Loss of Profilin3 Impairs Spermiogenesis by Affecting Acrosome Biogenesis, Autophagy, Manchette Development and Mitochondrial Organization

Abstract

Profilins (PFNs) are key regulatory proteins for the actin polymerization in cells and are encoded in mouse and humans by four Pfn genes. PFNs are involved in cell mobility, cell growth, neurogenesis, and metastasis of tumor cells. The testes-specific PFN3 is localized in the acroplaxome-manchette complex of developing spermatozoa. We demonstrate that PFN3 further localizes in the Golgi complex and proacrosomal vesicles during spermiogenesis, suggesting a role in vesicle transport for acrosome formation. Using CRISPR/Cas9 genome editing, we generated mice deficient for Pfn3. Pfn3^-/- males are subfertile, displaying a type II globozoospermia. We revealed that Pfn3^-/- sperm display abnormal manchette development leading to an amorphous sperm head shape. Additionally, Pfn3^-/- sperm showed reduced sperm motility resulting from flagellum deformities. We show that acrosome biogenesis is impaired starting from the Golgi phase, and mature sperm seems to suffer from a cytoplasm removal defect. An RNA-seq analysis revealed an upregulation of Trim27 and downregulation of Atg2a. As a consequence, mTOR was activated and AMPK was suppressed, resulting in the inhibition of autophagy. This dysregulation of AMPK/mTOR affected the autophagic flux, which is hallmarked by LC3B accumulation and increased SQSTM1 protein levels. Autophagy is involved in proacrosomal vesicle fusion and transport to form the acrosome. We conclude that this disruption leads to the observed malformation of the acrosome. TRIM27 is associated with PFN3 as determined by co-immunoprecipitation from testis extracts. Further, actin-related protein ARPM1 was absent in the nuclear fraction of Pfn3^-/- testes and sperm. This suggests that lack of PFN3 leads to destabilization of the PFN3-ARPM1 complex, resulting in the degradation of ARPM1. Interestingly, in the Pfn3^-/- testes, we detected increased protein levels of essential actin regulatory proteins, cofilin-1 (CFL1), cofilin-2 (CFL2), and actin depolymerizing factor (ADF). Taken together, our results reveal the importance for PFN3 in male fertility and implicate this protein as a candidate for male factor infertility in humans.

Keywords: acrosome biogenesis; autophagy; globozoospermia; male fertility; profilin 3; sperm biology.

FIGURE 1

Immunogold labeling of PFN3 in developing mouse spermatid during acrosome biogenesis. (A) cis- and trans-Golgi network; immunogold particles in cg and tg = black arrows. (B) Golgi phase; immunogold labeling highlighted in ag = circles, acp = white stars, and n = arrow head. (C) Cap phase; immunogold particles highlighted in av = asterisk. (D) Man immunogold labeling = white arrow heads. cg, cis-Golgi; tg, trans-Golgi; ag, acrosomal granule; n, nucleus; acp, acroplaxome; av, acrosomal vesicle. Scale bar = 500 nm. (A,B) are from the same image (ref.to Supplementary Figure 2A).

FIGURE 2

Generation and characterization of Pfn3-deficient mice. (A) Schematic representation of Pfn3 genomic locus with targeting sites of designed gRNAs. (B) Mating statistics of wild-type, Pfn3Δ254, Pfn3Δ41, and Pfn3Δ29 heterozygous and homozygous (n = 5/Pfn3-deficient mouse model) males. Successful mating of heterozygous and homozygous males with wild-type females was indicated by the presence of a vaginal plug at 0.5 dpc. (C,D) Relative weights of testes and cauda epididymis are comparable between all three genotypes of Pfn3Δ254 (n = 13). (E) H & E staining on Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testes section. (F) Sperm count comparison in Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– littermates (n = 13). (G) Eosin and nigrosine staining on biological replicates (n = 3) of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm. (H) Hypo-osmotic swelling test on biological replicates (n = 3) of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm. At least 200 spermatozoa were evaluated per sample.

FIGURE 3

Amorphous nuclear morphology of sperm cells in Pfn3-deficient mice. (A) Sperm cells stained with DAPI. (B) Area, (C) circularity, (D) length of hook, (E) bounding width, and (F) regularity of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm cells. (G) Head shape of Pfn3^+/+ sperm cells, two clusters for Pfn3^+/– sperm cells, and four clusters for Pfn3^–/– sperm cells.

FIGURE 4

Impaired acrosome biogenesis in Pfn3-deficient mice. Adult testis sections of Pfn3^+/+ (left panel), Pfn3^+/– (middle panel), and Pfn3-deficient (right panel) mice; the developing acrosome was labeled with PNA-FITC (green) and cell nuclei were stained with DAPI (blue); inset panel showed the Golgi/cap/acrosomal phase of the tubules displayed in the main panel (n = 3). In the Golgi phase, proacrosomal granule (green) labeled by PNA-FITC for (A) Pfn3^+/+, (B) Pfn3^+/–, and (C) Pfn3^–/– round spermatozoa. In the cap phase, acrosomal caps (green) stained for (D) Pfn3^+/+, (E) Pfn3^+/–, and (F) Pfn3^–/– round spermatozoa (white arrows show fragmented cap structures). In the acrosome phase, PNA-FITC-labeled acrosomal area on (G) Pfn3^+/+, (H) Pfn3^+/–, and (I) Pfn3^–/– elongated spermatids. Scale bar = 20 μm. (J) Ultrastructure analysis using TEM revealed the acrosomal structures of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm cells. Scale bar = 2 μm. (K) Immunofluorescence staining using PNA-FITC (green) on epididymal sperm cells of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– mice (n = 3). Scale bar = 20 μm. (L) TEM evaluation shows the percentage of malformed acrosomes in Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm cells (n = 2). (M) A graph represents the PNA-stained defective acrosome percentage of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm cells (n = 3). Two hundred spermatozoa were counted per genotype.

