Fig 2: A model illustrating the suggested roles of N-WASP and branched-actin polymerization in BTB dynamics.(Left) The BTB, composed of junctional and cytoskeletal structural and regulatory elements, provides a seal between the basal (down) and adluminal (up) compartments of the epithelium in seminiferous tubules. (Middle) At stage VIII of the epithelial cycle, the BTB undergoes restructuring, involving near simultaneous dismantling of the “old” BTB via endocytosis, to accommodate the apically directed transport of preleptotene spermatocytes between neighboring Sertoli cells, and assembly of a “new” BTB, to maintain the barrier. These events are mediated by the branched actin regulators drebrin E and Arp3, which localize to the BTB, and the nucleation promoting factor N-WASP, which activates branched-actin polymerization. The proposed role for the branched-actin arrays is to facilitate endocytic vesicle-mediated protein trafficking, so that integral membrane proteins at the “old” BTB can be recycled to the basal region and contribute to assembly of a “new” BTB. (Right) In N-WASPSC-cKO testes lacking Sertoli-cell N-WASP function, the branched-actin polymerization machinery can no longer mediate proper endocytic vesicle-mediated protein recycling, and the BTB, once dismantled, cannot be regenerated.

Fig 3: ABLIM1 knockout impairs actin polymerization. (A) ABLIM1-KO or control HCCLM3 cells were fixed and stained with primary antibodies against ABLIM1 proteins, and immunocomplexes were detected with Alexa488-conjugated secondary antibodies and Alexa594-conjugated phalloidin respectively. Nuclei were stained with DAPI. Fluorescence images were photographed by confocal microscopy. (Scale bar: 10 µm). (B) Cells with actin-rich regions were quantified by chi-square test from at least 50 cells per group. (C) The plasmids encoding GPF fused to the NLS epitope (GFP-NLS) and MKL1 tagged with mCherry were co-transfected into ABLIM1-KO or control HCCLM3 cells, followed by starvation for 2 hrs with 1%FBS media. Shown are representative live images from three independent experiments (Scale bar: 10 µm). (D) Subcellular localization of MKL1-mCherry in each group was quantified by chi-square test. n=50. (E) HEK293T cells transfected with two siRNAs against ABLIM1 were transfected with PGL433 and beta-actin Renilla plasmids. Luciferase activity was measured at 24 hrs post-transfection. The efficiency of ABLIM1 knockdown was examined by real-time PCR assay. The results shown are summarized from three independent experiments in duplicate. P-values were determined by Student's t-test. ***: p<0.0001; **: p<0.001. (F) Real-time PCR was performed to assess expression of Arp3 and c-Fos in ABLIM1-KO or control HCCLM3 cells. P-values were determined by Student's t-test. ***: p<0.0001. (G) The mRNA levels of Arp3 and c-Fos were examined by real-time PCR experiments in ABLIM1-KO cells transfected with plasmids encoding ABLIM1 (ABLIM1 OE) or control vector (Ctr). (H) The gene expression correlation of ABLIM1 and Arp3 or c-Fos in HCC tissues was analyzed from the TCGA dataset. Relative mRNA levels of ABLIM1 were respectively plotted against those of Arp3 or c-Fos. R represents a Pearson coefficient and the p-value is shown.

Fig 4: Both Ser 214 and Ser 431 residues of ABLIM1 are involved in the nuclear accumulation of MKL1 and activation of Arp3. (A) HCCLM3 cells transfected with a series of plasmids expressing ABLIM1 mutants were stained with anti-MKL1 antibodies, followed by incubation with Alexa488-conjugated secondary antibodies. Representative images were photographed using confocal fluorescence microscopy. (Scale bar: 20 µm). (B) The ratio of MKL1 immunofluorescence intensity in nucleus (N), nucleus and cytoplasm (N+C), or cytoplasm (C) among total cells was quantified by chi-square test. At least 50 cells were counted. ***: p<0.0001. (C) Real-time quantitative PCR was carried out to evaluate the relative mRNA levels of ABLIM1 and Arp3 in ABLIM1-KO cells transfected with indicated dominant negative mutants of ABLIM1. Differences between the different groups were quantified by one-way ANOVA test. (D) Schematic diagram illustrating a working model for ABLIM1 phosphorylation induced by Rictor involving in actin polymerization. In control cells, Rictor interacts with ABLIM1 and phosphorylates ABLIM1, thereby leading to actin polymerization and accumulation of MKL1 in the nucleus, where it promotes expression of Arp3. In HCC cells with Rictor knockdown, phosphorylated levels of ABLIM1 are dramatically decreased, which results in reduction of F-actin formation. Free G-actin can bind to MKL1 in the cytosol, resulting in its cytoplasmic retention. As a consequence, MKL1 fails to be translocated into the nucleus and activate Arp3 expression.

Fig 5: The branched microfilament regulators Arp3 and drebrin E are properly localized to the BTB in N-WASPSC-cKO tubules.(A) Immunoblotting data using lysates of testes from N-WASPSC-cKO and age-matched control mice demonstrate up-regulation in Arp3 and drebrin E expression in mutant mouse testes vs. the age-matched control testes with ß-actin served as a protein loading control. Histogram generated as in Figure 1. **, P<0.01. (B) Immunofluorescence analysis of Arp3 (green fluorescence, left) and drebrin E (green fluorescence, right) in the seminiferous epithelium of control (WT) and N-WASPSC-cKO mice. Yellow arrowheads mark the localization of either Arp3 or drebrin E at the BTB, which is near the basement membrane of the tunica propria (annotated by a broken white-line). Unlike basal ES proteins (see Figure 4) and apical ES proteins (see Figure 5), which were mis-localized in the seminiferous epithelium of mutant testes, these branched actin regulatory proteins remained properly localized to the damaged BTB. Cell nuclei were visualized by DAPI staining. Scale bars: 80 µm in the first two rows, and 240 µm in the third row.