-polarized epithelial cells. (B). Illustration of the precursor cell clusters of the male RS in the embryonic stages. PGC, primordium germ cells; SGPs, somatic gonadal precursors; msSGP, male-specific SGPs. (C-E) Images of stage S13 and S15 wt embryos, showing Ham staining in all SGPs. Abd-B and Sox100B label msSGPs, VASA labels PGCs. (F-G). Images of the wt and ham mutant embryos at S13 (F) and S15 (G), showing the reduction or absence of msSGPs in the hamSK1 mutant. Vasa is a germ cell marker, and Sox100B labels msSGPs in lime green.

An epithelial-specific Gal4 line in the ham locus

(A). The ham gene locus, indicating the fragments in the Gal4 transgenes. The one (hamRSGal4) that drives reporter gene expression in the RS epithelial cells is highlighted in orange. (B). Images of pupal genital disc and TE in animals with hamRSGal4 and UAS-mCD8GFP. Ham_V5 is stained in magenta, and DAPI in blue. (C). Images showing the presence of myotubes on the surface of the testis. The bar graph on the right shows the relative abundance of myotubes on the testis in wt (100%) and ham mutant condition quantified from five animals each. (D). Images showing the junction between the TE and SV, and myotubes are present even in the ham mutant. The white arrows indicate myotubes with two or three nuclei at the junction.

Ham regulatory activity and potential co-factors

(A). Principle component analysis of the RNA-seq datasets. (B). Principle component analysis of the Cut&Tag datasets. Note: some samples had poor sequencing depth, which were excluded from the PCA and later analyses. (C). The shot locus, showing the CUT&TAG tracks from the TE and GD (TG) mixed samples and the enhancer fragments in the luciferase assays. (D). Normalized luciferase activity to the empty reporter and then to the control vector that expresses GFP and RFP proteins. Statistical significance was calculated using unpaired t-test (****, p<0.0001; ***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant). (E). Significantly enriched motifs from Ham binding peaks in up-regulated genes (left) and downregulated genes (right). The dashed rectangles highlight MAD and ZNF263 motifs.

Split channel images for images in Fig. 5

Split channel images for images in Fig. 5

Expression of genes tested by SCRINSHOT in RNA-seq data

(A) Log2FC (hamSK1/Df versus wt/Df and hamPRΔ versus wt) in TE. (B) Log2FC(hamSK1/Df versus wt/Df and hamPRΔ versus wt) in GD.

Split channel images for images in Fig. 6

Additional genes tested in SCRINSHOT

(A). Images from SCRINSHOT, showing mRNA signal of the indicated genes in wt TE samples. DNA is marked by DAPI in grey. (B). Images from SCRINSHOT, showing mRNA signal of the indicated genes in wt TE samples at a later pupal stage (28-32 APF). DNA is marked by DAPI in grey. (C). Images from SCRINSHOT, showing mRNA signals of indicated genes in wt/Df and hamSK1/Df mutant GDs. (D). Images from SCRINSHOT, showing mRNA signals of indicated genes in GFP and ham RNAi GD samples. (E-F). Quantification of signal intensity for genes shown in C-D in two areas, the SV and EjD. (G). Quantification of signal intensity for genes that changed expression in ham RNAi TE. Statistical significance was calculated using unpaired t-test (***, p<0.001; **, p< 0.01; *, p<0.05; ns, non-significant).

Split channel images for images in Suppl._Figure 11

Split channel images for images in Suppl._Figure 11

Male fertility test for Toll-9

(A). Graphs of progeny number from male fertility test for hamRSGal4 driving Toll-9 RNAi in comparison to GFP RNAi. The number of tested males =3. (B). White-field image of Toll-9 RNAi adult male RS, showing morphological defects of the TE and the phenotype penetrance (11% of 37 animals). (C). Images of adult RS from control GFP RNAi andToll-9 RNAi males. hamRSGal4>UAS-mCD8GFP is used to label the testis terminal epithelium and the SV, and DAPI stains DNA. The panels below are higher magnification views on the selected SV area, highlighted with white dashed squares. White arrows indicate sperms in the control SV, which are absent from the Toll-RNAi sample. The bar graph shows the percentage of SV with sperms in GFP and Toll-9 RNAi animals (N=10).

PRDM16 and MECOM/PRDM3 expression in human tissues

(A). RNA expression data across human tissues through Cap analysis of gene expression (Fantom5 dataset from the Human Protein Atlas). (B). Expression in Single cell RNA-seq clusters in the indicated human tissues (from the Human Protein Atlas).