FIGURE 5

The acrosomal reaction (AR) using calcium ionophore. (A) Coomassie-stained sperm cells of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/–. Black arrow heads indicate successful acrosomal reaction took place. A white arrow indicates a crescent-shape acrosome on sperm head, indicating acrosomal reaction did not take place. (B) Percentage of acrosomal-reacted sperm for Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– (n = 3 biological replicates/genotype, ^∗∗∗p < 0.0005, Student’s t-test, one tail, paired).

FIGURE 6

Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testis sections stained for the cis- (GM130 antibody) and trans- (TGN46 antibody) Golgi compartment (green) and nuclei (DAPI, blue) (n=3). (A) WT and (B) heterozygous, and (C) Pfn3^–/– spermatozoa stained for cis-Golgi compartment (asterisks). (D) WT, (E) heterozygous, and (F) Pfn3^–/– spermatozoa stained for trans-Golgi network. Scale bar = 50 μm.

FIGURE 7

Changes in gene expression profile of Pfn3-deficient mice. (A) Heat map visualization of top 38 differentially expressed (DE) genes obtained by RNA-seq on Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testes. (B) Volcano plots displaying DE genes for Pfn3^+/+ vs. Pfn3^–/– [adjusted p-value < 0.05; log2 fold change of expression (LFC) > 1.5]. (C) DE genes obtained by RNA-seq were verified by qRT-PCR on Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testicular RNA (n = 3 biological replicates/genotype).

FIGURE 8

Disruption in autophagic flux and AMPK/mTOR signaling pathway of Pfn3-deficient mice. (A) Immunoblot analysis against ATG2A, LC3B, SQSTM1, mTOR, phospho-mTOR, and phospho-AMPKα on protein lysates from Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testes. (B) Immunohistochemical staining against phospho-mTOR on Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testis sections (top row). (C) Immunohistochemical staining against LC3B on Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testis sections (bottom row). Staining of testicular tissue sections from Pfn3^+/+ (left column), Pfn3^+/– (middle column), and Pfn3^–/– (right column) animals is shown. Scale bar = 100 μm.

FIGURE 9

Co-immunoprecipitation using anti-PFN3 and anti-TRIM27 antibody on testis lysates. (A) Silver-stained SDS-PAGE; lane 1: input control (protein lysate), lane 2: negative control (only antibody), lane 3: IP using PFN3 antibody, lane 4: IP using PFN3 antibody, and lane 5: flow through. (B) Western blot showed PFN3 and TRIM27 proteins. Lane 1: input control. Lane 2: immunoblot of reciprocal IP for TRIM27 (band observed at ∼58 kDa) and PFN3 protein levels (band observed at ∼14 kDa) using anti-PFN3 and anti-TRIM27 antibodies, respectively. Lane 3: flow through. Lane 4: negative control.

FIGURE 10

Immunoblotting against ADF, CFL1, and CFL2 on protein lysates from Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– testes.

FIGURE 11

Manchette structure stained by using α-tubulin antibody (green) on germ cell population isolated from (A) Pfn3^+/+, (B) Pfn3^+/–, and (C) Pfn3^–/– testes (n = 3/genotype). Nuclei were stained with DAPI (blue). Scale bar = 20 μm.

FIGURE 12

Flagellum analysis on mature sperm cells from Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– mice. Mito Red staining (red) of sperm flagellum on (A) Pfn3^+/+; (B) Pfn3^+/–; and (C–E) Pfn3^–/– sperm cells isolated from cauda epididymis. Scale bar = 10 μm. Ultrastructural analysis using TEM on (F) Pfn3^+/+; (G) Pfn3^+/–; and (H–J) Pfn3^–/– sperm cells isolated from cauda epididymis. Vacuolated mitochondria are shown by asterisk, and fibrous sheet that contained more than one axoneme–mitochondrial complex is shown by arrows. Scale bar = 5 μm. Surface analysis using SEM on (K) Pfn3^+/+, (L) Pfn3^+/–, and (M,N) Pfn3^–/– mature sperm cells isolated from cauda epididymis. Scale bar = 50 μm. (O) Statistical analysis of Mito Red-stained flagellum of Pfn3^+/+, Pfn3^+/–, and Pfn3^–/– sperm cells. Data is presented as mean ± SD using ANOVA (Tukey’s post hoc) (*p < 0.05).

FIGURE 13

Working hypothesis on the PFN3 role in acrosome biogenesis. (A) Schematic illustration of the Pfn3 presence in Golgi network responsible for proacrosomal formation associated with the autophagy mechanism. (B) Schematic illustration depicting the disrupted autophagy mechanism and acrosome formation in the absence of Pfn3. Black dots = PFN3.